David C. McClelland - Human - Motivation PDF

Preface
The psychology of motivation is a broad and loosely defined field. It covers

everything from detailed investigations of the physiological mechanisms involved
in animal drives to elaborate analyses of the unconscious motives behind abnor-
mal or symptomatic acts in a person to factor analyses of the motives people as-
sign to themselves to explain their behavior. Different textbooks and different
courses have been organized around these different areas of investigation. In this
book we will draw on all these sources of information and attempt to provide an
integrated view of the field by narrowing somewhat the focus of attention.
The book emphasizes how motives differ from other determinants of action
and how they relate to other motivation-type variables such as emotions, incen-
tives, values, causal explanations, and conscious and unconscious intents. It ex-
amines how motives are acquired, where they come from, and on what they are
based. Biological sources of human motives are reviewed, and this review intro-
duces the topic of natural incentives, or what is sometimes called intrinsic moti-
vation. Some selectivity is necessary in reviewing the large field of animal re-
search on motivation in order to focus on biological sources of individual
differences in human motive strength. Social sources of differences in motive
strength are also considered, including everything from the way parents rear
their children to educational interventions designed to change peoples' motives.
Such studies contribute not only practical information on how to develop mo-
tives, but also theoretical information on the nature of motives and how they
differ from other characteristics.
A major focus of the book is on how individual differences in the strength
of human motives are to be measured. The emphasis is on measuring motives in
associative thought or fantasy, a method that combines the sensitivity of Freud's
clinical analyses of motivation with the rigor of experimental psychology, since
the coding Systems for human motives are derived from the effects on associa-
tive thought of experimental arousal of the motives in question. However, alter-
native methods of measuring human motive strength are also reviewed, and all
methods are evaluated carefully in terms of acknowledged criteria for good mea-
surement. What comes out of this analysis is the importance of distinguishing
Downloaded from https://www.cambridge.org/core. University of Sussex Library, on 18 Apr 2018 at 10:06:09, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139878289.001
between more or less unconscious motives and conscious values as different de-
terminants of behavior.
Four major motive Systems—the achievement motive, the power motive, the
affiliative motives, and the avoidance motives are examined in detail. The large
body of research that has accumulated on how these motive Systems affect be-
havior is carefully reviewed and evaluated. The text illuminates not only how
motives serve to energize and Orient behavior and to promote relevant types of
learning; it also contains factual answers to such large questions as, Why do
great civilizations rise and then decline? or, in motivational terms, What is the
difFerence between the reasons for the commercial success of ancient Greece and
the organizational triumph of the Roman Empire? Why are some kinds of peo-
ple successful in business as managers while others are not? What have been the
motivational characteristics of U.S. presidents, and how are those characteristics
related to how they behaved in office? Why do nations make war? Can love re-
ally heal in the physical sense? The research that has been done on these motive
Systems not only provides information of theoretical importance; it also answers
questions of great practical and social significance.
Since motives are treated as only one of the determinants of action, some of
the usual topics in the field of motivation are discussed not in their own right,
but as part of other matters. Take, for example, aggression or the aggressive
drive or instinct, which is often a topic heading in books on motivation. In this
book the usual subject matter for the topic of aggression is dealt with in two
places—as a type of action characteristic of certain kinds of power-motivated
people (Chapter 8) and as a type of action that suggests the existence of a natu-
ral incentive that yields pleasure from having impact (Chapter 5). Other such
behavioral trends suggesting the presence of a motive, like pro-social or altruistic
behavior, are similarly treated. Pro-social behavior appears as characteristic of
people with a certain type of power motive (Chapter 8), and as indicating the
presence of a natural incentive to get pleasure from being with people based on
contact gratifications (Chapter 5) or from some kind of interpersonal exchange
(Chapter 9).
I have benefited greatly not only from the contributions of generations of
students and fellow workers in the field of motivation, but also from the specific
advice of Charles Cofer, Dan McAdams, and Thomas Srull, who read and com-
mented in detail on an early version of the manuscript. In addition, John W.
Atkinson, David G. Winter, Abigail Stewart, David Buss, and Richard Patten
have given me very useful feedback on particular chapters. Of course, none of
these people should be held responsible for my mistakes. I also owe a Special
debt to the National Science Foundation, which provided me with the funds not
only to complete parts of the research reported here, but also to spend a year
free of other academic duties to concentrate on completing the manuscript. Fi-
nally, the book could never have been completed without the devoted, patient,
and conscientious secretarial assistance of Kathleen McPherson and Samantha
George, for which I am very grateful.
David C. McClelland
Foreword
More than most textbooks in psychology, this book reflects the work, the
life, and the personality of its author. After forty prolific years of boldly original
research and theorizing on the topic of human motivation, David McClelland
has not produced a conservative, homogenized, and middle-of-the-road review of
the literature. Like Personality, McClelland's classic textbook on personality psy-
chology written over thirty years ago, this text takes some risks. First, the book
does not aim to review all of the important literature on human motivation;
rather, it seeks to explore in some detail a selected set of critical and intriguing
motivational issues. Second, the book does not merely summarize theories,
methods, and research findings pertaining to the scientific study of human moti-
vation; rather, it attempts a theoretical synthesis of its own based on the au-
thor's particular perspective on human motivation—a perspective that has devel-
oped through a number of stages during the last forty years.
David Winter (1982)—a Student and colleague of McClelland—has recently
traced McClelland's intellectual biography as a psychologist through six stages.
From his rigorous training within the behaviorist tradition of Clark Hüll at Yale
and his early research on verbal discrimination learning, McClelland moved to
the study of thematic measurement of psychological motives (such as the
achievement motive) in the late 1940s. The mid-1950s found him immersed in
the almost audacious investigation of how human motives—reflected in such un-
likely sources as children's readers and ancient urns—shape the economic devel-
opment of entire cultures. By the mid-1960s McClelland and his colleagues were
designing programs to facilitate the development of achievement motivation in
businesspeople; whereas the late 1960s and 1970s witnessed a shift away from
achievement to the study of the power motive and its myriad manifestations in
such phenomena as risk-taking behavior, patterns of leadership in groups, alco-
hol consumption, and war. In the most recent stage, McClelland's work has
moved into yet a new area: the relationship of psychological motives to physio-
logical functioning, especially with respect to the body's immune Systems and
thus sickness and health.
Over the shifting course of McClelland's intellectual journey, a number of
Downloaded from https:/www.cambridge.org/core. UCL, Institute of Education, on 13 Jan 2017 at 15:48:24, subject to the Cambridge Core terms of use, available at
https:/www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139878289.002
Foreword
salient themes have been expressed again and again. As a Student and colleague
of McClelland for only a short while, I have come to perceive eight recurrent
themes that aptly characterize his patented perspective on human motivation.
The careful reader should discern these eight in the pages that follow:
1. A fascination with Freud and the unconscious. For McClelland, like Freud,
the really important motives in human lives reside beneath the surface of
everyday awareness. To understand motivation, therefore, the psychologist
must tap into its subterranean source.
2. A commitment to measurement and quantification of human motives. Like
Clark Hüll and the behaviorists of his day, McClelland has little patience
for things that cannot be measured and transformed into numbers. Given
the well-known difficulties psychologists have encountered in measuring
Freudian constructs, Themes 1 and 2 often exist in a dynamic tension in
McClelland's work as he attempts to devise methods to measure what
some psychologists have claimed is unmeasurable.
3. An adherence to the measurement methodology of content analysis applied
to open-ended responses such as the stories that people teil McClelland is a
pioneer in the development of ways of interpreting the Thematic Apper-
ception Test (TAT) so as to quantify motive trends that exist beneath the
level of awareness. He has repeatedly argued that open-ended measures
such as the TAT, dream analysis, and the coding of myths and stories are
significantly more sensitive to unconscious motive trends than are
self-report measures such as questionnaires and rating scales.
4. An implied dimensional view of people. McClelland is a "trait theorist" in
the best sense of the term. He studies enduring and underlying dimensions
of the personality which motivate (energize, direct, and select) behavior
and experience. The three fundamental motive Systems which he has iden-
tified and measured in human lives concern (a) achievement/success, (b)
power/impact, and (c) affiliation/intimacy.
5. An interest in individual differences. McClelland tends to ask the question,
4
How do people differ?' more often than do most motivational psycholo-
gists. Motivational differences are measured via thematic coding of
open-ended responses, and the differences are often understood in terms of
the three major motive Systems.
6. A preoccupation with major questions of human adaptation. In his research
and his theorizing, McClelland does not shy away from the big questions
about human adaptation: Why do nations make war? Why are some peo-
ple successful and others not? Can love really heal? This leads him into
controversial, but fascinating, studies of healing, history, and mythology.
7. A belief that motives can be changed. McClelland and his colleagues have
developed systematic programs for altering people's motivational profiles.
The programs are seen as creative and cost-effective ways to help people
Foreword
change the internal psychological needs which energize, direct, and select,
their behavior. McClelland's methods have implications for psychological
interventions in business, industry, government, and therapy.
8. A concern about the welfare of society. Like B. F. Skinner and Erik Erik-
son, McClelland has shown a keen appreciation for the intricate interplay
of science and society through history. Laboratory findings are often
couched in terms of their meanings for contemporary society and the im-
minent prospects of human happiness and misery, peace and war. In ad-
dressing the future welfare of humans living on earth, McClelland does not
hesitate to scrutinize the motive patterns of entire nations and entire his-
torical epochs.
McClelland's perspective on human motivation distinguishes this text
from all others in the field. The author's creative blend of tough-minded
empiricism and tender-minded humanism has given birth to a textbook on
human motivation that should challenge, stimulate, and indeed motivate
the Student and the instructor alike.
Dan P. McAdams
Loyola University of Chicago
References
McClelland, D.C. Personality. New York: Holt, Rinehart and Winston, 1951.
Winter, D. G. David C. McClelland: An intellectual biography. In A. J. Stewart (Ed.), Motivation
and society. San Francisco: Jossey-Bass, 1982.
Contents
Preface iii
Foreword v
Part 1 Background 1
1 Conscious and Unconscious Motives 3
Motives as One of Three Major Determinants of Behavior 4
Conscious Intents 6
Unconscious Intents 15
Experimental Study of Unconscious Motives 22
Are Unconscious Motives Important? 25
2 Motives in the Personality Tradition 31
Motives as Reasons for What People Spend Their Time Doing 32
Motives as Reasons for Abnormal Behavior 35
Motives as Reasons for Creativity and Growth 40
Measuring Human Motives 42
Stages in Motivational Development 48
Other Views of Developmental Stages 57
Checking the Validity of the Motivational Stage Theory 58
Contributions of the Personality Tradition 64
3 Motivation in the Behaviorist Tradition 68
Thorndike's Studies of Animal Motivation 69
Hull's Model of How Drives Facilitate Adaptation or Learning 73
Excitatory Potential in Behavior Theory and the Meaning of the Term
Motivation 84
The Behaviorist Model of Motivation Applied to Humans by Spence and
Others 85
Reinterpreting the Behaviorist Studies of Human Motivation in Terms of
What Goes On in People's Minds 95
Comparison of the Psychoanalytic and Behaviorist Contributions to the
Study of Motivation 97
Limitations of the Behaviorist Model 99
Downloaded from https://www.cambridge.org/core. University of Sussex Library, on 13 Jan 2020 at 08:28:51, subject to the Cambridge Core terms of use, available at
https://www.cambridge.org/core/terms. https://doi.org/10.1017/CBO9781139878289
Contents
Part 2 The Nature of Human Motives 105

4 Emotions as Indicators of Natural Incentives 107
Early Attempts to Find a Biological Basis for Motives 108
Sign Stimuli in Ethology as a Basis for Natural Incentives 109
The Case for Natural Incentives in Humans 110
Emotions as Indicators of Natural Incentives 116
Positive Natural Incentives in Infants 122
Classification of Natural Incentives in Terms of the Primary
Emotions 124
Relation of Emotion to Motivation 128
5 Natural Incentives and Their Derivatives 130
How Natural Incentives Influence the Development of the Hunger
Motive 131
Natural Incentives, Emotions, and Motives 136
The Variety Incentive 139
The Impact Incentive 147
The Contact or Sexual Incentives 153
The Consistency Incentive 162
Interaction of Incentives 165
The Role of Cognition in the Development of Incentives 166
Symbolic Incentives, or Values 167
6 Measures of Human Motive Dispositions 172
The Motivational Sequence 173
Arousing Motives to Detect Their Unique Effects on Behavior 185
Measuring the Strength of Social Motive Dispositions 190
Alternative Measures of Motive Strength Evaluated According to the
Criteria for Good Measurement 199
Part 3 Important Motive Systems 221

7 The Achievement Motive 223
Measuring the Need for Achievement 224
Evidence That the Need for Achievement Score Measures a Motive 226
What Is the Incentive for the Achievement Motive? 227
How High Need for Achievement Affects Performance 238
Other Characteristics of People with a Strong Need to Achieve 246
Social Consequences of a Strong Need to Achieve 251
Relationship of the Achievement Motive to the Protestant Ethic and the
Rise of Capitalism 255
Other Influences AfFecting the Achievement Motive Early in Life 260
8 The Power Motive 268
Measuring the Need for Power 269
Evidence that the n Power Score Measures a Motive 271
Contents
Outlets of the Need for Power 280

Catharsis 292
Role of the Power Motive in Drinking 296
How Maturity Modulates the Expression of the Power Motive 302
Controlled and Impulsive Assertiveness in Organizational Behavior 312
Inhibited Power Motive Syndrome and Susceptibility to Illness 320
Origins of the Power Motive 325
9 The Affiliative Motives 333
The Meaning of Love 334
The Sexual Motive 335
Measuring the Sexual Motive in Fantasy 341
The Need for Affiliation 346
Characteristics of People with a Strong Need for Affiliation 348
The Intimacy Motive 359
The Affiliative Motives and Health 366
Origins of the Affiliative Motives 368
Relationship of Sexuality to Affiliation and Intimacy 369
10 The Avoidance Motives 373
Generalized Anxiety as a Motive 374
Fear of Failure 381
Measuring Fear of Failure in Fantasy 386
Comparison of Measures of Fear of Failure 391
Origins of the Fear of Failure 392
Fear of Rejection 393
Fear of Success 397
Fear of Power 404
Other Fears 408
Conclusion 410
Part 4 Contextual Effects on Human Motives 413

11 Motivational Trends in Society 415
Analyzing the Reasons for the Growth and Decline of Civilizations 416
The Collective Concern for Achievement, Entrepreneurship, and Economic
Growth 423
The Collective Concern for Affiliation and Civil Rights 438
The Collective Concern for Power 441
Historical Shifts in Collective Motive Levels 444
Origins of Collective Motivations 456
Difficulties in Interpreting Motive Trends in Society 464
12 Cognitive Effects on Motivation 473
Cognitive Influences on Motive Arousal 474
Motive-related Cognitions 488
Cognitions Affecting the Translation of Motivation Into the Impulse to
Act 504
xii Contents
13 How Motives Interact with Values and Skills to Determine What

People Do 514
Drives, Habit Strength, and Incentives as Determinants of Response
Strength 515
Motives, Expectancies of Success, and Values as Determinants of
Performance 516
How the Achievement Motive, Skill, and Achievement Values Affect
Performance 522
How Motives, Skills, and Values Jointly Determine Success as a Naval
Officer 527
Factors Influencing Affiliative Acts and Choices 527
The Distinction Between Motives and Intents 544
14 Motivation Training 547
Applying Expectancy-Value Theory to Improving Academic
Performance 548
Achievement Motivation Training for Entrepreneurs 553
Achievement Motivation Training in Schools 567
Origin Training in the Classroom 569
Power Motivation Training 577
15 Milestones in the Progress Toward a Scientific Understanding of
Human Motivation 587
Measuring Motives 588
Definition of a Motive 590
Accumulated Knowledge About Three Important Human Motive
Systems 595
Understanding How Motives Combine with Other Characteristics to
Determine Action 598
Some Issues Needing Further Clarification 601
The Relationship of Progress in Psychology to Its Role in Society 606
Bibliography 609
Acknowledgments 641
1
Conscious and
Unconscious Motives
• Motives as One of Three Major Determinants of Behavior
• Conscious Intents
Conflicts in Conscious Intents
Blocked Intent
Conscious Goal Setting
Is Conscious Intent Necessary for Learning?
Factors Underlying Conscious Intents
• Unconscious Intents
Forgetting a Proper Name Unintentionally
Unconscious Motives in Freud
Conclusions to Be Drawn from Freud's Approach to Motivational Analysis
• Experimental Study of Unconscious Motives
• Are Unconscious Motives Important?
• MOTIVES AS ONE OF THREE MAJOR
DETERMINANTS OF BEHAVIOR
What is the subject matter of motivation? From the commonsense point of view,
motivation refers on one hand to conscious intents, to such inner thoughts as, I
wish I could play the piano, I want to be a doctor, and I am trying hard to
solve this problem. On the other hand, looking at behaviors from the outside,
motivation refers to inferences about conscious intents that we make from ob-
serving behaviors. Thus, if we see a young girl perform a connected series of
acts such as walking into a room, drawing up the piano stool, getting out some
music, opening the piano, and starting to play, we infer that she wants to play
the piano. If she stops playing after a while, we infer that she no longer wants
to play the piano. As Marshall Jones (1955) put it in introducing the annual
volumes of the Nebraska Symposium on Motivation, the subject matter of moti-
vation has to do with "how behavior gets started, is energized, is sustained, is
directed, is stopped." Put another way, motivation has to do with the why of
behavior, as contrasted with the how or the what of behavior. We can observe
what the girl is doing, that is, playing the piano. Or we can observe how she is
doing it, that is, what motor skills she is using to play the piano. Or we can try
to determine why she is doing what she is doing.
Of course, when we make inferences from observing behavior about a per-
son's intent, we can arrive at a conclusion about the intent that difFers from
what the person feels her or his intent was. We refer to people's perception of
their wishes as conscious intents, and we infer that their wishes were unconscious
intents if they diflfer from the report of conscious wishes or if people cannot re-
port at all on their intents. The inferences we make about interests from observ-
ing behavior can be wrong; for example, we infer that the girl wants to play the
piano, whereas the same acts would be consistent with her feeling that she is
being compelled to practice.
It is very important to recognize at the outset that there are several kinds of
answers to the question why, only some of which deal with the problem of moti-
vation. A complete answer to the question why must include all the determi-
nants of behavior, not just the motivational ones. To distinguish among the de-
terminants of behavior, it is useful first to realize that any behavioral outcome is
a function of determinants in both the person and the environment. Fritz Heider
(1958) uses the example of a man rowing a boat across a lake to get to the
other side. Getting to the other side (the behavioral outcome) may be deter-
mined partly by the individual who is rowing or partly by wind currents blow-
ing on the boat. If the man did nothing and simply was blown across the lake,
we ordinarily would make no inferences about his motivation—about his desire
to get to the other side. On the other hand, if the day were perfectly calm and
he rowed vigorously, we would attribute the behavioral outcome to his intent to
go across the lake.
A number of recent experimental studies have dealt with the problem of
personal causation. In general, we do not think of people as causing an outcome
if we can find sumcient reason for the outcome in the external environment, as
Conscious and Unconscious Motives
in the example of the wind blowing the boat across the lake (Deci, 1975). How-
ever, if there are no environmental pushes or if such pushes would work against
a particular outcome, and if people have acted in ways that seem to prodüce the
outcome, we are even more apt to attribute the outcome to their actions.
For a concrete example of this principle, consider an autobiographical State-
ment made by Sigmund Freud (1910/1938), one of the most important contribu-
tors to the psychology of motivation, whose work will figure largely in this
chapter and the next: t4For psychoanalysis is my creation; for 10 years I was the
only one occupied with it and all the annoyances which this new subject caused
among my contemporaries has been hurled upon my head in the form of criti-
cism." Freud is saying that all the environmental forces were acting against the
creation of psychoanalysis. He got no help—only criticism—from his contempo-
raries. So it is proper to infer that he personally caused the creation of psycho-
analysis, since he persisted in the face of criticism. Or as Weiner and Kukla
(1970) have shown, "if one succeeds when others fail, or fails when others suc-
ceed, the outcomes are attributed to the person" (Weiner, 1980a). They found,
for example, that if a person succeeded in performing a task at which only 10
percent of the people succeeded, judges overwhelmingly attributed the success to
the person rather than to the characteristics of the task.
Once it has been decided that the person is responsible for an outcome,
when do we attribute motivation to the person? As Heider has pointed out,
common sense distinguishes between effort (the motivational factor) and ability.
A behavioral outcome is jointly determined by a person's efforts and ability to
perform the task. The outcome is also partly determined by the person's under-
standing of the Situation. Jones and Davis (1965) use the example of Lee Harvey
Oswald shooting President John F. Kennedy to illustrate how these three factors
interact to produce an outcome. Before we can infer that Oswald intended or
wanted to kill the President, we must know that he knew how to shoot a gun,
that is, that he had the ability and had not accidentally pulled the trigger. We
also need to know that he understood that the gun was loaded and that if he
pulled the trigger a bullet might enter the President^ head and kill him, as well
as that this would somehow fit into his ideas of what ought to be. These ideas
or expectations usually are referred to as cognitive variables.
Personal causation is made up of cognitions, skills, and motivations or in-
tents. Any general theory of action or of personality must take into account a
person's motives, skills or adaptive traits, and cognitions or Schemas (McClel-
land, 1951). These three types of variables interact in complex ways, as later
chapters will show, but the emphasis in this book is primarily on motivational
variables. More careful definitions of the determinants of behavior will be given
in later chapters. The purpose here is simply to give a general picture of the
types of variables psychologists have used to explain behavior.
Once again let us turn to Freud's autobiography to see how these factors in-
teract to produce an outcome in a concrete case. He said, "I had become a phy-
sician quite reluctantly, but was at that time impelled by a strong motive to help
nervous patients, or at least to learn to understand something of their condi-
tions. I had placed reliance on physical therapy and found myself helpless in the
Human Motivation
face of the disappointments with [it]" (Freud, 1910/1938). Freud made a State-
ment about his motives—his desire to help nervous patients—and a Statement
about a technique or skill (physical therapy) he tried that is considered to be an-
other separate determinant of his actions. More specifically, he tried to help a
patient he called Dora. He knew of an event in her life that he believed had
caused the outbreak of her illness but said, "I tried uncounted times to analyze
the experience, but all that I could receive to my direct demands was the same
scanty and broken description" (Freud, 1910/1938).
Only when he used his new technique of getting the patient to free associate
backward from the scene itself to earlier experiences was he able to understand
and solve the actual conflict. When he got to the root of the problem he discov-
ered it was sexual: 'The fact that a gross sexual . . . transference occurs in
every treatment of a neurosis . . . has always seemed to me the most unshake-
able proof that the forces of the neuroses originate in the sexual life" (Freud,
1910/1938). Freud was describing a third element that played an important part
in his treatment of a patient—namely, his understanding of the Situation or his
sexual theory of the origins of the neuroses. In other words, his treatment or its
behavioral outcome was a Joint function of his motivation to help, a particular
technique or skill he used (the free associative method), and his general under-
standing of the etiology of the neuroses.
As in this particular example, psychologists have shown that the personal
determinants of a behavioral outcome can be broken down into motivational
variables, skill or trait variables, and cognitive variables (beliefs, expectations, or
understandings). A general theory of behavior must include the contribution of
all three elements and their interactions, but this book will focus attention pri-
marily on motivation.
CONSCIOUS INTENTS
Consciously wanting something is an everyday experience. It will be called a

conscious intent to have, to get, or to do something. What people teil themselves
or others they want to do is closely related to what they will do, provided the
intent refers to the here and now. If a man in a clothing störe says he wants to
buy a shirt, the chances are very good he actually will go to the shirt depart-
ment and buy a shirt. If a woman in an automobile says she wants to get some
gas, her Statement of intent is excellent evidence that she will in fact buy gas
and not a shirt. Psychological studies have shown that conscious intents in the
here and now correlate about .95 with actions taken subsequently in the here
and now (Ryan, 1970; Locke & Bryan, 1968). As a recent example of this well-
known fact, consider a study reported by Smetana and Adler (1980). The inves-
tigators questioned a large number of women awaiting the results of a preg-
nancy test as to whether they did or did not intend to have an abortion if the
test was positive. The stated advance intention of the fifty-nine women with pos-
itive tests correlated .96 with whether they had the abortion or not.
The reason such intents predict actions so well in the here and now is that
they take into account not only motivation, but also the other determinants of
action, as Chapter 6 will show. That is, the environmental determinant is pres-
ent. The man is in the störe; the abortion is available. Also, the skills necessary
for performing an act like buying a shirt or going to see a doctor are available.
So are the cognitive determinants of the act: the customer understands what a
shirt or an abortion is for. Thus, conscious intents are not pure indications of
the motivation involved. They are a product of the motivation (including its un-
conscious aspects, considered later in this chapter) and other determinants of ac-
tion as well. Historically, however, intents have played an important role in the
way psychologists have studied motivation.
Besides demonstrating the obvious point that conscious intents influence ac-
tual choices, early psychologists investigated the strength of conscious intents.
Narziss Ach (1910) approached this problem by pitting an intent against a well-
practiced habit. He had subjects learn a number of pairs of nonsense syllables
that rhymed (for example, dak-tak). After the subjects had practiced learning
these pairs for some time, he would introduce a new set; for example, he would
ask the subjects to respond to the first nonsense syllable with its mirror image
(dak-kad). He wanted to measure the strength of the new intent by discovering
how many practice trials on the rhymed association task were necessary to
break through the new set. That is, if the first, or rhyming, task had been prac-
ticed only a few times, it was easy to maintain the new intent (mirror image
learning) without interference from the old one. As the number of practice trials
on the first task increased, however, errors from that type of learning crept
more readily into the new learning set, interfering with the intent to produce the
mirror image of the first nonsense syllable.
Ach thought he was measuring the strength of will by pitting it against an
old habit, but Lewin (1935) pointed out that there was an intent involved in the
first task also, and that really a conflict existed between two intents. He took the
Position taken by this book: A habit (for example, an associative link between A
and B) does not contain a motivational force of its own, as some association
theorists have argued. Rather, an association is an aspect of the determination of
a response that is conceptually distinct from the behavioral intent. Thus, in ana-
lyzing Ach's experiment, think in terms of a conflict between the old intent to
say dak-tak and the new intent to say dak-kad.
Conflicts in Conscious Intents

Lewin's interest in the conflict of intents led him and his students to do a num-
ber of studies on motivational conflicts. He introduced a very elaborate System
of notation for describing motivational forces in a psychological field, only a lit-
tle of which is relevant here. Table 1.1 illustrates Lewin's contention that the in-
tent, or psychological force, to perform an act was a product of two person vari-
ables (need and valence) divided by an environmental variable (psychological
distance). Need meant the desire for some end State; valence meant the reward
value of the end State; and psychological distance referred to all the difficulties
involved in performing a task or in adopting the means necessary to get to the
goal. Table 1.1 illustrates how this conceptual model explains the characteristics
of different types of motivational conflicts.
8 Human Motivation
Table 1.1.
LEWIN'S MODEL OF MOTIVATIONAL CONFLICTS (after Lewin, 1935)
_ . Need X Valence
Force toward an action = Psychological distance
Approach-approach conflict: deciding whether to stay home and write or go out to the
opera.
Needs fame (10) X Farne from book (3) 30
= 5.
Psychological distance in difficulty in writing (6) 6
Needs music (2) X Enjoyment from opera (5) 10

Psychological distance in going out to opera (2) 2
Avoidance-avoidance conflict: deciding whether to endure Jung's criticisms or the

rejection of the scientific world.
Needs scientific accuracy (10) X Jung's critique ( — 5) —50
Psychological distance in accepting Jung (5) 5
Needs acceptance of psychoanalysis (10) X Anti-Semitic rejection ( — 5) —50

Psychological distance in correcting Jung (5) 5
= -10.
Approach-avoidance conflict: deciding whether to teil the truth or avoid Breuer's

disapproval in writing up the Dora case.
Needs scientific accuracy (6) X Truth about Dora's sexuality (5) 30
Psychological distance in writing up Dora case (5) 5
Needs Breuer's friendship (5) X Breuer's disapproval ( — 6) —30

= —6
Psychological distance in writing up Dora case (5)
An approach-approach conflict is unstable and easily solved. The traditional

example of a donkey who starved because he was Standing equidistant from two
equally attractive piles of hay is incorrect. As a further example, suppose Freud
is trying to decide whether to stay home and work on his book or go to an
opera for the evening; both alternatives hold some attraction for him. If we as-
sign appropriate weights to the variables in Lewin's formula, we can equalize
the attractiveness of the two alternatives. On the one hand, he is an ambitious
man who needs fame (let us assign that a value of 10) and knows that he will
get some fame from Publishing this book (3); however, there is considerable dif-
ficulty involved in writing (6), which somewhat reduces the Overall attractiveness
of this alternative. The formula determines the attractiveness of Freud's staying

home and writing as (10 X 3)/6 = 5.
On the other hand, Freud may need music (2) less than fame and enjoy the
opera (5) somewhat more than writing, but the difficulty in going out to the
opera (2) is much less than the difficulty in writing. This increases the attrac-
tiveness of this alternative, making its overall attractiveness the same as staying
home and writing, or (2 X 5)/2 = 5.
Note, however, that moving in one direction or the other toward either al-
ternative immediately reduces the psychological distance, making that alternative
more attractive and solving the conflict. If Freud puts on his coat in preparing
to go out, he has reduced the psychological distance to the goal of going to the
opera and is likely to continue in that direction. If he Starts writing, however,
the difficulty associated with that alternative is reduced, making it the likely
choice for the evening. All the donkey has to do is accidentally move toward
one pile of hay or the other for that choice to be more attractive.
An avoidance-avoidance conflict tends to be very stable. Early in the history
of the psychoanalytic movement Freud was pleased to gain the support of an
energetic young Swiss psychiatrist, Carl Jung. Freud was very sensitive to the
scientific world's rejection of his sexual theories and feit some of the criticism
was motivated by anti-Semitism, since all the early psychoanalysts were Vien-
nese Jews. He feit that Jung, who was not Jewish, was a very important ally
and arranged for him to be president of the Psychoanalytic Association. How-
ever, Jung soon began to differ with Freud and introduce ideas of his own. This
put Freud into an avoidance-avoidance conflict. On the one hand, he was very
upset by Jung's new ideas, which he feit were wrong and would undermine or
dilute his most basic insights.
On the other hand, Freud wanted to avoid the criticism of the Community
by taking advantage of Jung's value as a non-Jewish supporter of psychoanaly-
sis. As the formulas in Table 1.1 show, Freud's moving toward continuing to ac-
cept Jung would decrease the psychological distance in the first alternative, mak-
ing the reality of Jung's deviationism even more painful. Thus, if he started to
do this he would immediately back off, as this alternative would become more
unpleasant than the other: (10 X —5)/5 = —50/5 = —10). On the other
hand, if Freud moved toward correcting Jung or even removing him from the
Psychoanalytic Association presidency, the reality of rejection by the scientific
Community would be even stronger. Having moved in this direction, Freud
would find it even more unpleasant and would back off again to increase the
psychological distance to this alternative. Thus, he would tend to vacillate, try-
ing to avoid first one unpleasant alternative and then the other. This in fact hap-
pened over a number of years as Freud tried to resolve this avoidance-avoidance
conflict. As Lewin pointed out, avoidance-avoidance conflicts are serious only if
a person cannot escape simply by going out of the field and avoiding both. In
this case Freud could not escape, because the action to be taken involved either
correcting or accepting Jung.
An approach-avoidance conflict also has Special characteristics. Consider the
case of Freud's deciding whether to teil what he thought to be the truth about
the sexual cause for Dora's neurosis or to avoid the disapproval of his mentor,
10 Human Motivation
Josef Breuer, which he was certain would follow if he published his findings.
Again, such a conflict is serious only if the same act or goal has both approach
and avoidance aspects. If two difFerent goals are involved, the person simply
avoids the negative one and approaches the positive one. However, here the
same act—writing up the Dora case for publication—would satisfy Freud's sci-
entific needs (approach) but earn Breuer's disapproval, which in the end had less
negative valence than his positive push to teil the truth about Dora's sexuality.
Other investigators have shown that as a conflicting goal in a Situation like
this is approached, the tendency to avoid its negative aspect grows stronger
more rapidly than the tendency to approach its positive aspect. Figure 1.1 illus-
trates the different slope of approach and avoidance gradients. In the original
study that demonstrated this difference, J. S. Brown (1948) placed some white
rats in harness in a runway and measured the strength with which they pulled
toward food or away from shock. One consequence of the difference in slopes is
apparent in Figure 1.1. If the approach tendency is very strong, it will get a
person very near the goal before he or she recoils in fear; this is not true if the
approach tendency is weak. Thus, if a man is still strongly attached to a woman
who has rejected him, he will do everything he can to get near her; just as he
gets in her presence, however, he will pull back in fear. His fear reaction will
be much stronger than it would be if he were not so attached to her, because in
that case the fear would have blocked him from approaching her much
sooner.
In reference to Figure 1.1, Miller (1951) says,
Avoid
o ——~^ ^-^___<>_Strong Approach

-C
03
O " 1
icy to Appi
Fear \Avoidance
Elicited
V-
T3
C
'S ——1 \
—_ Weak Approach \ .
ength
—-_______^\
Fear Elicited, \
Feared Near Far
Goal Distance from Feared Goal
Figure 1.1.
Graphic Representation of an Approach-Avoidance Conflict and of the Effect of Increasing the
Strength of the Motivation to Approach (Miller, 1951).
Conscious and Unconscious Motives 11
. . . When the point at which the gradients intersect is between the subject and the
goal, approach is stronger than avoidance. Therefore, the subject moves toward the
goal. When he passes the point of intersection, avoidance becomes stronger than ap-
proach; so he stops and turns back. Increasing the strength of the drive motivating
approach raises the height of the entire gradient of approach. Since this causes the
point of intersection to occur nearer the goal, the subject approaches nearer. Since
this nearer point is higher on the gradient of avoidance, more fear is elicited.
These deductions hold only for the ränge within which the two gradients inter-
sect. It is only for the sake of simplicity that the gradients are represented by
straight lines in these diagrams. Similar deductions could be made on the basis of
any curves that have a continuous negative slope which is steeper for avoidance than
for approach at each point above the abscissa.
Another consequence of the difference in slopes is illustrated by the example

of Freud's approach-avoidance conflicc given in Table 1.1. Given the equal
weights assigned the two tendencies in that case, it would be predicted that
Freud would not have written up the Dora case. As he started to write up the
Dora case, the threat of Breuer's disapproval would have grown stronger faster
than would the positive pull toward explaining the truth about Dora.
Blocked Intent
Lewin and his students also were interested in what happened to an intent when
it was interrupted, for it seemed to continue to influence behavior. For instance,
Lewin had observed that if he intended to mail a letter, it would continue to
"stay in his mind" even while he did other things until he actually had mailed
the letter. One of his students, Zeigarnik (1927), showed that tasks that were in-
terrupted tended to be better remembered than tasks that had been completed.
Other investigations dealt with what happens when an interrupted task can-
not be resumed. Two possibilities are that the person either finds a substitute
way of satisfying the intent or, if none is available, becomes frustrated and en-
gages in disorganized or regressive behavior. For instance, in one study children
were shown an attractive toy that was then covered up by a heavy shield with a
handle on top. The children showed their intent to get at the toy by trying to
lift the heavy cover off. If they could not succeed, they often complained or sat
down, cried, and did nothing, showing signs of regression to maladaptive forms
of behavior. If they were offered other toys, they might be satisfied with those as
substitutes. Much work has been done on the conditions under which children
will accept substitutes, regress, or show other forms of maladaptive behavior
such as aggression when an intent is interrupted.
Conscious Goal Setting

Probably the most influential work carried out by the Lewin group involved the
level of aspiration experiment (Lewin, Dembo, Festinger, & Sears, 1944). These
studies deal with the effects of conscious goal setting on behavior. Subjects typi-
cally are given a task to perform in a limited period, for example, a page of
12 Human Motivation
arithmetic problems to solve in five minutes. At the end of five minutes the sub-
jects observe or are told how many they have correctly solved and then are
asked to State how many they are going to try to correctly solve in the next
trial. The goal can be stated in terms of what they will "try for," the number
they hope to solve, the minimum number they would be satisfied with, and so
on. Then they work at the next page of problems, compare their Performance
with their goal, and set a new goal for the next trial.
Research questions have dealt with the factors that influence the goals
set—in particular the discrepancy between the goal and past Performance—and
also with the effects of various kinds of goals on subsequent Performance. It has
long been known that conscious goals people set for themselves influence their
level of Performance. For example, Figure 1.2, based on some early work by
Mace (1935), shows the effect on people's work Output of trying to do their best
versus trying to surpass a specific Standard for Performance set each day.
Clearly the intent to do better improves Performance, but intents that are tied to
specific goals for improvement over past Performance, as shown in the level of
aspiration experiment, are even more effective in improving Performance.
no -
Group A
100
90
J
03
V Group B
3 80
Q.
O
U
70
60
50
I I I 1 I I I I I 1 I
13 15 17
Figure 1.2.
Performance in Computation Under Two Instructions. Group A subjects (10) were instructed to
surpass a specific Standard prescribed for each day; group B subjects (10) were instructed to do
their best to improve, with no specific Standard (Mace, 1935, p. 21).
The relation of effort to Performance has been confirmed in a different type

of experiment: After failure, subjects are asked to explain their poor perfor-
mance. If they attribute their failure to lack of effort, they are more likely to do
better on the next trial (Weiner, Heckhausen, Meyer, & Cook, 1972). Locke and
Bryan (1968) attempted to see whether conscious goals have an effect over a
longer period of time. They asked College students what grades they hoped to
get or would try for in the Semester coming up and then correlated the goals set
with the grades actually obtained. In general, the correlations were significant
but low (around .30), indicating a slight tendency for students who aspired to
higher grades to get higher grades. However, it is not possible to be sure that
the intent was the causal factor here, because the ability the students had shown
in previous school work undoubtedly influenced the goals they set for them-
selves. Thus, better students would tend both to set higher goals and do better
the following Semester; it is not possible to infer that the higher goals were re-
sponsible for better Performance later on. However, Locke and Bryan realized
this difficulty and controlled for the ability of the students (for example, for the
grades they had obtained in previous Semesters); thus, it may reasonably be con-
cluded that trying for higher grades helps a little in getting better grades later in
school.
Is Conscious Intent Necessary for Learning?

If conscious goal setting facihtates Performance, can it be inferred that it is al-
ways necessary to produce Performance? Must people always be motivated to
some degree to do something before they will do it? This question has never
been completely answered to everyone's satisfaction, and we will come across it
again and again in different contexts throughout the book. For a time, associa-
tion theorists argued that simple conditioned responses were obvious examples
of learning without intent to learn. Suppose an experimenter sounds a tone just
before delivering an air puff to the eye that elicits a blink. After a few trials the
blink is conditioned to, or occurs in response to, the tone without the subject's
"wanting" to blink. Other theorists pointed out that the blink avoids the un-
pleasant puff to the eye, so it seems reasonable to infer that it occurs in response
to an intent—perhaps not very conscious—to avoid unpleasantness.
Early research on what was called incidental learning illustrates the com-
plexity of this type of analysis. Incidental learning research was designed to an-
swer the question of whether learning could occur without any intent to learn.
Conscious goal setting or processing of material clearly facihtates Performance,
as has just been shown, if only because it gets people organized around doing
things that will help them remember the material (Kintsch, 1977). But can
learning occur incidentally when there is no conscious intent to learn? A typical
experiment designed to investigate this problem was conducted by Jenkins
(1933). He asked subjects to memorize a list of words read to them by an exper-
imenter. After a number of repetitions he asked not only the subjects, but also
the experimenter, to recall as many words as possible. The subjects who in-
tended to learn the words recalled more than the experimenters, but the experi-
14 Human Motivation
menters also recalled some words, apparently without any intent to remember
them.
Many later, similar experiments conducted under a variety of conditions
confirmed the two basic facts that conscious intents facilitate learning and that
learning often occurs seemingly without conscious intent. Answering the ques-
tion of whether learning could occur without any intent, however, was difficult,
if not impossible, for it was not possible to prove there were no intents of any
kind present in instances of incidental learning. When experimenters were asked
to report what went on in their minds at the time, they often came up with in-
tents that were not part of their instructions but that were sufficient to explain
the learning that occurred. For example, the experimenters in the Jenkins study
might have wanted to give meaning to the boring task they were involved in by
forming the words into groups to make them easier to say, much as people
group telephone numbers to remember them better.
Factors Underlying Conscious Intents

These experiments, like the conditioned blink experiment, raised the question of
unconscious intents. Might not subjects be learning under the influence of a mo-
tive or intent of which they were unaware? A quite different tradition being es-
tablished in psychology stressed the vital importance of unconscious intents on
learning and memory.
However, this was not the only, or even the chief, difficulty with the psy-
chology of conscious intents. The real problem lay in the fact that focusing on
conscious intents did not sufficiently analyze the motivational processes involved.
Intents, whether conscious or unconscious, are themselves the outcome of other
motivational and nonmotivational variables, as was noted earlier. To focus on
the outcome of these variables (the resultant intent) without even considering
these variables is to engage in a kind of misplaced concreteness.
In performing a laboratory task, the subjects' motives may not be at all ob-
vious. The experimenter asks the subjects to do something, and they set to work
with a conscious intent to perform the task. Yet why do they comply and try
hard? Some may want to please the experimenter. Some may want to prove to
themselves that they are good learners. Some may fear looking foolish by not
being able to perform at all. When the subjects are asked to set a goal (that is,
to State their intent), they will be influenced (1) by these motive dispositions, (2)
by the extent to which they perceive that doing well and performing the task
satisfies these and other dispositions (the value of the task to them), (3) by the
perceived difficulty of performing the task (or the skill they have shown in per-
forming), and (4) by the fact that the task is there to be done (the environmen-
tal determinant).
The intent that results from such variables should not be confused with the
most distinctively motivational aspect of the process, namely, the motive disposi-
tions aroused in the Situation. Intent as the outcome of all the determinants of
behavior predicts choices in the here and now when the determinants are known
and specifiable, but it has much less value for predicting long-range behavioral
outcomes when the determinants are not so easy to specify. Thus, a student's
Statement that he or she intends to study right now will predict behavior fairly
well; the same type of Statement applied to "studying hard next year," however,
should be greeted with some skepticism, because it is not known what attractive
alternative courses of action the Student will face, whether or not the courses
taken will be interesting or not, and so on. Understanding such long-range
trends requires paying more attention to stable motive dispositions, which lead
to persisting in certain kinds of behavior (such as studying) in spite of a wide
Variation in the other determinants of action. The psychology of conscious in-
tents in its early form now is largely of historical interest only.
• UNCONSCIOUS INTENTS
If we move out of the laboratory into the clinic and listen to people talking
about themselves, the role of stable motive dispositions as determinants of imme-
diate choices or intents becomes more obvious. At the same time, however,
some of the motives reflected when people talk about themselves may be more
obvious to others than to the people themselves; that is, they appear to be un-
conscious of them. Freud (1910/1938) commented on his own motives as
follows:
I sacrificed unhesitatingly my budding popularity as a physician and a growing prac-

tice in nervous diseases because I searched directly for the sexual origin of their neu-
roses . . . but the silence which followed my lectures, the void that formed about my
person, and the insinuations directed at me, made me realize gradually that State-
ments concerning the role of sexuality in the etiology of the neuroses cannot hope to
be treated like other Communications . . . but as my conviction of the general accu-
racy of my observations and the conclusions grew and grew, and as my faith in my
own judgement and moral courage were by no means small . . . I was imbued with
the conviction that it feil to my lot to discover particularly important connections,
and was prepared to accept the fate which sometimes accompanies such discoveries.
This fate I pictured to myself as follows: . . . science would take no notice of
me during my lifetime. Some decades later, someone would surely stumble upon the
same, now untimely things, compel their recognition and thus, bring me to honor as
a forerunner, whose misfortune was inevitable. Meanwhile, I arrayed myself as com-
fortably as possible like Robinson Crusoe on my lonely island. When I look back to
those lonely years, . . . it seems to me like a beautiful and heroic era.
Freud describes himself as having the conscious intent of discovering the sexual
basis of neuroses in his patients in his daily practice. However, he explains his
persistence despite the attacks of others as being due to two stable motivational
dispositions in himself: a strong desire for fame as a scientific pioneer and a
weak fear of rejection by his peers (or a strong belief in himself). Like most
people, he has formed some clear ideas of his major motives in life. These dispo-
sitions not only are of an order of magnitude much greater than the temporary
16 Human Motivation
intents dealt with in the laboratory; they also are used to explain such intents,
much as Freud explains his daily decisions in terms of his dispositions.
Can we believe him? Our conclusion that he has a strong need for fame al-
ready involves an inference that goes somewhat beyond what he says. He insists
only that he desires to be scientifically correct, but we infer from the Robinson
Crusoe image and his dream of postmortem fame that he truly wants greatness.
His protests that he does not care what other people think of him are a little
too frequent. We suspect that if he really did not care, he would not mention so
often how much he is rejected. And it is Freud who taught us to be wary of be-
lieving what other people say about their motivational states or their motiva-
tional dispositions!
The major contributions of Freud's early work involved showing how the
motives that influence what people do in everyday life are often unconscious. He
was particularly ingenious in demonstrating that even the wildest thought se-
quences in dreams really were motivated by unconscious desires, which he could
expose by the technique of psychoanalysis. Here is a simple example:
"You are always saying that a dream is a wish fulfilled," begins an intelligent lady
patient. "Now I shall teil you a dream in which the content is quite the opposite, in
which a wish of mine is not fulfilled. How do you reconcile that with your theory?
The dream was as follows: / want to give a supper, but I have nothing available ex-
cept some smoked salmon. I think I will go Shopping, but I remember it is Sunday af-
ternoon, when all the shops are closed. I then try to ring up a few caterers, but the
telephone is out of order. Accordingly, I have to renounce my desire to give a supper.
(Freud, 1900/1938)
Freud explains that only she can find the explanation for the dream and that to
do so, she should relax and say whatever comes into her mind about the dream.
Her first associations have to do with the fact that her husband, who is a capa-
ble meat salesman, is growing too fat. He said he should undergo treatment for
obesity, rise early, keep to a diet, and "above all, accept no more invitations to
supper," which Freud points out contains the idea that her husband gets fat at
other people's dinners. The woman then blocks a little in her associations and
gives some irrelevant comments, which Freud attributes to unconscious suppres-
sion of what is about to come next. She then recalls a visit the day before to a
thin female friend her husband admires greatly. "Fortunately this friend is very
thin and lanky, and her husband likes füll figures." But the friend has spoken to
her about her desire to gain weight and asks Freud's patient, "When are you
going to invite us again? You always have such good food."
Now the meaning of the dream is clear. The thought that runs in the pa-
tient's mind and determines the nature of the dream may be summarized as fol-
lows: "Of course I am to invite you to dinner so that you will get fat and be
more pleasing to my husband! I would rather give no more dinners!" Why is the
dream about smoked salmon? Smoked salmon is her friend's favorite dish. The
dream is a simple expression of her wish not to make her friend plumper and
more pleasing to her husband. The important point to note is that the patient
was quite unaware of this intent when she related the dream, since she thought
it disproved Freud's theory of dreams as wish fulfillments.
Forgetting a Proper Name Unintentionally

Freud also illustrated with many examples how conscious intents often do not
bring about a desired result. In fact, the interest in his book The Psychopathol-
ogy of Everyday Life (1901/1938) lies precisely in the fact that most people feel
their conscious intents control everything they think, feel, and do. Let us exam-
ine an example drawn from Freud's own experience. Once when he was travel-
ing by train in Yugoslavia, he feil into conversation with a stranger. They talked
about visiting in Italy, and Freud asked if he by any chance had seen the fa-
mous frescoes of the Last Judgment in the dorne of the cathedral in Orvieto exe-
cuted by . Here Freud could not recall the name of the artist, Signorelli,
although he knew it perfectly well and recognized it as soon as someone eise
mentioned it. Instead he recalled the names of two other artists, Botticelli and
Boltraffio, the latter a painter who was much more obscure than Signorelli. In
other words, the forgetting could not be attributed to a lack of knowledge.
Therefore, it had to be attributed to intent. What unconscious intent could have
interfered with the conscious intent to recall the name? To find out, Freud ap-
plied his usual technique of free associating to the Situation in which he forgot
the name.
He first recalled that just prior to the discussion of the frescoes of the Last
Judgment in Orvieto, he had been discussing with the stranger the customs of
the Turks living in Bosnia and Herzegovina, through which they were passing.
Freud had related an anecdote told him by a physician friend about the Turks'
being surprisingly resigned at the prospect of death. If the doctor had to teil
them there was no hope for a man, they were apt to reply, "Sir, what can I say?
I know that if he could be saved, you would save him." Freud next recalled that
he also had remembered a second anecdote told him by the physician friend
that involved the sexuality of Turks. In contrast to their resignation at the
thought of dying, they were in utter despair over impotence. As one of them
said to his physician, 'Tor you know, sir, if that ceases, life no longer has any
charm" (Freud, 1901/1938). Freud refrained from telling this anecdote to his
companion "because I did not wish to touch upon such a delicate theme in con-
versation with a stranger." Freud inferred that these thoughts about death and
sexuality were connected with another concern about death and sexuality that
had been occupying his thoughts a good deal at the time, although it did not
enter into his consciousness during this episode. These thoughts concerned the
suicide of one of his patients in Trafoi over a sexual disturbance. Freud feit his
anxiety over his failure as a physician influenced the chain of his associations for
a reason that soon will become apparent.
Figure 1.3 shows how Freud explained his forgetting the name Signorelli.
The key to this explanation lies in the fact that the first part of the artist's name
—Signor—has the same meaning in Italian (sir) as the word Herr in German,
the language Freud was speaking. The word Herr (Signor) cued off anxiety-
18 Human Motivation
Freud's Explanation Forgotten Name Our Exploitation
Anxiety-ridden thoughts about death (Signor)elli: Artist who painted the—| Anxiety-ridden thoughts connected
and sexuality among Turks in Last Judgment in cathedral in with death of Moses and Last
(Her)azegovina and (Bo)snia Orvieto Judgment
Guilt over suicide of his patient Substitute Names Recalled:

for sexual reasons in Trafoi - (Bo)tticelli -*
- (Bo)ltraffio
a
Herr(s\r) in German has the same meaning as signor in Italian.
Figure 1.3.
Associative Connections in Freud's Forgetting a Proper Name.
ridden thoughts about death and sexuality, since it was associated with Her-
zegowina, in which the Turks lived and with two anecdotes, one relating to the
death of a patient, and the other relating to sexuality among the Turks, which
he had determined not to report.
Thus, Freud believed that his desire not to say something interfered with his
desire to say something associated with it via the connection of both associations
with anxiety over thoughts about death and sexuality. The substitute names of
artists he thought of confirmed this analysis. The first syllable in both of them
(Bo) refers to Bosnia, the other location in which the Turks liveJ. He regarded
this as evidence that Signor had cued off thoughts about Turks, since the substi-
tute names referred to another nonthreatening region associated with the Turks.
The elli of Botticelli obviously came from the last part of Signorelli, and the last
part of the obscure name Boltraffio he traced to his guilt over the suicide of his
patient at Trafoi. He believed that forces of both conscious suppression and un-
conscious repression were involved. He consciously suppressed the intent to teil
an anecdote about Turkish sexuality, but the screen name Boltraffio indicated to
him that his thoughts about death and sexuality among the Turks also had
touched on his guilt over the suicide of a patient in Trafoi.
Dozens of analyses of episodes like this convinced Freud and his followers
in the psychoanalytic movement that unconscious intents had to be taken into
account in explaining human behavior, particularly abnormal or unusual behav-
ior. Laboratory psychologists simply ignored the evidence or explained it away
as due to chance or associative interference. However, we must go further than
Freud did in analyzing this phenomenon, because we have argued that momen-
tary intents, whether conscious or unconscious, are themselves products of stable
motive dispositions and other factors. Freud analyzes this episode entirely in
terms of thoughts and feelings he had at the time, yet might not these in turn
have been shaped by his dominant motive dispositions? Figure 1.3 also indicates
that another explanation for Freud's forgetting the name Signorelli can be given
based on a fuller knowledge of his life story.
Unconscious Motives in Freud
A further look into Freud's life and deepest convictions can illustrate his teach-
ing that unconscious motives shape even the most ordinary acts, such as forget-
ting a name. It also can illustrate the methods he used in making such analyses.
Freud was strongly committed to rationality, to the scientific method, and to
making psychology into a natural science. Yet as David Bakan (1958) makes
clear, the psychoanalytic approach Freud evolved had much in common with an
ancient Jewish religious tradition that was anything but scientific and rational.
This tradition involves mystical ideas about the nature of God and goes back at
least to the first Century A.D. By the eleventh Century it was referred to as Kab-
bala, a mystical tradition of textual and oral Communications involving the reve-
lation of meanings hidden in the scriptures. The Kabbalists searched for hidden
meanings. They found wisdom in explaining secrets just as Freud explained the
secret meaning of dreams and symptomatic acts. Kabbalistic thinking played a
role in messianic movements that swept the Jewish Community in the seven-
teenth and eighteenth centuries. Charismatic leaders appeared and convinced
many Jews that the messiah had come who would lead them cut of the bonds of
oppression and into a life of joy and union here and now with God.
The movements were branded as heretical by Orthodox Jewish rabbinical
leaders, but the enthusiastic antirational tendencies they engendered survived in
Chassidism, a movement that centered originally on Baal Shem-Tov, a man of
magical healing powers and much folk wisdom, although technically he was an
"ignoramus," that is, not trained in Jewish law and tradition. To nineteenth-
century Viennese Jews of Freud's circle, Chassidism had a number of bad asso-
ciations. It evoked thoughts of heresy, unseemly enthusiasm, the weird mystical
ideas of uneducated people, and rejection of the scholarly rationality characteris-
tic of Orthodox Judaism. It is small wonder that Freud wanted no connection
made between what he was doing and the traditions of Jewish mysticism.
As noted already, however, the tradition emphasized, above all, penetrating
secrets. So did Freud. The method used for discovering secret wisdom was de-
scribed by a thirteenth-century Jewish mystic, Abraham Abulafia, as "skipping
and jumping." By this he meant almost exactly what Freud meant by free asso-
ciating on a text. In this way a person can penetrate beyond the outer meaning
of a word of scripture (what Freud called the manifest content) to its inner
meaning (Freud's latent content) much as one removes the shell of a nut to get
at its kernel (Scholem, 1966).
The most important book in the mystical tradition came to be the Zohar,
prepared in part by Moses de Leon in the thirteenth Century. It conceived of the
Torah (Jewish scripture) as a living organism. The Torah has a head, a body, a
heart, a mouth, and so on (Scholem, 1966). Thus, the idea is clearly present that
uncovering the truth about the Torah also uncovers the truth about man. Un-
covering the truth about people by penetrating their thoughts and dreams is ex-
actly what Freud thought he was doing. To Bakan it hardly seemed an accident
that Freud chose the name Dora (pronounced like Torah) as the name for his
earliest, most famous case for illustrating the method of his psychoanalysis.
20 Human Motivation
Furthermore, the sexual metaphor is used throughout the Zohar for inter-
preting the ultimate meaning of life. 'The soul, according to the Zohar, has an
unquenchable yearning to be united with its source in God. This union is char-
acteristically discussed in the metaphor of sex. . . . Thus human sexual relations
become symbolic vehicles of divine acts. . . . Freud's use of the idiom of sexual-
ity as the basic one for the expression of all the deeper and more profound
Problems of mankind is entirely in the spirit of the Kabbala" (D. Bakan, 1958).
In this tradition, the ultimate form of secret knowledge is uncovering the funda-
mental sexuality of the universe. Mystical union with God is represented, again
using the sexual metaphor, as union with the Holy Shekinah, or the female, ma-
ternal aspect of God (D. Bakan, 1958). The Hebrew word for "knowledge" used
in this context also means "sexual knowledge," as is clear in the King James
translation of the Hebrew biblical text when it states in Genesis that Adam
"knew" his wife Eve, that is, "knew" her sexually.
Finally, the Jewish mystical tradition was antiestablishment, just as psycho-
analysis was. It represented a revolt against patriarchal rabbinical authority;
Freud, as an apostate Jew, often feit himself in a similar rebellious frame of
mind. Although he remained identified with the Jewish Community, he rejected
Judaism as a religion. In fact, he conspicuously violated the First Command-
ment brought down by the father of Judaism, Moses, from the mountain: "The
Lord thy God is a jealous god, thou shall have no other gods before me." Freud
filled his Consulting room with statues and other images of gods from many
other religions. There is ample reason to believe, as Bakan points out, that
Freud was somewhat ambivalent about his rebellion and that he specifically
feared the wrath of Moses.
Freud describes at length his feelings when Standing in front of the huge
statue of Moses by Michelangelo in a church in Rome. Moses has returned from
the mountain with the tablets containing the Commandments only to discover
that the Jews are worshiping false gods. Looking at the seated Moses, Freud
thinks to himself, "In his first transport of fury Moses desired to act, to spring
up and take vengeance and forget the tablets (the Law) but he has overcome the
temptation and he will now remain seated and still in his frozen wrath." Freud
also says he feit "of the mob upon whom Moses' eye is turned." The text makes
it clear that Freud feit he himself had some reason to fear the wrath of Moses.
What Freud does not mention is that as visitors enter the long, narrow
church of San Pietro in Vinculi, where the statue of the seated Moses is placed
at the back on the right side, they must pass on the left a vivid reminder of the
death and destruction involved in the Last Judgment for all people. The entire
wall is a huge collection of skeletons, skulls, and bones together with the Grim
Reaper, which serves to remind the observer as vividly as possible of the Last
Judgment. Was Freud to be punished for his revolt against Moses, the symbol
of Orthodoxy—not just as an apostate Jew, but as a psychoanalyst arguing for
freedom against the excesses "of the Mosaic type law which made for neurosis"
(D. Bakan, 1958)?
The meaning of this deep-seated conflict in Freud's life is revealed by a visit
to the Sistine Chapel, which contains some of the most famous paintings in
Rome. Certainly Freud knew about these paintings, as art and Rome had fasci-
nated him from boyhood. The paintings suggest another reason why Freud for-
got the name of the artist responsible for the frescoes of the Last Judgment in
the cathedral in Orvieto. The dominant theme in the Sistine Chapel once again
is the Last Judgment, represented in a huge mural painted by Michelangelo at
the front end of the room. However, as a person enters the room through a
small door at the rear, the first painting seen is the Death of Moses by Signorel-
li, the very painter whose name Freud forgot in connection with another repre-
sentation of the Last Judgment. Furthermore, right next to this scene is one rep-
resenting the punishment of the Cores (maidens), by Botticelli, whose name
Freud thought of as a substitute for Signorelli.
Thus, it might be argued that the name Signorelli had a negative signifi-
cance for Freud that went far beyond its casual link with some talk about Turks
living in the neighborhood through which he was passing on a train. Signorelli
also painted one of the most famous pictures of the death of Moses, who, Freud
had argued, had been killed by apostate Jews like himself. In Freud's view that
was why Moses had never seen the Promised Land. Thus, he had reason to fear
the Last Judgment and to be anxioüs about any associations involving his pro-
found revolt against religious Orthodoxy, as represented by Moses. Freud's
theory of the Oedipus complex dealt with the guilt a son feels over his incestu-
ous desires for his mother and his desire to kill his father. Symbolically Freud
was worshiping false gods (that is, the Shekinah represented as female in the
Jewish mystical tradition), killing Moses, and defying rational authority. No
wonder the name of the renowned painter (Signorelli) of the scene representing
the death of Moses had an unusually negative connotation for him—so much so
that he was unable to recall it when thinking about the Last Judgment in an-
other connection.
To return to Figure 1.3, another anxiety-ridden thought can be added that
could have blocked Freud's ability to recall the name of the painter of the Last
Judgment in the cathedral in Orvieto: Signorelli was associated with painting the
death of Moses, which evoked Freud's anxiety over his rebellion against Mosaic
authority.
So we have succeeded, using Freud's approach, in relating his momentary
unconscious intent to avoid thinking of a name to a deeper, stable motive dispo-
sition in his life involving guilt and anxiety over his rebellion against established
patriarchal authority.
Conclusions to Be Drawn from Freud's Approach to Motivational Analysis

What are we to make of this kind of motivational analysis? Has it any more
than a historical interest as an explanation of one of the main traditions that
has shaped modern motivational psychology? The following conclusions seem
justifiable:
1. What people say about their motives should not be taken at face value.
Even in ordinary life, but particularly in dreams and symptomatic acts, people
often do not know what their motives are. When they do assign motives to their
22 Human Motivation
acts, these motives, on further analysis, may turn out to be incorrect. Even
Freud's motivational analyses should not be taken at face value: by examin-
ing his life a little further, we found a motivational explanation for his for-
getting a name that was somewhat different from the explanation he was sat-
isfied with.
2. Associative thought is extraordinarily fluid and easily influenced by con-
scious or unconscious motivational forces. What is striking about many of
Freud's analyses is how quickly and automatically a thought or associative chain
is deflected by a motivational influence, whether it leads the dreamer to imagine
the telephone is out of order or leads Freud to forget the name of a famous
painter. As Freud himself noted on another occasion, the overall impression that
these observations create is that we are not complete masters in our own house-
hold. The mind functions quite easily and efficiently without our being in con-
scious control of it. And fantasy or free association, over which we exercise little
control, reveals most obviously and easily motivational forces of which we may
be unaware.
3. The ultimate motivation is not necessarily sexual, as Freud argued, al-
though it is understandable why sexual motives were so important to him if he
was drawing on Kabbalistic, mystical insights into the ultimate nature of man.
Many others who associated with him early, like Carl Jung and Alfred Adler,
insisted on the importance of other nonsexual motives, as Chapter 2 will discuss.
In analyzing Freud's own life, we had to resort to power rather than sexual mo-
tivational dispositions. On the one hand, he showed a strong need for power in
his dreams of fame and in his break with orthodox scientific and religious tradi-
tion; on the other hand, he clearly feared the power of others, from Moses to
his scientific peers. The power motive will be discussed in Chapter 8.
4. The method of uncovering unconscious intents by analyzing free associa-
tions is not completely convincing. Whereas it often yields interesting explana-
tions of otherwise puzzling behaviors, the chance exists that the explanation has
been invented through free associations to please the psychoanalyst. What
criteria are there for deciding whether an interpretation is correct? The ordinary
ones are that the interpretation account for a number of otherwise inexplicable
facts and that it satisfy the person who arrives at the interpretation. Thus,
Freud was satisfied with his account of forgetting the name Signorelli, but we
were not and pursued the matter further to locate a somewhat different explana-
tion. We need a less-questionable method of establishing the truth about the in-
fluence of unconscious intents.
EXPERIMENTAL STUDY OF UNCONSCIOUS MOTIVES

Experimental psychologists interested in the influence of unconscious intents
have studied the effects of known unconscious intents on behavior, particularly
fantasy. They then are not caught in the logically awkward position of trying to
reason backward from effects to causes. For example, they have sometimes in-
duced unconscious intents under hypnosis. A hypnotized woman is told that
after she is awakened from the trance State she will meet and talk with another
person she finds dull and uninteresting. She also is told she will have forgotten
when she awakes whatever went on during hypnosis. She meets the stranger,
Starts talking with him, and after a while asks him to "shut the bore," thus dis-
playing exactly the kind of slip of the tongue that Freud claimed unconscious
intents produced.
Since whether the hypnotized person has really forgotten what she has been
told can be questioned, the matter has been pursued further in experimental set-
tings. Some overweight and normal weight undergraduates were asked to write
imaginative stories in response to pictures shown on slides. One of the pictures
showed two men and what was clearly a piece of steak, as was reported by
nearly all subjects. However, one overweight subject who had reported in an-
other Situation that she was struggling hard to keep from overeating told a story
about two men and a snake they had found on an island. There was no hint in
the manifest content of the story that she had recognized the object in the slide
as anything edible, but her fear of overeating had apparently led her to perceive
a dangerous object that rhymed with the word steak.
Other studies have aroused intents experimentally to see whether their ef-
fects on fantasy are what Freud said they would be. Since Freud argued that
sexual motivations were particularly likely to distort the associative flow, Clark
(1955) arranged to have slides of nude pinup girls shown to a group of male un-
dergraduates in a psychology class as part of a study they were told was to de-
termine what features of the body were associated with attractiveness. Another
control group of subjects was shown an equivalent number of slides of landscape
scenes and buildings they were told were part of an investigation on factors af-
fecting aesthetic judgment. Both groups of subjects then wrote imaginative sto-
ries based on pictures they were told were part of another study on creative
imagination. All these stories were coded for the presence of manifest sexual im-
agery or for sex symbolism without knowledge of which group of slides the sub-
jects had been exposed to. Manifest sexual imagery included references to sexual
intercourse, kissing, dancing, fondling, or being in love.
Clark presumed the young men exposed to the nude slides would be more
sexually aroused than those exposed to pictures of buildings. However, the
amount of explicit or manifest sexual imagery in the stories from the aroused
group was significantly lower than in the stories from the control group. This
might be taken as evidence for Freud's position, namely, that unconscious sexual
intents automatically elicit censorship, which tends to block out conscious
thoughts about sex. Otherwise, why would young men who had just been made
to think about sex subsequently think less often about sexual matters? Is it pos-
sible that they became bored or satiated with the subject? To be sure that this
was not the case, and that they actually were more sexually aroused, it was nec-
essary to conduct two further studies.
In one, the experiment was repeated by Clark (1955), but at a fraternity
beer party rather than in the classroom. The nude slides or the landscape scenes
were introduced accidentally by a stooge as something he owned and was show-
ing to add to the fun of the party. Under party conditions the amount of sexual
24 Human Motivation
imagery in stories written after both the control and nude slides were shown
was significantly higher than under classroom conditions. Apparently the alcohol
and party setting together increased thoughts about sex, but the average amount
of sexual imagery in the stories after seeing the nude slides was significantly
higher than after seeing the landscape slides. This finding later was confirmed by
Kaiin, Kahn, and McClelland (1965) using an attractive female model as the
sexually arousing Stimulus in a controlled party setting. Sexual arousal of young
males decreases the manifest sexual content of thought under conventional social
conditions but increases it in party settings when alcohol is served. This con-
firms Freud's contention that unconscious (or even conscious) sexual intents
often inhibit or disturb associative thought processes.
Freud also argued that the unconscious intents continue to operate by pro-
ducing Symptoms (screen names, apparently chance acts, and disguised wishes)
from which can be inferred the presence of unconscious intent. This means that
sexual arousal in the young men, while blocked from direct expression, would
show itself through more thoughts of a disguised sexual nature. Therefore, Clark
scored all the stories he obtained for sexual symbolism, using the criteria that by
now are Standard in the literature (see, for example, Gutheil, 1939; Hall, 1953).
In general, sex Symbols refer to objects that share key characteristics with sexual
body parts (round objects = breasts; a long object = penis; a window or door
as an entry point = vagina; and so on). In Clark's scoring, classic Symbols had
to be involved in some action that could be regarded as symbolically sexual. For
example, if in response to a picture of a person silhouetted in a window, a per-
son wrote "that this was a boy looking out of his dormitory window," this was
not scored as symbolic. If the person described a thief "climbing up to go
through the window in order to steal the jewels in the house," this was scored
t4
as referring symbolically to sexual intercourse" (Clark, 1955).
Figure 1.4 plots the sexual symbolism score against the manifest sex score
for young males under the sexually aroused and controlled conditions. Note first
of all that the overall amount of sexual symbolism was considerably and signifi-
cantly higher for the sexually aroused subjects than the control subjects. That is,
the dotted line is higher than the solid line at every point in the two curves.
This is what Freud would have predicted: sexual arousal not only blocks mani-
fest expression, but it also promotes disguised expression of the intent. Note
next that on the right side of the graph, if the manifest sex score is low, the sex
symbolism score tends to be much higher in the aroused group. This also is
consistent with Freud's interpretation of a blocked intent seeking an indirect
expression.
However, one finding shown in the figure is not consistent with Freud's ar-
gument: subjects whose stories showed a great deal of manifest sexual content
also included much sexual symbolism (the left side of the graph). This is consis-
tent with a theory advocated by Hall (1953), who has maintained that "sexual
symbolism in dreams is not a disguised expression of sex, the purpose of the dis-
guise being to smuggle the content past the censor, but rather a means of repre-
senting as clearly as possible a particular conception of sexuality that the
dreamer has in mind. One of several cogent reasons that Hall has for offering
2.00
1.50
p
Lon\ro\ Group
0.50
0.00 1 1
High Moderatc Low
Manifest Sex Score
Figure 1.4.
Sexual Symbolism Score as a Function of Manifest Sex Score for the Aroused and Control Groups
of the Nonalcohol Condition (Clark, 1955).
his alternative theory is that in his collection of dreams he often found both
manifest and symbolic expressions of sexual activity in the same dream or in the
same sequence of dreams" (Clark, 1955).
Why disguise sexual content if in the next moment it is revealed in manifest
form? According to the results shown in Figure 1.4, both Freud and Hall ap-
pear to be correct. For some subjects sexual arousal apparently causes anxiety,
which blocks manifest sex but allows disguised expression of the sexual arousal.
For other subjects sexual arousal does not arouse anxiety, and the sexual drive
expresses itself in both manifest and disguised or metaphoric terms. The general
conclusion can be drawn, however, from Figure 1.4 that a sex symbolism score
is a better indicator than overt sexual content of the presence of aroused sexual
intents or motives for both types of people.
• ARE UNCONSCIOUS MOTIVES IMPORTANT?

These studies demonstrate that what Freud said did happen can happen. How-
ever, this is not the same as saying it happens very often. As a practical matter,
need we worry about the way unconscious motives influence our everyday be-
havior, or can we assume that such phenomena occur only under very unusual
circumstances? Gordon Allport (1953) was very concerned that the preoccupa-
26 Human Motivation
tion with unconscious motives that developed in psychiatry following Freud

might engender "a kind of contempt for the 'psychic surface' of life. The indi-
vidual's conscious report is rejected as untrustworthy . . . the individual loses
his right to be believed." Allport recognizes unconscious motives have to be
taken into account for some people, particularly neurotic people, but generally
speaking such motives can be ignored and what people say their motives are can
be believed:
The patient should be assumed insightful until he has proved otherwise. If you asked
a hundred who go to the icebox for a snack why they did so, probably all would an-
swer, "Because I was hungry." In 99 of these cases we may—no matter how deeply
we explore—discover that this simple, conscious report is the whole truth. It can be
taken at its face value. In the lOOth case, however, our probing shows that we are
dealing with a compulsive overeater, with an obese seeker after infantile security
who, unlike the majority of cases, does not know what he is seeking—perhaps his
mother's bosom—and not the leftover roast. In this case—and in a minority of all
cases—I grant that we cannot take the evidence of his overt behavior, nor his ac-
count of it, at their face value.
Is Allport right that we can ignore unconscious motivations except for a few
aberrant cases? Combs (1947) carried out a study that sheds some light on this
issue. He had students write out füll autobiographies and also write twenty sto-
ries to the pictures in Murray's Thematic Apperception Test (TAT). The idea
behind the test is to sample fantasies or daydreams of the type Freud was inter-
ested in to find out what kinds of motives were on people's minds when they
were not consciously reporting on their own desires. (See Chapters 2 and 6 for a
fuller description of the TAT.) Combs then coded all the desires that appeared
either in the TAT or the autobiography according to about twenty different cat-
egories. First he ranked the frequency with which various desires appeared in
the autobiography and in the TAT stories. The correlation between these two
rank Orders was .74, indicating that there was considerable agreement between
the direct and indirect measures as to the frequency with which various motives
were mentioned.
Combs then observed whether a particular motive was mentioned by any
given individual in the TAT stories alone, in the autobiography alone, or both.
Table 1.2 illustrates some of his most important results. It shows that a number
of desires were mentioned exclusively in the TAT. They included "to die" and
"to atone," "to have sexual relations with," "to be accepted," and "to avoid
blame." Allport clearly overestimated the frequency with which undesirable mo-
tives appear in consciousness. Even among normal people, a number of desires
do not regularly appear in the conscious reports of motives. These desires do ap-
pear, however, in fantasies produced in the TAT. An area of overlap lies "in the
milder, or expected desires of the average individual for security, response, rec-
ognition and pleasant occupation" (Combs, 1947). The TAT deals more often in
"socially unacceptable and violent kinds of motivation." The autobiography, on
the other hand, "emphasizes the socially acceptable, normal and expected more
Table 1.2.
NUMBER OF INDIVIDUALS SHOWING SIGNIFICANT DIFFERENCES FOR EACH ITEM OF
THE DESIRES LIST IN ANALYSIS OF TAT AND AUTOBIOGRAPHY (after Combs, 1947)
Appears Appears
Appears in TAT in
in TAT and Autobiography Number of In Favor
Item Only Autobiography Only Difference Of
Sexual relation with 20 2 3 17 T
To die 17 0 0 17 T
To atone 17 1 0 17 T
To avoid struggle 16 3 0 16 T
To be with 15 4 2 13 T
To abandon 16 1 4 12 T
To avoid death 9 0 0 9 T
To protect 10 1 1 9 T
To know 12 9 3 9 T
To be consoled 17 4 8 9 T
To defy Convention 8 1 0 8 T
To care for 17 4 10 7 T
To have sensuous experience 12 4 5 7 T
To have child 8 6 2 6 T
To punish 10 9 5 5 T
To control 12 6 7 5 T
To assist 12 6 7 5 T
To have means to 15 4 10 5 T
To believe in 12 3 8 4 T
To maintain relations with 14 9 10 4 T
To do as admired person
wishes 15 7 11 4 T
To play 14 7 10 4 T
To do one's duty 12 9 10 2 T
To keep 9 0 6 2 T
To have mental peace 9 12 8 1 T
To be attractive to 11 9 10 1 T
To be safe 11 8 11 0 Neither
To maintain Status quo 11 4 11 0 Neither
To belong 11 7 11 0 Neither
To avoid illness 4 1 6 2 B
To be married 9 9 11 2 B
To be loved 6 14 8 2 B
To be helped 9 8 11 2 B
To avoid restriction 9 12 11 2 B
To accomplish 10 11 13 3 B
28 Human Motivation
Table 1.2. (Continued).
Appears Appears
Appears in TAT in
in TAT and Autobiography Number of In Favor
Item Only Autobiography Only Differe nee Of
To be respected 10 9 13 3 B
To avoid pain 4 1 7 3 B
To avoid blame 7 11 12 5 B
To be accepted 10 8 16 6 B
To overcome a handicap 7 2 15 8 B
Totais 456 227 286
Note: Significance is defined here as any desire whose frequency for an individual is greater than the average frequency for the entire groups
of subjeets.
often than the TAT" (Combs, 1947). Any general theory of motivation must
take both conscious and unconscious motives into aecount.
NOTES AND QUERIES
1. Philosophers always have divided the mind into several parts, just as we ar-
gued in this chapter that three different aspects of the mind must be taken
into aecount in explaining behavior—motives, cognitions (understandings
and beliefs about what is going on), and skills. One of the oldest controver-
sies concerns the relative importance of motivation and reason. In the Mid-
dle Ages, church fathers argued that nihil volitum nisi cognitum (nothing is
wanted unless it is understood). In other words, understanding, a cognitive
variable, produces desires. The Freudian revolution seemed for a time to re-
verse the order of importance in insisting that cognitions were largely
shaped by desires. Today psychology once again asserts the importance of
reason, or the cognitive variables (Abelson, 1981). What is the best evidence
for thinking that a motivational or dynamic component that is important
in its own right as a determinant of behavior must be attributed to the
mind?
2. Plato divided the mind into three parts: the desiring element, the reasoning
element, and the spirited element (which became known as the will for
other scholars). What has become of the will in our threefold division of the
mind? Has it been left out, or is it an aspect of some other component? If
strength of will cannot be measured the way Ach measured it, how can it
be measured?
3. If setting a conscious goal reflects motivational and nonmotivational deter-

minants of behavior, how can a pure measure of the motivational determi-
nant alone be obtained? Why are dreams supposedly purer measures of mo-
tivational determinants?
4. When you first wake up, write out as completely as you can a dream you
had that night. Later try free associating to it, as Freud suggested, to see if
you arrive at a satisfactory explanation for why you dreamed as you did.
5. Make a list of some of the goals that are most important to you. Try to
rank them in order from most important to least important. After you have
completed the list of goals, write down different things you have spent time
thinking about lately. Again rank them from those you have spent the most
to the least time thinking about. Then try to match up what you have been
thinking about with your goals. Is there a close fit? If there have been
things you have been thinking about that do not seem related to your goals,
can you explain your thoughts in terms of other goals? If certain goals are
not related to any of your thoughts, what does that mean? For a reference
on method, see Klinger, Barta, and Maxeiner (1980).
6. Laboratory studies of motivation seem dull compared with the analyses of
motive trends in people's lives like the analysis of Freud's motives. Some
psychologists focus entirely on the experimental, and others on the clinical,
psychodynamic approach to motivation. What are some of the advantages
and disadvantages of each approach?
7. Just because people write imaginative stories that contain more references to
socially undesirable motives than their autobiographies do, does that mean
people actually are influenced more often by such motives? How can you
teil that the motives a person attributes to someone in a story also charac-
terize the person writing the story? Freud's female patient accepted his in-
terpretation of the motivation behind her dream, but is acceptability a safe
criterion for establishing the truth of an interpretation? What other criteria
are there for establishing that the motive attributed to a character in a story
also belongs to the person writing the story?
8. Write some stories to pictures of the type used in the TAT. Try to deter-
mine if the concerns the characters have in the stories are in any way re-
lated to your concerns. Be careful not to be misled by superficial differences
between your life and the lives you describe in the stories. For example, a
woman wrote a story about a young man who led an irresponsible, dissolute
life in which he wasted his inheritance. She feit this story in no way related
to her because (1) it was about a man and she was a woman; and (2) she
was not leading a dissolute life: quite the contrary, she feit she was leading
a careful, well-controlled life, so how could his motivations be hers? When
she was asked if she were concerned about waste, however, her face lit up
and she replied, "But of course. That is what I worry about above every-
thing eise."
When you write the stories, put aside analytic thoughts, relax, and write
spontaneously whatever comes into your mind. Obviously, if you worry
30 Human Motivation
about what your writing will reveal or attempt to control what you write
about, your stories will not give you a very accurate picture of what you are
concerned about. Save the stories you have written. You will have opportu-
nities to analyze them later, after methods of coding such stories have been
explained in subsequent chapters.
2
Motives in the
Personality Tradition
• Motives as Reasons for What People Spend Their Time Doing

McDougall's Classification of Motives
• Motives as Reasons for Abnormal Behavior

The Freudian Motivational Trilogy: Sex, Aggression, and Anxiety
Bad Desires and Good Desires
Jung's View of Motivation
Basic Motives and Later Psychoanalytic Theorists
• Motives as Reasons for Creativity and Growth
• Measuring Human Motives

Murray's Study of Motives
Cattell's Identification of Motives Through Factor Analysis
• Stages in Motivational Development

Stages Reflected in a Freudian Dream
Stages Reflected in a Murderer's Dream
Implications of the Dream Analyses
• Other Views of Developmental Stages
• Checking the Validity of the Motivational Stage Theory

Hermeneutics
Experimental Studies of the Psychoanalytic Theory of Motivational Stages
• Contributions of the Personality Tradition
31
Downloaded from https://www.cambridge.org/core. University of Cambridge, on 19 Mar 2018 at 09:10:00, subject to the Cambridge Core terms of use, available at
s, "tudents of personality have been primarily interested in motivational disposi-
tions in the individual. They have asked, What motives are there? How many
are there? What are the most important motives? How do we know what mo-
tives a person has? This chapter's goal is to provide an overview of the answers
to such questions that motivational theorists have given and, in particular, to
consider how motives might be measured in people. Let us begin with the phe-
nomena that have led theorists to assign motive dispositions to people.
• MOTIVES AS REASONS FOR

WHAT PEOPLE SPEND THEIR TIME DOING
To a considerable extent there is a theory of motivation to go with every field of

human endeavor. We observe that people do various things frequently and infer
that therefore they must want to do them. People eat; therefore, they must want
to eat. Some people do well in school, so we infer they have a need for aca-
demic success. Children play; therefore, they must have a need to play. Some
people save; therefore, there must be a saving drive. People in business often
work hard, and since business is organized around making a profit, economic
theorists from the time of Karl Marx to the present have assumed they work be-
cause of the profit motive. In fact, a modern theorist like John Kenneth Gal-
braith (1967) has not hesitated to write a chapter in a book on economics enti-
tled "The General Theory of Motivation" based on his observations of the
various goals of economic enterprise.
Similarly, those who study politics infer that people have a desire to domi-
nate one another, since they often fight political battles or actual wars in which
people kill one another. By way of contrast, those who observe family life find
there is a need to nurture and protect human beings.
What are we to make of all of this confusion? Is it possible to bring any
order out of chaos and arrive at a list of a few basic human motives? Some the-
orists always have been skeptical about such a possibility. In thinking about the
problem, U.S. personality theorist Abraham Maslow (1954), for example, con-
cluded that "we should give up the attempt once and for all to make atomistic
lists of drives or needs. . . . If we wished, we could have such a list of drives
containing anywhere from one to one million drives, depending entirely on the
specificity of analysis."
Anthropologists have been particularly likely to take this point of view,
since they have observed that different cultures value different things, and there-
fore that what people want obviously varies from culture to culture. They rea-
son that it would be impossible to discover a set of motives of importance that
applied to all cultures. Some clinicians working with individuals have feit the
same way. Each person appears to have a uniquely diiferent set of motivational
needs. Why not just spend time discovering them and stop worrying about
whether or not they are similar to those discovered in other individuals? As
U.S. personality theorist Gordon Allport (1937) was fond of saying, there are no
32
Motives in the Personality Tradition 33
simple sovereign motives common to all men. Each person is unique, just as
each culture is unique.
Before we give up altogether trying to find motives of general significance,
let us examine more critically the process of identifying motives that has just
been described. First, to say that because people do something they must want
to do it is little more than animistic thinking. It is like saying that because a
plant grows it wants to grow, or because the apple falls it wants to fall. As
Heckhausen (1980) and others have long argued, such a naming process is tau-
tologous. It adds nothing of scientific value to our understanding of what is
going on unless we have some independent way of measuring the alleged motive
behind the behavior.
Second, we must remember, as Chapter 1 pointed out, that motives are only
one determinant of behavior. If we consider a particular behavioral outcome
such as eating, the hunger drive is only one of the reasons that explain it. Peo-
ple also eat because they know how to eat (the skill or habit determinant) and
because it is time to eat or they think it is good for them to eat (the value deter-
minant). In fact, they may eat for these reasons when they are not hungry at
all. Thus, we cannot automatically infer a motive to perform an act from the
Performance of the act itself. Above all, we cannot infer exactly what general
motive lies behind performing the act, since the same act can be motivated by
several different drives. Thus, if students study hard in College, we do not know
whether it is because they want to please their parents, because they want to do
a good Job, or because they want to earn social recognition. To infer simply that
they have a "studying motive" is to commit the "naming" fallacy, which if pur-
sued would create a long list of motives so specific they would have little value
in explaining a wide variety of behavioral phenomena.
If we keep in mind that along with motives, values also influence behavior,
we can escape the conclusion that there can be no general motives, since cul-
tures differ so much in what they consider important. It is primarily cultures
that influence values, and it is still possible to think in terms of a general human
motive, such as the need for achievement—for doing something better (see
Chapter 7)—which is expressed differently depending on variations in what the
culture values. For example, mainland U.S. Citizens appear to be more oriented
toward individualistic success or achievement compared with native Hawaiians,
who appear to value interpersonal relations more (Gallimore, 1981). Thus, it is
possible that the achievement motive might be more engaged for Hawaiians in a
collaborative Situation than it would be for other U.S. Citizens. Thus, some of
the proliferation of motive concepts derives from a confusion between values
and motives. What people value is in no way simply related to their motives, as
will be shown later.
As another example of this fact, consider the so-called profit motive. Karl
Marx argued over one hundred years ago in the Communist Manifesto that it
was the profit motive that drove the capitalist class around the world seeking its
advantage. Modern U.S. business executives agree it is the profit motive that
drives their behavior. All this means is that such people believe the profit
34 Human Motivation
motive drives their behavior. That is, they value making a profit. It does not
follow that profit actually is what motivates them to energetic entrepreneurial
behavior. A belief in the profit motive certainly helps determine behavior. It
may make business executives decide they are in favor of a tax cut to stim-
ulate business.
However, it is fairly easy to demonstrate that it cannot be profit that moti-
vates the sustained energetic behavior of business executives for two simple rea-
sons. First, as Andrew Carnegie discovered somewhat to his surprise years ago,
business executives often continue to work as hard as ever long after they have
made as much money as they can possibly use. Something other than profit
must be driving them. Second, there is no profit for the individual business exec-
utive in the Communist System as it has developed in the Soviet Union, yet
many Soviet business executives behave just as energetically as their Western
counterparts. The emphasis on profit, therefore, represents a kind of misplaced
concreteness that easily can be shown not to be a very good way of describing
what motivates business executives. Many of the motives that have been attri-
buted to people in everyday life turn out to represent the naming fallacy, mis-
placed concreteness, or a confusion between what people value or believe in and
the motives that energize their behavior.
M cDougall's Classification of Motives

An early attempt was made by psychologist William McDougall (1908, 1932) to
avoid these diflficulties and separate motives from other human characteristics.
He argued that certain behavioral tendencies (that is, propensities) were inherit-
ed, instinctive, and "common to men of every race and age." He defined an
instinct as consisting of (1) a tendency toward selective perception of certain
Stimuli (a hungry person perceives food objects more readily than other objects),
(2) a corresponding emotional excitement experienced on perceiving the object
(the root of the instinct), and (3) the activation of a tendency to seek a goal.
McDougall (1908) said, "Every instance of instinctive behavior involves a know-
ing of some thing or object, a feeling in regard to it, and a striving towards or
away from the object." Put simply, his idea was that certain actions or objects
innately give rise to emotional excitement, which leads to goal-directed activities.
These goal-directed activities could be called motives, although he called them
instincts or propensities. Other activities or objects did not create emotional ex-
citement and hence did not qualify as motives.
McDougall went further and tried to identify the main instinctive propensi-
ties that drive normal behavior. Some typical examples from his last list of eigh-
teen propensities (McDougall, 1932) are summarized in Table 2.1. His emphasis
on the instinctive basis of all motives occurred during a historical period in aca-
demic psychology, when the behaviorist view was dominant and stated that
nearly all human characteristics are learned and not based on inherited or innate
tendencies. Therefore, McDougall had little influence at the time. Most psychol-
ogists, while willing to admit that a few physiological needs, such as hunger,
were innate, feit that more complex motives, such as the need for Submission or
Table 2.1.
SOME INSTINCTIVE PROPENSITIES (MOTIVES) IDENTIFIED BY
McDOUGALL (1932)
1. Food-seeking propensity: to seek (and perhaps störe) food.

2. Disgust propensity: to reject and avoid certain noxious substances.
3. Sex propensity: to court and mate.
4. Fear propensity: to flee; to cower in response to violent impressions that inflict or
threaten pain or injury.
5. Curiosity propensity: to explore stränge places and things.
6. Protective and parental propensity: to feed, protect, and shelter the young.
7. Gregarious propensity: to remain in Company with fellows and, if isolated, to seek
that Company.
8. Self-assertive propensity: to domineer, to lead, to assert oneself over, or display
oneself before one's fellows.
9. Submissive propensity: to defer, to obey, to follow, or to submit in the presence of
others who display superior powers. . . .
13. Acquisitive propensity: to acquire, possess, and defend whatever is found useful and
attractive. . . .
acquisition, were acquired and not innate in everyone. It was not until much
later, when ethologists introduced the notion of "sign Stimuli" or natural incen-
tives (see Chapters 4 and 5), that some of McDougall's ideas came back into
vogue. On the other hand, he did provide a beginning taxonomy of human mo-
tives based on normal behavior that influenced later students of human motives
like Henry A. Murray and Raymond Cattell, whose contributions will be de-
scribed later in this chapter. McDougall's descriptions of the characteristics of
motives have continued to be part of their definition ever since.
MOTIVES AS REASONS FOR ABNORMAL BEHAVIOR

Another approach to discovering basic motives has been to pay attention to
what leads people to behave in unusual or abnormal ways. This makes sense if
behavior is jointly determined by motives, skills or habits, and values or cogni-
tive Schemas. Suppose a hypothetical Student suddenly stops studying—an un-
usual or abnormal act for her. We know that two of the determinants of study-
ing behavior are in place: she knows how to study (she has been studying), and
she values it. Thus, we infer that something has happened to her motivation and
try to find out what motive is responsible for her not doing what she would nor-
mally do.
The most systematic theories of human motivation, in fact, have been devel-
oped by clinicians who have tried to determine why people behave in abnormal
36 Human Motivation
ways. These clinicians have had ample opportunity to build and check such the-
ories in the course of psychotherapy as they listened to the waking and dream-
ing thoughts of patients hour after hour and day after day, sometimes for years.
To a surprising extent clinicians have tended to agree that there are a few basic
human motives, although they have not always agreed on just what they are.
This approach has the obvious advantage of paying close attention to what peo-
ple actually are thinking and doing, but it also has the important disadvantage
of using data that are drawn almost exclusively from the minds of people who
are disturbed enough to need psychotherapy. Thus, it is not surprising that the
motives clinicians have discovered tend to be negative. They are conceived as
powerful urges that, if not controlled or moderated, can produce mental and be-
havioral disorders such as depression, psychosis, and neurosis.
The Freudian Motivational Trilogy: Sex, Aggression, and Anxiety

Freud concluded on the basis of his studies of dreams and free associations, as
described in Chapter 1, that the major human needs were—to oversimplify a
bit—to obtain sexual satisfaction (in the broadest sense); to express aggression;
and to reduce the anxiety and suffering that result from conflicts involving the
first two drives, from conflicts of those drives with the demands of society, or
from threats to survival. Freud viewed all three of these motives negatively as
potentially producing suffering or illness. In this respect he was following a tra-
dition set more than two thousand years earlier in the West by Plato. In Book 9
of The Republic, Plato has Socrates speak of the unnecessary pleasures and de-
sires that
. . . are active during sleep . . . when the rest of the soul, the reasoning, gentle, and
ruling part of it is asleep. Then the bestial and savage part, when it has had its fill of
food or wine, begins to leap about, pushes sleep aside and tries to go and gratify its
instincts. You know how in such a State it will dare everything, as though it were
freed and released from all shame and discernment. It does not shrink from attempt-
ing incestual intercourse in its dream, with a mother or with any man, god or beast.
It is ready for any deed of blood.
Like Freud, Plato identified the major motives observable in dreams as in-
volving sex (including incest) and aggression, and he also labeled them as un-
lawful or disruptive in the sense that they would fall short of "no extreme of
folly or shamelessness." He contended earlier that these unlawful desires were
"probably innate in everyone" and that they could be "disciplined by law and
by the better desires with the assistance of reason." Freud did not really believe
that there are any "better desires," but he did agree that the instinctual, unlaw-
ful desires that he consigned to the unconscious id had to be controlled either
by society (law) or by reason. In fact, he spoke of psychoanalysis, the instru-
ment he devised, as the means for the progressive conquest of the id through
reason (Freud, 1927a).
Bad Desires and Good Desires
The distinction between bad or unlawful desires leading to sickness or sin and
good desires is present in various religious traditions that emphasize that the
good desires are the means of Controlling or disciplining the bad desires. In
Christian writings the sins related to excessive sexual or aggressive desires are
mentioned prominently, but so is the virtue of loving kindness, as in St. Paul's
famous passage on charity in his second letter to the Corinthians. Or, as
Thomas put it in his epistle, "God is love and he who abides in love abides in
God, and God abides in him." In Dante's dramatization of Christian theology
in The Divine Comedy, the sins deriving from the unlawful desires are vividly
described in the Inferno. Reason in the form of Dante's guide, Vergil, helps
Dante understand how bad desires lead to sin and punishment as Vergil leads
him through hell and purgatory, but Dante sees that he can only attain heaven
and overcome these unlawful desires through the divine gift of grace, which con-
fers the virtues of loving kindness, peace, and harmony with all.
In Buddhism the scenario is similar, although the concepts are different. To
Buddha the most obvious fact of life was suffering. He viewed desires as the
cause of suffering and therefore concluded that desires should be abandoned as
soon as possible. He proposed as a method of disposing of desires the disciplined
technique of meditation, which after years of practice enables a person to live
without desires—or at least without unlawful or ordinary desires. This mode of
coping with bad desires functions much as reason functioned for Plato, Dante,
or Freud. However, in folk Buddhism, as Gombrich (1971) has pointed out, the
desire to be loving and kind has become a virtue to be practiced as a means of
eradicating some of the more unlawful and selfish desires, just as it is in Chris-
tianity. Even in the theology of Buddhist meditation, loving kindness is a desir-
able stage on the way to giving up all desires. In Buddhism, therefore, the desire
to get rid of desires is a good desire that continuously motivates monks as they
pursue arduous meditation practices. In fact, Buddha's last words were said to
be something like "strive on mindfully." Clearly, there are good desires in this
religious tradition as well as in the Christian tradition.
Freud did not believe in good desires such as kindness, perhaps because his
patients, being ill, were unable to show them, or perhaps because of his general
pessimism about human nature. In fact, he interpreted loving kindness as really
representing other self-serving motives. Take, for example, the loving kindness
Christians believe Jesus showed in dying on the cross to help sinners attain sal-
vation. Freud (1918/1938) explained this understanding of the Crucifixion as a
manifestation of deeper unconscious motives in all of us. He started with the as-
sumption that all little boys wish to sleep with their mothers, are jealous of their
fathers, hate their fathers, wish to kill them, but feel guilty over this wish be-
cause they also love their fathers. Their guilt over their incestuous and aggres-
sive wishes normally leads them to want to atone in some way, and in the end
they do so by identifying as closely with their father as possible. Thus, the ap-
peal of Jesus' death on the cross is precisely that a son like themselves is killed
as a means of atoning for their hidden, unconscious incestuous and aggressive
38 Human Motivation
desires. Jesus' Crucifixion helps males get rid of their guilt. Thus, people inter-
pret his act as representing loving kindness, but in fact it is perceived as such
only because of their own guilt. In this Situation, as in so many others Freud wrote
about, the master motives turn out to be sex, aggression, and anxiety or guilt.
Jung's View of Motivation

Other clinicians working at first within the Freudian tradition disagreed about
the major forces that guided human behavior. Carl Jung, a Swiss physician of
Protestant background, was an early follower of Freud. As Chapter 1 men-
tioned, Jung was welcomed by Freud because he was eager to extend the influ-
ence of psychoanalysis beyond the largely Jewish circle of analysts that clustered
around him in the early days in Vienna. Jung also studied the dreams of his pa-
tients; however, he tended not to interpret them in terms of sexual urges, but in-
stead to see them as representative of universal themes or archetypes.
For instance, in Jung's (1961) autobiography he describes a powerful dream
he had as a young boy of going Underground and seeing a giant phallus sticking
straight up from a throne. As he gazed at it, he heard his mother's voice say,
"That is the man eater," which he associated with a prayer she had taught him
to say before he went to bed every night. The prayer spoke of being protected
by, taken in by, or "eaten" by Lord Jesus to protect him from Satan, who
wanted to devour him. A Freudian might want to interpret Jung's vision as an
oedipal dream in which he is identifying with or admiring the phallus, which he
wishes he could use on his mother. However, she wams him of its dangers, be-
cause it represents the father—both admired and feared—who will, like Lord
Jesus, take him in or eat him. Jung enlarges rather than reduces the meaning of
the vision and sees it as reflecting the conflict between the loving Lord Jesus
(seen as above ground in the sunshine) and the Underground Lord Jesus associ-
ated with death and dying. Jung was well aware from associating with his fa-
ther, a pastor, that at funerals and in the cemetery the dead were seen as being
taken in or protected by the Lord Jesus.
Thus, sex for Jung becomes not the root motive, but a means of dramatiz-
ing the real motivational conflict over loving and fearing the Jesus figure, which
in turn can be traced to a universal image, or archetype, involving father fig-
ures. Dreams were used by Jung not so much to diagnose basic human motives
as to enlarge the patient's understanding for therapeutic purposes. Thus, Jung
has little to say directly about what the basic human motives are except to
argue that every individual has a fundamental instinctive drive to push toward
individuation or self-realization, a theme that recurs in the writings of other
therapists as well.
Basic Motives and Later Psychoanalytic Theorists

Of the many psychoanalysts who came after Freud, space permits us to mention
only a few who made distinctive motivational contributions. Alfred Adler (1917)
was another of Freud's early disciples who eventually broke with him because
he came to see the power drive as more central than the sexual drive. Adler feit
that all children developed a striving for superiority because they all experienced
the weakness of being small and less capable than adults. Thus, everyone is mo-
tivated primarily by the desire to compensate for weakness, to become stronger
and more important in the eyes of others. This evolves into "social interest," a
legitimate motive in its own right, according to Adler. The search for superior-
ity in this form is similar to the drive for self-realization as described by Jung
and others.
Freud always resisted yielding the primacy of the sexual instinct to other
drives, as was advocated by Jung, Adler, and others, for several reasons. As
Chapter 1 pointed out, Freud appears to have drawn much of his inspiration for
psychoanalysis from the Chassidic tradition, which expresses the ultimate mean-
ing of life in sexual terms. Even more important, the theory of sex was one of
Freud's critical links with biology and the theoretical orthodoxy of nineteenth-
century natural science. To cast it aside, as he told Jung, was to slide into "the
black tide of mud of the occult" (that is, antiscience). The sexual instinct was to
be understood developmentally as a recapitulation of phylogeny, and recapitula-
tion theory held a prominent position among Darwinians—those developing the
theory of evolution—at the turn of the Century (Sulloway, 1979). According to
the theory, the stages of development of the sexual instinct in the individual are
the same as the stages of its development in the history of humankind.
Despite his insistence on the primacy of the sexual drive, Freud also recog-
nized the great importance of the aggressive instinct and the drive for power.
When Freud wrote about his own motives in the history of the psychoanalytic
movement, it is clear from the quotations in Chapter 1 that he seemed often to
be motivated primarily by a drive for power or glory that would not brook com-
petition from others so far as his views on the ultimate nature of human beings
were concerned.
Later therapists in the psychoanalytic tradition, such as Karen Horney and
Harry Stack Sullivan, emphasized the importance of the other basic motive in
the Freudian trilogy, anxiety. Horney (1945) feit basic anxiety comes from "the
feeling a child has of being isolated and helpless in a potentially hostile world.
A wide ränge of adverse factors in the environment can produce this insecurity
in a child: direct or indirect domination, indifference, erratic behavior, disparag-
ing attitudes, too much admiration or the absence of it." As a result of this
basic insecurity, children develop motives either to move toward people (to gain
protection, warmth, and support), to move away from people (to be self-
sufficient or unassailable), or to move against people (to show they are more
powerful than other people or to pay them back for past unjustices). At the root
of all motives, however, lies anxiety.
In Sullivan's (1953) view anxiety derives from basic physical needs such as
food or oxygen, but more importantly from a tension transmitted empathically
from mother to child. The infant develops a "self-system" as a response, which
exists as an "organization of experiences" designed to minimize anxiety. Anxiety
is not the only motive for Sullivan, although it is the first. The need for inter-
personal intimacy arises developmentally in preadolescence as a motive powerful
40 Human Motivation
enough to preempt anxiety, and impel the organism to transcend the self-system,
risk anxiety, and seek "chumship" in order not to experience loneliness, which
is worse than anxiety. Note that the new motive is another form of insecurity.
Sullivan's and Horney's views happened to coincide with those arrived at out of
the animal behavior tradition (described in Chapter 3), in which anxiety and
anxiety reduction came to be viewed as the master motive.
MOTIVES AS REASONS
FOR CREATIVITY AND GROWTH
In contrast to the psychoanalysts, who focused on the negative motives that lead
to illness because they inevitably conflict with one another or society's demands,
American psychotherapist Carl Rogers (1942) found evidence in his patients of a
basically constructive motive—the drive to become self-actualized. According to
Rogers (1951), "the organism has one basic tendency in striving—to actualize,
maintain, and enhance the experiencing organism." This rush toward self-actual-
ization is instinctive and present in everyone, much as Jung and Adler had sug-
gested. Rogers also feit two other important needs are acquired in growing up.
One is the need for positive regard from others or, more simply, the need for
love and acceptance by others. It develops from the fact that children learn that
some things they do are approved and others are disapproved by significant oth-
ers. They come to want the satisfaction that comes from approval. They also
consistently act to defend themselves against disapproval, which indicates they
are developing another need—the need for positive self-regard. Notice that all
three of Rogers' needs are positive—the need for growth, for love, and for self-
respect.
Thus, Rogers was taking a Step in the direction advocated by Abraham
Maslow (1954, 1967, 1968), who pointed out that there was a bias in the basic
motives attributed to human beings that resulted from the fact that the people
being studied were nearly all sick or unhappy. Thus, the motives discovered
seemed predominantly negative. Maslow argued that the picture should be bal-
anced by just as careful study of very healthy people. He too feit there is a basic
"impulse toward growth" in all human beings, which is perhaps weak and rela-
tively easily frustrated, but nevertheless is in every individual. He found strong
evidence for the existence of such a drive toward growth in unusually active and
healthy people he studied, such as Eleanor Roosevelt, Abraham Lincoln, and
Albert Einstein. Maslow did not deny that there are negative needs, but as Plato
did, he distinguished between negative, or "deficit," needs such as those for
being loved and for self-respect and what he later came to call the "metaneeds,"
which were what Plato would have called the "better desires"—namely, those
for justice, goodness, beauty, and order.
The best-known contribution of Maslow (1954) to motive theory is his clas-
sification of human needs in a hierarchy running from basic physiological needs
to self-actualization needs (Table 2.2). The classification was a response to the
problem so cogently stated by Allport, who argued that whereas it might be ac-
Table 2.2.
MASLOW'S HIERARCHY OF NEEDS (after Maslow, 1954)
Lower, or Deficit, 1. Physiological needs (need for food, water, and sex): homeostatic and organic.
Needs 2. Safety needs (need for security and protection from pain, fear, anxiety, and
disorder); need for order, lawfulness, and discipline.
3. Need for belongingness and love (need for love, tenderness, and togetherness).
4. Esteem needs (need for achievement, respect, and approval).
Higher, or 5. Self-actualization needs (need for self-fulfillment, for realizing one's potential,
Growth, Needs for understanding and insight).
curate to describe an infant as primarily motivated by physiological needs such

as hunger and the relief of discomfort and anxiety, it certainly is inaccurate to
think of mature adults as motivated primarily by such needs, even in the most
derived or symbolic form. Allport (1937) argued for the "functional autonomy
of motives," meaning that higher motives, such as the need for self-esteem, de-
velop later in life independently of the lower physiological needs.
Maslow solved the problem by including both kinds of needs in his hierar-
chy and insisting that higher needs can emerge only as lower needs are satisfied.
He used the image of primitive people to explain this. To survive, primitive peo-
ple had to first satisfy their physiological needs—for food, water, and sex (to in-
sure survival of the race). As they met these needs enough to survive, their next
concern was for safety—from the saber-toothed tiger or from the elements (heat,
cold, and floods). Therefore, primitive people lived in caves or built shelters. As
their security needs were satisfied, they were free to establish tender relation-
ships with fellow human beings. As they were assured of love, they were able to
work on satisfying needs for achievement and self-esteem. Underlying this devel-
opment, however—even in primitive people—was a push toward growth, indi-
viduality, or self-actualization. People do not rest content as their lower needs
are satisfied but always are pushing on to higher things. Even if all the lower
needs are satisfied,
. . . we may still often (if not always) expect that a new discontent and restlessness
will soon develop, unless the individual is doing what he, individually, is fitted for.
A musician must make music, an artist must paint, a poet must write, if he is to be
ultimately at peace with himself. What a man can be, he must be. This need we
may call self-actualization. . . . It refers to man's desire for self-fulfillment, namely,
to the tendency for him to become actualized in what he is potentially. (Maslow,
1954)
Lower needs are characterized by deficits—such as the lack of food and

water—which push the organism to become active and to seek substances that
will bring the organism back into homeostatic balance. Higher needs are less
42 Human Motivation
urgent and pull rather than push the person toward positive goals. They develop
later in life than lower needs, but they also are "instinctoid"—that is, inborn in
everyone. However, they are weaker and can be prevented from developing by
the urgency of working to satisfy lower needs. Needs affect behavior only when
they are unsatisfied. As they are satisfied, the person is free to pursue higher
needs. Satisfying lower needs removes tension and gives peace and satisfaction.
Satisfying higher needs leads more directly to joy and a feeling of personal
accomplishment.
From his study of outstanding creative individuals presumably functioning
at higher levels in his hierarchy, Maslow concluded that being able to satisfy
lower needs and seek self-actualization leads to greater biological efficiency
(sleeping better, eating better, longer life, and less sickness) and many desirable
human traits such as spontaneity, reality orientation, spirituality, ability to dis-
tinguish between means and ends, creativity, autonomy, and democratic values.
His theory gained wide acceptance in the mid-twentieth Century because it fit in
well with the liberal view that poor and oppressed peoples were prevented from
functioning at higher levels because poverty and oppression forced them to
spend their time trying to satisfy physiological (for example, food) and safety
(for example, housing) needs, leaving them no time or energy to develop self-
respect or their own potential. His theory was a major force in the development
of the humanistic psychology movement in the United States, which stressed the
primacy of the higher needs he had defined as opposed to the lower needs that,
if unsatisfied, could lead to mental illness and despair.
Maslow's view can be criticized on the grounds that he picked creative indi-
viduals to study who showed the characteristics he thought such people should
have. For example, he did not choose to study Richard Wagner, a creative mu-
sician who showed almost none of the characteristics Maslow valued. In his the-
orizing he also tended to neglect the nature of the environmental conditions that
arouse the various needs in the hierarchy. Nor did he undertake the empirical
investigations necessary to show that the needs really do form a hierarchy. Oth-
ers have looked into this matter, as Chapter 10 will show, but what is most im-
portant about Maslow is that he gave psychologists a positive way of thinking
about motives to contrast with the negative view deriving from the psychoana-
lysts, who learned about motives primarily from studying patients in need of
therapy.
MEASURING HUMAN MOTIVES

The motivational theorists so far mentioned provided psychology with a concep-
tual framework in terms of which to think about basic human motives, as well
as a language of human motivation, so to speak. None of them, however, made
a systematic attempt to measure the motives they thought were important. Thus,
to a certain extent, they are guilty of the naming fallacy mentioned earlier. They
observed that people thought and acted in certain ways and inferred that there-
fore they must have wanted to think and act in those ways. The only way to es-
cape such a fallacy is by measuring the motives in question.
An early attempt at measuring human motives was made by W. H. Sheldon

(1942), who was primarily interested in the varieties of human physique, which
he called somatotypes. He had concluded that human bodies could be rated on
three basic characteristics—namely, the extent to which fat predominated, or en-
domorphy; the extent to which muscle predominated, or mesomorphy; and the
extent to which skin predominated, or ectomorphy. He then argued that certain
temperament traits (which are in part motivational) derive from each of these
characteristics of the body. Sheldon developed temperament rating scales, which
included motivational variables. Thus, the viscerotonia scale, which was associ-
ated with endomorphy, contained items like "love of physical comfort" and "so-
ciophilia," or love of being with people. The somatotonia scale, which was asso-
ciated with mesomorphy, included items like 'love of physical adventure," "love
of dominating," and "lust for power." The cerebrotonia scale, which was associ-
ated with ectomorphy, contained items like "love of privacy" and "sociopho-
bia," or dislike of being with people.
The motive measurements are very crude, but they do represent attempts to
be more specific about some of the motives clinicians had talked about. And
Sheldon identified two of the motive types that have cropped up again and again
in psychological literature, namely, the contrast between those who love risk and
physical adventure (somatotonics) and those who love privacy (cerebrotonics).
The former also have been called extraverts and the latter, introverts, following a
distinction first popularized by Jung and later forming a major orientation in the
work of the British psychologist H. J. Eysenck (1947). The distinction appeared
again in the work of Zuckerman (1974), who has measured the extent to which
people seek excitement and Sensation, as extraverts do, or seek to avoid it, as in-
troverts do. The work on extraversion-introversion will be discussed again in
later chapters, although its place in the psychology of motivation is not clear be-
cause it has been conceived primarily as a trait—a consistent way of behav-
ing—rather than as a difference in motivational orientation.
Such theorists have attempted to delimit the bewildering variety of possible
human motives by anchoring them to a more limited set of biological character-
istics. This strategy was pursued vigorously by those who concentrated on ani-
mal motivation, as the next chapter will show.
Murray's Study of Motives

A systematic approach to measuring human motives was introduced by Henry
A. Murray (1938) at Harvard University, who sought ways of assessing the mo-
tives clinicians had found to be important. His method was eclectic: he set out
to measure in a variety of ways any and all of the motives previous theorists
had feit to be important. This method had several advantages from the point of
view of advancing knowledge of human motives.
First, Murray primarily was interested in understanding and measuring
human motives as contrasted with other aspects of personality such as traits,
habits, or skills, Thus, he realized that whereas investigators could look for con-
sistency from one act to the next so far as a trait like assertiveness was con-
cerned, motives often would find quite different behavioral outlets in different
44 Human Motivation
people. For example, a desire for power might be satisfied by a person's watch-
ing sex and aggression movies on television, arguing with other people, or join-
ing organizations to feel more powerful as part of a larger group. The people
who often argued, however, would not necessarily be more likely to watch sex
and aggression programs on television or join organizations, so any attempt to
build a measure of a "need for Power" by adding together such activities would
be likely to fail. This simple fact is well known to clinicians, who often see
Symptoms as alternative manifestations of basic needs, but it has not been well
understood by personality theorists, who keep looking for consistency in the be-
havioral manifestations of motives. As Murray pointed out, motives are neces-
sary to explain inconsistencies in personality and traits are necessary to explain
consistencies.
Second, because Murray saw motives as expressing themselves in a variety
of alternative ways, he stressed that their manifestations should be studied in
every possible way—through autobiographies, through behaviors in laboratory
experiments, through reveries associated with music, through dreams, and
through questionnaires about sentiments and attitudes—but above all through a
Special instrument he devised and called the Thematic Apperception Test
(TAT). Of all the measurement techniques Murray and his collaborators invent-
ed, this is the one that caught on, particularly in clinical psychology. In effect, it
is a method of systematically collecting the kind of free associations Freud and
his successors had used to draw inferences about motives in their clinical work.
In the TAT, subjects are asked to teil imaginative stories after viewing a series
of twenty pictures suggestive of key emotional complexes in the life of the
individuals, such as the relationship of father to son or mother to daughter.
The first ten pictures represent typical dramas from normal life (for example,
father-son or mother-son relationships) and the second ten, more fantastic scenes
to elicit more deeply repressed unconscious associations. The resulting "day-
dreams" are subjected to the same kind of analysis carried out by psychoana-
lysts, which will be illustrated later in the chapter.
The third advantage of Murray's method was that the subjects in his re-
search were small groups of College undergraduates who were studied intensively
over the entire four years they were in College. Thus, he collected as much ma-
terial as psychotherapists often collect, but on relatively normal individuals, so
the bias toward seeing motives as leading to illness could be avoided. Further-
more, he gathered enough material over time to study consistencies and incon-
sistencies in motives and the extent to which they were changed, aroused, or
weakened by events that occurred during the four years. However, there is some
bias in the fact that the subjects were all male and quite intelligent and articu-
late compared with the general run of individuals.
Fourth, to measure motives properly, Murray realized he had to begin by
carefully defining each of the motives he was trying to measure. He gathered to-
gether a diagnostic Council of experts representing a variety of motivational tra-
ditions to decide what the main motives were and how they should be defined
and measured. Perhaps his greatest contribution, in addition to the TAT, is a
vocabulary of human motives that has shaped work in the field ever since. Table
2.3 lists some of the most important needs he measured, together with their
definitions.
Obviously, Murray's approach does not reduce all human needs to one basic
motive force (such as Jung's, Rogers', or Maslow's drives for self-realization), or
to two or three basic motive forces (such as Freud's sex, aggression, and anxiety
needs). Murray also does not fall into the opposite difficulty of listing so many
needs that working with all of them becomes impractical. Science is an econo-
mizing enterprise. lts purpose is to explain as much as possible of the bewilder-
ing variety of events with as few basic constructs and laws as possible. In this
respect the field of human motivation is no different: the goal is to find the few-
est motives that will account for the most human behavior. Obviously no set of
human motives can account for everything a particular individual thinks or
does. Only three motives in Murray's list have been the subject of extensive in-
vestigation since his time: the need for achievement, the need for affiliation, and
the need for power (or dominance), as later chapters of this book will describe.
The fifth advantage to Murray's method was that his Council of experts
solved the problem of measuring the strength of an individual's need in all its
manifestations by asking each judge to rate the need based on all information
available, and then by having the ratings discussed in a group to arrive at a
final rating that would represent a synthesis of the viewpoints of all the judges.
This procedure gave the Council greater confidence that it was arriving at the
truth about a person's needs, but as a measurement technique it has important
drawbacks. A difficulty with this procedure lies in the group process by which
the final rating is reached. One judge may have been more influential than an-
other in arriving at the final rating; it does not follow, however, that this judge
was necessarily nearer the truth, nor does it follow that the opinion of a group
of judges is better than the opinion of the "best" judge by some outside criteri-
on. Another difficulty lies in the fact that it is impossible to reconstruct exactly
what thoughts or actions the judges were using in arriving at their ratings. It is
not even possible to be sure that most, or even any, of the judges were closely
following the motive definitions in making their judgments. Methods of dealing
with these difficulties through more precise coding definitions have been devel-
oped (see Chapter 6).
Catteü's Identification of Motives Through Factor Analysis

Like Murray, Raymond Cattell (1957, 1965) has sought to identify and measure
the major human motives. He too recognized that motives were only one of the
determinants of behavior (along with abilities and temperament traits), and he
too obtained many different presumed indicators of different motives. However,
instead of using human judges to synthesize them all into a numerical rating for
the strength of a particular motive in a person, as Murray had done, Cattell em-
ployed factor analysis to arrive at a final score for a motive for a person. Factor
analysis is essentially a statistical method for extracting the factors that account
for covariation among various measures.
46 Human Motivation
Table 2.3.
ILLUSTRATIVE LIST OF MURRAY'S NEEDS (Hall & Lindzey, 1957, after Murray, 1938)
Need Brief Definition

Abasement To submit passively to external force. To accept injury, blame, criticism, or punishment.
To surrender. To become resigned to fate. To admit inferiority, error, wrongdoing, or
defeat. To confess and atone. To blame, belittle, or mutilate the seif. To seek and enjoy
pain, punishment, illness, and misfortune.
Achievement To accomplish something difficult. To master, manipulate, or organize physical objects,
human beings, or ideas. To do this as rapidly and as independently as possible. To
overcome obstacles and attain a high Standard. To excel oneself. To rival and surpass
others. To increase self-regard by the successful exercise of talent.
Affiliation To draw near and enjoyably cooperate or reciprocate with an allied other (an other who
resembles the subject or who likes the subject). To please and win affection of a cathected
object. To adhere and remain loyal to a friend.
Aggression To overcome Opposition forcefully. To fight. To revenge an injury. To attack, injure, or
kill another. To oppose forcefully or punish another.
Autonomy To get free, shake off restraint, break out of confinement. To resist coercion and
restriction. To avoid or quit activities prescribed by domineering authorities. To be
independent and free to act according to impulse. To be unattached, irresponsible. To
defy Convention.
Counteraction To master or make up for a failure by restriving. To obliterate a humiliation by resumed
action. To overcome weaknesses, to repress fear. To efface a dishonor by action. To
search for obstacles and difficulties to overcome. To maintain self-respect and pride on a
high level.
Defendance To defend the seif against assault, criticism, and blame. To conceal or justify a misdeed,
failure, or humiliation. To vindicate the ego.
Deference To admire and support a superior. To praise, honor, or eulogize. To yield eagerly to the
influence of an allied other. To emulate an exemplar. To conform to custom.
Dominance To control one's human environment. To influence or direct the behavior of others by
Suggestion, seduction, persuasion, or command. To dissuade, restrain, or prohibit.
Exhibition To make an impression. To be seen and heard. To excite, amaze, fascinate, entertain,
shock, intrigue, amuse, or entice others.
Harmavoidance To avoid pain, physical injury, illness, and death. To escape from a dangerous Situation.
To take precautionary measures.
Infavoidance To avoid humiliation. To quit embarrassing situations or to avoid conditions that may
lead to belittlement: the scorn, derision, or indifference of others. To refrain from action
because of the fear of failure.
Nurturance To give sympathy and gratify the needs of a helpless object: an infant or any object that
is weak, disabled, tired, inexperienced, infirm, defeated, humiliated, lonely, dejected, sick,
mentally confused. To assist an object in danger. To feed, help, support, console, protect,
comfort, nurse, heal.
Order To put things in order. To achieve cleanliness, arrangement, organization, balance,
neatness, tidiness, and precision.
Table 2.3. (Continued).
Need Brief Definition

Play To act for "fun" without further purpose. To like to laugh and make jokes. To seek
enjoyable relaxation of stress. To participate in games, sports, dancing, drinking parties,
cards.
Rejection To separate oneself from a negatively cathected object. To exclude, abandon, expel, or
remain indifferent to an inferior object. To snub or jilt an object.
Sentience To seek and enjoy sensuous impressions.
Sex To form and further an erotic relationship. To have sexual intercourse.
Succorance To have one's needs gratified by the sympathetic aid of an allied object. To be nursed,
supported, sustained, surrounded, protected, loved, advised, guided, indulged, forgiven,
consoled. To remain close to a devoted protector. To always have a supporter.
Understanding To ask or answer general questions. To be interested in theory. To speculate, formulate,
analyze, and generalize.
An example of how Cattell worked will make it easier to understand his

method. He started with the supposition, just as McDougall and Murray had,
that there might be a motive disposition for affiliating with others, which he la-
beled Gregariousness. One way to find out if people are strong in such a motive
is to ask them to introspect on their desires in the area by answering yes or no
to Statements such as "I want to belong to a sociable club or team of people
with congenial interests." But Cattell realized that such self-reports are influ-
enced by response sets—sentiments or other factors that have nothing directly to
do with the motive in question. Thus, he also sought a number of other indica-
tors of Gregariousness such as the knowledge people had about clubs available
in the area, how many clubs the people would join or had joined, and the emo-
tional response of the people (as reflected in imperceptible sweating affecting the
galvanic skin response) when presented with a Stimulus phrase such as 'Today
too many people are joining clubs."
The next question to be answered by factor analysis is whether all of these
different indicators covary, or "hang together." That is, do people who say they
want to join clubs also react emotionally to the Statement that too many people
are joining clubs? Cattell argues that it can be concluded that there is a motive
called gregariousness only when such a series of indicators related to a common
goal can be shown to covary, preferably by the objective method of factor analy-
sis, which is based on the statistical intercorrelation of indicators. He defines a
motive (or Erg, in his terminology) in part as starting "various courses of action
(equivalents) which cease more completely at a certain common, definite, con-
summatory goal activity than at any other. The common goal character is de-
monstrable by, among other methods, a factor-analytic proof of functional unity
in attitude-action courses that can be perceived (or demonstrated) by the psy-
chologist to lead to a common goal" (Cattell, 1957).
48 Human Motivation
Cattell has carried out dozens of factor analytic studies on dozens of such
indicators in thousands of subjects in an effort to define the major motive dispo-
sitions and find objective measures of them. He has arrived at a list similar to
Murray's and McDougalPs and includes such Ergs (motive energizers) as curios-
ity, sex, gregariousness, protection, self-assertion, security, hunger, anger, and
disgust. His method appeared to be particularly promising, because it provided
an objective, statistical method of "proving" the existence of a motive and of
combining different indicators into a Single factor score for the motive, as con-
trasted with Murray's intuitive method of attaining these same goals.
However, the approach also has its drawbacks. First, what comes out of a
factor analysis depends entirely on what measures are put into it, which in turn
depends on the psychologists' judgment as to what is relevant or important to
measure; this in turn depends on their intuitive understanding of.the field. The
results may show whether the psychologists' intuitive understanding was correct,
but they do not guarantee that they put in everything of importance. Second,
factor analytic results for different samples of subjects from whom the same
measures are obtained often reveal differences as to what indicators define an
Erg that can be reconciled only in a speculative, intuitive, or superficial way.
Third, and most important, the factor analytic score is a composite of many
different indicators that can be classified as belonging under a common label
only in a very loose way. Statistical precision may lead to conceptual confusion:
How could this behavioral indicator that loads high on the Gregariousness fac-
tor really indicate a desire to affiliate with people? And why does this other in-
dicator, which ought to indicate Gregariousness, fail to load on the factor?
Cattell has dealt with such questions in imaginative ways but has failed to
convince many psychologists that factor analysis provides the ideal way of solv-
ing such problems. Instead, as we shall see, the trend has been toward more
careful conceptual and measurement definitions of a few motives that then are
studied in detail rather than toward construction of the overall picture of all
motives and other personality variables sought by Murray and Cattell in differ-
ent ways.
• STAGES IN MOTIVATIONAL DEVELOPMENT

So far we have reviewed motives identified by personality theorists and some
preliminary attempts to measure them. However, such an overview leaves out
how personality theorists have identified the presence of motives in a concrete
case. Generally they have been most convincing in doing so when they have an-
alyzed fantasies, dreams, or more generally, thought processes. To give a better
idea of how this approach works in practice, considerable space will be devoted
to illustrating how it can be used to make inferences about a person's motives.
The context will be one of explaining more about the Freudian view of stages in
motivational development, since this has had a wide impact on the field, particu-
larly as it has been elaborated by Erik Erikson (1963).
Freud concluded not only that there were a few basic human motives, but
also that one of them, the sexual drive, developed in certain stages. In observing
how the sexual instinct manifested itself, he followed the strategy of linking
stages of development to biological events. He noted first of all that the most
striking character of sexual activity is "that the impulse is not directed to
other persons but that the child gratifies himself on his own body" (Freud,
1905/1938). He observed that "the first and most important activity in a
child's life, the sucking from the mother's breast (or its substitute)" (Freud,
1905/1938), has acquainted the child with the pleasure to be obtained from the
lips, which Freud referred to as the oral erogenous zone. Sucking and getting
food are normally combined, but sucking becomes pleasurable in its own right:
in Freud's view the sexual instinct gratifies itself first in the region of the mouth,
or the oral zone. In more general psychological terms, the oral stage represents
a period of dependency in which gratification comes from intake—from taking
something in from the outside.
Next, according to Freud, children begin to get gratification from the anal
zone—from holding on to and letting go of feces. This provides pleasure in the
simple biological sense, but it also represents a period in which children learn
self-control. They are now becoming independent in the sense that they alone
can control whether they hold on or let go.
In the third stage, children discover the pleasure to be obtained from ma-
nipulating the genitals. In more general terms, they are learning to get pleasure
from assertiveness, from actually doing something active to produce pleasure. In
psychoanalytic theory this period is referred to as the phallic stage to distinguish
it from the later, truly genital stage, which derives pleasure from the sexual in-
tercourse between a man and a woman and is characterized by mutuality.
Stages Reflected in a Freudian Dream

Using this elementary introduction to Freud's conception of psychosexual stages,
let us look at the method used by clinicians to analyze two dream sequences to
see how it enables them to make inferences about motivational stages, particu-
larly as they have been described more fully by Erikson (1963). The purpose of
such an exercise is not only to give a better idea of stages in motivational devel-
opment, but also to give a concrete demonstration of the modes of analysis typi-
cally used to arrive at motivational inferences from fantasy material.
The first sequence consists of two dreams reported by Freud, which were
later analyzed in some detail by Erikson (1964), who also made use of Freud's
associations to the dreams. The following is Freud's first dream, as described by
Erikson (1964):
I went into a kitchen in search of (um mir geben zu lassen) some pudding. Three
women were Standing in it, one of them was the hostess (Wirtin) of the inn and was
twisting (drehen) something about in her hand, as though she was making Knödel
(dumplings). She answered that I must wait till she was ready. (These were not defi-
nite spoken words.) I feit impatient and went off with a sense of injury (beleidigt—
insulted).
50 Human Motivation
Obviously, the dream has to do with the oral stage, with being fed. The En-
glish translation, Erikson points out, is misleading, because what the German
states is not that Freud goes actively in "search of " food, but that he goes into
the kitchen in order for someone to give him some food. He is dependent on
others to be fed. The fact that the dream has to do with being fed is not surpris-
ing, since Freud had it after going to bed without his supper; the way in which
the oral crisis is described, however, is of key importance.
Before interpreting it further, let us review Freud's associations to it. First,
he remembers a novel he read when he was thirteen years old (incidentally, this
is a critical point in a Jewish boy's life when he has his Bar Mitzvah represent-
ing his initiation into Jewish manhood) in which the hero went mad, calling out
the names of three women who had been responsible both for his greatest joys
and sorrows. Freud then recalls the three Fates, who in Greek mythology were
responsible for spinning people's destiny. Finally he realizes that the hostess is
his mother: "love and hunger, he reflects, meet at a woman's breast" (Erikson,
1964). His associations so far have to do with women Controlling people's fate.
He next recalls an incident that further illustrates that men do not have
control over their fate with women. He recalls that a man he was with noticed
how attractive his former wet nurse was and remarked that it was too bad he
had not been big enough to take advantage of his opportunities at the time.
Finally, Freud recalls an episode when he was six years old; his mother was
trying to illustrate for him the meaning of the biblical passage "dust thou art to
dust returneth." She rubbed her hands together vigorously until the skin peeled
off in little black rolls to convince him of the truth of the Statement. As Erikson
(1964) notes, "here it is important to see that the very origin of man, and, in
fact, the origin of living matter is at stake, and that the mother, the source of
life-giving food and of hope-giving love, herseif demonstrates the fact that her
very body is created of dead matter, of earth and of dirt" (p. 182). Freud's disil-
lusionment with women seems to be complete.
Now we are in a better position to interpret this dream in terms of the
characteristics of the oral stage as they have been outlined by Erikson in Table
2.4. In his search for food, the dreamer goes to the kitchen, where the food is
provided, and meets women, who normally provide him with food. Then he is
told he must wait. In the oral stage that is all the baby can do—wait initially
for the mother to provide nourishment somewhat magically. Whether the
mother provides nourishment and love in a predictable manner or not causes the
development of the sense of trust or mistrust that characterizes the psychosocial
equivalent of the oral psychosexual stage, according to Erikson. As he puts it
elsewhere, "it is clear then that the Optimum total Situation implied in the
baby's readiness to get what it is given is his mutual regulation with the mother
who will permit him to develop and coordinate his means of getting as she de-
velops and coordinates her means of giving" (Erikson, 1963). The ability of the
growing child to develop trust and hope depends on this mutual coordination.
If the coordination fails in some critical sense, the child distorts reality in
an attempt to get basic oral gratification and becomes either delusional, as in
some types of schizophrenia; addictive, an attempt to get oral gratification
Table 2.4.
CHARACTERISTICS ASSOCIATED WITH PSYCHOSEXUAL STAGES (after Erikson, 1964)
Related
Psychopatho- Related
Psychosexual Psychosocial Rudiment of logical Elements of
Stage Organ Mode Stage Ego Strength Mechanisms Social Order
I. Oral-sensory- Incorporative Basic trust Hope Psychotic, Cosmic order
cutaneous versus mistrust addictive
II. Muscular-anal- Retentive- Autonomy Will Compulsive, Law and order
urethral eliminative versus shame impulsive
and doubt
III. Phallic- Intrusive Initiative Purpose Inhibitive, Ideal prototype
locomotor versus guilt hysterical,
phobic
through drugs; or depressed, as when mistrust develops to such a degree that a

person gives up trying to adapt. Thus, from a psychoanalytic point of view,
traumas at the oral stage are the most serious in that they cause the most fun-
damental breaks with reality, leading to such major disorders as schizophrenia
and depression.
What about Freud's reaction in the dream and his associations to it? He
does not want to wait, goes off with a sense of injury, and in his associations
makes it very clear that men cannot depend on women, who are "too perish-
able, too mortal, and too dangerous" (Erikson, 1964).
Instead, in the next part of the dream he turns away from women toward
men and toward establishing his own identity. The second part of the dream is
as follows:
In the second part of the dream one lone man appears, and no women. After some
altercations, a stranger and the dreamer become "quite friendly with each
other"—which "ends the dream." (Erikson, 1964)
In associating to this dream Freud first thinks of a number of names of men

reminiscent of food. This reminds him of his own name, which has inspired
some crude jokes. As Erikson points out, in German the two most obvious re-
lated names are Freudenhaus (house of joy, or whorehouse) and Freuden-
mädchen (girl of joy, or prostitute). Again, Freud seems to be casting aspersions
on women with his name. As Erikson (1964) sums it up, "If your own mother
is made of earth or dirt, or worse, and if your own name is like a curse, you
cannot trust mother, origin, or fate: you must create your own greatness and,
indeed, all the dreamer's associations concerning men converge on the great Vi-
52 Human Motivation
enna Institute of Physiology in which, so the dreamer says, he spent the happi-
est hours of his Student days" (p. 183). Freud recalls a poem by Goethe about
yearning daily evermore for the "breasts of wisdom," or knowledge, which,
since it is related to the real world, gives people power to control what happens
to them. This differs from the oral period, in which they can only wait.
Freud then remembers cocaine, a substance he had introduced for anesthetic
purposes before its addictive powers were known. This suggests there are also
dangers in incautious intake. Thus, the role of mistrust is precisely to warn the
growing child against taking everything in heedlessly, but obviously the danger
of too much mistrust is that it inhibits taking anything in at all. Thus, Freud
feels the dream reminds him that he should take advantage of opportunities
even though there may be some drawbacks in doing so, as in the case of co-
caine. As Erikson (1964) sums it up,
. . . the second part of the dream, then, emphasizes the turn from dependence to
self-help, from women to men, from perishable to eternal substances, and ends with
a friendly affiliation with a man with a pointed beard—a paternal teacher figure. . . .
If the first part of this dream, evoked by hunger, goes back to the actuality of the
first stage of life, the second part leads (as I think all successful dreams do) forward
again; for it obviously promises to the sulky dreamer autonomy from women and
participation in the world of intellectual skills.
To return to the stage model presented in Table 2.4, Freud has thought
himself forward in his dream from Stage I, where the issues have to do with
waiting and taking in, to Stage II, where the person has progressed to issues of
autonomy and self-help. Failures in development at Stage II lead to compulsive
or impulsive disorders. People either become too willful in holding on or at-
tempting to control everything in an obsessive-compulsive way, or they give up
and act always on impulse, or they alternate between the two modalities. Erik-
son also makes the point that just becaue Freud is thinking in his dream about
the issues of the oral stage, it should not be inferred that he has in any way "re-
gressed." Instead, his hunger reminds him of one of life's major themes and the
way he has successfully dealt with it by moving from dependence to autonomy,
or from Stage I to Stage II.
Stages Reflected in a Murderer's Dream

There are other examples of progression through stages in a dream. In his book
In Cold Blood, Truman Capote (1965) gives a vivid account of the lives of two
young men with criminal records, Perry Smith and Dick Hickok, who brutally
murdered a farmer and his family. What caught the attention of the press and
Capote was the fact that the murder, particularly the first murder of Mr. Clut-
ter, the farmer, seemed to be entirely without a motive. Perry and Dick had
broken into the farm at night in the hopes of stealing some money, but they had
found none. They had no intention of harming the family, but having wakened
the people and tied everyone up, they hung around for several hours trying to
decide what to do. In attempting to explain what then happened, Perry said in
his confession after he was caught, "I didn't want to härm the man. I thought
he was a very nice gentlemen. Soft-spoken. I thought so right up to the moment
I cut his throat." Even the psychiatrists from the Menninger Clinic who exam-
ined Perry could not really find a motive for the murder. They noted that he
had a history of impulsive acts of violence and concluded that somehow some-
thing in the Situation had elicited a similar act.
Fortunately from the point of view of motivational analysis, Capote provides
a very detailed record of Perry's dreams and associations. What follows is a
dream Perry reports to Dick:
As long as you live, there's always something waiting, and even if it's bad, and you
know it's bad, what can you do? You can't stop living. Like my dream. Since I was
a kid I've had this same dream, where I'm in Africa. A jungle. I'm moving through
the trees toward a tree Standing all alone. Jesus, it smells bad, that tree; it kind of
makes me sick, the way it stinks. Only it's beautiful to look at—it has blue leaves
and diamonds hanging everywhere. Diamonds like oranges. That's why I'm
there—to pick myself a bushel of diamonds. But I know the minute I try to, the
minute I reach up, a snake is gonna fall on me. A snake that guards the tree. This
fat son of a bitch living in the branches. I know this beforehand, see? And Jesus, I
don't know how to fight the snake. But I figure, well, I'll take my chances. What it
comes down to is I want the diamonds more than I'm afraid of the snake. So I go
to pick one, I have the diamond in my hand, I'm pulling at it, when the snake lands
on top of me. We wrestle around, but he's a slippery sonofabitch and I can't get a
hold, he's crushing me, you can hear my legs cracking. Now comes the part it
makes me sweat even to think about. See, he Starts to swallow me. Feet first. Like
going down in quicksand. (Capote, 1965, p. 92)*
From Freudian analyses of dreams and the experimental studies of sex sym-
bolism reviewed in Chapter 1, we know sexual thoughts and desires often dis-
guise themselves in dreams. The method of disguise involves replacing the sexual
objects or desires with other objects or desires that have similar attributes. Thus,
to decode the first part of Perry's dream, we must ask what has all the stränge
attributes he describes. That is, what is it (1) that is enormously valuable
("beautiful," "diamonds") that Perry wants desperately ("more than I am afraid
of the snake"); (2) that is in a stränge place (Africa, jungle); (3) that smells bad
("makes me sick, the way it stinks"); (4) that is round and attractive and hang-
ing "like oranges"; and (5) that if he takes it a snake falls on him and Starts to
swallow him.
It is not difficult to decipher this in terms of the Standard oedipal triangle.
Perry is reaching in his dream for his mother's breasts (which are seen as enor-
mously valuable to him, round and hanging, to be plucked and taken like a
treasure; this already makes us think he desperately needs them or his mother's
love). The whole scene takes place in a jungle, which normally represents the
unconscious in dreams. And this valuable prize is guarded by a snake that falls
*From IN COLD BLOOD by Truman Capote. Copyright © 1965 by Truman Capote. Reprinted by permission
of Random House, Inc. and Hamish Hamilton Ltd.
54 Human Motivation
on him when he reaches for it. The snake can readily be seen as the father fig-
ure, which punishes Perry for his incestuous wishes.
Some details in the dream, however, make sense only if we know a little
more about Perry's life history. Why should the tree smell bad and make him
sick? Perry's mother was a full-blooded American Indian and his father Irish.
They were at first expert rodeo performers ("Tex and Flo"), and Perry's happi-
est early childhood memories were of his mother as "a lean Cherokee girl riding
a wild horse, a bucking bronco." Unfortunately, they had to retire because of
ailments, and his mother "took to whiskey," left her husband, abandoned her
children, and became so promiscuous that she would sleep with any man who
would drink and dance with her beforehand. Perry had what he called "odious"
childhood memories of watching his mother sleeping with stränge men. When
she drank she often got sick, so in the end she "strangied to death on her own
vomit." Hence the bad smell in Perry's dream reinforces our belief that it is
really about his mother.
Perry loved his father and, when he was young, often tried to run away
from his slovenly mother to be with him. However, they also fought. Describing
one such battle, Perry says, "Dad, though, he's slippery, a smart wrestler."
Again, the fact that he uses the same imagery to describe the snake reinforces
our interpretation of the snake as a father figure. Notice, however, that the
snake is swallowing him and crushing his legs. Perry had in fact broken his legs
in a motorcycle accident some years earlier, an event that had a profound effect
on his image of himself as "a real man," which had always been shaky.
If we refer back to Table 2.4, we can see that, in terms of motivational
stage theory, this part of the dream has to do with taking initiative, with acting
on the world in Stage III fashion. However, Perry knows ahead of time that this
initiative will fail, and in fact it does fail and he is crushed for making the at-
tempt—crushed at least symbolically in the phallic region. Even more important
is the image of being sucked down into quicksand, which suggests going even
deeper into the unconscious, or regressing back to an earlier stage.
The account of Perry's dream continues:
Dick said, "So? The snake swallows you? Or what?"

"Never mind. It's not important." (But it was! The finale was of great impor-
tance, a source of private joy.) He'd once told it to his friend Willie-Jay; he had de-
scribed to him the towering bird, the yellow "sort of parrot." Of course, Willie-Jay
was different—delicate-minded, "a saint." He'd understand. But Dick? Dick might
laugh. And that Perry could not abide; anyone's ridiculing the parrot, which had
firstflowninto his dreams when he was seven years old, a hated, hating half-breed
child living in a California orphanage run by nuns—shrouded disciplinarians who
whipped him for wetting his bed. It was after one of these beatings, one he could
never forget ("She woke me up. She had aflashlight,and she hit me with it. Hit me
and hit me. And when the flashlight broke, she went on hitting me in the dark.")
that the parrot appeared, arrived while he slept, a bird "taller than Jesus, yellow like
a sunflower," a warrior-angel who blinded nuns with its beak, fed upon their eyes,
slaughtered them as they "pleaded for mercy," then so gently lifted him, enfolded
him, winged him away to "paradise."
As the years went by, the particular torments from which the bird delivered
him altered; others—older children, his father, a faithless girl, a Sergeant he'd known
in the Army—replaced the nuns, but the parrot remained, a hovering avenger. Thus,
the snake, that custodian of the diamond-bearing tree, never finished devouring him
but was itself always devoured. And afterward the blessed ascent! Ascension to a
paradise that in one version was merely "a feeling," a sense of power, of unassailable
superiority—sensations that in another version were transposed into "a real place.
Like out of a movie. Maybe that's where I did see it—remembered it from a movie.
Because where eise would I have seen a garden like that? With white marble Steps?
Fountains? And away down below, if you go to the edge of the garden, you can see
the ocean, terrific! Like around Carmel, California. The best thing, though—Well,
it's a long, long table. You never imagined so much food. Oysters. Turkey. Hot
dogs. Fruit you could make into a million fruit cups, and, listen—it's every bit free.
I mean, I don't have to be afraid to touch it. I can eat as much as I want, and it
won't cost a cent. That's how I know where I am."*
The next part of the dream confirms this impression. It has to do with
Stage II, the retentive-eliminative stage. Perry had wet his bed for years and had
been punished severely for it in the orphanage. The adjustments related to Stage
II, therefore, also had not been properly resolved. In fact, he was both compul-
sive and impulsive. On one hand, he was obsessively neat and clean, and one of
the things he disliked most about his companion in crime was that he was slop-
py. On the other hand, as the psychiatrist had noted, there were instances in his
life in which he had suddenly "let go," just as he had let go in wetting his bed.
What happens next? Once more a problem at a given stage in development
is resolved by regressing to an earlier stage. The wonderful parrot bird comes,
avenges him on the nuns who persecuted him, and lifts him to an oral paradise.
In his dream he revels in all the wonderful kinds of food available. And there is
no waiting, as there is in Freud's dream. "It's every bit free. . . . I can eat as
much as I want, and it won't cost a cent." Perry has regressed from Stage III to
Stage II, and then back to Stage I, where he rests totally happy in the kind of
loving support, symbolized in terms of food, that he never got in real life. In-
stead of thinking himself forward up the developmental scale, as Freud did, he
thinks his way backward to a dream of escaping to paradise.
There were hard feelings between Perry and Dick. Perry originally had
teamed up with Dick because he had doubts about his own ability to be tough
and strong and thought Dick was the image of the strong man. However, he
found that Dick was weak: "Just then it made my stomach turn to think that I
had ever admired him, lapped up all that brag . . . I meant to make him admit
that he was a phony and a coward." Perry had given a knife to Dick, trying to
persuade him to kill Mr. Clutter so that there would be no witnesses. Dick,
however, would not or could not do it. So Perry took the knife and later said,
"But I didn't realize what I had done till I heard the sound. Like somebody
drowning. Screaming under water."
What happened before the murder is very like the sequence in Perry's
*From IN COLD BLOOD by Truman Capote. Copyright © 1965 by Truman Capote. Reprinted by permission of
Random House, Inc. and Hamish Hamilton Ltd.
56 Human Motivation
dream. He is faced with the crisis of taking initiative. He has doubts about his
ability to take initiative, just as in the dream. He tries to get Dick to take the
initiative and fails. What is the way out of this crisis? Unfortunately for Mr.
Clutter, the Situation suggests a possibility very like the escape route Perry has
always followed in his dream: regressing to a Stage II mode of action, which is
associated with ultimate escape to an oral paradise. Mr. Clutter is tightly bound
and his mouth taped shut. This suggests a Stage II crisis (holding on versus let-
ting go) Perry has resolved in the past by suddenly letting go, which he does
this time by cutting Mr. Clutter's throat in a peculiarly liquid way. To his un-
conscious mind (since he did not consciously intend to härm Mr. Clutter) it
seems the only, as well as the habitual, way out of a crisis in order to escape to
the oral paradise or find the sunken treasure, dreams of which had led him to
plan the robbery in the first place. Thus, the dream sequence provides a clue to
explaining what happened in real life, which was so puzzling to Perry himself
and to all those associated with the murder.
Implications of the Dream Analyses

Several conclusions can be drawn from our interpretation of Freud's and Perry's
dream sequences:
1. They vividly illustrate several sources of pleasure or concern that seem to
be quite general, at least in thought patterns. They can be conceived of as gen-
eral goals: (a) to receive or take in food, love, and support; (b) to be autono-
mous and regulate one's own behavior; and (c) to take initiative or be success-
fully assertive. The next, or genital, stage is not illustrated by the dream
sequences or listed in Table 2.4, although Erikson (1963) elaborated on it, along
with later developments in the life cycle. Its goal is mutuality or generativity.
2. These general goals appear to be arranged in a developmental sequence.
Freud arrived at them by tying them to sources of erogenous satisfaction on the
body surface that seem to be the focus of attention at progressively later periods
in the child's life. As Erikson has generalized this idea and as we observed the
sequences in these dreams, progression or regression occurs from one stage to
the next in adults without reference to the parts of the body on which they are
presumably based. Freud thought forward from the oral stage to the stage of au-
tonomy, and Perry thought backward from the stage of initiative to the stage of
autonomy to the stage of oral dependency.
3. Psychological health and psychopathology appear to depend on how suc-
cessfully a person negotiates these stages one after the other. Perry had not suc-
cessfully negotiated the oral crisis involving, in Erikson's (1963) terms, "the mu-
tual regulation with the maternal source of supply." His yearning for his
mother's breast, so vividly portrayed in his dream, may not have been so much
sexual as it was a desire for love and support. However, the disruption was not
so great that he became psychotic. He did progress somewhat through to the
stage of autonomy, but here too he ran into problems, as he did when he made
tentative attempts to achieve the assertive goals of Stage III. He was half crip-
pled at every stage, so to speak, and tended to resolve his problems by regress-
ing back to a goal of an earlier stage. Freud, on the other hand, successfully ne-
gotiated the oral crisis and moved on to the stage of autonomy.
4. These developmental stage goals and how they are negotiated appear to
have very wide effects on the personality—not only in the realm of psychopa-
thology, but also in the realm of a person's whole outlook on life. Thus, Erikson
relates the stages to such virtues as hope, will, and purpose and to such ele-
ments of the social order as a person's philosophy of life, the importance of law
and order, or ideal prototypes (see Table 2.4). And in both Freud's and Perry's
cases, but particularly in Perry's we were able to show that the sequence of
thoughts about these issues is related to what the men actually did in the real
world.
OTHER VIEWS OF DEVELOPMENTAL STAGES

The psychoanalysts have not been the only ones to describe goals characteristic
of different periods in the life of the growing child. Even Jean Piaget's descrip-
tion of cognitive stages in development can be understood in motivational terms.
In discussing how children play a game like marbles, Piaget (1932) identifies
first a stage of conformity, in which the children are busy learning how to play
the game. Their goal at this stage is to play the game correctly according to the
rules, that is, to conform. When they are very young they cannot even compre-
hend that the game could be played differently. This corresponds in Freudian
terms to Stage I—the intake stage—in which children are totally dependent on
others for support (Freud) or information (Piaget). Obviously the two theorists
place the intake stage at difFerent ages—emotional intake in the first year of life
(Freud) and information intake at the age of five or six, when children play
games (Piaget). That is, the stage goals are functionally equivalent and occur
roughly in the same sequence, but not at the same age.
Next, according to Piaget, as children mature cognitively, they begin to real-
ize that the rules of the game can be changed. They no longer are bound by
what others teil them the rules are but begin to invent rules of their own or
challenge the rules. In Freudian terms they are now at Stages II and III, having
become autonomous and assertive in their own right.
Finally, according to Piaget, they become intellectually capable of empathy
—of understanding the viewpoint of another player in the game rather than just
their own viewpoint. Once they are able to assume the role of another, they
want to. They then are able to design or play games that take their own and
other viewpoints into account. In Freudian terms they are in Erikson's fourth
stage, which is characterized by mutuality.
The Freudians used the sequence of pleasure in erogenous zones and rela-
tions to parents to define the developmental stage goals, whereas Piaget used
characteristics of intellectual development to define very similar stage goals.
More recently, other scholars such as Loevinger (1976) and Kohlberg (1969)
also have defined and measured stages of conformity, autonomy, and mutuality
that parallel somewhat the Freud-Erikson stage sequence, although there also
58 Human Motivation
are important differences. Thus, the Freud-Erikson sequence obviously has great
appeal as a model of the hierarchy of goals pursued in conceptions of the ma-
turing process.
In fact, parallels also exist in the world's principal religions, all of which de-
scribe first a stage of people's being good by conforming to Standards set by oth-
ers because they do not know any better. But this does not represent ideal matu-
rity, because people must freely choose to be good on their own. They must
question the rules and become somewhat autonomous and assertive until they
have developed enough to choose the good voluntarily and knowledgeably. This
period of self-development eventually leads to the highest level of maturity, in
which the goal is to "lose the seif," the sense of which has been so laboriously
gained, in the Service of others. The progression is from conformity to self-
realization to altruism. Once again the agreement on developmental stage goals
is considerable, although the basis for arriving at them is quite different.
CHECKING THE VALIDITY OF

THE MOTIVATIONAL STAGE THEORY
How do we know that the mode of analysis used in Erikson's interpretation of
Freud's dream or our interpretation of Perry's dream has yielded valid findings?
How can we be sure Perry's dream really is about his mother and father?
Maybe he just likes diamonds and oranges and dislikes snakes. Academic psy-
chologists have always been extremely suspicious of this way of measuring and
identifying motives, because there does not seem to be any way to find out
whether an interpretation really is correct. If there is no way a hypothetical in-
terpretation can be falsified or proved wrong, it is not a genuine scientific hy-
pothesis. Could not someone eise come along and interpret Perry's dream in
some other way that was equally plausible? How then could we decide between
the two interpretations?
Academic psychologists initially reacted to this dilemma in one of two ways.
They either left this kind of analysis wholly outside the field of scientific psy-
chology, or they attempted to design experiments in the laboratory that would
check some of the hypotheses in a falsifiable way. This section will review some
of these recent experiments, but first let us consider a third alternative, which
argues that the experiments are unnecessary.
Hermeneutics
In recent years scholars have argued that a careful distinction must be made be-
tween understanding the meaning of a text, a dream, or a life, and proving by
independent investigation that the interpretation is really correct. In their view,
understanding comes about through the correct application of hermeneutics, the
science of interpretation, which involves fitting pieces of information together
into a meaningful whole, much as a person would fit pieces into a Jigsaw puzzle
(Radnitzky, 1973; Steele, 1979). In Steele's words, "analysis is complete when

the circle of meaning is constructed, when the puzzle is finished. Meaning re-
sides in the articulation of the parts with each other and the whole. . . . The
puzzle fitting together precisely is vital, for any incompleteness creates doubt.
. . . The original text of the dream—its solution, the latent meaning—exists
solely as a product of interpretation; just as the analysand's life story with
its gaps filled in and its distortions corrected is a product of the analytic
encounter."
Understanding is a product of the analyst's personality and view of the
world as determined by the particular historical period in which he or she lives.
It is also the product of the exchange between the analyst and the text as the
analyst tries out various interpretations and corrects them. This is even more
true of the therapeutic encounter in psychoanalysis, in which understanding is
the product of a dialogue between the analyst and the analysand: "Reality
comes to reside in the ever expanding circle of meaning between analysand and
analyst" (Steele, 1979). Thus, Freud (cited in Steele, 1973) used the method of
hermeneutics in interpreting dreams and life histories, but he was very worried
about its subjectivity: " 'Interpret!' A nasty word! I dislike the sound of it; it
robs me of all certainty. If everything depends on my interpretation, who can
guarantee that I interpret right?"
In Steele's view, therefore, Freud made the mistake of trying to bolster his
interpretations by appeals to "objective" facts of biology and history outside
those obtained by the hermeneutic method. Objective facts, if they He outside
the circle of meaning, are irrelevant. Thus, it does not matter that a patient re-
members a childhood event that turns out never to have occurred. It is the
memory of the event that is relevant to understanding. The soundness or credi-
bility of an interpretation of a dream text or a life depends on the same criteria
used in evaluating the credibility of an interpretation of facts obtained in a series
of experiments.
In each case there are a number of events a theory is constructed to explain.
In both instances, whether the theory is good or not depends on whether the
rules of logic or reasoning are carefully followed. Is the interpretation internally
consistent? Does it maximize understanding the parts in terms of the whole?
Does it account for all the facts—in this case, the images in the dream? Is there
an alternative explanation that will fit all the facts as well? If something is in-
consistent with the interpretation, does the interpreter shift his or her stance and
explain it away?
The analysis of Perry's dream is a crude example of hermeneutics. It at-
tempts to account for a number of images in the early part of the dream se-
quences—the diamonds, the fact that the dream is in a jungle, the oranges, the
snake swallowing him feet first, and so on. However, things are left out. Why
are they oranges and not apples, which could symbolize breasts just as well? Is
it because of his mother's bronzed, yellowish skin as a full-blooded American
Indian? Even more significant, what explains the tree that is "standing alone"?
Or the parrot? Why is it yellow?
Obviously, a much more detailed explication de texte could be made, and it
60 Human Motivation
would be made by someone engaged in a serious application of the science of

hermeneutics. Such a thorough investigation should result in a convincing dem-
onstration that the interpretation given, and only that interpretation, could ac-
count for all the images in the dream sequence.
As has been noted, advocates of hermeneutics argue that the appeal to facts
outside the text to validate an interpretation is unnecessary. Thus, it would not
be proper to go beyond the text to bring in the fact that Perry's mother often
vomited to account for some of the dream imagery. Even more important, we
should not attempt to predict actual behavior from interpretations of a textual
sequence, as we attempted to do in going from Perry's dream to his murder of
Mr. Clutter. To these theorists that would be simply to confuse two different
"texts," so to speak (the dream sequence and the murder sequence), each of
which would have its own interpretation. However, there seems to be no power-
ful reason for leaving out any source of information in attempting to arrive at
correct scientific inferences. Thus, it does seem to lend credibility to the inter-
pretation of Perry's dream sequence to bring in facts about his mother, his fa-
ther, and the murder he committed apparently without a motive.
Experimental Studies of the Psychoanalytic Theory of Motivational Stages

Experimentally oriented psychologists never have been satisfied with hermeneuti-
cal demonstrations of psychosexual or psychosocial stages of motivational devel-
opment as outlined in Table 2.4. They have feit that gifted interpreters could
make sense out of almost any dream sequence and could manage to find, if they
were creative enough, the particular kind of sense they were expecting to find.
Therefore, efforts have been made to check the theory by other, more conven-
tional methods.
However, extensive reviews of these studies by Sears (1943) and Kline
(1981) have not led to convincing proof by these methods that motives exist and
develop as the Freudians have argued. As Silverman (1976) pointed out, most of
the studies have attempted to discover whether certain types of unconscious
wishes are more common among those who have a personality disorder, which,
according to psychoanalytic theory, derives from the conflicts in those uncon-
scious wishes. For example, psychoanalysts have argued that male homosexuals
have a strong incestuous wish to sleep with their mothers, which is accompanied
by castration anxiety (like Perry's fear of having his legs broken). They deal
with this by denial, so they lose all sexual interest in the opposite sex. But it has
been diflficult to show that incest wishes and castration anxiety are more com-
mon in male homosexuals than in male heterosexuals. In fact, psychoanalytic
theory does not unequivocally State that there should be a difference. Rather, it
argues that while incest wishes and castration anxiety are universal, they have a
peculiarly strong motivating impact on male homosexuals.
To return to Erikson's analysis of psychosocial stages, everyone goes
through an oral period, but how a person negotiates the period affects how
strong the motives deriving from that period will continue to be throughout the
individual's life-span. Freud dreamed about oral wishes and so did Perry, but we
must conclude that the oral wishes had a much stronger motivational impact on
Perry than they did on Freud, who dreamed about leaving them behind rather
than returning to them.
The most systematic attempt to put Freudian motivational theory to the ex-
perimental test has been made by Silverman (1976). He has designed a proce-
dure that makes it theoretically possible to check the specificity of the link be-
tween conflicts characteristic of a motivational stage and adult personality
disorders associated in psychoanalytic theory with those conflicts. His method
presents visual Stimuli so rapidly that they cannot be recognized, and it shows
that the Stimuli that evoke the stage conflicts supposedly characteristic of differ-
ent disorders affect only people with those disorders, and not others. The
method permits so much greater precision in the conclusions drawn than other
approaches that it is worth reviewing in some detail. Some examples will show
how it works in practice.
Psychoanalysts have argued that disturbances in the oral stage of motive de-
velopment can lead to serious distortions of reality, and particularly to schizo-
phrenia, as we pointed out in discussing Table 2.4. More specifically, schizophre-
nia often develops out of the second part of the oral stage, in which the infant
learns to take "pleasure in biting on hard things, and biting through things and
in biting pieces off things" (Erikson, 1963). This is the beginning of the child's
understanding of the difference between good (being close to Mommy) and evil
(losing Mommy as she withdraws in anger from the biting). As Erikson (1963)
puts it so strikingly,
Our clinical work indicates that this point in the individual's early history can be the
origin of an evil dividedness, where anger against the gnawing teeth, and anger
against the withdrawing mother, and anger with one's own impotent anger all lead
to a forceful experience of sadistic and masochistic confusion leaving the general im-
pression that once upon a time one destroyed one's unity with a maternal matrix.
This earliest catastrophe in the individual's relation to himself and to the world is
probably the ontogenetic contribution to the Biblical saga of paradise, where the first
people on earth forfeited forever the right to pluck without effort what had been put
at their disposal; they bit into the forbidden apple, and made God angry. (1963)
So Silverman argued that the oral aggressive wish should be particularly dis-
turbing to schizophrenics. He aroused the wish unconsciously by presenting oral
aggressive Stimuli in a tachistoscope so fast (for four milliseconds) that subjects
could not consciously report what they saw. He knew from prior research that
Stimuli presented for this exposure time in fact did register and have effects,
even though they could not be consciously reported.
Silverman (1976) measured the effect of such an unconsciously arousing
Stimulus by coding the thoughts of the subjects after its exposure for pathology
("illogical, loose, or unrealistic thinking," called primary process thinking). One
of the oral aggressive Stimuli he used was the phrase Cannibal eats person, a
Stage I Stimulus (see Table 2.4). He found that thought pathology, or primary
process thinking, was greater among schizophrenics after being exposed to this
62 Human Motivation
Stimulus than after being exposed to control Stimuli such as the phrase people
are walking or a picture of a man reading a newspaper. Furthermore, the differ-
ence in the pathology of thought did not occur if the cannibal eats person Stimu-
lus was presented slowly enough so that the schizophrenics could recognize it.
In other words, the wish had to be unconsciously aroused to disturb thought, as
psychoanalysts would argue.
Perhaps it is not oral aggressive wishes that disturb schizophrenics, but
rather any aggressive wish. To find out, Silverman tested another group of
schizophrenics, using the arousing Stimulus phrase murderer stabs victim, a
Stage III (intrusive) Stimulus. It did not increase primary process thinking in
schizophrenics.
As was noted, the oral aggressive wish provokes anxiety precisely because it
is associated with the mother's withdrawal, and so Silverman also presented sub-
liminally the Stimulus phrase / am losing Mommy. This increased thought pa-
thology in schizophrenics, as it should have if the oral aggressive wish were as-
sociated in the schizophrenic's mind with the mother's withdrawal. The phrase
destroy Mother with a picture of someone about to Stab an elderly woman had
the same effect. Furthermore, the opposite phrase Mommy and I are one re-
duced pathological thinking in schizophrenics, presumably because it reduced
the anxiety connected with the basic oral aggressive wish by suggesting an un-
ending source of oral supplies, such as Perry dreamed of.
Silverman demonstrated the specificity of the motivational issue involved by
showing that the Stimulus phrase Daddy and I are one did not reduce primary
process thinking in schizophrenics. The advantage of this mode of studying the
motivational complexes involved is that it permits testing and falsifying all sorts
of alternative formulations of the motivational stage conflicts involved.
Table 2.5 summarizes the findings for thought disorder in schizophrenics,
who are the victims of Stage I conflicts, and compares the effects of similar and
different Stimuli on the reactions of people who have negotiated in an unusual
way Stage II (stutterers) or Stage III (homosexuals). The table shows how
Silverman can carefully check the specificity of the hypotheses in regard to the
relationships involved by varying the Stimulus inputs and determining their ef-
fects on people with different types of problems.
The extent to which alternative hypotheses to explain the results can and
have been checked is much greater than in other research of this type. For com-
parison, consider the results obtained with stutterers, who, according to psycho-
analytic theory, have a disorder associated with not negotiating the second, or
anal, psychosexual stage successfully. That is, their key conflict represents
"wishes and inhibitions relating to the expelling and retaining of feces" (Silver-
man, Klinger, Lustbader, Farrell, & Martin, 1972; Silverman, Bronstein, &
Mendelsohn, 1976). They stutter precisely because they are in conflict as to
whether to let go (that is, speak) or hold on (that is, not speak). Therefore,
Silverman aroused motivational conflicts in various ways associated with Stage I
(for example, by a picture of a lion roaring), with Stage II by a picture of a dog
defecating or the phrase go shit with an accompanying picture, and with Stage
III by the phrase fuck Mommy with a picture suggesting sexual intercourse. He
examined stuttering behavior after very rapid exposure to these Stimuli and
Table 2.5.
EFFECTS OF AROUSING MOTIVES CHARACTERISTIC OF STAGE CONFLICTS IN
INDIVIDUALS WITH DISORDERS ATTRIBUTED TO THOSE STAGES (after Silverman, 1976;
Silverman, Bronstein, & Mendelsohn, 1976; Silverman, Klinger, Lustbader, Farrell, & Martin, 1972;
Silverman, Kwawer, Wolitsky, & Coran, 1973)
Type of Disorder and Measure Used

Schizophrenia, Stage I Stuttering, Homosexuality,
(Oral Conflicts): Stage II Stage III
Examples of Stimuli Thought Disorder (Anal Conflicts): (Phallic Conflicts):
Rapidly Presented^ Reaction Stuttering Reaction Sexual Feelings Reaction
Stage I: Oral
Cannibal eats person or
picture of Hon roaring
Unrecognized Increased Weak increase
Recognized3 No change No change
/ am losing Mommy Increased
Destroy Motherh Increased
Mommy and I are one Decreased No change
Recognizeda No change
Daddy and I are one No change
Stage II: Anal
Go shitb or picture of Increased
dog defecating
Stage III: Phallic
Murderer stabs victim No change
Fuck Mommyh No change No change Increased homosexual
and decreased
heterosexual feelings (no
effect on heterosexuals)
Control Stimuli (for No change No change No change
example, People are
walking and pictures of
a man reading)
a
Stimulus was presented in a tachistoscope too rapidly (for four milliseconds) to be recognized, except where indicated.
b
Phrase was accompanied by picture showing the action; both were subliminal.
other neutral Stimuli. As Table 2.5 makes clear, the Stimuli suggesting Stage II
conflicts increased stuttering most significantly in stutterers.
To return to what motivates male homosexuals, who negotiate the third, or
phallic, stage of motive development in an unusual way, Silverman, Kwawer,
Wolitzky, and Coran (1973) presented to them the Stimulus fuck Mommy subli-
minally, sometimes with an additional suggestive picture to increase the anxiety
64 Human Motivation
they supposedly feel over their particularly strong, guilt-ridden incest wishes.
The investigators found the Stimulus increased homosexual and decreased het-
erosexual feelings among male homosexuals as reflected in self-ratings, but it
had no effect on heterosexuals. Thus, Silverman (1976) concludes he was able to
demonstrate "that this kind of sexual orientation involves (in part) a flight from
incest." That is, among male homosexuals the anxiety about incest with the
mother has generalized to all women, and the sexual urge is diverted to men.
However, the same phrase had no efFect on thought pathology among schizo-
phrenics or on stuttering in stutterers, indicating that the incest conflict was par-
ticularly crucial for male homosexuals.
So far Silverman's experimental approach has been applied only to the links
between the unconscious motive conflicts that presumably are connected with
certain types of psychopathology according to psychoanalytic theory. This ap-
proach has made considerable progress in demonstrating the specificity of some
of these links, although some still are skeptical about Silverman's results and
claim they cannot be replicated (Allen & Condon, 1982; Heilbrun, 1982).
The psychopathological implications of motivational stage theory represent
only one of its aspects. What also needs to be studied using a technique like
Silverman's is whether there is a hierarchy of stages, as the theory claims. That
is, if there is blocking or frustration at a higher stage, will there be a tendency
to regress to the next lower stage in thought among those with behavior disor-
ders like Perry, and not among normal individuals? If a motivational conflict
representing Stage II, for example, is presented to normal people, will they be
more likely to think themselves through to Stage III, as Freud did, demonstrat-
ing their greater maturity? Furthermore, we need to know much more about
characteristics of normal individuals that are not so extreme as to be psycho-
pathological, but that might represent only mildly unresolved motivational con-
flicts representative of a particular stage. Above all, we must have some way of
measuring the extent to which individuals possess motivational orientations
characteristic of various motivational stages. A technique for doing this has been
invented by Stewart (1973) and will be discussed in Chapter 8.
• CONTRIBUTIONS OF THE PERSONALITY TRADITION

Most people tend to explain everything others do in terms of motives invented
on the spot. Psychologists studying personality have progressed well beyond this
"naming fallacy" by identifying a limited list of key human motives and gener-
ally distinguishing them from other personal characteristics such as traits and
abilities. Above all they have provided a vocabulary of adult human motives in
terms of which to describe individual lives. Those who were required by their
occupation as therapists to make sense out of disordered lives have made the
greatest contribution to describing the motives and motive conflicts that account
for what people do or fail to do. They have not agreed entirely on what the key
motives are, but there has been a widespread consensus that these motives often
are unconscious and not known to the individual, and that any final short list of
human motives certainly would have to include needs for aggression or power,
for love (or sexuality in the broadest sense), for relief from anxiety and insecuri-
ty, and for some kind of mastery or self-actualization.
The major limitation of personality theory has been its inability to measure
the motives in question, which has made it difficult (if not impossible) to test
rival theories and clarify relationships among motives and presumably connected
behaviors or Symptoms. Both Murray and Cattell launched major research pro-
grams designed to remedy this shortcoming. They invented many ways of mea-
suring a wide variety of motives and then tried to integrate the information col-
lected about a person's motives through either a diagnostic Council of judges or
factor analysis. Neither integrative approach succeeded completely in providing
definitions of motives that were conceptually clear in the measurement sense.
However, the efforts of Murray and Cattell created a much better understanding
of the measurement problems involved and made further advances possible.
The most successful method of assessing motives in clinical settings has in-
volved the analysis of dreams, free associations, or fantasies. It is believed this is
true because certain key motives, such as sexual and aggressive urges, are re-
jected as bad; continue to function out of consciousness; and show up only in
dreams and fantasies, which are not as subject to censorship as conscious
thoughts. As Chapter 6 will discuss, motives are also best measured in dreams
because other determinants of behavior are less influential in shaping fantasies,
so they give clearer indications of motivational—and only motivational—reasons
for what a person is doing. Whatever the reason, the analysis of fantasy has pro-
vided the most convincing evidence of the existence of key human motives and
the best theory currently available about the stages in their development, as out-
lined by Freud and expanded by Erikson.
Efforts to check the validity of the information on motives obtained through
analysis of fantasy had been made with limited success until recent investiga-
tions by Silverman improved the methods employed. Whatever the outcome of
these validity checks, it seems likely, in view of the experience of personality
theorists, that the best way to measure human motives will involve some use of
fantasy material.
NOTES AND QUERIES

1. To arrive at a short list of the most important motives, most theorists
(McDougall, Freud, Sheldon, and Maslow) try to link certain behavioral
trends to inherited or innate biological characteristics. Is there any other
way to determine what the most powerful and pervasive human motives are
likely to be?
2. Motives are defined by most theorists in terms of behavioral trends.
McDougall speaks of persons' domineering, avoiding, or seeking the Com-
pany of their fellows; Murray, of influencing or submitting passively to oth-
ers, or drawing near to others; and Cattell, of acting gregariously. However,
in Chapter 1 we repeatedly pointed to the dangers of trying to infer motives
66 Human Motivation
from behavioral trends, since other factors play a role in determining ac-
tions. Is there any alternative way to define motives? Could they be defined
in terms of end states—that is, states of being—rather than in terms of the
means of getting to the end states? Try to make a list of end states.
3. Murray's definitions of needs contain many elements. How could we be sure
all the elements listed belong under a given need? For example, "to excel
oneself " and "to rival and surpass others" are part of the definition of the
need for Achievement. How can we teil whether these characteristics go
with another need for Achievement characteristic like "to accomplish some-
thing difficult" rather than, for example, with the need for Exhibition, de-
fined in part as "to be seen and heard"?
4. The following is a list of the attitude items that load high on one of the mo-
tivational factors ("Ergs") Cattell (1957) has identified by statistical inter-
correlation of a large number of such items. Try explaining as consistently
as you can why each item would indicate a need or wish for something. Try
to formulate a definition of the motive in question that would include all
these items.
Loading
I like to take an active part in sports and activities. .5
I would rather spend free time with people than by myself. .4
I have no wish to disagree with authorities. .3
I do not enjoy hunting and fishing. .3
What are some of the difrlculties in trying to infer motives from attitudes?
5. Academic psychologists have been very critical of the kinds of psychoana-
lytic interpretations of dream sequences discussed in this and the preceding
chapter. For example, Eysenck (1965) concludes, "As has often been
pointed out, the complexities of psychoanalytic reasoning effectively pre-
clude any scientific testing of these theories." Do you agree? What reasons
are there for agreeing or disagreeing with Eysenck?
6. Research has demonstrated that there is little relationship between people's
attitudes toward authority figures and the way thev write about authority
figures in imaginative stories produced for the TAT (Burwen & Campbell,
1957). Psychologists have concluded (see Mischel, 1968) either that TAT
stories give no information of value about people or that people are just too
inconsistent to characterize. What other inference might be drawn from this
fact? Think, for example, of the way Perry thought of his father in real life
and in his dream.
7. Do you think Silverman's work demonstrates that adult psychopathology is
derived from early traumas in psychosexual development? What other expla-
nation can you give for the fact that a phrase such as go shit—uniquely, out
of all the emotional phrases used—when presented too rapidly to be recog-
nized, increases stuttering in stutterers? How would you determine if child-
hood traumas were actually involved? Perry was exposed to some rather se-
vere toilet training. Why did he not stutter?
8. A comparison of Freud's and Perry's dream sequences suggests that people's

thinking themselves forward or backward through psychosexual stages
might indicate greater or less maturity. Can you think of a way in which a
measure of a person's maturity might be obtained in this way? You might
want to consult Stewart's method of measuring the Freud-Erikson stages, as
described in Chapter 8.
9. In commenting on the interpretation of dream sequences like Perry's,
Eysenck (1957b) remarks that "while it may be interesting at times, it has
not produced a single fact which could be regarded as having scientific va-
lidity. Everything is surmise, conjecture, and interpretation; judgements are
made in terms of what is reasonable and fitting. This is not the method of
science . . . [in which] you State a definite hypothesis, make certain deduc-
tions from the hypothesis, and then proceed to carry out experiments to
prove or disprove your theory. This is the scientific method, and that is pre-
cisely what is missing in all the work." What is the difference between the
scientific and the hermeneutic method? Is there a difference between check-
ing a hypothesis against a fact (for example, a dream image) and against the
result of an experiment?
3
Motivation in the
Behaviorist Tradition
Thorndike's Studies of Animal Motivation

Drives as a Means of Insuring Survival
HulPs Model of How Drives Facilitate Adaptation or Learning

Drive as a Strong Stimulus
Does Animal Learning Occur in the Absence of Drive?
Cue Characteristics
Response Characteristics
Reward Characteristics
The Incentive Variable
Excitatory Potential in Behavior Theory and the Meaning of the Term
Motivation
Excitatory Potential for Operants More Indicative of Motivation
The Behaviorist Model of Motivation Applied to Humans by Spence and

Others
How a Strong Drive Can Interfere with Performance on Complex Tasks
Interference Effect Due to More Experiences of Failure
How the Presence of Others May Increase Drive and Interfere with Performance on
Complex Tasks
1
Reinterpreting the Behaviorist Studies of Human Motivation in Terms of
What Goes On in People's Minds
> Comparison of the Psychoanalytic and Behaviorist Contributions to the Study
of Motivation
Similarities
Differences
> Limitations of the Behaviorist Model
68
• THORNDIKE'S STUDIES OF ANIMAL MOTIVATION
At the very moment Freud was discovering the motives behind the dreams of
his patients in Vienna, psychologists in the United States were pursuing a radi-
cally different approach to understanding motivation. They feit that reports of
inner states of mind were unreliable and therefore could never form the basis of
an objective science of psychology patterned after the natural sciences. The
idea is nicely expressed in a recent physiology textbook (Vander, Sherman, &
Luciano, 1975): "Conscious experiences are difficult to investigate because they
can be known only by verbal report. Such studies lack objectivity . . . in an at-
tempt to bypass these difficulties scientists have studied the behavioral correlates
of mental phenomena in other animals."
It was this line of reasoning that led the U.S. psychologist Edward L.
Thorndike to begin studies of motivation and learning in kittens, dogs, and
chickens in the 1890s. He placed animals in boxes made out of orange crate
slats with a door that opened when a string was pulled or a button inside was
turned from the vertical to the horizontal position. Thorndike (1899) summa-
rized his observations as follows:
A kitten, three to six months old, if put in this box when hungry, a bit of fish being
left outside, reacts as follows: it tries to squeeze through between the bars, claws at
the bars, and at loose things in and out of the box, reaches its paws out between the
bars, and bites at its confining walls. Some one of these promiscuous clawings,
squeezings, and bitings turns round the wooden button and the kitten gains freedom
and food. By repeating the experience, the animal gradually comes to omit all the
useless clawings, etc., and to manifest only the particular impulse (e.g., to claw hard
at the top of the button with the paw, or to push against one side of it with the
nose) which has resulted successfully. It turns the button round without delay when-
ever put in the box.
There need be no such congruity between act and result. If we confine a cat and
open the door and let it out to get food only when it scratches itself, we shall after
enough trials find the cat scratching itself the minute it is put in the box.
Several aspects of this procedure commended themselves to psychologists

interested in building a science of behavior. It was objective: anyone could ob-
serve what the cat was doing, and in time its actions even could be recorded
mechanically. It was potentially quantitative: the number of hours the animal
had gone without food could be recorded, as could the number of minutes the
animal took to get out of the box. It was experimental: the number of hours the
animal went without food could be varied, as could the size of the food reward
or the nature of the response required to get out of the box. Above all, this pro-
cedure led to results that could be interpreted without recourse to such seem-
ingly vague concepts as purpose and reason, which tended to be invoked when
human beings were involved.
At this time in the United States, psychologists were beginning to define
psychology as the study of such behavior, that is, of concrete observable and
recordable acts rather than inner wishes, thoughts, and expectations, which they
69
70 Human Motivation
did not know how to measure objectively. Thus, it seemed perfectly natural to
study animals, which could show the same type of acts as humans could but
under more controlled conditions. Psychologists further assumed that whatever
principles were found to govern the behavior of lower animals would apply to
humans.
Thorndike was interested in the effect of hunger on the activities of animals
in the box. He used the term impulse to describe what hunger led the animals
to do in preference to terms like motive or desire, which he feit implied the sub-
jective or conscious experiences he was trying to ignore for the moment. This
practice has been followed by psychologists in the behaviorist tradition ever
since, except they later substituted the word drive for impulse.
Thorndike (1899) observed first of all that a hungry animal was much more
active than one that was not hungry. In his words, a hungry cat "will claw and
bite and squeeze incessantly." He stated that the starting point for the kitten
was "discomfort from confinement or lack of food" and that this discomfort was
relieved when the kitten got out of the box.
Whereas Thorndike correctly reported that "discomfort from confinement"
was more characteristic of cats than other animals put in the box, this type of
drive was not investigated further by the early behaviorists. They focused on the
hunger drive, because it produced similar effects across species; they wanted to
establish the most general laws they could, ignoring species' differences, particu-
larly because they might raise the fundamental issue of whether the human spe-
cies might differ in important ways from the animal species being studied.
The fact that getting to the food reward led the animal to make the correct
response to get out of the box more and more quickly was formulated by
Thorndike (1911) into what he called the Law of Effect, which he stated as
follows:
Of several responses made to the same Situation, those which are accompanied or
closely followed by satisfaction to the animal will, other things being equal, be more
firmly connected with the Situation, so that, when it recurs, they will be more likely
to recur; those which are accompanied or closely followed by discomfort to the ani-
mal will, other things being equal, have their connections with that Situation weak-
ened, so that, when it recurs, they will be less likely to occur. The greater the satis-
faction or discomfort, the greater the strengthening or weakening of the bond.
Thorndike is particularly careful to avoid referring to such subjective states

as "feelings of pleasure" or attributing intelligence to the animal or the idea that
a response will lead to satisfaction. Thus, he defines a satisfying State of affairs
as "one which the animal does nothing to avoid, often doing such things as at-
tain and preserve it. By a discomforting or annoying State of affairs is meant one
which the animal commonly avoids and abandons" (Thorndike, 1911). Infer-
ences about satisfaction or dissatisfaction are made according to the actions the
animal takes. Furthermore, the automatic nature of the strengthening of the re-
sponse is made clear by the example of the kitten learning to make a response
like scratching itself, which does not in the real world lead to opening the door.
Motivation in the Behaviorist Tradition 71
This emphasizes the fact that the animal's learning the response occurs automat-
ically, without its necessarily understanding a connection between the response
and the reward.
So far Thorndike has identified two functions of drive: it energizes behavior
and selects out certain responses that lead to reward; that is, it causes learning.
Another function of drive is also easily observable. The animals in the box typi-
cally spend more time on the side on which the food is placed. Their biting,
squeezings, and clawings are not purely random. They are oriented toward the
goal, or food reward. Thus, the early behaviorists observed that drive has three
important functions: it energizes, Orients, and selects behavior.
No one has seriously questioned these functions (see Melton, 1941), al-
though they have inspired much discussion about the terms to be used to de-
scribe what Thorndike and the early behaviorists discovered. Some have been so
impressed by the similarity between animal drives and human motives that they
use the terms more or less interchangeably, as we shall in this book. Others
have elaborated further what Thorndike refers to as a satisfying State of affairs.
It is described as rewarding or reinforcing from the objective point of view,
since it rewards or reinforces the response that leads to it. From the subjective
point of view, it is commonly called an incentive, since the anticipation of it
leads to making the appropriate response.
Later work stressed that the directive, or orienting, function of drive shows
up as much in increased sensory or perceptual sensitivity to certain Stimuli as it
does in a more focused set of activities. Some scholars, however, feit that this
directive function of drive should be attributed to its cue characteristics, or to
the associations it evokes, rather than to drive itself (see Farber, 1954).
Finally, B. F. Skinner (1938) and his followers believed it was not at all
necessary to talk about drives in the organism, since the learning to get out of
the box that Thorndike observed could be explained solely in terms of the re-
wards or reinforcers provided the animal under various conditions. As we take
up these matters, however, remember that Thorndike's observations are as true
today as when he first made them: drives or motives serve to energize, to Orient,
and to select behavior.
Drives as a Means of Insuring Survival

In the functional behaviorist tradition, drives were closely linked to biological
needs, because they led the organism to learn to do things it needed to do to
survive. Obviously, an organism needs such things as food, air, and water to
live. Thus, it is adaptive if the lack of these substances causes disturbances in
the body that lead to activities that will remove the disturbances. Drives were
thought to be like thermostats that turn on the furnace when the temperature
gets too low. Early studies seemed to show this was the way the hunger drive,
in particular, functions. For example, Richter (1927) demonstrated that when
rats were hungry they turned an activity wheel many more times than when
they were satiated. Furthermore, if they were allowed to eat whenever they
wanted to, they were much more apt to eat when they were most active. Hun-
72 Human Motivation
ger was associated with activity, which was associated with eating, which de-
creased the activity.
Warden (1931) and a group at Columbia University designed an obstruction
box to measure the strength of various biologically based drives. The animal was
placed in a starting box and had to cross an electrified grid to get to what was
needed. The strength of the drive was measured by the number of crossings an
animal would make when the grid was electrified at a certain level of shock.
The researchers plotted the number of crossings against days of food or water
deprivation. They found that maximum strength of drive, as measured by the
number of crossings, occurred after one or two days of food and water depriva-
tion and then tended to fall off as the animal got weaker. See illustrative data
for the white rat in Figure 3.1. Furthermore, as the figure shows, the strength of
the rats' drive to get food and water was somewhat higher than the drive of a
male rat to cross the grid to get to a female rat in heat. The strongest drive of
all was for a mother rat to cross to get to her young. It was reasoned that the
sexual and maternal drives obviously were important, because without them the
species would not have survived.
Earlier it had been discovered that stomach contractions actually increased
as people went without food for some hours (Cannon & Washburn, 1912). Air-
filled balloons that subjects swallowed clearly showed evidence of such contrac-
tions. This finding was historically very important; it suggested that biological
• Maternal Drive
20 -
15 - /
o
U
-• Sex Drive
3.
Ö
10
\ \
Thirst Drive Hunger Drive
6
3
1 1 1 1 1 1 1 1 1
0 2 4 6 8
Deprivation (Days)
Figure 3.1.
Effect of Increasing Drive Strength on Willingness of White Rats to Endure Pain to Get to Various
Goal Objects (after Zimbardo, 1979, after Warden, 1931).
needs might create strong sensations inside the body, which could be the basis
for the discomforts that activities were designed to remove. Later research
showed that stomach contractions are not the main factor governing the ten-
dency to eat. Blood sugar level is much more important (see Mayer, 1955;
Thompson & Campbell, 1977). If it falls below a certain level people feel hungry
and eat more; if it rises above a certain level they feel füll and eat less. It is dif-
ficult to imagine what sensory end organ would be strongly stimulated by a de-
creased blood sugar level, but this discovery had not been made at the time the
behaviorist model of drive was constructed. Thus, many behaviorists concluded
that drive could be conceived as strong sensations of the sort derived from the
hunger cramps that developed from going without food.
HULL'S MODEL OF HOW DRIVES FACILITATE

ADAPTATION OR LEARNING
Most psychologists, following Thorndike's lead, devoted their attention to trying

to understand how drives facilitated learning ultimately in the interest of surviv-
al. Like Thorndike, they wanted to describe the process in a completely objec-
tive way without recourse to terms suggesting subjective states of mind, such as
feelings, hopes, fears, and purposes. Edward Tolman (1932) used such terms but
was careful to objectify their meaning. For example, he feit purpose could be de-
fined simply by observing behavior in an animal that persisted until some end
State was reached.
Clark Hüll (1943) tried to go even further and delete such terms as purpose
altogether from a scientific vocabulary by demonstrating that what was called
purpose could be explained in purely mechanical stimulus-response terms. He
thought it was best for psychologists to think of an organism—whether human
or animal—as a machine, and he conceived of himself as an engineer trying to
create "a truly self-maintaining robot" (Hüll, 1943). In this way Hüll could
avoid using vague mental states of mind to resolve difficulties in explaining be-
havior, for "the temptation to introduce an entelechy, soul, spirit, or demon into
a robot is slight." Therefore, as a model for the organism, he created in his
fancy a motorized machine that ran about on little wheels. It contained an oil
tank and a proboscis—something like a flexible gas pump hose or an elephant's
trunk—that could move about freely. He equipped the robot with a liquid level
indicator that, when the oil fuel supply got below a certain point, would activate
the machine, and particularly the proboscis, so both would move about much
more vigorously.
Hüll imagined the robot was moving about in a landscape where there were
little pools of oil in various places. He reasoned that by chance (not by fore-
thought!), in waving about, the proboscis might come in contact with one of
these pools of oil. When it did so, because of a mechanism built into it, it would
automatically start sucking up the oil into the tank until the liquid level indica-
tor rose and turned off the mechanism causing the increased activity. Hüll then
equipped his robot with a memory device that after this experience would lead it
74 Human Motivation
more quickly to a pool of oil the next time its fuel supply ran low. This device
corresponded to Thorndike's Law of Effect. It simply insured that whatever re-
sponse the robot was making when it sucked up oil was more likely to occur the
next time it ran low on fuel.
What was important to Hüll was that no one could conceive of a robot say-
ing to itself, "Aha, I see when I am running out of fuel I'd better look around
for some pools of oil so that I can replenish my supply." The whole process was
conceived in purely objective, mechanical terms. Hüll wanted psychology to be a
natural science, like nineteenth-century physics. Thus, his task as a psychologist
was to identify the key variables in psychology analogous to such variables as
mass, velocity, and time in the physics of his day and to State the empirical rela-
tionships among these variables in quantitative form, just as the laws of physics
were stated.
Hull's approach appealed greatly to many psychologists in the 1930s and
1940s. It was objective. It was quantitative. It promised to develop a natural sci-
ence like the other natural sciences that had so successfuUy improved the under-
standing of the physical world. Above all, it appeared to explain some very
complex mental phenomena in terms of a few simple concepts and their pre-
cisely stated relationships. The task of psychology was reduced to manageable
size: it consisted of working out the basic principles of behavior and then apply-
ing them to explain all sorts of complex phenomena, from psychotherapy to
human aggression and war. It is not feasible, nor is it any longer appropriate, to
describe Hull's principles and their application in any detail. It is important,
however, to understand how his model of motivation worked, for it has influ-
enced thinking about the problem ever since.
Figure 3.2 presents a simplified version of the model as it was developed at
Yale University by Hüll and others associated with him, including Miller and
Dollard (1941), Mowrer (1950), and Spence (1956). It represents in mechanical
or objective terms what goes on in a motivational sequence, such as when a
woman walks down the street, feels hungry, sees a restaurant, enters, and eats.
As she walks along she is exposed to certain cues, or Stimuli, in the environ-
ment, such as the sights of various storefronts and other pedestrians, or the
sight and sound of automobiles in the street. These cues are represented by the
Symbols Su S2, and so on. Now suppose her "hunger stat," analogous to a ther-
mostat, is tripped off. This provides a new source of Stimulation corresponding
perhaps to hunger pangs, which are internal and represented by the symbol SD,
for drive Stimulus. It is also represented by an increase in the line at the bottom
of Figure 3.2 representing an increased drive tension. The woman also is making
responses (represented as Rl9 R2, and so on) as she proceeds down the street:
she is walking, looking at storefronts, and so on.
Suppose she sees a restaurant (S2) at the moment her drive increases. There
now is a new Stimulus complex (S2 + SD); in the past it was associated with re-
sponses such as entering a restaurant and eating that have led to drive reduc-
tion, which has strengthened these responses. Therefore, on this occasion the
S2 + SD complex evokes entering the restaurant (R}), which in turn leads to
eating (defined as R„f or the goal response, since it leads to a reduction in the
Cues (Stimuli, or S)
Environmental: S
Internal (Drive):
Responses (R)
(Walking, Entering Restaurant, Eating)
Drive (Hunger)
I
Increase Decrease
li
Sj} Drive Stimulus ( h u n e e r ) .
Rt Cioal responsc.
Figure 3.2.
How Cues, Responses, Drive, and Drive Reduction (Reward) Combine to Produce Behavior
Motivated by Hunger.
drive). Drive reduction automatically reinforces the connections between the

Stimuli and responses in this sequence, so that the next time the cue complex of
hunger plus sight of restaurant oecurs, the response of entering and eating will
oeeur more promptly and efficiently. Note that the complex behavior of a
human being has been explained without recourse to any concepts other than
those used to explain the behavior of Hull's self-maintaining robot.
Furthermore, the model can explain such a purely mentalistic coneept as
purpose in mechanical terms. Note that the drive Stimulus (SQ) is associated
with the goal response (Rg), which is regularly followed by drive reduction.
Therefore, the bond between SQ and Rg is strengthened; every time SD oecurs it
should tend to evoke Rg. However, it cannot elicit the füll response of eating,
since no food is present. Therefore, it elicits a fractional antieipatory re-
sponse—namely, salivating (rg)—rather than Rg, the füll goal response. This no-
tion came from Ivan Pavlov's conditioning experiments with dogs, in which he
found that a bell associated with feeding a little later would, in time, evoke sali-
vation before the food actually was presented.
Furthermore, since every response produces some sensations, the antieipa-
tory salivary response (rg) also produces its own Stimulation, sg, that is, rg -• 5g.
This third type of Stimulation combines with S2 and with SD (sight of the res-
taurant) in the model in Figure 3.2 to make up the total Stimulus complex that
leads to the responses of entering the restaurant and eating. It is crucial here to
76 Human Motivation
observe that the hunger drive (SD) evokes a fractional anticipatory goal response
with its attendant Stimulation (sg) that moves forward from the end of the se-
quence to the beginning and becomes a determinant of what follows. In com-
monsense terms, a man thinks about eating and the relief of his hunger pangs
that goes with eating. Therefore, he enters a restaurant to eat; that is his pur-
pose. In the behaviorist model, however, there is no need to invoke a mentalistic
concept like purpose: it is represented by the fractional anticipatory goal re-
sponse, which is set up automatically from previous experiences with eating and
which also automatically moves forward in the stimulus-response sequence to
the beginning whenever SD occurs.
Drive as a Strong Stimulus

Miller and Dollard (1941) most explicitly generalized the model of drive built
on hunger pangs. They State the following:
A drive is a strong Stimulus which impels action. Any Stimulus can become a drive
if it is made strong enough. The stronger the Stimulus, the more drive function it
possesses. The faint murmur of distant music has but little primary drive function;
the infernal blare of the neighbor's radio has considerably more. While any Stimulus
may become strong enough to act as a drive, certain Special classes of Stimuli seem
to be the primary basis for the greater proportion of motivation. These might be
called the primary or innate drives. One of these is pain. Pain can reach stabbing
heights of greater strength than thirst, the pangs of extreme hunger, and the sore
weight of fatigue which are other examples of powerful innate drives. The bitter
sting of cold and the insistent goading of sex are further examples. (Miller & Dol-
lard, 1941)
Note that they say that a person in all such instances seeks relief from Stim-
ulation or discomfort. There is no such thing as an interest in pleasure per se.
Even what looks like pleasure seeking really is motivated by the desire for ten-
sion reduction. For example, people ride roller coasters to get what they call
"pleasure," but they simply are inducing tension to get more satisfaction from
reducing it. Even the pleasure from something like tickling derives not from the
sensations as they are induced, but from their dying away or fading out after-
ward. Such interpretations may seem farfetched in light of many human experi-
ences, but they were essential to a theory that maintained drives could only de-
rive from strong Stimulation.
Drives also can be acquired. The most important acquired drive in the Mil-
ler and Dollard model is fear or anxiety. Pain is a strong drive Stimulus that
leads to many activities, some of which are associated with escaping the pain. In
time, therefore, any cues associated with the pain get connected with the sensa-
tions and responses associated with it, so that in time the cues have the ability
to elicit what accompanies the pain before the pain actually occurs. Since the ef-
fects of the pain are strong, what the cue elicits serves as a new type of strong
Stimulus or a secondary, or learned, drive, commonly called anxiety.
An experiment reported by Miller (1948) illustrates the way this happens.

Miller placed rats in a white compartment, the floor of which was a grid that
could be electrified. The rats were given a series of brief electric shocks every
five seconds in the compartment; a door then was opened, which allowed them
to escape into another compartment painted black. After the rats had been
shocked a few times in the white compartment, most quickly learned to escape
from it into the black compartment as soon as they were placed in the appara-
tus. Notice, however, that the white compartment had no drive Stimulus value
in the beginning. It acquired drive value because it was associated with the pain
of electric shock. The evidence for this fact was that after experiencing shock,
the rats would run out of the white compartment even when they were not
shocked. The only way this could be explained was to assume they were escap-
ing an acquired strong Stimulus that derived from the learned anticipation of
strong shock sensations.
Next Miller put a barrier between the white and the black compartments
that could be let down only if the rats managed to turn a wheel that dropped
the door between the two compartments. Would the fear of the white compart-
ment—the secondary, or acquired, drive—be sufficient to motivate the rats to
learn a new habit, namely, wheel turning to escape the fear? Figure 3.3 shows
the results. Note that the animals were learning to turn the wheel more and
more quickly as they got out of the white box that way in successive trials.
Since they no longer were getting shocked, the only way this result could be ex-
plained in terms of the model is that there was a new acquired drive—namely,
fear or anxiety—responsible for the learning.
2 4 6 8 10 12 14 16
Trials with Wheel Functioning to Open Door
Figure 3.3.
Rats' Learning the First New Habit, Turning the Wheel, During Trials Without Primary Drive.
This figure shows the progressive increase in the average speed with which animals ran up to the
wheel and turned it enough to drop the door into the nonshock compartment during the sixteen
trials when they no longer received shock in the first compartment (Miller, 1948).
78 Human Motivation
From such an analysis it was but a short Step to the inference that anxiety
could well be the master motive in human behavior (Mowrer, 1950), since
human beings are subjected to so many discomforts they might wish to avoid
before they occur. Thus, in this model it is fear that drives people to accumulate
money in order to avoid the discomforts that come from lack of money. It is
fear that leads people to associate with and seek the approval of others because
of the suffering they must endure if others punish and disapprove. Freudian the-
ory also emphasized the importance of anxiety as a master motive, and this was
just beginning to be widely understood, if not entirely accepted, at about this
time in the United States. Thus, the behaviorists were particularly gratified to
think they had found an objective means of showing how and why anxiety was
so important in human life.
Does Animal Learning Occur in the Absence of Drive?

Some psychologists associated with E. C. Tolman at the University of California
at Berkeley in the 1920s had reported a phenomenon that seemed to throw into
question Thorndike's Law of Effect and the model of learning constructed by
Hüll and his associates. Rats were first allowed to explore a maze without food
reward. After they had explored the maze on several days, food was placed in
the goal box. Almost immediately the rats showed great improvement in getting
from the starting box to the goal where the food was. They ran the maze more
quickly, making fewer errors. As Figure 3.4 shows, "Group I was given a food
reward on every trial. In Group II, the food reward was not introduced until
the seventh day (at point Z). In Group III the food reward was not introduced
until the third day (at point X). Both Group II and Group III showed a sub-
stantial decrease in errors after the first rewarded trial" (Weiner, 1980a).
Apparently the rats were learning something even when they were not being
rewarded. They showed signs of what was called latent learning. How could this
be, according to the model of learning that stated associations were formed by
rewards or reductions in drive? Tolman concluded that the Law of Effect re-
ferred to Performance rather than to learning. In other words, the animals had
learned "what went with what"—which turns in the maze led where—through
sheer contiguity. Neither reward nor drive reduction was necessary for this kind
of learning. Instead, rewards determined how the rats used the information they
had acquired to get from the Start to the goal.
However, the mechanism of acquired drive allowed the Hullians to explain
this apparent contradiction in terms of their model. They only needed to assume
that rats did not like surprises, that stränge or unfamiliar Stimuli would produce
discomfort, and that to reduce such discomfort the rats would learn their way
around the maze. This was the explanation given for exploratory drive; its
evidence was the tendency of rats when put in a maze to explore all its aspects
(W. I. Walker, 1959). In this view, novel Stimuli would serve as mild drive Stim-
uli, which the rat would learn to reduce by exploring the maze. The explanation
for latent learning, then, was that the learning had occurred in the Service of the
exploratory drive rather than the hunger drive, and when hunger and food re-
3.0
2.5
o 2.0
in
ot 1.5
1.0
0.5
0 1 2 3 4 5
Number of Days
Figure 3.4.
Latent Learning Phenomenon. Errors made by rats in running a maze when food was available from
the start (Group I) or introduced at Points X and Z (Groups II and III) (Weiner, 1980a, after
Blodgett, 1929).
ward were introduced the learning simply transferred to another drive Stimulus
(see Montgomery & Segall, 1955). The argument is involved, but it seemed
reasonable and saved the Hullian model from what could have been a fatal
objection.
Cue Characteristics
In the behaviorist model, as Figure 3.2 illustrates, cues, or environmental Stimu-
li, determine where and when a response will be made. For a hungry woman,
the sight of a restaurant trips off the response of entering it to eat. In the same
way, a dinner bell elicits going into the dining room, and a factory whistle
evokes the response of leaving work. Cues also determine which response will be
made, since they get associated with the particular response that happens to re-
duce a drive. Our hypothetical hungry woman enters the restaurant when she
sees it. She does not throw a rock through the window, because entering has
been associated with satisfying the hunger drive and throwing a rock has not.
However, if cues are responsible for directing or channeling behavior, how can
we say that drives direct or Orient behavior?
Some have argued that, strictly speaking, drives do not direct behavior (see
Farber, 1954), although the cues associated with them, like any other Stimuli,
80 Human Motivation
have the capacity to direct behavior. Different drives have different cue values.
The sensations produced by thirst are quite different from those produced by
hunger. These unique sensations produced by different drives may get uniquely
associated with different means of reducing each drive—by drinking in the case
of thirst and eating in the case of hunger.
The Stimulus characteristics of drives were particularly emphasized by Estes
(1958), who analyzed data showing that response strength varied as a function
of the relationship between drive level during training and during testing. Sup-
pose rats were trained to go down a runway to get food after ten hours of food
deprivation. Now suppose the rats are deprived of food for twenty hours and
placed in the same runway, but this time without any food at the end. How
long will the rats continue to run? How long before the running response is ex-
tinguished altogether? Oddly enough, the rats will not run as many times as
they would have if the hours of deprivation were the same as those under which
they had been trained, even though they are hungrier during testing. Estes'
(1958) explanation is that "changing the level of deprivation between training
and test has the effect of dropping out some of the drive cues that had become
conditioned during training." This means the running response will be less
strongly evoked during testing, because responses learned to one type of cue do
not occur as strongly to different, although similar, cues.
Hüll called this phenomenon the Stimulus generalization gradient. It is well
known in conditioning studies that if, for example, the salivary response of a
dog is conditioned to a sound, so that every time the sound occurs the dog sali-
vates, a change in the sound—for example, replacing it with one of a different
pitch or lesser intensity—will evoke a weaker salivary response. The response
generalizes to other Stimuli similar to the one to which it was conditioned, but
the capacity of Stimuli to evoke the response declines regularly the more dissimi-
lar they are to the original conditioned Stimulus. In other words, there is some
transfer of learning to different Stimulus situations, and this principle applies to
the transfer of learning across drive Stimulus situations. The same principle of
transfer of learning was invoked to explain latent learning in the maze discussed
in the previous section. In that case, the learning associated with the exploratory
drive Stimulus was supposed to have transferred to the hunger drive Stimulus.
It was also generally believed that stronger drives tended to flatten the Stim-
ulus generalization gradient (see Miller, 1948). That is, the stronger the drive,
the greater the tendency to respond to Stimuli a little different from the original
conditioned Stimulus. Rosenbaum (1953) showed that experimentally induced
anxiety tended to increase Stimulus generalization. A school phobia illustrates
the principle. As one child gets closer to school, he gets more and more afraid
until, by the time he is a block away, he is too frightened to go any further.
However, if another child has an even stronger phobia, she might respond to
Stimuli present three blocks away from the school just as strongly as the first
child did to Stimuli one block away. In extreme cases she might not even leave
the house. Looked at in another way, the anxiety drive appears to be sensitizing
the children to more and more cues related to the cue originally connected with
it (the school). They perceive even streets far away from the school as school re-
lated. Perceptual sensitization was not particularly studied by the behaviorists,

because they focused more on action than perception.
Response Characteristics
In the model presented in Figure 3.2, a response must occur before it can be
"stamped in," or learned. Scared children cannot be rewarded for entering the
school if they are so afraid they never make the response of entering the school.
Where do these responses come from? Some are innate, such as when a rat
jumps up as electric shock is applied to its feet. Others occur randomly if they
are part of the animal's repertoire, as when a cat is first placed in one of Thorn-
dike's boxes.
Especially important is what Hüll called the habit family hierarchy. That is,
an animal or person normally enters a Situation with some responses much more
likely to be evoked than others because of past experience in similar situations.
These responses are conceived as being arranged in a rank order in which the
strongest one occurs first, then the next strongest, and so on. Reward changes
the rank order of responses in this Situation. The cat may respond first by trying
to stick its paw through the slats, but this response is not rewarded; eventually
the response of turning the button, which is rewarded, assumes the highest place
in the habit family hierarchy in the box Situation. It is important to remember,
however, that a reward increases the likelihood of recurrence of any response as-
sociated with it, no matter how irrelevant it actually may be to producing the
reward. Thus, the cat may learn to scratch itself, even though scratching did not
directly open the door.
The behaviorists recognized that so far as people were concerned, many of
the important responses were not overt actions and therefore could not be ob-
served. People obviously think as well as act; they have ideas and expectations.
They use language to try out responses before performing them visibly. The be-
haviorists cut themselves off from studying such responses directly, largely by
working with animals, which could not talk. That was justified on two grounds:
First, it prevented psychologists from using poorly defined mentalistic concepts,
as Hüll had argued. Second, it was assumed that it was easier to work out the
principles of behavior with overt acts and that these same principles then could
be applied to any kind of behavior, including covert or symbolic acts such as
thoughts.
Reward Characteristics
Rewards were defined as anything that reduces a drive, but the behaviorists rec-
ognized that many objects or situations could be shown to have reward value in
the sense of facilitating learning even though just what drive or drives they were
reducing was not always clear. Thus, a child might collect stamps or marbles,
and it would be difficult to determine just what strong Stimuli these objects were
reducing. Perhaps the child had been teased for not having had any marbles and
collected them to reduce the acquired anxiety drive resulting from teasing. Once
82 Human Motivation
it had been determined that marbles were important to the child, however, they
could be used reliably to get him or her to learn things. This line of reasoning
led Skinner (1953) and his associates to discard the notion of drive altogether
and to argue that it was sufficient simply to list what would serve as a reward
or reinforcer for various types of learned behavior.
According to Hullian theory, objects could acquire reward value by being
associated with primary drive reduction. They were called secondary, or ac-
quired, rewards to contrast them with primary rewards, like food, that were in-
nately rewarding. An early and very influential experiment by Wolfe (1936) il-
lustrated how objects could acquire secondary reward value, even for animals.
He taught chimpanzees to use poker chips to get food. When they inserted a
chip in a vending machine, it would provide them with a piece of banana or a
grape. The chimps even learned that chips of different color had different value:
a blue chip might produce five grapes and a white chip, only one. Wolfe demon-
strated that the chimps would work almost as hard to get the chips as they
would to get food rewards, even if the rewards were delayed after the chimps
put the chips in the vending machine to get food. How hard the chimps would
work depended on the value of the chips and on how many chips they already
had. Furthermore, they hoarded chips and competed to pick them up if a num-
ber were thrown into a cage where two chimps were present. In other words,
the chips had definitely acquired a reward value they did not have before they
were associated with producing food, which resulted in primary drive reduction.
The analogy with the way humans work for and collect money was inescap-
able. Miller and Dollard (1941) put it this way:
During the course of his socialization, the individual learns that the possession of
money is the means of gratifying many different needs and that the lack of money is
a signal that he may have to bear the uncomfortable goading of many undesirable
desires. Some of the drives upon which the desire for money is based are primary
drives, such as hunger and cold. Others may be secondary, such as anxiety. Because
of the number of different drives supporting the need for money, it is a rare occasion
when the individual is without any primary motivation to summate with and acti-
vate his need for money."
Secondary rewards even generalize from the Situation in which they are
learned to another Situation. Estes (1949) demonstrated that a sound associated
with drinking had a reinforcing effect even when rats were hungry but not
thirsty. In this way, lucky charms acquire reward value. A person might have
been wearing something when almost hit by a car, so that it is associated in his
or her mind with escape from pain. Thus, the person wears the same item on
any threatening occasion to reduce the anxiety that arises.
The Incentive Variable

In the first set of principles published by Hüll (1943), excitatory potential, or
the tendency to act, was a Joint function of drive strength and habit strength, or
the number of hours of food deprivation and the number of practice trials in
running a maze. However, it was apparent from the very beginning that size of
reward also affected the tendency to make a response, even though drive was
held constant. This point is made dramatically in Figure 3.5. In this experiment,
the strength of the tendency to act was measured by the speed with which rats
traversed a runway to get to food. They ran much faster to get sixteen pellets of
food than to get one pellet of food and ran still faster to get 256 pellets of food,
even though in all these conditions the hours of food deprivation were held
constant. Furthermore, if the amount of food reward shifted part of the way
through the experiment, so those formerly getting 256 pellets now received only
sixteen, speed of running declined sharply. However, the speed increased if the
amount of the food reward went from one to sixteen pellets.
To take into account the influence of the size of the incentive on response
strength, Hüll added a variable to his equation, generally labeled K (incentive).
In his revised principles (Hüll, 1952) it read as follows:
Excitatory potential = Drive X Habit strength X Incentive,
or SER = D X sHR X K.
That is, on the basis of the data then available, Hüll thought it likely that the
incentive value (K) would multiply with the product of habit strength (
Freshifi Postshifl
4.0
3.0
1 0
1.0
Trials
Figure 3.5.
Speed of Running in a Long Runway as a Function of Amount of Reinforcement. For the first
nineteen trials different groups were given 1, 16, or 256 pellets of food (acquisition data for the
one-pellet group are not presented); after trial twenty, all subjects were given sixteen pellets (Weiner,
1980a, after Crespi, 1942).
84 Human Motivation
and drive (D) for much the same reason as he thought these two variables
would multiply with each other to determine response strength. If any of the
variables in the equation is reduced to zero, there should be no excitatory poten-
tial ($ER) for the behavior in question. If the animal has zero drive, it will not
run, no matter how practiced it is or how huge the reward. It also will not be
able to run a maze without error—no matter how strong its drive or the incen-
tive—if it has had no practice, indicating zero habit strength. Likewise, no mat-
ter what the level of drive and habit strength, if there is no reward or incentive,
there should be no action.
Kenneth Spence (1956, 1958a, 1958b) disagreed. He was a major exponent
of the view that incentive value had to be added to the equation, but feit that it
summated with drive as follows:
S E R = (D + K ) X SHR.
Spence feit this was a more likely way for the two variables to combine because
he conceived of the influence of K as occurring largely through the fractional
anticipatory goal response mechanism (rQ — sQ) discussed earlier. A large food
reward would produce an rg — sg distinctively different from the one produced
by a smaller food reward. The different kinds of rQ — sQ would move forward in
time and combine additively with the drive Stimuli present from the drive State,
and the sum of the two sets of Stimuli would multiply with habit strength to
produce excitatory potential. Research investigations never conclusively settled
the question of which of these two formulations was more nearly correct.
EXCITATORY POTENTIAL IN BEHAVIOR THEORY AND

THE MEANING OF THE TERM MOTIVATION
Hull and Spence referred to the product of habit, drive, and incentive as "excit-
atory potential," which is analogous to what Thorndike called the "impulse to
act." Several recent texts (Atkinson & Birch, 1978; Klein, 1982; Weiner, 1980a)
have dropped this terminology and substituted the word motivation for excit-
atory potential. Thus, in their terminology, anything that increases the tendency
to act can be described as a motivator or as increasing motivation. In this sense,
habits increase motivation, as do expectations and motives or drives. For exam-
ple, suppose we are trying to find out what "moves" a person to go into a res-
taurant at 6:00 P.M. At least three factors contribute to the tendency to perform
this act: the fact that the person usually eats at this time and place (habit), the
fact that the person expects that food is likely to be present in the restaurant
(expectancy), and the fact that the person is hungry.
Atkinson and Weiner refer to habits and expectancies as increasing "motiva-
tion" to perform the act, as "moving" the person to perform it. We will use the
term motivation only to refer to the aroused motive, that is, to hunger in the ex-
ample just given. This is because to describe a habit as a motivator or as con-
tributing to motivation not only departs from traditional usage; it also contra-
dicts common sense and introduces conceptual confusion, particularly in study-

ing human motives. At that level we must distinguish between a mo-
tive—conceived as an individual disposition or trait (the general tendency to be
hungry all the time)—and motivation—conceived as a motive disposition
aroused at a particular moment in time (how hungry the person is here and
now). If the term motivation already has been used to describe any factor that
moves a person to act, it is no longer available to describe an aroused motive.
For all these reasons, we will use the term motivation to describe an aroused
motive and such terms as excitatory potential or the impulse or tendency to act
to describe the effect of the sum total of all determinants of action, following
Hull's example. (See Tables 6.1 and 12.2 in Chapters 6 and 12, respectively.)
Excitatory Potential for Operants More Indicative of Motivation

Hüll built his model of motivated behavior by starting from the conditioning ex-
periment in which a Stimulus is associated with a response followed by a re-
ward. This meant he used, as measures of response strength, latency (the time
between Stimulus and response), the amplitude of the response, the regularity of
response to the Stimulus, or the resistance to extinction (the number of times a
response occurs on presentation of the Stimulus when the reward is not present).
Skinner (1938), however, pointed out that there are many responses that
occur naturally for which a Stimulus cannot readily be identified; for these re-
sponses, he said, such measures of response strength are unavailable. He called
these responses operants. They appear to occur spontaneously, and the measure
of their response strength is simply frequency of occurrence rather than latency
or other such measures of the strength of a stimulus-response connection. Skin-
ner focused attention on observing the rate with which a rat would press a bar
to get food in a small box; in this case the "Stimulus" that evokes bar pressing
is not easily identifiable. He found that the frequency of the bar-pressing re-
sponse was readily influenced by variables of the incentive type or reinforcers
and the schedule with which they were provided to the rat. Thus, for Skinner,
variables of the motivational type were particularly important in influencing the
strength of operant behaviors. In fact, Skinner (1966) flatly stated that "operant
behavior is essentially the field of purpose." The distinction between respon-
dents, where a Stimulus can be identified, and operants, where it cannot, is im-
portant, and we will return to it in Chapter 13. It appears that the response
probability of an operant is a more sensitive indicator of motive strength than
are the measures of respondent response strength, such as latency, amplitude,
and resistance to extinction.
THE BEHAVIORIST MODEL OF MOTIVATION APPLIED

TO HUMANS BY SPENCE AND OTHERS
Behaviorists since Thorndike had studied animals exclusively to get the simplest,
most objective picture possible of how motivation influenced learning. Having
86 Human Motivation
developed a model of motivation from studying other animals, they tried to

apply it to human behavior. One way to do this was simply to assume complex
human motives had to be explained in terms of the principles they had estab-
lished (Korman, 1974). The way behaviorists accounted for the desire for
money, as described earlier, is a good example of how they proceeded. Money is
a secondary reward that has acquired its value because of its association with
many different kinds of drive reduction. Unfortunately, there was no real way to
check this assertion, nor was there any attempt to study the "money drive" as it
functioned in human behavior. It was enough simply to show how it could fit
into the model of motivation that had been constructed (see J. S. Brown, 1961).
This was obviously unsatisfactory, because it left human motives altogether
unstudied. Then Kenneth Spence and his associates, particularly Janet Taylor
Spence, to their great credit, undertook a whole series of investigations using
human subjects; the investigations were designed to test directly the applicability
of the model to human motives. In one group of studies they investigated the
conditioned eye blink (Spence, 1958b). In the experiment a sound is followed
after a brief interval by an air puff to the eye, which elicits an eye blink because
the puff is aversive. After a few trials the sound evokes the eye blink before the
puff occurs, and the eyelid response is said to be conditioned. A conditioned eye-
lid response is one that occurs to the sound cue before the aversive air puff is
applied.
Kenneth Spence and his associates were interested in two motivational influ-
ences on the curve of acquisition of the conditioned eyelid response. The first
was the intensity of the air puff, which corresponded to variations in induced
drive strength. They reasoned that if the air puff were stronger, the drive Stimu-
lus and the anxiety conditioned to it also would be stronger, and the condi-
tioned eyelid response should be learned faster because the drive was stronger.
This is exactly what happened, as Figure 3.6 shows. Note that the pair of learn-
ing curves for the stronger air puff is above the pair of curves for the weaker air
puff throughout the learning trials. That is, the stronger air puff elicited a higher
percent of conditioned responses throughout than the weaker air puff.
Notice also that the curves tend to draw apart, as they should if drive and
habit strength are multiplying, as Hüll and Spence claimed. A little arithmetic
will show why this should be so. In the following numerical example it is as-
sumed that the weaker air puff has a D value of 1 and the stronger air puff, a
D value of 2, and that habit strength (H) has the value of the trial number indi-
cating the number of times the conditioning has been carried out. If we multiply
D X H, the difference in the predicted response strengths on the second trial
would be 1 X 2 = 2 versus 2 X 2 = 4. On the tenth trial, however, the rela-
tive values would be 1 X 10 = 10 versus 2 X 10 = 20. The difference between
10 and 20 is much greater than the distance between 2 and 4, which is exactly
the way in which the curves draw apart in Figure 3.6. If the two factors (D and
H) simply summed rather than multiplied, the curves would stay a constant dis-
tance apart throughout their entire course.
Spence and his associates also were the first to be interested in studying in-
dividual differences in the strength of a motive. Once again they worked with
the anxiety motive, and Taylor developed a self-report measure of Symptoms of
1.5 Pounds, HiA

t/5
1>
70 _ t
1/5
C
O
G. 60 — J» 1.5 Pounds, LoA
QC
iditio ned
50
40
o / y^m^^ 0.25 Pounds, HiA
o 30
»tage
/i<y^ ^ ^ ^ . ^ ^ ^ ^ 0 . 2 5 Pounds, LoA

20
<D
<U
CL 10 ~ 4
n 1 1 1 1 1 1 1 1
1 2 3 4 5 6 7 8
Blocks of Ten Trials
S. = Conditioned Stimulus (light, 1.5 footcandles).
S =Unconditioned Stimulus (air puff).
S c .-S w = 450 milliseconds.
Figure 3.6.
Performance in Eyelid Conditioning as a Function of Test Anxiety Score (A) and Intensity of
Unconditioned Stimulus (Air Puff to Eye) (after Spence, 1958b).
manifest anxiety, the Taylor Manifest Anxiety Scale (MAS). In other words,
they recognized that not only could anxiety be increased experimentally, but
also that individuals might vary as a result of their life histories in the extent to
which they had developed a generalized type of anxiety drive. The scale score
consisted of the extent to which individuals agreed with items like "I worry
quite a bit over possible misfortunes." Spence and his associates divided their
subjects into those who agreed very much with items like this and those who
agreed very little with them. The learning curves for those scoring very high
(the high anxiety, or HiA, subjects) and for those scoring very low (the low anx-
iety, or LoA, subjects) are shown separately in Figure 3.6.
It is evident that this type of drive also interacts with habit strength over
the course of learning in the same way induced anxiety (from the air puff) does.
The high anxiety subjects conditioned more rapidly than the low anxiety sub-
jects at both levels of induced anxiety, represented by weaker or stronger air
puffs. The strength of the air puff can be conceived of as an index of aversive in-
centive in the Situation, since it can be assumed that the eye blink is learned to
avoid the unpleasant puff, and the stronger the puff, the greater the drive reduc-
tion from avoiding it and the stronger the incentive to blink. If so, the curves
should give some indication as to whether drive (D) and incentive (K) multiply
88 Human Motivation
(as Hüll thought) or add together (as Spence thought) in multiplying with habit
strength (H) as represented by the number of practice trials. If they multiply,
the curves for the high anxiety subjects, for example, at two incentive levels
should be much farther apart and show greater Separation with increasing H
than if D and K summate. Unfortunately, the results are ambiguous: The curves
do not seem as far apart as Hull's assumptions would suggest they should be.
Thus, Spence's model may seem more reasonable, but there is no way to judge
for certain how far apart the curves should be.
How a Strong Drive Can Interfere with Performance on Complex Tasks

In the conditioned eyelid experiment, one response is clearly dominant, namely,
the eye blink. In more complex learning situations, however, many responses are
potentially relevant. In some cases the response most likely to occur (the one
highest in the habit family hierarchy) may be the wrong one for solving a prob-
lem. Yet according to the behaviorist model, drive should facilitate the appear-
ance of the strongest response, even if it is the wrong one. It follows that in-
creases in drive strength might delay learning of complex tasks because they
would increase the likelihood of making a wrong or irrelevant response at the
beginning of learning.
Haner and Brown (1955) carried out a very simple experiment that demon-
strated how increasing drive could make it more likely that an irrelevant re-
sponse would occur. They had subjects play a game in which the objective was
to place thirty-six marbles in small, round holes in a board in a limited period
of time. The experimenter signaled when the time was up by pushing a lever,
which dropped all the marbles down below the board. The strength of drive was
manipulated by indicating that the time was up after various numbers of mar-
bles had been put in place. That is, it was assumed that frustration would be
greater if the time were up after the subjects had put nearly all the marbles in
place (thirty-two out of thirty-six) than if they had put only half of them in
place (eighteen out of thirty-six). In drive theory, frustration is an important
source of tension, or drive strength. After the trial was ended by dropping the
marbles below, the subject had to push a plunger to get ready for the next trial.
As frustration or drive strength increased from the subjects' being interrupted
nearer and nearer the goal, they tended to push the plunger more and more vig-
orously. However, vigor of plunger pushing is in no way related to doing better
at the task, and under certain conditions such responses might interfere with
solving a complex problem.
An experiment in learning paired associates was conducted to test this hy-
pothesis (Spence, Farber, & McFann, 1956). Subjects were presented with a
word like harren, to which they had to learn to say a response word whenever
harren recurred. In one experiment the response words were highly associated
with the Stimulus words. For example, fruitless was the response to be associ-
ated with harren. In such instances the correct response would be assumed to be
fairly high in the subject's habit family hierarchy and likely to occur in any
case. In another experiment the response words were totally unrelated to the
Stimulus words. Now a word like grouchy would have to be associated with har-
ren. In this instance, more dominant responses to harren, like fruitless or arid,
would be incorrect; according to theory, however, greater drive strength should
strengthen them, just as it strengthens all responses that tend to occur in the Sit-
uation, so learning should be slowed. Thus, the prediction was that greater drive
strength should facilitate the learning of highly associated pairs where the domi-
nant responses were appropriate, but it should inhibit the learning of unasso-
ciated pairs where the dominant responses were incorrect.
The measure of drive strength was whether the subject scored very high or
very low on the Taylor MAS. Some of the early results confirm the prediction
made by the theory, as Figure 3.7 shows. The high anxiety, or high D, subjects
learn the highly associated pairs faster than the low anxiety subjects, but they
learn the unassociated pairs less quickly than those low in anxiety.
Sarason, Mendler, and Craighill (1952) obtained similar results in a different
test Situation and explained them a little differently. They asked subjects to place
digits under associated Symbols, with either of two sets of instructions. In one
case they told the subjects that they could not be expected to finish substituting
all the digits for Symbols in the time allotted, and in the other they told the sub-
jects that the average College Student should find it fairly easy to complete the
test within the time limit given. They also divided the subjects into those who
scored high or low on self-reports of the amount of anxiety experienced in test-
ing situations. As the results in Table 3.1 shows, the highly anxious subjects did
Low Anxiety (D)

•2 100 High Anxiety (D)
cd
1
< 80
s
o
U 60
S 40
Low Interpair
20
Association
4-5 6-7 8-9 10-11

Pairs of Trials
Figure 3.7.
Paired Associate Learning as a Function of Anxiety or Drive Level and Degree of Interpair
Association (after Atkinson & Birch, 1978, after Spence, Farber, & McFann, 1956).
90 Human Motivation
Table 3.1.
EFFECT OF INDUCED PRESSURE ON PERFORMANCE FOR SUBJECTS HIGH AND LOW IN
TEST ANXIETY (data from Sarason, Mandler, & Craighill, 1952)
Mean Number of Digit Symbol Substitutions Made on Trial One

When Not
Subject When Expected to Expected to Finish Difference
a
Classification Finish (A) (B) (A - B) Pn difference
Subjects high in 28.7 29.8 -1.1 .09

anxiety
Subjects low in 32.7 29.8 + 3.1 .003
anxiety
a
Probability of the difference occurring by chance.
less well under the pressure of being expected to finish, and they made fewer
substitutions than when they were not expected to finish. The subjects low in
anxiety performed better when expected to finish than when not expected to
finish.
To explain this difference, the authors again draw on the principle that a
stronger drive strengthens all responses more. In the case of the high anxiety
subjects, the increased drive strength induced by Performance pressure strength-
ens task-irrelevant responses (for example, interfering thoughts about doing
badly) that frequently have been evoked in the past for these subjects. Therefore,
their Performance is poorer when they are expected to finish. In contrast, for the
low anxiety subjects an increase in drive strength evokes task-relevant responses
that have led to success in the past. So they perform better when they are ex-
pected to finish.
In this case the responses strengthened by drive are in the subject rather
than in the material to be learned, as they were in the Spence studies. Other re-
search has shown that highly anxious subjects generally do worse when faced
with complicated or difficult tasks. For example, Tennyson and Woolley (1971),
using the Spielberger measure of the State of anxiety at the moment (Spiel-
berger, Gorsuch, & Lushene, 1970), found that high anxiety subjects made many
more errors on a difficult task than did low anxiety subjects. The task was de-
ciding whether a line of poetry illustrated trochaic meter or not; it was arranged
so some of the identifications were very easy and others were difficult. In the
easy identifications, the high anxiety subjects did better than the low anxiety
subjects because of their higher drive level. However, when drive was pushed
still higher, by requiring the subjects to work on a difficult task, it facilitated the
appearance of task-irrelevant responses (for example, worries about how well
they were doing), which interfered with Performance. This suggests that task de-
mands can first facilitate Performance and then, as they increase still further, in-
terfere with Performance—a phenomenon known as the Yerkes-Dodson law,
which will be discussed in Chapter 6.
Interference Effects Due to More Experiences of Failure

Weiner (1966) noticed an obvious fact about the Spence experiments that also
would explain Tennyson and Woolley's results. Generally speaking, the tasks at
which the highly anxious subjects did more poorly were more difficult, and the
ones at which they did better were easier (see Figure 3.7). On more difficult
tasks, subjects experience more failure as they go along, and perhaps it is the ex-
perience of failure that makes highly anxious subjects do worse rather than the
effect of increased drive on irrelevant responses.
Weiner designed an experiment to test this hypothesis. He employed the
same sets of paired associates used in the Spence experiments, but he severed
the connection between experiences of success and failure and task difficulty as
follows:
Subjects learning an easy list of paired associates were told that they were perform-
ing poorly relative to others. In this manner the easy task was paired with a failure
experience. Subjects learning a difficult paired associates task were told that they
were doing well relative to others, thus pairing the difficult task with a success expe-
rience. If the differential reactions to success and failure experiences are the essential
determinants of behavior in this Situation, then on the easy task highly anxious sub-
jects experiencing failure should perform worse than subjects low in anxiety experi-
encing failure. On the difficult task, highly anxious subjects experiencing success
should perform better than subjects low in anxiety experiencing success. The experi-
mental design consequently provides a definitive test of the alternative explanation of
the Spence et al. data. (Weiner, 1966)
Weiner's results, as shown in Table 3.2 and confirmed in a subsequent ex-

periment, conform to his prediction. Subjects high in anxiety do better than sub-
jects low in anxiety after success, even on the difficult paired associate task. This
was the same task on which they did worse when they were given no informa-
tion as to how well they were doing and inferred they were not doing very well
because they kept failing at it (see Figure 3.7).
This finding is very important, but it does not resolve all questions arising
from this type of research. Spence (1958a) reports that with certain types of
more difficult tasks, the more anxious subjects do better, contrary to Weiner's
prediction, although this result could not be regularly obtained. If we do not ac-
cept the Hull-Spence explanation of why high anxiety subjects do better on easy
tasks or on those in which the dominant response is correct, how is this fact to
be explained? One possibility is that failure increases task-irrelevant responses in
high anxiety subjects and success quiets them. This would explain why highly
anxious subjects do better on easy tasks at which they are successful than on
difficult tasks at which they fail; however, it would not explain why the highly
92 Human Motivation
Table 3.2.
MEAN NUMBER OF TRIALS TO SUCCESSFUL LEARNING OF PAIRED
ASSOCIATES (after Weiner, 1966)
Subject Classification Easy List (Failure) Difficult List (Success)

Subjects high in anxiety (//) 9.55 14.85
Subjects low in anxiety (L) 7.08 20.10
Difference (H - L) 2M -5.25
p < .10 p < .10
Note: Low score indicates faster learning. Subjects high in anxiety do better than subjects low in anxiety after
success, even on a difficult task, but they do worse even on an easy task if it is associated with failure.
anxious subjects do better than less anxious subjects on easy tasks, since less
anxious people should have no task-irrelevant responses to interfere with their
Performance. We are led to infer, as Atkinson and Birch (1978) argue, that
other motives must be higher for the highly anxious subjects. Perhaps subjects
high in anxiety have a stronger need for social approval than subjects low in
anxiety, which leads them to do better when they are not bothered by task-
irrelevant responses.
Later research has tended to emphasize the importance of other motives
present in Performance situations. For example, Tables 3.1 and 3.2 might bring
up the question of why the subjects low in anxiety do better after being put
under pressure. Atkinson argues that the only way to understand this is in
terms of knowing something about the achievement motivation of the subjects
(see Chapter 7). This is because challenges increase Performance among subjects
high in achievement motivation, particularly if they are also low in anxiety. At-
kinson reinterprets the results of these experiments in terms of the Joint eifects
of the fear of failure, as measured by test anxiety, and the achievement motive,
as Chapter 7 will show. Subjects low in anxiety are often high in achievement
motivation, so setting challenging or explicit goals for them may improve their
Performance for a reason quite unrelated to the drive level of anxiety. Certainly
other motives in a Situation need to be taken into account, but no one can ques-
tion the importance of the attempt Spence and his associates made to apply the
behaviorist model of motivation to individual differences in motive strength at
the human level.
How the Presence of Others May Increase Drive and

Interfere with Performance on Complex Tasks
Just as the line of investigation started by Spence had begun to run its course,
Zajonc (1965) came up with a human application of the behaviorist model of
motivation in a quite different area. He reviewed a considerable body of research

evidence that showed that the presence of another member of a species often in-
creases response strength. For example, if a chicken is allowed to eat its fill until
it stops eating, and then another chicken is introduced and Starts eating the
common food supply, the first chicken will begin to eat again. It looks as if the
second chicken has increased the hunger drive level of the first chicken. The
phenomenon is generally known as social facilitation. Zajonc found evidence
that social facilitation improves Performance of simple tasks but can decrease
Performance of more complex tasks. He saw this as another application of
Hull's model, since the presence of another, if it increases drive, should facilitate
Performance if there is a simple dominant response and should inhibit perfor-
mance if there are many competing responses that would be strengthened by the
increased drive.
Zajonc and Sales (1966) designed the following experiment to test the hy-
pothesis. They flashed nonsense syllables briefly on a screen, which they told the
subjects were Turkish words that the subjects were to learn to pronounce when
they saw them. A given word was presented one, two, four, eight, or sixteen
times. In short, some words were seen and pronounced much more often than
others. Interspersed among the presentations were some blurs, which the sub-
jects were told were words they could not see very well but were to guess at
anyway. These were called pseudorecognition trials. The average number of
words that had been exposed different numbers of times that were called out in
the pseudorecognition trials is plotted in Figure 3.8. As the general upward
slope of the curves indicates, the subjects were more likely to call out words to
the blurs they had seen more often.
Some of the subjects performed this task alone, and others while two other
students watched them—in what was called the audience condition. If we sup-
pose that adding two watchers increases drive, the two curves would seem to
conform to the prediction made from Hull's model. The subjects under condi-
tions of high drive (audience condition) produced more of the dominant re-
sponses (those exposed sixteen times) than did the subjects with lower drives
(alone condition). However, the reverse is true for the infrequently exposed
words, where there is no clearly dominant response. It is assumed here that high
drive would make it more likely that subjects would respond with any word
high in their hierarchy, so they would be less apt to respond with a word that
had been seen only once than a subject performing alone at a condition of lower
drive.
However, more complications arose as others joined in investigating this ap-
plication of the Hüll model. Cottrell, Wack, Sekerak, and Rittle (1968) showed
that it was not just the presence of others that produced this effect, but rather
what Cottrell called evaluation apprehension—the expectation that the audience
would evaluate Performance. He repeated the Zajonc and Sales experiment using
"Turkish" words, except that the two others present were blindfolded, and the
subjects were told they were sitting there to adapt their eyes to the dark. Under
these conditions the curve obtained was exactly parallel to that obtained for sub-
jects when they were alone.
Cottrell et al. (1968) also used the paired associate technique used by
94 Human Motivation
12
Facilitated Subjects,
10
High Drive ^_
1=
z
Trial Block IV
0 J_ J_ 1 L
1 2 4 8 16
Words Seen Words Seen
Less Often More Often
Training Frequency (Habit Strength)
Figure 3.8.
Average Number of Responses of Different Frequency Classes in Separate Trial Blocks of the
Pseudorecognition Series (after Zajonc & Sales, 1966).
Spence and others to test the application of the model for the anxiety drive, and
they obtained the same results. That is, the presence of an audience tended to
improve Performance on the paired associates of high association value where
there was a dominant response, and it impaired Performance on the low associa-
tion pairs where there was no dominant response. Cottrell (1972) argues that
"through experience the individual learns to anticipate subsequent positive or
negative outcomes whenever others are merely present and not overtly doing
anything that has motivational significance for him. It is these anticipations,
elicited by the presence of others, that increase the individual's drive level." In
other words, Cottrell agrees that drive level is increased by an audience and pro-
duces the effects expected from the Hüll model, but he is defining the nature of
the drive differently from the way Zajonc defined it.
Geen and Gange (1977) have reviewed these experiments and showed that
all kinds of influences can result from the presence of others, depending on the
nature of the Situation. Others may simply provide more information on what to
do. Goldfish swimming in a group learn a response faster than when swimming
alone. Others may increase competitiveness, or fear that the other person will
take what one wants. An audience may increase fear of disapproval, which leads
subjects to make more conventional—and also more common or dominant—
responses. The presence of others simply may distract people so they can con-
centrate less on the task. It clearly is too simple to think of the mere presence
of another as having a universal effect on drive level.
• REINTERPRETING THE BEHAVIORIST STUDIES OF

HUMAN MOTIVATION IN TERMS OF WHAT GOES ON
IN PEOPLE'S MINDS
A serious limitation of studies in the behaviorist tradition is that they cut them-
selves off from determining what is going on in the subjects' minds during an
experiment. That is why the behaviorists started working with animals in the
first place. Instead, as different conditions produced different effects, they were
forced to infer what was going on in the subjects' minds, whether animal or
human. If a satiated chicken Starts eating again when a second chicken eats its
food, the experimenter may infer that the hunger drive is aroused in the first
chicken or that it "feels" more competitive. However, there is no way to be sure
what is going on in the chicken's mind, although experimenters can carry out
further experiments designed to test whether it is the competitive urge or hunger
that is aroused.
The research on the audience eflfect in humans has been carried out accord-
ing to the animal model; it tries to infer what is happening in subjects' minds
from studying their behaviors in different situations. Compared with using ani-
mals, however, using human subjects has an important advantage: it is entirely
possible to find out what is going on in their minds. Fortunately, Kawamura-
Reynolds (1977) has carried out an experiment that does just that. She first
demonstrated that the subjects in her experiment showed the Standard effect
supposedly caused by the increase in drive due to an audience; that is, subjects
in the audience condition (higher drive) learned difficult paired associates of low
association value (with no appropriate dominant responses) more slowly than
they did in the alone condition (lower drive). The explanation, according to the
behaviorist model, is that the increased drive due to the audience strengthens
task-irrelevant associations when there is no dominant association, so the sub-
jects take longer to learn the difficult associations.
Kawamura-Reynolds then did something different. After the subjects had
learned the associations, she had them write imaginative stories to pictures,
using a technique fully described in Chapter 6. This procedure was to find out
what the subjects had on their minds or were thinking about in the alone and
audience conditions. She used six picture cues: Three were designed to evoke
Standard or dominant responses in everyone. One had to do with thirst and
drinking; one, with hunger and eating; and one, with art and aesthetic activity.
Three other pictures were ambiguous and designed to evoke stories relating to
achievement or affiliation. Kawamura-Reynolds' first prediction was that if the
behaviorist model was correct, and the audience had increased nonspecific drive,
it should increase the likelihood of simple, dominant responses being made in
the audience condition compared with the alone condition. She obtained this re-
sult, as Table 3.3 shows. More of the subjects in the audience condition gave
96 Human Motivation
Table 3.3.
NUMBER OF SUBJECTS SHOWING DOMINANT IMAGINATIVE THEMES IN RESPONSE TO
THREE STRONGLY CUED PICTURES AND STRONG NEED FOR AFFILIATION RELATIVE TO
NEED FOR ACHIEVEMENT IN RESPONSE TO THREE WEAKLY CUED OR AMBIGUOUS
PICTURES (after Atkinson & Birch, 1978, after Kawamura-Reynolds, 1977)
Need for
Strongly Cued Pictures Affiliation Ambiguous Pictures
Dominant Relative to Need
Themes Audience Ahne for Achievement Audience Ahne
One or more 25 17 High 22 14
None 11 19 Low 14 22
X2 = 3.66* X2 - 3.52b
a
/7 < .03 in the direction predicted by drive theory.
h
p < .04 in the direction predicted by cognitive theory of instigating force and dynamics of action.
more conventional or dominant themes, such as a drinking story to a thirst cue,

than subjects in the alone condition.
However, Kawamura-Reynolds, following Cottrell's argument, also had pre-
dicted that the presence of the audience would evoke evaluation apprehension,
which would show up in the stories as a greater concern for affiliation, as con-
trasted with a concern for achievement (both scored according to the Standard
procedures described in Chapter 6). That is, it has been established that subjects
with a stronger need for Affiliation (abbreviated n Affiliation) generally have a
greater need to be liked and approved of. Thus, if the audience in fact increases
the need for approval or evaluation apprehension, as Cottrell argues, it should
show up as an increased need for Affiliation. This is what Kawamura-Reynolds
found, as the right side of Table 3.3 illustrates. More subjects in the audience
condition than in the alone condition had a higher n Affiliation score relative to
their need for Achievement (n Achievement) score. Furthermore, she found that
it was the subjects who were relatively higher in their affiliation concern who
produced more dominant or conventional responses to the strongly cued pic-
tures. In other words, it may be the speciflc motive aroused by the audience—
rather than a nonspecific drive—that on one hand produces more conventional
or dominant responses in simple situations and on the other produces anxiety
that interferes with Performance in complex situations.
What is important about this study is that it introduces a methodology that
enables experimenters to check directly their inferences about what is going on
in subjects' minds. This should greatly speed up understanding of situations like
these, in contrast to what can be gained from dozens of experiments in the be-
haviorist tradition aimed at trying to pin down inferences about what subjects
are thinking or wanting.
A COMPARISON OF THE PSYCHOANALYTIC AND

BEHAVIORIST CONTRIBUTIONS TO THE
STUDY OF MOTIVATION
Similarities
It is useful to compare Freud and Hüll as representatives of the psychoanalytic
and behaviorist traditions in psychology, respectively, because their characteris-
tics strongly influenced these traditions and in the end epitomized them. On the
surface they seem very different—Freud the cultured product of European civili-
zation who saw wealthy patients in his Viennese Consulting room, and Hüll the
American son of an illiterate father who ran white rats in mazes in New Haven,
Connecticut. As far as the science of psychology was concerned, however, they
basically agreed. Both were strongly antireligious and feit psychology could
never become a natural science unless it got rid of such obviously absurd no-
tions as God, the soul, sin, and loving kindness. They were both thoroughgoing
determinists, believing that the scientific study of people could be modeled after
nineteenth-century physics and that they could explain the notion of "free will"
in purely mechanistic terms. They both opposed studying conscious experience,
although for different reasons—Hüll because he thought such study would lead
inevitably into mentalistic or religious concepts, and Freud because he was fasci-
nated by the power of unconscious processes.
Both Freud and Hüll believed psychology could become a quantitative sci-
ence, as physics was, and that it should be based on physiological processes, al-
though neither used physiological information directly in their work. Both were
close reasoners and scorned sloppy thinking; Freud's analyses of clinical cases
read like the careful unraveling of a complicated plot in a detective story in the
Sherlock Holmes tradition. Hull's model for thinking was geometry, with its axi-
oms, propositions, and careful derivations. Both intensely disliked armchair
speculation because it involved loose thinking, was often moralistic, and was not
based on observable behavioral facts.
Darwin's theory of evolution greatly influenced both Freud's and Hull's
thinking about motivation. They concluded that organisms were endowed with
motives to help them survive. The hunger motive enabled the individual to sur-
vive, and the sexual instinct enabled the race to survive. Freud and Hüll came
up with surprisingly similar conceptions of how drives affected the organism:
they energized behavior. Hüll spoke of hunger activating the organism and
Freud of the energy provided by the libido, or sexual drive. They both believed
drives were tension inducing or disturbing and that reward or pleasure was pri-
marily relief from tension. Drives also had aims (Freud) or goals (Hüll), which
focused behavior in one direction or another. Finally, drives were responsible for
98 Human Motivation
fixation of Symptoms (Freud) or learning (Hüll). Since Freud was not particu-
larly interested in problem solving, he did not emphasize learning very much; he
thought of Symptoms as fixated (Hüll would say learned) responses produced by
a strong drive.
It is small wonder, in view of these similarities, that Freud's contributions
to psychology have always been more acceptable to the behaviorists than, for ex-
ample, Jung's. Jung did not share many of these values. He was overtly reli-
gious; he was not averse to speculation, particularly about mystical as opposed
to mechanical matters; and he was not so thoroughgoing a determinist.
Differences
Freud and Hüll both were empirically oriented toward behavioral facts and ob-
servations, but in quite different ways. Hüll emphasized the experiment and
what could be learned about motivation by systematically varying external ex-
perimental conditions. Freud focused on the case study, on the pattern of moti-
vation that would explain a number of symptomatic behaviors in an individual.
This approach lent itself to the study of individual differences in motives much
more than did the experimental approach. However, an adequate study of moti-
vation involves an understanding of how the person (Freud's orientation) and
the Situation (Hull's orientation) interact to produce behavior.
Another difFerence was in the relative emphasis they gave to induction and
deduction. Freud was obsessed with factual observations—with inferring the mo-
tivational meaning of particular dream contents or symptomatic acts. By induc-
tion he arrived at generalizations from all these observations and interpretations,
but since he incessantly made new critical observations, his generalizations kept
changing; therefore, it is difficult to find "the" Freudian System. He kept chang-
ing his mind as he got new and better insights from further clinical studies. This
led to great creativity and originality. Freud's books have been continuously
mined by psychologists ever since he wrote them for ideas on what is important
to investigate in psychology.
In contrast, Hüll developed a tight, logically related, simple System of prin-
ciples that could be used to explain the most complex aspects of behavior, much
as a few laws in physics explain many complex natural phenomena. Once these
principles had been established, the only task left for psychology was to apply
them. Facts that did not fit the principles were a nuisance; they seldom led to
new insights or ideas. Instead, they were "explained away" by some more elabo-
rate deduction from the basic principles. What Hüll gained in systematic think-
ing, therefore, he lost in originality, and what Freud gained in originality, he
lost in systematic thinking. Good science requires both approaches and a bal-
ance between induction and deduction, as well as between adhering to a System
and being open to new ideas.
Hüll was interested in objective, observable acts and Freud, in Symptoms,
thoughts, and dreams. The acts of particular interest to behaviorists were those
that produced learning or solved problems, whereas maladaptive acts, Symptoms,
or drives interested Freud. As was just noted, trying to carry out motivational
studies of action in the behaviorist tradition without knowledge of dreams or

what people are thinking about is like trying to paste up wallpaper with one
hand tied behind your back. There is no reason to exclude the kinds of behav-
iors Freud was interested in from experimental studies of motivation. They have
been excluded out of ignorance about how thinking can be studied objectively
and out of fear that such an approach would reintroduce mentalistic, subjective,
or spiritual concepts. Fortunately, recent studies of cognitive factors in human
psychology have begun to correct this behaviorist bias.
However, the Freudian emphasis on dreams and Symptoms led to a neglect
of the human adaptive capacities. Later psychoanalysts attempted to overcome
this deficit with a study of the ego, although for the most part without the
benefit of the knowledge U.S. psychologists had accumulated about the learning
process.
LIMITATIONS OF THE BEHAVIORIST MODEL
Hull's view of drive was attacked early and effectively by those who demon-
strated that motivation did not always involve tension reduction. Sheffield and
Roby (1950) showed that a nonnutritive sweet taste from Saccharin had reward
value for rats. That is, the rats would learn to get the sweet taste even though
Saccharin could not reduce tension or the hunger drive. Sheffield, Wolff, and
Backer (1951) reported that male rats would cross an electric grid to copulate
with a receptive female even though they were always interrupted before orgasm
so there was no drive reduction. Harlow, Harlow, and Meyer (1950) found that
rhesus monkeys would work to disentangle a mechanical puzzle even in the ab-
sence of primary drive reduction. It was difficult to imagine what kind of sec-
ondary reward in the Situation could account for their persistent attempts to dis-
entangle the puzzle. In fact, the introduction of food reward for working the
puzzle correctly interfered with the monkeys' Performance. Furthermore, at
about this time, Olds (1955) was beginning his very important series of studies
showing that there appeared to be "pleasure centers" in the brain: rats would
press a bar to get a slight electrical Stimulation in some portions of the brain
and not others. All these experiments and many others pointed to the existence
of many sources of motivation other than those related to Stimulus or drive re-
duction. They will be discussed in the next two chapters.
Another criticism of the Hullian theory of drive was that it excessively fo-
cused on survival needs, such as hunger, thirst, and pain avoidance. Thus, mate-
rialistic needs were primary, and all other so-called higher needs were second-
ary, derived from, and presumably less urgent than primary drives. Even
humanistically oriented psychologists like Maslow and Allport, as noted in
Chapter 2, bought the notion that materialistic biological needs are primary, al-
though they then declared that the higher needs were independent of their bio-
logical origins.
The idea is not new. The sixteenth-century French satirist Francois Rabe-
lais, in his Fourth Book of the Heroic Deeds and Sayings of Pantagruel, made
100 Human Motivation
the point vividly in picturing Sir Gaster (that is, Sir Stomach) as the ruler of the
world, or
. . . foremost master of arts in the world. . . . Into whatever Company he may go,
there can be no discussion there of first rank or preference; for he always goes before
all others, even though they be kings, emperors, or the Pope himself. . . . To serve
him, all the world is busy, all the world labors; and as recompense, he does this for
the world: he invents for it all arts, all machines, all trades, all implements, and all
refinements. . . . He invented water-mills, wind-mills, hand-mills and myriad other
contrivances, to grind his grain and reduce it to flour, invented yeast to ferment the
dough and salt to give it flavor. . . . If a sea or river stood in the way of its trans-
portation, he invented boats, galleys and ships.
Is the stomach the master of all arts? Do material needs take precedence over
all others, which in the end derive from them? The behaviorist position implied
that biological needs were primary.
Even the prima facie case for such a proposition is not as strong as one
might think in view of its wide acceptance. Certainly in history individuals and
groups have often acted in ways that were quite contrary to the satisfaction of
their material wants. Martyrs have gone to the stake for principles. Christians
were willing to be eaten by lions rather than deny their Christianity. Whole na-
tions have embarked on idealistic adventures that have deprived them of mate-
rial pleasures. A miscalculation in U.S. foreign policy at the time of the Bay of
Pigs disaster in Cuba was based in part on the assumption that material incen-
tives would take precedence over more idealistic social values. It was known at
the time that the Cuban people were suffering from severe shortages of food and
other material goods under the Communist Castro government. It was therefore
assumed by U.S. policy makers, many of whom had been taught in psychology
courses that the hunger drive was primary, that the Cubans would be ready to
revolt against an oppressive regime that was not satisfying their material wants,
and that an invasion from the outside was all that was needed to touch off the
revolt. History proved these assumptions incorrect. The Cuban people were un-
doubtedly in physical need. Other incentives and motives proved more impor-
tant to them, however, and they did not join in the attempt to overthrow the
government. At the very least it would have to be concluded that the case for
the primacy of material wants is not proved.
This viewpoint could well be mistaken; motives could develop around bio-
logical needs in the same way they develop in general out of certain built-in ten-
dencies to approach and avoid certain experiences. This viewpoint is elaborated
in the next two chapters, which maintain that all motives, including hunger, are
learned out of experiences with certain natural incentives. Some motives may be
stronger than others for a variety of reasons, but there is no convincing evidence
that motives based on biological survival need take precedence over all others in
human adults.
The final major limitation of the behaviorist model of drive has already been
referred to several times: its approach to human motivation, as opposed to ani-
mal motivation, was scholastic rather than empirical, except for the flurries of
research started by Spence and Zajonc summarized earlier. The approach was
scholastic in the sense that explanations for human desires and needs, such as
the desire for money, were simply derived from basic principles and never actu-
ally studied (J. S. Brown, 1961). Once a researcher had shown that the need for
money could be explained by the fact that money was likely to be associated
with the reduction of various primary and secondary drives, there was nothing
more to be done. Such a conclusion was not a hypothesis that could be realisti-
cally tested in any way; thus, it was not a scientific hypothesis, but instead was
a scholastic deduction from first principles.
It is ironic that a movement that was begun to eliminate mentalistic con-
cepts ended up attributing mental states to people that could not be checked
empirically. Behaviorists like Hüll, Spence, or Skinner often use mentalistic con-
cepts or refer to inner states of mind to explain what they mean. They apologize
for doing so and defend themselves by saying they must use such terms instead
of more objective ones to be understood easily. What they fail to recognize is
that in the meantime methods have been developed for coding and counting
thought contents, and these methods are just as objective as those used in ob-
serving overt actions. These methods, as illustrated by the Kawamura-Reynolds
experiment, are explained in Chapter 6 and throughout this book. They clearly
must be used to Supplement the behaviorist contribution.
The concepts in the behaviorist model were too limited in scope, but that
should not blind us to its contributions. It demonstrated the great importance of
the motivational experiment, which must be used to test alternative hypotheses.
The behaviorist model discouraged psychologists from using vague, ill-defined
concepts to explain what is going on. It taught them that relationships among
well-defined variables must be explicitly and carefully stated so they can be
checked for their validity through observation and experiment.
NOTES AND QUERIES

1. The behaviorist model objectified purpose through the concept of the frac-
tional anticipatory goal response. After a rat had been fed in a maze, it was
supposed to Start salivating when it was put back in the maze, indicating
that it was already anticipating eating as much as it could with no food
present. The increased salivation could be interpreted as a sign that it was
"thinking about eating." Does this seem equivalent to "wanting to eat"? No
one actually checked to see if rats did salivate more in this way. What
would be the implications of finding or not finding that they salivated more
when they were put back in a maze after being fed in it?
2. How would you measure the strength of the "money motive" in humans?
How would you determine whether such a motive is based on relief of ten-
sions, discomforts, and anxieties, or on something eise? How would the mo-
tive in question be identified in people at a time before they began using
money?
3. The behaviorist model explained that rats' latent learning of maze pathways
(see Figure 3.4) was actually the result of learning in the Service of the ex-
ploratory drive. Thus, the effort to show that learning could occur where no
motivation was present appeared to fail, just as similar examples of inciden-
tal learning were explained away as being due to motives human subjects
themselves brought to laboratory experiments (see Chapter 1). Can you
think of a way of testing whether learning can occur in the absence of moti-
vation? If motivation must always be present to explain learning, is it a fal-
sifiable hypothesis that motivation is necessary for learning? If the hypothe-
sis is not falsifiable, is it a scientific hypothesis? What is the difference
between stating that motivation facilitates learning and stating that motiva-
tion is necessary for learning?
4. Is it necessary to assume that drives are based on strong Stimuli and re-
wards on Stimulus reduction or tension reduction? What eise could drives be
built on?
5. The chapter argues that acquired, or secondary, rewards in humans often
appear to be much stronger than the primary rewards on which they are
based. This is seldom if ever true in animals. For example, no matter how
much experience Wolfe's chimpanzees had with token rewards, they still
preferred and worked harder for grapes and pieces of banana than for the
tokens. Why then would Christians allow themselves to be eaten by lions,
giving up all primary rewards from going on living in favor of secondary, or
acquired, rewards from sticking to their beliefs? Can you think of circum-
stances that might make acquired rewards stronger than primary rewards
(see McClelland, 1942)?
6. Why should behavior occurring in the absence of an identifiable Stimulus
(that is, operant behavior) reflect more sensitively motivation or purpose
than behavior occurring in response to a Stimulus?
7. According to the way motivate is used in this book, is the term properly
used in the sentence "One way to motivate people to cooperate is to in-
crease communication among them"? If not, how would you rephrase the
sentence to reflect the terminology adopted here?
8. Sometimes people cannot recall the name of someone they know very well
no matter how hard they try. Then somewhat later, when they no longer
are trying to recall it, the name pops into their head. How can this phenom-
enon be explained by the behaviorist model of how drive multiplies with re-
sponse strength to produce a response? If you do not like its explanation,
how would you explain the phenomenon?
9. Do you think subjects high in anxiety drive would be more likely than sub-
jects low in anxiety drive to forget names, as described in the previous ques-
tion? Why or why not?
10. Why do you think subjects low in anxiety do not do as well on easy tasks
as subjects high in anxiety? How does your answer fit with various theoreti-
cal positions?
11. Do you perform better or worse when someone is watching you? Does it
matter what you are doing? How do your reactions fit with various explana-
tions given in the chapter for audience effects?
12. Why, in your opinion, should subjects high in the need for Affiliation or for
social approval have given more dominant or conventional responses than
those low in this need to highly cued pictures in the Kawamura-Reynolds
experiment (see Table 3.3)? What if subjects high in the need for Achieve-
ment also had given more such responses than those low in the need for
Achievement? What would the implications of such a result be for the At-
kinson versus the Spence explanation of the results of this experiment?
13. Why do you think so much hostility existed between psychology and reli-
gion in the early part of the twentieth Century? How do religious views of
motivation differ from psychological views of motivation?
14. Chapter 1 argued that Freud's cultural heritage influenced his view of
human motivation. Make a case for the hypothesis that Hull's background
also shaped his view of human motivation.
15. One of the main objections to the behaviorist model of motivation was that
it was too simplistic in defining satisfactions and in assuming that people's
behavior was wholly determined by the pleasure principle or reward and
punishment. Gordon Allport (1946), for example, summarizes these objec-
tions with the following example: "What happens in normal human learn-
ing, beyond the infant stage, is that experiences of satisfaction serve as indi-
cators, which, valuable as they are to the individual, are not dynamically
decisive. If I am trying to become a writer and am downcast by a rejection
slip, I may thereupon cease the style of work I was attempting . . . yet I
may be so sure of myself (in my ego-structure) that I will persist in the face
of bad news." Allport is arguing that rewards and punishments do not seem
to control behavior very closely among human adults the way they do
among Thorndike's kittens or Hull's rats. Why not? What does Allport
mean by ego-structure? Could personal motive dispositions of the type dis-
cussed in Chapter 2 be part of the ego structure and explain why individu-
als react differently to bad and good news, as well as why what they do is
less controlled by such external events?
4
Emotions as Indicators
of Natural Incentives
• Early Attempts to Find a Biological Basis for Motives

• Sign Stimuli in Ethology as a Basis for Natural Incentives
• The Case for Natural Incentives in Humans
Sensory Incentives
Opponent Process Theory
Intrinsically Satisfying Activities
• Emotions as Indicators of Natural Incentives
Phylogenetic Basis for Emotions
Emotions as Primary and Universal
Hormonal Patterns and Emotion
Brain Reward Systems and Emotions
• Positive Natural Incentives in Infants

• Classification of Natural Incentives in Terms of the Primary Emotions
• Relation of Emotion to Motivation
107
EARLY ATTEMPTS TO FIND
A BIOLOGICAL BASIS FOR MOTIVES
Motives as we experience them are bewilderingly complex. We want to graduate
from College. We would like to be respected by others. We want to be loved. We
would like to get married. We would like some excitement in our lives. Or per-
haps we would just like to be able to study harder. Where do these motives
come from? Are they instinctive, as McDougall, Freud, and Maslow argued (see
Chapter 2)? Are they simply the product of our learning to satisfy certain bio-
logical needs like hunger, as the behaviorist model argued (see Chapter 3)? Do
we learn to want these things because our society teaches us that we should
want them? Or are there some deeper drives guiding human action that shape
society?
Initially many theorists like McDougall had found it easy simply to reason
that certain fundamental urges were biologically built in, or instinctive. The be-
haviorists, however, were skeptical. They feit such a hypothesis was vague and
impossible to test empirically. How could one prove that a drive for power, for
example, was instinctive rather than built on social learning? Since the behavior-
ists were oriented entirely toward overt behavior (rather than inner "urges"),
they not unreasonably asked what were the fixed action patterns, shown by all
human infants (prior to learning), that would signal the presence of an innate
power need. They recognized that lower animals showed such fixed action pat-
terns, which therefore could be called instincts and could be used to explain the
way birds build nests or cats pounce on mice. However, human beings show lit-
tle evidence of such fixed action patterns. All they can do at birth is sneeze,
vomit, suck, and display a few other innate reflex responses. Where is the drive
for power or for love to be found in such simple innate responses? More com-
plex motives had to be the product of learning, or so these theorists argued.
As Chapter 3 showed, learning theorists found in the so-called primary
drives like hunger and thirst a biological base on which all the more complex
motives could be built through learning. In the simplest version of the theory
they argued that the only innate biological given was strong Stimulation or ten-
sion and the innate satisfaction that comes from reducing it. The idea was ap-
pealing in its great simplicity: The organism started out in life with a few bio-
logical sources of tension such as hunger pangs, pain, and various types of
strong Stimulation, relief from any of which was automatically rewarding. The
organism acquired more complex drives and rewards through association with
these primary determinants of action.
As we have seen, however, this solution to the problem was too simple. It
led to derivations of secondary, more complex motives that were theoretical and
difficult to confirm through actual observation and experiment. Above all, it
failed to account for the fact that organisms often seek Stimulation as well as
Stimulation reduction. Various attempts were made to shore up the Stimulus- or
tension-reduction theory and restore its obvious appeal. For example, Fiske and
Maddi (1961) argued that it was better to think of the organism as seeking cer-
tain optimal levels of activation. If it was below the optimal level at any point,
108
Emotions as Indicators ofNatural Incentives 109
it would seek Stimulation. If it was above the optimal level, it would seek Stimu-
lation reduction. There was nothing wrong with the idea as such, except it was
too general and did not predict with enough precision in advance what Stimula-
tion children or adults would seek in a given Situation.
Tomkins (1962) also tried to expand the notion of Stimulus intensity to
cover what he called "the density of neural firing." In this biologically based
model it was not only the intensity of Stimulation that mattered, but also the
rate of increase or decrease of Stimulation and the number of different sources of
Stimulation. Thus, if the rate of increase was very fast, the person would react
with startle. A somewhat slower rate of increase combined with a lack of recog-
nition of the source of Stimulation would give rise to fear, and a still slower and
less dense increase in Stimulation would give rise to interest-surprise. A sudden
reduction in dense firing would give rise to joy-happiness-pleasure. Thus,
changes in a few aspects of Stimulation as represented by neural firing could
generate quite different emotions, and for Tomkins different emotions were the
basis of different motives.
One difficulty with both these theories is that they deal in terms of a "con-
ceptual nervous System." It has been hard to identify activities in the brain and
nervous System that correspond to what the theories refer to, and therefore they
have been hard to test.
SIGN STIMULI IN ETHOLOGY

AS A BASIS FOR NATURAL INCENTIVES
What directed attention away from such overly general theories was the prog-
ress being made by ethologists like Lorenz (1952) and Tinbergen (1951) in iden-
tifying the sign Stimuli that released instinctive behaviors in lower animals. For
example, it had long been known that male sticklebacks (a kind of fish) often
will attack other male sticklebacks in the springtime. From this it was inferred
that sticklebacks have an aggressive "instinct." However, Tinbergen showed con-
clusively that it was certain sign Stimuli—namely, the red belly of the other
sticklebacks, which appeared only at certain times of the year—that released the
attack response. By making dummies of a fish in various sizes, shapes, and col-
ors, Tinbergen showed it was particularly the red belly that elicited the attack
response. He and other ethologists reported dozens of similar instances in which
particular responses like following, attacking, pecking, and so forth were re-
leased by very specific sign Stimuli.
Note the important shift in emphasis here from what had been the focus of
attention in earlier instinct theory. Then it was the instinctive behavior that was
of primary interest. Now the focus was on perception and on the sign Stimuli
that released the behavior. Although the ethologists paid attention to action pat-
terns displayed, they described more the Variation in the acts preparatory to the
consummatory act. For instance, the sight of the red belly of the male stickle-
back releases aggressive behaviors that vary in speed, direction, and form as the
other male attacks.
Earlier instinct theory had been rejected as inapplicable to humans on the

grounds that humans did not display fixed invariant patterns of behavior, as
lower animals did. Now, however, the question arose as to whether humans
might not be affected by sign Stimuli in more subtle ways, just like lower ani-
mals. That is, human beings may be "turned on" innately by certain sign Stimu-
li, which they experience just the way the male stickleback does when "turned
on" by seeing the red belly of another male stickleback. However, the nature of
the behavior the "turn on" elicits in humans may be less specific or more vari-
able than it is for other species. In fact, it may be something like affective, or
emotional, arousal in humans, which through learning leads to the adoption of
specific behaviors to sustain a positive emotional State or dissipate a negative
one. Also, the nature of the sign Stimulus itself may be much less specific in hu-
mans than in lower animals, being more like a pattern or sequence of Stimuli or
response-produced sensations. In fact, it is easy to imagine human beings seek-
ing out these "turn on" sign Stimuli through a flexible variety of behaviors
rather than through a fixed invariant pattern of responding being released by a
general type of sign Stimulus. For the sake of convenience, let us refer to these
possible "turn on" sign Stimuli as natural incentives and review some of the rea-
sons for thinking they exist.
• THE CASE FOR NATURAL INCENTIVES IN HUMANS

Sensory Incentives
It has long been realized that certain sensations are innately pleasant and
sought, whereas others are unpleasant and avoided. For example, Engel (see
Woodworth, 1938) published curves showing that practically all concentrations
of sweet Solutions are considered pleasant, whereas sour, salty, and bitter solu-
tions are considered pleasant in weak concentrations and unpleasant in strong
concentrations, as shown in Figure 4.1. As the widespread use of perfume testi-
fies, certain smells are pleasant and others unpleasant. In fact, some smells and
tastes are so unpleasant they lead to the facial expression of disgust, which
seems part of a general nausea response preliminary to getting rid of an offend-
ing substance.
Certain types of colors, sounds, and physical contact seem to be innately
pleasing, whereas others are displeasing. For instance, softness and smoothness
are more pleasant than stiffness, roughness, and coarseness (Berlyne, 1967).
Aside from these qualitative differences, psychologists beginning with Wundt
(1874) have long noted that as Stimulus intensity in any modality increases from
low to moderate levels the effect is pleasant, whereas if it increases to still higher
levels the effect is unpleasant. Thus, tones and lights of moderate intensity gen-
erally are sought out over tones and lights of great intensity or very low intensi-
ty. A typical result confirming this curve is shown in Figure 4.2, which was ob-
tained after asking subjects to judge the pleasantness of a spot of light as it
-100%
Stimulus Concentration
Figure 4.1.
Preponderance of "Pleasant" or "Unpleasant" Judgments in Relation to the Concentration of a
Sapid Solution. The Ordinate gives the percentage of "pleasant" minus the percentage of "unpleas-
ant" judgments. The abscissa is proportional to the concentration, the füll length of the baseline
Standing for 40 percent cane sugar, 10 percent salt, and 0.004 percent quinine sulfate, all by weight
(Engel, cited in Woodworth, 1938).
varied in intensity. The subjects observed the spot in the center of a translucent
eggcup fixed to one eye. The illumination of the eggcup total field, or "Ganz-
feld, " was set either fairly low or fairly high, so the eye could be adapted to low
or high illuminations. When the eye was adapted to a low level of illumination,
increases of intensity of the spot were at first judged pleasant and then unpleas-
ant, just as the Wundt curve predicted. However, when the eye was adapted to
a bright light, any intensity of the spot less than this adaptation level was
judged pleasant. Thus, the shape of the hedonic curve to increasing Stimulus in-
tensity depends on the adaptation level of the sense organ, as well as the upper
and lower physiological limits of its sensitivity.
Intense Stimulation can lead to pain sensations, particularly if it involves
sensory end organs in the skin or on internal body surfaces. Thus, extremes of
heat and cold produce painful sensations, as do electric shock to the skin, mus-
cle cramps, or bladder distension. Pain can be regarded as a natural negative in-
centive in the sense that it usually evokes vigorous behavior designed to escape
from the pain. Learning occurs very quickly in response to this incentive. A
child quickly learns that touching lighted electric bulbs leads to pain and avoids
touching them in the future to reduce the anxiety or fear associated with think-
ing about doing it (an anticipatory goal response, in the language of Chapter 3).
Such phenomena led early theorists such as Miller and Dollard (1941) and
54 A
53 : - \ /** J
52
f \
51 V
/ \
50
A
\ /
49 /
48
47
46 /
45 _ "Ganzfeld' AL 2 = 90 Volts (High) \
44
M 43 - "Ganzfeld" /1Z., = 30 Volts
(Low)\
\
5 .1 42
S I 41
| ^ 40 | 1 . 1 t 1 I i V
20 30 40 50 60 70 80 90 100 110 120 130
AL] AL2
Lamp Voltage Values Representing Increasing
Intensity of Spot Illumination
Figure 4.2.
Hedonic Tone Judgments for Discrepancies in Spot Illumination Above and Below Low {ALX) and
High (AL2) "Ganzfeld" Illuminations. A red light was used; ten subjects made four judgments at
each lamp voltage value (Alpert, 1953, cited in McClelland, Atkinson, Clark, & Lowell, 1953).
Mowrer (1950) to conclude that all motives were based on the need for pain
avoidance or tension reduction. No one doubts that some motives are built on
these negative incentives, but there is ample evidence that not all motives de-
pend on them. For example, McMurray (1950) reported the case of a girl who
apparently developed quite normally without ever having feit any pain. How
could she develop into a normal adult if pain reduction is the fundamental basis
of all motivation?
An important recent discovery for motivation theory is that the body con-
tains a natural mechanism for moderating the effects of excessive pain. The
brain can release a number of endogenous opiates or morphinelike substances,
which produce analgesia and sensations of warmth and pleasure. This can be
demonstrated by the use of naloxone, a drug that blocks the action of endoge-
nous opiates. After the administration of naloxone, subjects report much greater
pain from strong Stimuli, indicating that the morphinelike substances are no
longer moderating or counteracting its effects. Thus, it appears that pain not
only acts as a negative incentive, but also releases substances that can serve as
positive incentives, namely, the endogenous opiates (Olson, Olson, Kastin, &
Coy, 1980).
Opponent Process Theory

Richard Solomon (1980) has accumulated evidence showing not only that pain
often is followed by pleasure, but also that pleasure often is followed by pain.
From these observations he has developed a theory that "for some reason the
brains of all mammals are organized to oppose or suppress many types of emo-
tional arousals or hedonic processes, whether they are pleasurable or aversive,
whether they have been generated by a positive or negative reinforcer" (Solo-
mon, 1980). That is, either a positive or negative affective process automatically
sets in motion its Opponent process. In other words, Solomon has identified an-
other source of natural incentives in the way in which affective experiences auto-
matically influence each other.
Table 4.1 illustrates the way these Opponent processes develop over time
after repeated exposure. In the sauna-bathing example at the top of the table,
bathers feel nothing in particular before the first experience, during it they feel
very uncomfortable from the heat, and afterward they feel relief. After repeating
the unpleasant experience, perhaps because they believe it is good for their
health, the picture changes. Bathers find the experience quite exciting, and they
feel exhilarated or very good afterward. It is as if the repeated experience of
heat in time more dependably elicits the endogenous opiates that give a strong
feeling of being "high" afterward, much like the feeling an injection of morphine
would produce. Many have suggested that a rather painful experience like jog-
ging also produces a high after a time, perhaps also because of the release of en-
dogenous opiates.
Solomon's Opponent process theory also works the other way. As shown by
the example at the bottom of Table 4.1, a duckling quickly gets imprinted on its
Table 4.1.
CHANGES IN AFFECT BEFORE, DURING, AND AFTER EACH STIMULATION
FOR THE FIRST FEW AND AFTER MANY EXPERIENCES (after Solomon, 1980)
Period Affect During First Few Affect After Many
Sauna bathing
Before Resting State Resting State
During Pain, burning Hot, exciting
After Relief Exhilaration
Resting State Resting State
Social attachment in ducklings
Before Contentment Some distress
During Excitement Following
After Distress Intense distress
Contentment Some distress
mother or a moving mother Surrogate soon after birth. In the duckling's first
contacts with the mother Surrogate, it shows no Special emotion before contact,
some excitement during contact, and some distress followed by a return to
its normal State after contact. After a number of contacts, however, the duck-
ling shows some distress beforehand in the absence of the mother Surrogate,
following behavior when the sign Stimulus is present, and intense distress after
the sign Stimulus has disappeared. In other words, the positive affect associated
with contact and following the mother Surrogate builds up its Opponent process
in time—namely, strong negative affect, which occurs when the Stimulus
disappears. That negative affect, however, was not present before the experience
of positive attachment. As Solomon (1980) points out, "this new motivational
problem would never arise for the ducklings if they never saw a mother
Surrogate."
The Opponent process theory can be used to explain many unusual motiva-
tional phenomena, such as drug addiction. Why do people get addicted to a
drug like heroin, which produces a strong positive "rush" after several experi-
ences? According to the theory, the repeated positive experiences build up the
negative Opponent process so strongly that the intense distress following with-
drawal of the drug forces the person, like the duckling searching for the mother
Surrogate, to seek the drug that will remove the distress—and also cause an
even greater distress after its removal in an intensifying vicious circle. Solomon's
theory also explains in the same way the intense distress of lovers who are sepa-
rated from each other. Their many positive experiences together have built up a
strong negative Opponent process, which erupts as soon as they are no longer in
each other's presence.
Why do people engage in sport parachute jumping, which produces strong
anxiety in the first few jumps? According to Solomon's theory, the anxiety auto-
matically builds up a positive Opponent process, which gives the jumper an in-
creasingly positive rush during and after the jump. Solomon's theory provides a
ready explanation for some phenomena that are otherwise difficult to explain.
For example, it accounts for the fascination that the thrills and chills of amuse-
ment parks have for children and even adults. People are purposely undergoing
frightening experiences they ought to avoid, but in fact undertake because they
feel exhilarated afterward. This exhilaration is far more than just the tension re-
lief suggested by earlier drive theory. It represents positive pleasure, which
seems to come automatically from the frightening affective experience.
Solomon's theory is too new to have been thoroughly checked. It remains to
be seen whether all positive or negative affective experiences automatically pro-
duce their opposites afterward. One difficulty with the theory is that whereas
Solomon has produced curves showing how the Opponent processes theoretically
affect each other, it is as yet not clear what brain or other physiological mecha-
nisms underlie the curves. Once these mechanisms have been identified, it will
be possible to check whether these curves have an empirical basis. The one pos-
sible physiological mechanism involved, which has already been mentioned, is
that in time negative affect, or pain, increases the release of endogenous Opiates,
which produce pleasure. However, this would occur whether the pain were re-
moved or not, whereas the Opponent process theory indicates the greatest plea-
sure occurs after the removal of the negative affect. No similar mechanism has
been suggested for explaining the way in which positive experience might auto-
matically release or build up some physiological substance—perhaps another
neurohormone—associated with negative affect.
Whatever the final judgment on Solomon's theory, it calls attention dramati-
cally to the fact that there appear to be strong positive and negative natural in-
centives that influence behavior in important ways.
Intrinsically Satisfying Activities

There is another, very simple basis for believing in natural incentives. Psycholo-
gists have long noted that animals and human beings often do things for "intrin-
sic" reasons. Monkeys like to spend their time looking through a window at a
monkey in another cage, at a toy train, or even at another part of the laboratory
(Butler, 1954). Children spend an enormous amount of time just "playing." Deci
(1975) has accumulated much evidence showing that offering people extrinsic re-
wards such as food or money for doing these intrinsically interesting things
often decreases their tendency to do them. He had subjects spend time putting
together a three-dimensional spatial relations puzzle (much like Rubik's cube)
into various configurations. The subjects enjoyed doing it and, in fact, would re-
turn to playing with the puzzle when they had free time. However, if he paid
them a dollar for each puzzle correctly solved, they would not return to the ac-
tivity when they were free to do what they liked and were no longer being paid.
Lest we infer that the subjects reasoned there was no point in doing the
puzzles if they were no longer going to be paid, it is worth remembering that
Harlow, Harlow, and Meyer (1950) reported exactly parallel results for monkeys
in a study done a generation earlier. Monkeys, like people, enjoy manipulating
things. But when Harlow's monkeys were given a food reward for correctly put-
ting together a mechanical puzzle, they abandoned playing with the puzzle on
their own. Extrinsic rewards seem to interfere with intrinsically satisfying activi-
ty. Thus, it is hard to argue that liking to manipulate things is somehow based
on the satisfaction of primary drives like hunger. Yet these studies do not make
much progress toward identifying what experiences are intrinsically satisfying.
They simply call attention to the fact that some undoubtedly are.
Language learning is another vivid example of what appears to be an intrin-
sically satisfying activity guided in some way by natural incentives. R. Brown
(1973) noted that while it is clear all children learn to talk in a way that ap-
proximates what they hear older people around them saying, no one really
knows why they improve their Speech. At first it was assumed that children's
speech was shaped by rewards and punishments from parents or others. If chil-
dren speak incorrectly, they will not be understood, and therefore they will cor-
rect their speech so they can be understood. Or if they speak incorrectly their
parents will correct them. This is a typical "application" of the behaviorist
model of motivation. However, actual observational studies show that parents,
in fact, do not correct grammatical mistakes very often and that children are
understood perfectly well even when they speak without the use of such words
as the article the. Almost no one bothers to correct children if they do not use
the, as in get ball. However, all children in time learn to use the if adults
around them use it. The inference seems inescapable that it is not selection pres-
sure from the environment that improves the speech of children, but some as yet
unidentified "natural incentives" or "turn ons" that are intrinsically satisfying
when the children match their response to the one they hear.
Csikszentmihalyi (1975) has studied in some depth what he calls autotelic
activities—things that are enjoyable in themselves. In his words, "in a world
supposedly ruled by the pursuit of money, power, prestige, and pleasure it is
surprising to find certain people who sacrifice all those goals for no apparent
reason: people who risk their lives climbing rocks, who devote their lives to art,
who spend their energies playing chess" (Csikszentmihalyi, 1975). He extensively
interviewed and observed people who devoted a good deal of time to rock climb-
ing, dancing, playing chess, and playing basketball and found that it did not
seem reasonable to conclude that they spent their time that way for ulterior rea-
sons related to primary drives or basic personality needs. Instead, they found as-
pects of the activities themselves enjoyable.
Even more persuasive is Csikszentmihalyi's study of "microflow" activities,
or little things people seem to enjoy doing when they are not doing anything in
particular. He found that people spend much time humming, whistling, or sing-
ing to themselves; watching people or things; chewing; walking; touching, rub-
bing, or fiddling with objects; playing a musical instrument; doodling; or simply
browsing, Shopping, or joking with others. While some of these activities some
of the time undoubtedly are in the Service of major motives, much of the time
they seem to be "incentive driven" in the sense that some experiences attained
this way are innately satisfying. Csikszentmihalyi does not make much progress
toward identifying what these natural incentives are, but his work points
strongly to the conclusion that such natural incentives do exist.
• EMOTIONS AS INDICATORS OF
NATURAL INCENTIVES
Phylogenetic Basis for Emotions
In the course of evolution, the human brain evolved slowly from lower forms.
In particular, the human forebrain, or neocortex, underwent great expansion as
man evolved out of lower primates and mammals. This is the part of the brain
associated with thinking, reasoning, and imagining. However, the older, more
primitive parts of the brain still exist encapsulated, so to speak, in the cortex.
Paul MacLean refers to what he calls the triune brain, because there are three
brains in one in humans—namely, the oldest, or reptilian, brain; the next oldest,
or paleomammalian brain; and the large, late-developing neomammalian brain,
which includes the neocortex (see Figure 4.3). The limbic System, or early mam-
malian brain, mediates the affective states, or emotional experiences of desire,
anger, fear, sorrow, joy, and affection (MacLean, 1975).
Emotions as Indicators of Natural lncentives 117
Figure 4.3.
Evolutionary Human Brain Expansion in Hierarchie Fashion in Three Basic Patterns: Reptilian,
Paleomammalian, and Neomammalian (MacLean, 1975).
Therefore, it makes sense to think of our affective experiences as represent-

ing a more primitive level of brain funetion that can be modified and modulated
enormously by cognitive events taking place in the cortex of the brain. In lower
animals with lesser cortical development, speeifie sign Stimuli release speeifie be-
haviors via the older brain centers. In humans, releasing Stimuli are more vari-
able and more modified by the cortex; however, the affective core—to use a
phrase employed by McDougall long ago (1908)—remains a key to understand-
ing what human beings find satisfying or not. In short, emotions, as well as mo-
tives built on them, appear to be mediated by a different and older part of the
brain than are cognitions or associations.
Unfortunately, at present it is not possible to be much more speeifie about
how this happens. One possibility is that there are certain natural incentives (or
sign Stimuli) in humans that give rise to affective states in the older brain sys-
tem; these in turn get connected with speeifie actions but do not lead automati-
cally to them, as older instinet theory contended. Another possibility is that the
natural incentives themselves must be defined in much more general and chang-
ing terms than is common when describing lower animals. We will return to
this matter later.
Emotions as Primary and Universal

Emotions are more primitive than cognitions in the sense that they are mediated
by the older portions of the brain. They are universal—everyone experiences
emotions—inescapable, and at times very intense. They act like amplifiers, as
Tomkins (1962) has pointed out. Certain body requirements, such as the need
for oxygen, are normally met easily without the involvement of affect or emo-
tion: the accumulation of carbon dioxide in the lungs automatically leads a per-
son to breathe more rapidly and take in more oxygen, thus bringing the body
into homeostatic balance, much as HulPs model of a mechanical robot assumed.
However, the effect of hunger is different: It does not automatically lead to the
eating response that relieves the hunger. Furthermore, if an infant is not fed, the
hunger cues lead to widespread signs of emotional upset and distress, which am-
plify the effect of the hunger per se. Evidence for this conclusion is readily ob-
tained simply by picking up and holding a crying baby, which can reduce the
emotional distress without reducing the hunger.
In Tomkins' view, affect promotes motivation in several ways. It amplifies
biological signals, but even in the absence of such signals (1) affect increases the
variety of the consummatory responses that will serve to reduce or increase mo-
tivation, and (2) affect engages the learning capacity of the organism in finding
ways of reducing or increasing motivation. That is, breathing is the automatic,
unlearned consummatory response to the lack of oxygen: no other response will
do, and no learning is necessary to teach the organism to make the response.
However, when affect is aroused—whether it be distress or happiness—the or-
ganism is energized to reduce or enhance it in various ways at various times and
places, and its learning capacity will be engaged in finding these instrumental
activities.
Whereas for a time psychologists overlooked the primary, compelling aspect
of affect in their concern for cognitive variables like memory, expectations, and
attitudes, Zajonc (1980) recently summarized evidence that affective judgments
like pleasant-unpleasant or good-bad are more "primary" than cognitive judg-
ments like seen before-not seen before. Affective judgments are more effortless,
inescapable, immediate, irrevocable, and difficult to verbalize, although they are
easy to communicate and understand. It is as if the older part of the brain is re-
sponding more quickly or immediately before the newer, cognitively oriented
part of the brain has time to function. For example, under some conditions peo-
ple say they like Stimuli better to which they have been exposed previously. (See
the discussion in Chapter 5 on the variety incentive.) That is, they like "old"
Stimuli better than "new" ones. Zajonc (1980) reports a number of studies that
show old Stimuli are liked better than new ones even when they are not yet rec-
ognized as having been seen before. Recognition is less sensitive to previous rep-
etitions than affect is. Thus, the affective response of like-dislike precedes and is
more "primary" than the cognitive recognition response of seen before-not seen
before.
The universality and "demanding character" of emotions suggest they may
be innate responses to certain natural incentives or sign Stimuli. Unfortunately,
not much progress has been made in defining just what "elicitors" give rise to
different emotions. Some feel there are unlearned elicitors (Eibl-Eibesfeldt, 1970),
such as the sign Stimuli in lower animals; others have argued that sign Stimuli
are not as specific in humans as in lower animals and that social learning pro-
duces the main elicitors for various emotions (Ekman, 1972; Tomkins, 1962).
This summary by Klinger (1977) is typical: "The emotions are 'primary' in the
sense that they are unlearned reactions to certain basic kinds of situations. For
instance, anger may be an unlearned reaction to interference with a goal-
directed sequence of behavior. However, people learn a wide variety of new situ-
ations are equivalent to the basic situations for which the primary emotional re-
actions are programmed. They then respond to these new situations as well." A
possible sign Stimulus or natural incentive is mentioned ("interference with goal-
directed sequence of behavior"), along with the indication that it can change
with experience. This is about as far as theorists have been able to go in identi-
fying precisely what the natural incentives or sign Stimuli are for different types
of emotions.
However, extensive studies of the facial expression of emotions suggest there
are a limited number of basic, or primary, emotions, at least as reflected in fa-
cial expressions, and that each is uniquely associated with a particular set of
evoking situations (or sign Stimuli). The evidence for these conclusions comes
from a series of photographic studies of facial expressions of infants in the
United States and other cultures (Ekman, 1972; Izard, 1979). While experts do
not completely agree, there is a general consensus that at least six difFerent emo-
tions can be distinguished in all babies from all cultures, indicating that the
emotions are innate. They are joy-happiness-pleasure, sadness-distress, anger-
excitement, disgust, fear, and interest-surprise (see Figure 4.4). Furthermore,
these facial expressions are readily recognized by members of a culture very dif-
ferent from the one to which the person expressing the emotion belongs. People
in the United States can recognize a happy Japanese or New Guinean as easily
as they can recognize a happy American. In short, the language of the emotions
is universal and easily understood. This further Supports the conclusion that dif-
ferent emotions are expressing the innate response of all human beings to certain
sign Stimuli, which are "turning on" the organism in different ways. We will re-
turn to trying to identify these sign Stimuli later.
Hormonal Patterns and Emotion

Different emotions are not only accompanied by distinctive facial expressions;
there is increasing evidence that they also are associated with different hormonal
profiles. Initially attention was focused on the general physiological excitement
that appeared to accompany all emotions. It was argued that emotion aroused
the organism, readying it for "fight or flight" (Cannon, 1915). Heart rate, blood
pressure, and muscle tension all increase with emotional excitement such as fear
or anger-excitement. Selye (1956) popularized the notion of a general adaptation
syndrome, which included a number of physiological responses that appeared to
occur in all types of emotionally arousing situations, or Stresses.
The brain operates through three somewhat different types of effector Sys-
tems: the skeletomuscular nervous System, the autonomic nervous System, and
the endocrine System. Some of the more obvious general characteristics of all
Figure 4.4.
Two Readily Recognizable Facial Expressions of Emotions in Infants. (Left) The expression of
anger-excitement is shown on the face of a fourteen-month-old girl. (Right) The expression of
interest-surprise can be seen on the face of a four-month-old girl (Izard, 1979).
emotions are mediated through the first two of these Systems. Thus, any type of
emotion-arousing Stimulus—whether an electric shock or a possible sexual part-
ner—produces at first some startle reactions, mediated through the skeleto-
muscular System, and some physiological change, such as increased heart rate or
sweating of the palms, initiated by the autonomic nervous System.
It has not proved possible to differentiate clearly what emotion—and there-
fore what natural incentive—is involved on the basis of these reactions alone
(Lacey, 1967). However, careful work by Mason (1975) and his collaborators
has suggested that it may be possible to detect different emotional states from
their profiles of endocrine response, just as it is possible to do so from differ-
ences in facial expressions. Most of the work has been done on fear and anger-
excitement, because they are relatively easy to arouse and study in the laborato-
ry. Whereas both fear and anger-excitement are accompanied by the release of
Cortisol from the cortex of the adrenal gland and adrenaline (or epinephrine)
from the medulla of the adrenal gland, Mason has found some evidence that
anger-excitement also is associated with an increase, and fear with a decrease,
in sex hormone release. In his words, there are "preliminary and limited indica-
tions that different, relatively specific emotional states may be correlated with
different, specific patterns of multiple hormonal responses" (Mason, 1975). If

this is true, there is a physiological basis for the qualitative differences in feel-
ings and facial expressions associated with different types of emotional arousal,
for hormones are closely associated with feeling states. Thus, in thinking about
how natural incentives might affect the body, we need not be limited to concep-
tions of general arousal but may conceive of natural incentives leading to quali-
tatively different emotions mediated by different hormonal patterns.
Brain Reward Systems and Emotions

A quite different line of investigation also suggests that there may be different
types of affective arousal, or emotion, mediated by the brain. Olds and Milner
(1954) first demonstrated that rats would press a bar to get electrical Stimulation
in some parts of the brain but not in others. Ever since, investigators have been
pinning down more and more precisely just where in the brain electrical Stimu-
lation is rewarding and where it is not (Olds, 1977). Attention first focused
on the neurons for which the catecholamines serve as neurotransmitters
across the synapse from one neuron to the next. It is known that the nerve im-
pulse travels in the nervous System via the release of tiny biochemical packets
released from one neuron and picked up by receptors on the next, as well as
that the biochemicals involved in different neuronal Systems are different. It
was first thought that the neurons served by a catecholamine transmitter,
noradrenaline (a precursor of adrenaline), were the ones that, when electrically
stimulated, would lead to the most active self-stimulation by rats (see Stein,
1975).
Later evidence strongly suggests that this effect, if it occurs, is indirect, and
that electrical Stimulation of another set of neurons is much more closely associ-
ated with self-stimulation, for which the neurotransmitter is dopamine, another
catecholamine (Wise, 1980). Some investigators believe the dopamine circuits
may mediate the reward System for the brain, but there is considerable evidence
for independent sources of self-stimulation. The strongest such evidence comes
from studies showing that opiates like morphine enhance self-stimulation
(Adams, Lorens, & Mitchell, 1972). The brain also produces its own endogenous
opiates, the enkephalins and endorphins, which could serve as rewards if re-
leased by emotional arousal of various types. Furthermore, there is some evi-
dence that self-stimulation occurs in the vicinity of neurons where the neuro-
transmitter is serotonin (Phillips, Carter, & Fibiger, 1976), as well as in the
prefrontal area, where the neural Substrate has not yet been identified (Corbett,
La Fernere, & Miliver, 1982).
The assumption underlying these results is that the artificial introduction of
electrical Stimulation in different parts of the brain mimics the natural electrical
Stimulation that would occur from various types of psychological arousal, as in
various emotional states. It would be useful to know whether various types of
what we have been calling natural incentives give rise to different biochemical
neurotransmitters that have been found to be rewarding. But the most that can
be said at this point is that the existence of different types of reward Systems in
the brain provides a biological basis for a variety of types of incentives that
might release the neurohormones found to be rewarding.
POSITIVE NATURAL INCENTIVES IN INFANTS
There is much evidence that natural incentives guide human behavior, that they
involve affective arousal, and that a limited number of different types of affective
arousal (or emotions) may be elicited by specific natural incentives or sign Stim-
uli. Affective arousal may be thought of as positive or negative in the sense that
it either facilitates or inhibits activity intrinsically. What is needed now is more
information, particularly on the positive natural incentives, since the negative
ones have been more fully studied in the behaviorist tradition. Unfortunately,
such information is hard to come by, particularly because it is difficult to distin-
guish what is natural from what is learned, since humans begin learning even
before birth, in the womb.
Let us observe a one-year-old boy, Peter. One evening he was left free to
crawl about on the floor among a group of adults who were seated on the floor
and singing. Peter was very happy and active. He crawled everywhere, exploring
new areas, pushing objects over, grasping other objects, poking a candle, sound-
ing an unused musical instrument. In fact, the adults were understandably ner-
vous from time to time that he would break something, push over a lighted can-
dle and Start a fire, or hurt himself as he pushed his way around the environ-
ment. To generalize about what all this activity had in common, it might be
said Peter was seeking to have impacts on the environment.
At one point in Peter's exploration, he came across a threshold into the
next room that was raised about two or three inches off the floor. He struggled
to get over it and managed to do so with some difficulty. As soon as he had
successfully negotiated the barrier to the other side he turned back and smiled,
reversed his direction, and proceeded to climb back over the obstacles. He re-
peated this process six or seven times, smiling after each success. In this case, he
seemed to be attracted by some kind of challenge. At the simplest level he
seemed to be intrigued by doing something slightly different from what he had
been doing all along, namely, crawling unimpeded in all directions. We shall call
this the variety incentive, although much of the work on it has focused on a
search for variety and new sensations rather than on new types of actions, as in
this example. Also, from time to time Peter would crawl to his mother and cud-
dle up against her, snuggling into a pillow at his mother's side. We can think of
such behavior as released by contact incentives.
At a somewhat more systematic level, Cicchetti and Sroufe (1976) investi-
gated the effectiveness of thirty different types of Stimulation in producing smil-
ing in young infants. The smiling response represents a sign of positive affect
that was not differentiated further into the distinctions among different types of
positive affect made by students of facial expressions in infants. Among the most
effective situations for producing a smile in the last five months of the first year
of life were the following:
Situation Percentage of Babies Smiling
Allow the baby to grasp yarn, then 43% at 9 months

tug it three times, trying not to pull it
away from the infant. Pause and
repeat.
Put a cloth in your mouth, and lean 50% at 8 months

close enough for the baby to grasp it.
Allow the baby to pull the cloth out
and replace it if this is his or her
tendency.
These would seem to be examples of the impact incentive: the baby is getting
pleasure from grasping or pulling.
Next, consider these examples:

Say aah, starting low, then crescendo- 55% at 11 months
ing to a loud voice, with an abrupt
cutoff, then a six-second pause.
Lift the baby slowly to a position 50% at 10 months
overhead, with the baby looking down
back.
Place the baby in a high chair or in- 60% at 12 months

fant seat. Crawl across the baby's field
of vision, not toward the baby. Stand,
return to starting point.
Obtain the baby's attention. Hold a 80% at 10 months

human mask up so the baby can see
it. Place the mask in front of your
face, lean slowly to within one foot of
the baby's face, and pause two sec-
onds. Lean back slowly and remove
the mask slowly.
What these situations have in common is a change in what the baby expects. In
the case of the sound, an expectation is built up from the crescendo, which is
abruptly changed by the cutoff. Similarly, when the baby sees the adult crawling
rather than Standing, or with a slightly different face than what the baby is used
to, the changes in what is expected produce smiling. These seem to be clear ex-
amples of the variety incentive.
Finally, consider these examples:

Give four quick pecks on the baby's 50% at 8 months
bare stomach.
Say lyrically, ' T m gonna get you" 75% at 8 months
("Fm" quite protracted), while leaning
toward the baby with your hands
posed to grab. Then grab the baby
around the stomach.
Focus the baby's attention on your fin- 60% at 10 months
gers. Walk your fingers toward the
baby; then give the baby a poke in the
ribs.
These would seem to be clear examples of contact incentives that involve tactile
and kinesthetic Stimulation. The ' T m gonna get you" procedure also involves
the variety incentive, since some surprise is involved. In contrast, a number of
procedures used by Cicchetti and Sroufe (1976) evoked smiling only rarely if at
all from the babies, including blowing gently at the babies' hair for three sec-
onds, whispering "Hi, baby, how are you?" in the babies' ear, or letting the ba-
bies observe themselves in a mirror. These procedures do not involve any of the
three major incentives just identifed. Cicchetti and Sroufe did not attempt to
generalize about what kinds of procedures are most efFective in eliciting smiling,
since their interest was primarily in the efFectiveness of various Stimuli in pro-
ducing smiles as a function of the infants' cognitive development. Chapter 5 will
discuss this further, but in general the observations of Cicchetti and Sroufe sup-
port the inference that there are at least three general types of Stimulation that
may serve as positive incentives for human infants.
• CLASSIFICATION OF NATURAL INCENTIVES

IN TERMS OF THE PRIMARY EMOTIONS
Several clues about the nature of some of the primary natural incentives have
been scattered throughout the chapter. The work on emotions suggested that
there might be a limited number of incentives, each primarily associated with
one of the six different emotions that have been reliably difFerentiated in facial
expressions. Observations of children either learning a language or responding to
Stimulation have suggested what some of the natural incentives might be. Table
4.2 attempts a preliminary classification that fits these bits of information togeth-
er. The six primary emotions, as identified from facial expressions, are listed on
the left side of the table and classified as positive or negative. Opposite each
emotion are listed the subjective feeling states that accompany it and a possible
natural incentive that gives rise to it. In some cases the fit seems quite good.
For example, the evidence for the existence of a positive natural incentive for
moderate variety in Stimulation is excellent, as Chapter 5 will show, and the fa-
cial expression of interest-surprise is equally well documented and associated
with changes in the variety incentive.
The nature of the incentive or sign Stimulus for anger-excitement is less
clearly defined, but Chapter 5 will review evidence pointing to the importance of
an impact incentive—a pleasure that derives from having impact on the world at
first physically and later psychologically. This incentive is accentuated not by in-
terference with any goal-directed sequence, as Klinger (1977) has suggested, but
by interference only with an impact or manipulation attempt. This leads auto-
matically to an intensification of effort, which becomes recognizable as the emo-
tion of anger-excitement. This emotion in time may develop into aggression;
however, aggression is not itself a primary emotion, but a behavioral trait
strongly influenced by learning and all sorts of cognitive variables (for example,
thoughts about injuring or the intent to injure). Anger-excitement is considered
a positive emotion because it is associated with approach, although its social ef-
fects often are regarded negatively.
Table 4.2.
CLASSIFICATION OF PRIMARY EMOTIONS AND NATURAL INCENTIVES
THAT RELEASE THEM (after Ekman, 1971)
Primary Emotion Associated Natural Incentive

(Facial Expression) Subjective States or Moods Characteristics
Positive emotions
Interest-surprise Feeling curious or Variety
exploratory
Anger-excitement Feeling strong, excited, seif Having impact
as causal agent
Joy-happiness-pleasure Feeling loved, loving, Contact
peaceful, happy
Negative emotions
Fear Feeling fearful or anxious Pain
Disgust Feeling repelled or rejecting Naturally unpleasant
sensations (e.g., bitter tastes
and bad smells)
Sadness-distress Feeling unhappy or Major lack of consistency
grieving over loss with expectation
In Table 4.2, the emotion of joy-happiness-pleasure has been linked primar-

ily to the contact sensations, which relate to sexuality, although obviously some
kind of pleasure can derive from any of the positive incentives, as the research
on infant smiling demonstrates. The basis for associating joy-happiness-pleasure
primarily with the contact incentive lies not only in the desire to have a neat
classification, but also in research that suggests this connection may exist as pos-
tulated. Underwood, Moore, and Rosenhan (1972) asked some children to think
of things that made them happy and other children to think of things that made
them sad. Afterward the children, who were tested individually, were told they
could help themselves to some pennies available in a Container nearby, but that
they could also leave some if they desired for other children who could not par-
ticipate in the experiment. The children who had been thinking happy thoughts
left more pennies for the other children than the children who had been think-
ing sad thoughts or who had done no Special emotional thinking. In other
words, the emotion of joy-happiness-pleasure was directly connected with a con-
cern for others, which we are supposing develops out of the contact incentives.
To further test whether this is so would require seeing if children who were
made happy by introducing mild variety also would be more altruistic, or
whether those who had been thinking happy thoughts would do other things
more, such as hit a ball harder, which Table 4.2 would associate with the im-
pact incentive.
The negative emotions fit less well with associated natural incentives, al-
though the connection of fear to the negative incentive from pain sensations is
clear enough. The sensations giving rise to disgust are so specific it is hard to
see how they can play much of a role in developing more complex motivational
patterns. Naturally, pleasant and unpleasant sensations provide a basis for pref-
erences in taste and smell, but at present no one has argued that they enter into
the development of important social motives.
The poorest fit is between the emotion of sadness-distress and what Chapter
5 will call the consistency incentive. The problem derives in part from the fact
that the emotion occurs in the absence of consistency rather than in its pres-
ence. The variety incentive places a premium on moderate degrees of novelty,
but it is equally clear that there is a very powerful incentive for things to match
up to expectation. The consistency incentive underlies language learning when
children repeat the dog rather than just dog because they have heard the dog all
the time. That is, when children say dog, they realize the inconsistency with
what they have heard, which arouses distress; this, in turn, leads children to
correct the mistake and say the dog.
The inconsistency incentive underlies the search for reduction of uncertainty
and confusion—the need to produce order and predictability in a constantly
changing world. The drive-reduction theorists were right in arguing that there is
an incentive to reduce tension, conflict, and uncertainty. However, they were
wrong in arguing it was the only incentive. The emotional sign that consistency
with expectation has not occurred is sadness-distress, as reflected in facial ex-
pression, according to this line of reasoning. It is also possible to think of Solo-
mon's negative Opponent process as a major source of the inconsistency incen-
tive, in which a strong positive affect no longer occurs after a person has be-
come accustomed to it. This type of inconsistency is particularly likely to give
rise to distress.
The classification in Table 4.2 must be regarded as a very preliminary at-
tempt to suggest how some order might be introduced into a field characterized
by considerable confusion. It can be questioned on many grounds. For example,
not everyone agrees perfectly on which emotions are primary. Plutchik (1980)
extends the list to eight primary emotions by separating interest from surprise
and adding acceptance, which in his theory is associated with affiliating, whereas
joy-happiness-pleasure is associated with cooperating and sexuality. Plutchik
may turn out to be right, but we have followed the simpler classification primar-
ily because acceptance does not lead to an identifiable facial expression and
because interest and surprise appear to be on a continuum so far as facial ex-
pressions are conceived. Our analysis places greater weight on infant facial ex-
pressions than Plutchik's does, because we are searching for natural—that is,
universal and innate—incentive characteristics. Thus, the fact that certain facial
expressions are universal in infants and readily recognizable leads us to believe
that they may be cued off by sign Stimuli or incentives the observer can readily
identify. Thus, we have not included additional emotions mentioned by Plutchik
(1980) or Izard (1979) because of doubts about the universal recognizability of
the facial expressions of these emotions in infants. There obviously are many
more emotions than those listed that result from blends of natural incentives or
are the product of later learning.
What is more, a very respected tradition in psychology is based on an influ-
ential experiment by Schachter and Singer (1962), which argues that there are
no primary, unlearned emotions. Rather, there is only an unlearned State of af-
fective physiological arousal, which is converted into various emotions by the
subject's cognitive understanding of the Situation. Chapter 12 examines this
work at length and finds it does not support the conclusion that there are no
fundamentally different affective states.
Chapter 5 will attempt to describe and define more carefully some of the
natural incentives listed in Table 4.2. One general point about them needs to be
made now. This chapter has used the term sign Stimuli throughout, which is
adopted from the ethologists to describe the external events that release innate
affective responses or emotions. This describes fairly well what elicits the variety
incentive: moderate environmental changes in what is expected elicits interest-
surprise. In describing the baby Peter's behavior, however, we noted that some
of this Variation was response produced: It was his climbing over the threshold
and the perceived novelty of what he had done that surprised and pleased him.
He was not a passive responder to Stimulus change; he produced it.
This is even truer of the impact incentive, for which it is hard to identify a
sign Stimulus that is not response produced. Peter enjoyed the sensations he ex-
perienced when he banged, grasped, or pushed something. Perhaps, as has been
noted, interference with grasping or pushing serves as an additional sign Stimu-
lus to evoke more attempts to have impact, but it is the sensations from having
impact that are sought as the primary source of pleasure.
• RELATION OF EMOTION TO MOTIVATION
Some theorists (Izard, 1979; Tomkins, 1962) consider emotions to be motiva-

tions in the sense that they influence behavior. As we pointed out in Chapter 3,
however, the term motivation is used by many people as being equivalent to de-
termination. That is, emotions—like habits, expectations, and so on—can be
shown to determine or influence behavior. But in the restricted sense in which
the term is used in this book, emotions do not motivate behavior: only motives
motivate behavior. Emotions are not motives, but they are an important part of
motivational Systems: they indicate the presence of natural incentives. Further-
more, motives presumably are built on natural incentives, as the next chapter
will attempt to make clear. Emotions also accompany motives and amplify their
effects on behavior. For example, they intensify the reactions to success or fail-
ure in satisfying a motive. They provide the affective "charge" that makes moti-
vational Systems so powerful and persistent in shaping behavior. They involve a
more primitive level of brain functioning that continues to influence what people
say and do often without their knowing what is going on, that is, without ade-
quate cognitive representation in the later developing cortical association areas
of the brain.
NOTES AND QUERIES

1. Sign Stimuli in animals tend to be quite specific and differ from one species
to another. For example, whereas it is the sight of the red belly of another
stickleback that releases an attack response in the male stickleback, it is the
smell of urine of stränge males that releases intermale aggression in mice
(MacKintosh & Grant, 1966). Several species of animals attack intruders in
their territory (Lorenz, 1966). It has often been inferred that humans, like
other animals, have an "aggressive instinct," but the question always has
been, Like which other animals? Are we most like lions, mice, or deer?
What are the sign Stimuli that "turn on" the aggressive response in hu-
mans? What kinds of sensations, perceptions, or situations are most likely to
elicit anger-excitement in you? How would you determine if such "turn
ons" were natural (part of the biological inheritance) rather than learned?
2. Solomon's Opponent process theory explains why many people undergo
painful experiences, like a sauna bath, to get the pleasure it automatically
induces. Can you think of any painful experiences in your own life that
have not been followed by pleasure? What are the implications of such in-
stances for the theory?
3. It often has been observed that people learn languages more easily when
they are young. Can you think of an explanation for this fact that would in-
volve the Operation of a natural incentive? Why might it be easier for people
to get pleasure from producing the sounds they hear in childhood than to
do so later on?
Emotions as Indicators of Natural lncentives 129
4. How would you define the natural incentive involved in chewing gum? How
could you determine whether it is "natural" or acquired? If you consider it
a subtype of a more general natural incentive, can you think of other things
gum chewers also would be likely to do or enjoy?
5. Why does the discovery of "pleasure centers" in the brain suggest the pres-
ence of natural incentives? Would you expect to find the same pleasure cen-
ters in humans that have been found in rats? Why or why not?
6. If the procedures initiated by Cicchetti and Sroufe represent sign Stimuli
that release pleasure in human infants, how would you explain the fact that
in no instance did all the infants respond with a smile to a particular
procedure?
7. Can you think of any natural incentives that have been left out in Table
4.2? What about Sensation pleasures—for example, from sights, sounds, or
tastes? Can you think how possible natural incentives involved might lead to
the development of motives that shape the lives of musicians, artists, or
gourmets?
8. Izard (1979) has pointed out that a greater number of different emotions
can be distinguished in higher than in lower forms of life (that is, say, in
human beings in comparison with Salamanders or even dogs and cats). It is
the face in particular in humans that reflects many different emotions. Izard
argues that this is adaptive, because emotions help the organism adapt and
survive, and the more differentiated the emotions are, the more flexibly the
organism can adapt. If the motives are built on the natural incentives that
elicit different emotions, as this chapter suggests, does this mean there will
be a greater number of different types of motives in humans than lower ani-
mals? Does the fact that lower animals have more built-in responses to spe-
cific sign Stimuli than humans do suggest that, among human beings, mo-
tives may be a more flexible way of incorporating natural sign Stimuli into
means of adapting successfully?
5
Natural Incentives and
Their Derivatives
• How Natural Incentives Influence the Development of the Hunger Motive

• Natural Incentives, Emotions, and Motives
• The Variety Incentive
Individual Differences in Seeking Variety
The Variety Incentive and the Achievement Motive
• The Impact Incentive
Distinction Between Impact-Anger and Aggression
Brain Reward and the Impact Incentive
Impact Incentive Derivatives
• The Contact or Sexual Incentives
Imprinting and Attachment
Sexual Arousal from a Distance
Derivatives of the Contact Incentives: Succorance and Nurturance
• The Consistency Incentive
Cognitive Dissonance
Motives Related to the Consistency Incentive
• Interaction of Incentives
• The Role of Cognition in the Development of Incentives
• Symbolic Incentives, or Values
130
• HOW NATURAL INCENTIVES INFLUENCE THE
DEVELOPMENT OF THE HUNGER MOTIVE
Much has been learned about the factors that affect eating in animals and hu-
mans. The purpose of this chapter is not to review all this knowledge, but to
learn from it as much as we can about the way motives develop out of sign
Stimuli and the behavior they release, or what we have called natural incentives.
The advantage of using hunger as the model for this purpose is that everyone
agrees that some of the sign Stimuli involved produce innate affects and that
enough research has been done on eating in animals and humans to show how
motives might develop out of these innate affects through learning.
To oversimplify somewhat, three types of sign Stimuli influence eating: (1)
sign Stimuli arising from nutritional deficiency, particularly low levels of avail-
able blood sugar (Mayer, 1955; Mayer & Marshall, 1956), which increase eating;
(2) sign Stimuli arising from the palatability or tastiness of food; and (3) sign
Stimuli arising from satiety, such as a füll stomach or high blood sugar level,
which decrease eating. Ordinarily the mechanisms involved operate automati-
cally to make sure the body has enough energy (represented as available blood
sugar) to do its work. A person's going without food lowers blood sugar, which
among other effects stimulates a portion of the midbrain that mediates affect
(that is, the lateral hypothalamus), which elicits eating if food is present (Gross-
man, 1960). This was discovered by injecting insulin, which lowers blood sugar
level and leads to increased eating. If food is tasty, it also stimulates the lateral
hypothalamus (Burton, Mora, & Rolls, 1975), leading to an increased release of
insulin; this lowers the blood sugar level, which stimulates the lateral hypothala-
mus and thus leads to more eating. Eventually a füll stomach, higher blood
sugar levels, and several other factors stimulate another portion of the hypothal-
amus and the liver to inhibit eating. Many physiological mechanisms are in-
volved in the control of food intake (see Carlson, 1977), and their exact influ-
ence is still not fully understood. All that need concern us here is that sign
Stimuli built into the organism automatically and in various ways produce the
responses (eating or not eating) that regulate food intake and maintain the
organism.
But food is not always present, the way oxygen nearly always is, when a
person or animal needs more of it. Furthermore, some objects that are eaten and
swallowed are good for the organism (that is, increase the blood sugar level) and
some are not; some may even be harmful or poisonous. So an individual must
learn how to get the right kind of food. And learning associated with eating be-
gins the moment the human infant is born and can no longer get nutrition di-
rectly from the mother (Marquis, 1941). Furthermore, learning about what to
eat occurs easily and rapidly. For example, many studies have shown that ani-
mals very readily learn to avoid the taste of a substance that has produced ill-
ness (Garcia & Koelling, 1966), but not to avoid a taste paired with electric
shock. That is, the animal seems to be preset to learn an association more easily
that would keep it from eating poisonous foods than an association involving the
pain of electric shock. Furthermore, animals quickly learn to prefer a diet con-
131
taining a substance they need (such as salt), presumably because they are avoid-
ing the old distress-producing salt-free diet and have found a new one that re-
duces the distress (Garcia, Hankins, & Rusiniak, 1974).
At what point do we begin to think in terms of motivation and motives? So
long as a System works automatically without important learning having taken
place, there seems little reason to speak of a motive being present. A lack of ox-
ygen stimulates breathing; low available blood sugar stimulates the human infant
to suck from the mother's breast. These mechanisms are often referred to as ho-
meostatic in the sense that when a disturbance occurs, it sets in motion the re-
sponse that removes the disturbance. The term motive is not ordinarily applied
to such sequences but is reserved for instances in which important learning has
taken place in connection with the affect associated with natural incentives.
At what point do we conclude learning has produced a hunger motive? A
conditioned taste aversion might be described as an "avoidance motive," but it
seems preferable to reserve the term motive for a more general type of goal.
Human infants at birth must above all learn what they want—what goals or end
states are associated with proper eating. Assume that the absence of food in the
presence of low blood sugar levels arouses the emotion of sadness-distress (since
eating is impossible) and that eating tasty food arouses the emotion of joy or
pleasure. These emotions provide much of the energy behind what becomes a
learned motive either (1) to relieve the distress by finding a way to eat or (2) to
feel pleasure from eating. In both cases the goal defining the motive is the
same—to eat—although in the first instance eating has been associated with re-
lief of distress and in the second, with pleasure from eating. The way the two
types of natural incentives may develop into motives is diagramed in Figure 5.1.
The connection between low available blood sugar and readiness to eat via
Stimulation of the lateral hypothalamus is what can be called a natural incen-
tive—an innate connection between a sign Stimulus and an affective response in
the midbrain, which releases an impulse to act in a certain way if certain other
conditions occur (chiefly the presence of food). Since food is not always in-
stantly available to human infants, they experience the negative affect that auto-
matically arises when the impulse to eat is activated and they cannot eat. In
time this builds an association between the low available blood sugar cue and a
threefold complex of events—namely, the absence of food, the accompanying
distress, and eating to relieve the distress. This complex may be described as an
affectively charged anticipatory goal response (eating to reduce the negative af-
fect) or a goal State (since in some cases what response produces the relief of
distress is not as clear as in the case of eating).
Thus, a motive may be defined briefly as a learned, affectively charged antic-
ipatory goal State aroused by various cues—initially by the natural sign Stimulus
of low available blood sugar in the case of eating, but later by other cues such
as the time of day. A motive activates the organism to learn the instrumental re-
sponses necessary to bring about the goal State.
The strength of this type of hunger motive presumably is largely determined
by the frequency, intensity, and variability of the condition of "no food" being
present when the organism is ready to eat. That is, the more the low available
Natural Incentives and Their Derivatives 133
Natural Incentive A
Lateral hypothalamus
A. Innate connections
Sign Stimulus: Affective arousal Impulse to act

low available blood sugar (LBS) Eating readiness "released"
If no food —Negative affect
If food-Relief by eating
B. Acquired motive to relieve sadness-distress by eating
Precondition: Low LBS frequently and variably associated with absence of food
Cue:
Motive: LBS • Anticipated absence of food
Other sources of sadness-distress

(tail pinching in rats, anxiety in humans) • Accompanying sadness-distress
Eating to relieve sadness-distress

(anticipatory, affectively charged
goal response or goal State)
Activates instrumental behavior
that permits eating
Natural Incentive A'
Lateral hypothalamus
A. Innate connections
Sign Stimulus: tasty food Affective arousal Impulse to act

Eating readiness "released"
If tasty food present —Pleasure
in eating
B. Acquired motive to enjoy eating tasty food
Precondition: Various types of tasty food frequently and variably available

Cues: Anticipated presence of tasty food
Motive: Sight, smell of food, mealtime
Accompanying pleasure
Eating to feel pleasure (anticipatory,

affectively charged goal response
or goal State)
Activates instrumental behavior

that permits eating tasty food
Figure 5.1.
How Two Types of Eating Motives Develop from Natural Incentives.
blood sugar cue is associated with intense negative affect under a variety of con-
ditions, the stronger this type of hunger motive will be. Such a conclusion is
supported by a number of studies showing that variable deprivation of food in
infant rats is associated with unusually strong tendencies to get and eat food in
adult rats (see, for example, J. M. Mandler, 1958). The hoarding is explained on
the grounds that early deprivation of food increases the frequency and intensity
of the negative affective arousal that occurs when the animal is ready to eat but
no food is present. Thus, the natural cue, low blood sugar, as well as other asso-
ciated cues, gives rise to a much stronger hunger motive to relieve distress by
eating. The anticipatory goal State (the reduction in distress by eating) is much
more affectively charged than it would be without the experience of variable
food deprivation. Therefore food, when it appears and permits eating, is associ-
ated with a much greater relief of distress. It has acquired a much greater re-
ward or incentive value because of the experience of deprivation, and it contin-
ues to maintain this extra value even in adulthood when plenty of food is
present. Thus, the animal hoards food because it was associated early in life
with relief of distress. In much the same way some people are presumed to have
developed an intense interest in food and eating as a means of relieving the dis-
tress they experienced early in life in connection with eating.
Furthermore, the distress need not derive completely from lack of food to
produce this effect. As Figure 5.1 points out, other sources of distress that occur
during eating also can increase the motive to eat to relieve distress. If rats' tails
are mildly pinched while they are eating, they eat more (Rowland & Antelman,
1976). This is surprising, because fear normally inhibits eating. The presumption
is that the tail pinching creates some negative affect that eating reduces. This
source of distress, like the distress from food deprivation, acts to increase the
strength of the goal response of eating to relieve distress.
A similar finding has been reported for certain kinds of overweight human
subjects (Schachter, Goldman, & Gordon, 1968). Subjects who were overweight
and of normal weight were given various foods to eat—for example, chocolate
chip Cookies (McKenna, 1972)—when they were in a State of high or low anxi-
ety. The subjects of normal weight ate fewer Cookies when the anxiety was
strong than when it was weak. Fear inhibited their eating. The obese subjects,
however, ate equal numbers of Cookies whether the anxiety was strong or weak,
despite independent evidence that they are more emotional and respond more
nervously to threats of various kinds. A reasonable explanation for their behav-
ior is that anxiety has been associated with eating in the past for them, so they
have a stronger motive than do subjects of normal weight to relieve distress by
eating. Certainly eating can be used to relieve distress (M. C. Jones, 1924), as
every mother knows who nurses or gives a bottle to calm an upset baby who is
not even hungry. Perhaps obese people have developed a stronger motive to eat
to relieve distress because of various experiences of negative affect associated
with eating in the past. This would also explain why they eat more than is nec-
essary to supply their bodies with the energy needed as signaled by low blood
sugar levels.
Natur'al Incentives and Their Derivatives 135
The sccond type of natural incentive illustrated in Figure 5.1 involves tasty
food as a sign Stimulus. More palatable food activates the lateral hypothalamus
more (Burton, Mora, & Rolls, 1975), which produces insulin release and more
eating. An animal will eat much more of the same food if its taste is changed
several times (Le Magnen, 1956). It is assumed that eating tasty food produces
strong pleasure, so if people have this type of experience frequently, they may
acquire a motive characterized by the anticipation of eating tasty food and the
pleasure associated with doing so. In this case the releasing sign Stimuli or cues
that arouse the motive are external—involving, for example, the sight, smell,
and taste of more palatable food—rather than internal, as was the previous mo-
tive complex built on the internal sign of lowered blood sugar, plus the distress
caused by its absence.
Schachter (1971a) and his associates have collected a large amount of evi-
dence demonstrating that certain overweight people are strongly influenced in
how much they eat by such external cues. Compared with people of normal
weight, the obese people consumed more of a good-tasting milk shake and less
of a bad-tasting one (Decke, 1971). They ate more than people of normal weight
when they were led to believe it was dinnertime by a clock that had been set
ahead in the laboratory (Schachter, 1971a). They ate more when food cues
were obvious and less when they were not obvious. For example, if nuts were in
their shells they ate them less often; if they were unshelled they ate them more
often than people of normal weight, who ate them equally often under both
conditions.
One way to explain these results is to assume that the approach eating mo-
tive outlined in the lower half of Figure 5.1 is stronger in some overweight peo-
ple as compared with people of normal weight. It might be inferred that more
attention was given to providing them tasty food when they were young, so the
external cues of sight, time of day, and so on, were more strongly associated
with the joys of eating. Since we have concluded that these overweight people
also might have a stronger motive to eat to reduce distress, obviously the eating
modality should have much more importance for them than for other people,
which is certainly the case. The presumption is strong that their concern with
eating derives from affective experiences associated with eating, particularly
early in life. This is precisely the line of reasoning advanced by Freud and his
followers, who believed on the basis of clinical case studies that early affective
experiences associated with eating—or the oral intake modality—lead people to
fixate on or be oriented toward that modality throughout life.
Unfortunately, no studies of humans have been reported that correlate affec-
tive feeding experiences in infancy with the strength of hunger motives in adult-
hood or with obesity, so the hypothesis cannot be directly checked. For that
matter, there is no direct evidence that obese people have stronger hunger mo-
tives, using the methods for assessing the strength of human motives described
in subsequent chapters. Nevertheless, the behavior of the types of overweight
people studied by Schachter and his associates is consistent with the theoretical
model of the way two types of hunger motives develop as outlined in Figure 5.1.
NATURAL INCENTIVES, EMOTIONS, AND MOTIVES

A natural incentive in humans has three components, as Figure 5.1 makes
clear—namely, (1) a sign Stimulus, (2) a State of central affective arousal, and
(3) what is usually called a consummatory act—a type of response that, like eat-
ing, is released by the sign Stimulus and the State of central arousal (see Cofer,
1972). It is called consummatory because it is aimed at satisfying whatever gave
rise to it. The red belly of a male stickleback elicits aggressive behavior in an-
other male stickleback that is aimed at the red belly. This is confusing, because
the sight of the red belly, which initiates the behavior, is not identical with the
consummatory experience, which comes from attacking the red belly. Nor is the
sight of tasty food, which produces lateral hypothalamic Stimulation and eating,
exactly the same as the experience of eating tasty food, which is the consumma-
tory affective experience sought. Sometimes what Starts an incentive sequence
is similar to the experience that consummates it, and sometimes the two are
different.
Table 5.1 summarizes the relationships involved in a natural incentive se-
quence and shows how the motives treated in subsequent chapters develop from
the natural incentives to be discussed in this chapter. The table outlines what is
to come. Each of the incentives mentioned will be discussed in detail later in
this chapter and the motives based on them, in later chapters. Notice that in the
case of the avoidance motives, the sign Stimulus differs from the consummatory
experience sought. Inconsistency, conflict, and pain are sign Stimuli that produce
negative emotions and release acts designed to produce the consummatory expe-
rience of consistency, relief, or fear reduction. Pain and the inconsistency sign
Stimuli also may be referred to as negative incentives, although their positive as-
pect (consistency or relief) is used here to label them.
In the case of approach motives, the sign Stimulus that sets up the natural
incentive sequence initially may be the consummatory experience itself. Eating
tasty food produces central affective excitement, which leads to eating more of
the food and endows the sight and smell of tasty food with the power through
learning to evoke an anticipatory affective goal State, thus forming the positive
hunger motive. In what follows, we will use the term natural incentive to refer
primarily to the consummatory experience and will label the incentive in terms
of that experience.
Table 5.1 also shows the type of emotion as shown in facial expressions,
which indicates the presence of a particular consummatory experience and the
motive that is presumably acquired in connection with each type of consumma-
tory experience. Consider the variety incentive first. As this chapter will show,
there is considerable evidence that mild Variation in Stimulation is pleasant and
sought and produces an emotion characterized by the facial expression of
interest-surprise (see Chapter 4). The consummatory experience of pleasure from
variety may occur by chance from changes in the environment, without the or-
ganism having done anything to bring it about. However, by the very nature of
the Stimulation, if the organism does nothing but passively experience it, the
Stimulation will become expected or boring and will lose its appeal. The con-
Table 5.1.
OUTLINE OF RELATIONS BETWEEN NATURAL INCENTIVES, ASSOCIATED EMOTIONS, AND MOTIVES
Natural Incentives
Emotion as Shown in Motive Acquired in
Sign Stimulus Facial Expression 'Consummatory Acts ' - Connection With
f
Consummatory
Experience of
Approach motives
Sight or smell of tasty Tasty food Pleasure Eating Hunger motive
food
Small variations from Mild variety Interest-surprise Exploratory behavior Achievement motive
expectancy (Chapter 7)
Having impact, threats, Having impact Anger-excitement Asserting, banging, and Power motive
Naturail Incentives cmd Their Derivatives

interference with having so on (Chapter 8)
impact
Touching, hearing, or Contact, rubbing, Sexual excitement, feel Seeking and exchanging Affiliative motives
seeing another person snuggling, rocking loved, loving, joy contact with others (Chapter 9)
Avoidance motives
Inconsistency, conflict Consistency, conflict Sadness-distress Seeking to relieve distress
reduction and increase consistency
Large variations from
expectancy Avoidance motives
(Chapter 10)
Pain Relief Fear Seeking to avoid fear
Unpleasant taste Disgust Spitting out

summatory act of exploring the environment is the way in which the organism
normally produces the consummatory experience from mild Variation in Stimula-
tion. As the experience of variety diminishes, the organism does something to
increase it. Here what started the sequence (the experience of mild variety) is
also the experience sought by the consummatory act of exploratory behavior.
And out of these experiences the achievement motive that will be discussed in
Chapter 7 presumably develops.
In the case of the impact incentive, a baby cannot have the consummatory
experience without doing something. The baby is naturally active, and in the
course of pushing things around and grasping and dropping things, the baby
feels the pleasurable excitement that is a built-in response to the experience of
having impact. Having an impact experience is the sign Stimulus for having
more such experiences in an ever increasing cycle that eventually results in the
facial emotion called anger-excitement (although, as we shall see, the term anger
has a negative connotation that it does not deserve in its most primitive, un-
learned form).
Note that here the consummatory experience or natural incentive is response
produced. The child produces a Stimulus—for example, a bang—that releases
more of the same type of behavior. Out of these impact experiences develops the
power motive, discussed in Chapter 8. The role of external sign Stimuli in releas-
ing the impulse to have impact is not altogether clear. There is some evidence
that in humans, threats (as from stränge or threatening gestures) and interfer-
ence with having impact innately release anger-excitement (see Morris, 1967).
These actual or potential interferences with having impact produce a negative
emotional State somewhat different from pleasurable anger-excitement. In time
the person seeks to release this distress through having more impact in much
the same way as the absence of food develops a motive to eat to relieve distress,
as outlined in Figure 5.1.
Contact experiences, such as those from snuggling, rocking, and cuddling,
lead to sexual excitement or, more simply, to joy-happiness-pleasure as reflected
in facial expressions. Such experiences are obtained normally by mutual ex-
change with another person, usually the mother or primary caretaker in the case
of human infants. Eventually they endow cues from seeing or hearing such a
person with the capacity to evoke anticipations of joy-happiness-pleasure from
contact with such people; these develop into the affiliative motives characterized
by being loved and loving, as will be discussed in Chapter 9.
Finally, mismatches between expectation and reality generate negative affect
—reflected in the facial expressions of sadness-distress and fear—which generate
consummatory acts designed to reduce the sadness-distress or maintain consis-
tency. As noted already, in this instance there is a lack of fit between the experi-
ence that Starts the natural incentive sequence (for example, inconsistency or
pain) and the consummatory experience that satisfies or ends it (relief from the
removal of inconsistency or fear). In general, experiences of negative affect from
inconsistency or fear lead to the avoidance motives, which will be discussed in
Chapter 10.
The outline in Table 5.1 leaves out motives associated with certain emotions
represented in facial expressions—disgust, for example—which may develop into

important human motives but on which no empirical work has been done. This
only emphasizes the fact that the classification in Table 5.1 is preliminary and
tentative. Its primary function is to suggest possible connections among natural
incentives, emotions, and motives that deserve further study and to provide a
framework in terms of which to organize the presentation of a large body of
knowledge.
• THE VARIETY INCENTIVE
It has long been known that animals seek at least moderate amounts of novelty.
White rats typically will explore a new environment unless the novelty is too
strong, in which case they will freeze or even make vigorous attempts to escape
(Fiske & Maddi, 1961). According to Berlyne (1967), "higher animals often find
access to Stimulation gratifying and properties known to raise arousal—such as
novelty, surprisingness, complexity—may enhance the reward value of extero-
ceptive Stimuli. Even pain can apparently be rewarding in small doses." Dember,
Earl, and Paradise (1957) painted the walls of the two loops in a figure eight
maze with two patterns of stripes. In one loop the stripes were vertical; in the
other, stripes of the same width were painted horizontally. On the second day of
exposure to this Situation, sixteen of seventeen rats spent most of their time in
the loop with vertical stripes, presumably because it provided more variety and
Stimulation when they ran past the stripes.
E. L. Walker (1964, 1973) extended this study by dividing a maze into four
quadrants, each of which varied progressively in degree of complexity. Typically
the animals started out by spending most of their time at a low level of com-
plexity, but as they grew more and more familiär with it they moved to a higher
level of complexity. This is reminiscent of the way humans with a strong
achievement motive seek out more and more difficult (complex) tasks (see Chap-
ter 7). Rats also have been shown to alternate the pathways they take to food if
the pathways are comparable (Heathers, 1940) or to prefer a path to food with
a barrier in it over an unobstructed path (Festinger, 1943).
McClelland, Atkinson, Clark, and Lowell (1953) attempted to systematize
such observations under the general principle that moderate discrepancies from
adaptation level would be found to be pleasing or rewarding, whereas large dis-
crepancies would be unpleasant and be avoided. A completely familiär Stimulus
would be hedonically neutral or boring in this formulation, whereas slight varia-
tions along any of a number of dimensions would attract attention and be pleas-
ant. Sudden large shifts from what the organism was adapted to would evoke
startle, discomfort, and avoidance.
A simple experiment by Haber (1958) illustrates how the principle works.
He had subjects place their hands in a bücket of water until the hands were well
adapted to its temperature. Then, at the same moment, the subjects removed
both hands from the bücket and immersed each in separate buckets of different
temperatures, with the instructions to withdraw the hand from the bücket that
was more uncomfortable. Haber found that temperatures either one degree Cel-
sius warmer or colder than the adaptation temperature were considered more
pleasant than temperatures that differed more from the adaptation temperature,
producing what has been called the butterfly curve (see Figure 5.2). Further-
more, he found he could shift the hedonic curve upward or downward by
changing the temperature of the adaptation bücket. Thus, if the hands were
adapted to thirty-three degrees Celsius, a temperature of thirty-two degrees Cel-
sius was considered pleasant, but if the adaptation level was thirty-four degrees
Celsius, a temperature of thirty-two degrees Celsius was considered unpleasant.
Of course, these changes take place within certain absolute limits, since all tem-
peratures above or below certain levels are unpleasant. The phenomenon has
been experienced by everyone who has tried to drink coffee and eat ice cream at
the same time. If the mouth is adapted to the higher temperature of coffee, the
ice cream can produce a painful reaction—a reaction much more painful than
it produces without the coffee, although the temperature of the ice cream is
constant.
A similar type of curve has been obtained by Maddi (1961) along the di-
mension of expectedness-novelty. Subjects were provided with booklets consist-
ing of a number of pages stapled together; on some of them appeared numbers
and on others, sentence stems like In winter or The stranger. The subjects were
instructed to turn the pages, and when they came to a stem they were to com-
plete it with a sentence in any way they liked. They were told that the num-
-15 -7 -3-101 3 + 15
Adaptation
Level
Discrepancy frorn Adaptation Level (in Degrees Centigrade)
Figure 5.2.
Typical Affective Preference Curve as a Function of Temperature Discrepancies from Level to
Which Hands in Water Are Adapted (after Haber, 1958).
bered pages were "included as spacers between the stems so that they might rest
when not completing sentences" (Maddi, 1961). Some of the subjects were asked
to predict whether a number (N) or a stem (S) was Coming up next. For one
group of subjects the number of pages with numbers and stems was quite regu-
lär, following an NNS pattern—that is, every third page contained a stem—and
the subjects were asked to predict whether a number or stem was Coming up
next. For another group the ordering of numbers or stems had no regularity.
The sentence completions produced by the subjects were objectively coded for
affective tone along a scale from 1 (if they contained strongly negative affect) to
5 (if they contained strongly positive affect).
Figure 5.3 shows the results. The affective tone of the sentence completions
for those experiencing regularity Starts out neutral, gradually increases toward
positive affect up through the third or fourth repetition of the NNS sequence,
and then declines again. The affective reaction of these subjects to the first stem
was neutral, but as they became more certain of what to expect, they reacted
with positive affect when the stem occurred. However, later in the series, when
they were absolutely sure a stem would occur, the occurrence of the stem no
longer evoked positive affect in the sentences they wrote. For the subjects expe-
riencing no regularity, there was no change in the affective tone of their sen-
tence completions over time.
After the eighth repetition of the NNS sequence, subjects experiencing regu-
4.0
Prediction Group
No-Regularity Group
2.0
1.0
1 2 3 4 5 6 7 8
Series
Figure 5.3.
Mean Affect Scores as a Function of Number of Series Experienced. These series constituted
environmental regularity for the prediction group (Maddi, 1961).
larity were exposed to a ninth stem representing either zero discrepancy from
the previous series (that is, NNS), a moderate discrepancy over expectation (ei-
ther NS or NNNS), or a large discrepancy from expectation (either S or
NNNNS). As Figure 5.4 shows, for subjects experiencing regularity, moderate
discrepancies over expectation produced a more positive affect to the ninth than
to the eighth stem, whereas a much larger discrepancy from expectation pro-
duced much less of a positive reaction or even a slightly negative one. Again, no
marked difference in affect was displayed by those who had experienced no
regularity.
The occurrence of events that are moderately different from what people ex-
pect is pleasing, whereas they react to the too familiär or the too novel either
indifferently or negatively. As noted earlier, Zajonc (1980), using an entirely dif-
ferent technique, demonstrated that familiär items are more positively evaluated
than totally new ones, although he does not explore the part of the curve in
which too much familiarity breeds boredom. However, the experience is a com-
mon one in everyday life: At first you may listen to a new piece of music more
or less with indifference until you become somewhat familiär with it. As the fa-
miliarity grows you confirm your partial expectations of sound sequences and
rhythms and react positively. If you listen to the piece a hundred or a thousand
times, however, it can become positively unpleasant.
Kagan, Kearsley, and Zelazo (1978) report a number of studies that show
Prediction Group
1 2
Treatments
Figure 5.4.
Mean Affective Difference Scores Associated With Zero, One, and Two Degrees of Change in Order
of Events (Maddi, 1961).
infants in the first year and a half of life pay more attention to and act more
pleased by moderately discrepant Stimuli. In their view, infants first are able to
form some Schema (for example, of the human face), to which they can compare
a discrepant Stimulus (say, a human mask without any eyes or with the features
scrambled). At an early age (say, three months) the mask with scrambled fea-
tures evokes considerably less attention than it does at thirteen months, because
at the earlier age children have not yet formed a good enough Schema of the
human face to compare the discrepant Stimulus with it (see also Spitz & Wolf,
1946).
A carefully designed experiment by Hopkins, Zelazo, Jacobson, and Kagan
(1976) illustrates this point and also that what infants pay attention to or like
depends on what they are habituated to. Infants seven and a half months old sat
on their mothers' laps viewing objects exposed in a chamber in front of them.
The objects were made of papier-mäche, as Figure 5.5 illustrates. The first four
were red with half-inch black stripes. When the infants pressed the lever in front
of them it resulted in an illumination two and a half seconds long of a Stimulus
object in one of the viewing compartments. During habituation they were al-
lowed to keep illuminating the object to view it for at least three minutes or
until they began to tire in doing so. They were habituated either to object A or
object D in the series.
A switch then was thrown so a bar press would illuminate a comparison ob-
ject in a second viewing chamber. This object also was available for viewing for
a minimum of three minutes. The crucial question was which of the comparison
objects the infants would be most interested in viewing, as indicated by the
number of times they pressed the bar to get a look at it. As the results in Figure
A B C D E
Figure 5.5.
Stimuli Used in Infant Habituation Study (Hopkins, Zelazo, Jacobson, & Kagan, 1976).
5.6 show, they were most interested in the object that was moderately different
(second-level moderate discrepancy) from the Standard object to which they had
been habituated. If they had been habituated to object A they were most inter-
ested in object D , and if habituated to object D they were most interested in ob-
ject A. Lesser discrepancies from the familiär object (object B after object A, or
object C after object D ) produced less interest, as did the most discrepant Stimu-
lus, object E. In short, the results are the same as for the M a d d i experiment:
moderate discrepancies from expectation are pleasurable; very small or very
large discrepancies are not.
These studies define familiarity-novelty not in terms of repetition of the
same event, as in the M a d d i experiment, but in terms of visual/spatial di-
mensions. Berlyne (1967) summarized evidence that the variety incentive is
appealing across a n u m b e r of dimensions—familiarity-novelty, expectedness-
surprisingness, clarity-ambiguity, and complexity-simplicity. (See also Berlyne &
Madsen, 1973.) However the dimensions are defined, moderate variety is re-
sponded to positively, whereas extreme variety produces negative reactions and
no variety produces indifference.
W h a t is particularly noteworthy about the variety incentive is that it con-
10.5 3.5
9.0 3.0
7.5 2.5
6.0 2.0
4.5 1.5
3.0 1.0
.2 <55 1.5 0.5 ,O
0 8.
0 S
-1.5
-3.0
-0.5 -S
I
|
-4.5 -1.5 U
-6.0 -2.0 §
Reinforced Instrumental Responses
-7.5 -2.5
-9.0 J | | L -3.0
No Change Minimal First-Level Second-Level Unrelated
Discrepancy Moderate Moderate Discrepancy
Discrepancy Discrepancy
Figure 5.6.
Changes in Responsivity from the Last Two Minutes of Viewing the Standard Object to the First
Two Minutes of Viewing the Discrepant Comparison Object (after Hopkins, Zelazo, Jacobson, &
Kagan, 1976).
tains both unlearned and learned components. What appears innate or invariant
is the degree of novelty or unexpectedness that produces the hedonic curve.
Learning also plays a crucial role, because what is expected—or the Situation to
which the person is adapted—completely depends on experience or learning. All
the natural incentives appear to operate this way: a certain type of experience
innately yields the hedonic curve, but what produces that type of experience
changes quickly through learning.
Individual Differences in Seeking Variety

Zuckerman (1974) has developed a sensation-seeking scale indicating that some
people seek much more variety or Stimulation than others. It consists of items
such as / would like to try parachute jumping; I like to have new and exciting
experiences and sensations even if they are a little frightening, unconventional, or
illegal; I like "wild," uninhibited parties; and / get bored seeing the same old
faces. In each case the subject has an opportunity to check one alternative or
another one indicating the opposite preference. Obviously, many other variables
besides the sensation-seeking incentive might influence the way subjects answer
questions like these, but in a general way the scale does indicate real differences
in the amount of Stimulation people seek.
Figure 5.7 provides one of the clearest examples of this fact. Male volunteer
subjects spent either eight hours of sensory deprivation or eight hours of con-
finement in social Isolation with some pictures and music provided. The extent
High Sensation
Seekers in
Isolation
Low Sensation
Seekers in Isolation
Low Sensation
Seekers in Stimulation
1 i 2 I 3 ' 4 I 5 I 6 I 7 I 8 '
Base Line Hours of Confinement
Figure 5.7.
Mean Periods of Movement per Hour of Confinement (after Zuckerman, 1974).
of bodily movements was measured "by an air mattress connected to a pressure

transducer" (Zuckerman, 1974). Clearly those scoring higher on Sensation seek-
ing in fact moved much more in either condition, presumably to stimulate them-
selves more. In another study Zuckerman measured scalp electrical potentials
evoked by different intensities of visual Stimuli for subjects high and low in Sen-
sation seeking. He found the high Sensation seekers (as measured by one of his
scales) were augmentors. That is, the brighter the visual Stimulus, the greater
the electrical response in the brain measured from scalp electrodes. Those not
seeking Sensation did not show this effect. Apparently, Sensation seekers can tol-
erate more such Stimulation than those low in Sensation seeking.
Maddi, Charlens, Maddi, and Smith (1962) also developed a measure of in-
dividual differences in seeking novelty. They compared imaginative stories writ-
ten after exposure to a very monotonous tape recording with those written after
exposure to a tape containing many interesting and surprising Statements. The
differences in the content of the two sets of stories were worked into a scoring
System called "the desire for novelty." The measure is most successful in picking
out those who are bored (those low on the scale) and does not correlate with
other measures of curiosity or of creativity.
Beswick (1965) was more successful in deriving a measure of the extent of
human curiosity. He too developed a scoring System for imaginative stories,
which covered such characteristics as Wonder-Interest, Perceptual Instrumental
Acts, Exploratory Role Behavior, and Intrapsychic Responses of Excitement,
which were found to differentiate the stories written after exposure to curiosity-
arousing experiences versus ordinary experiences. Beswick found that the test-
retest reliability of the measure was high (r = .71 after six months); also, the
measure is valid in that it correlates both with self-descriptions of orderliness
and with behaviors associated with interest in novelty, as in girls' agreeing with
an item like Novelty has the greatest appeal for me. He theorizes that being curi-
ous involves wanting to order experience on the one hand, and on the other
hand being willing to experience disorder so as to bring it into order. Beswick
found some support for the belief that his TAT measure tapped this characteris-
tic in the fact that students who scored high in it took more books out of the li-
brary to read for pleasure, as well as the fact that they were more apt to sponta-
neously make independent inquiries of a minister about religious matters than
those who scored low in the characteristic. He does not think of curiosity as a
motive, but it is one in the sense that the variety incentive has more appeal for
some individuals than others.
Using quite another measure, Liam Hudson (1966) has found he can distin-
guish between people who seek primarily novelty and those who seek primarily
order. He categorized individuals according to whether they scored high or low
on an intelligence test and on open-ended tests of creative fluency, such as list-
ing as many uses of a paper clip as possible. He compared those who scored
high on the intelligence tests and low on fluency tests (called convergers) with
those who did better on the fluency tests than they did on the intelligence tests
(called divergers). The divergers much more often were students of the arts,
whereas the convergers were science concentrators. In comparison to the con-
vergers, the divergers were more liberal and nonauthoritarian, more emotional,
gave more responses involving humor or personal violence, and did much better
on tests involving words rather than numbers or spatial relations. Science ap-
peals to convergers because it involves explaining everything new in an orderly
way, whereas the arts allow divergers to express their interest in novelty,
change, and exciting new forms.
The Variety Incentive and the Achievement Motive

McClelland, Atkinson, Clark, and Lowell (1953) argued that the variety incen-
tive was the basis for the achievement motive (see Chapter 7) and possibly for
all motives, which it is not. For example, there is no evidence that people high
in the need for Power or Affiliation are especially "turned on" by variety or
moderately challenging tasks the way those high in the need to Achieve are (see
Chapters 8 and 9). Since variety is a natural incentive, it can add spiee to any
consummatory experience; however, that is not the same as saying variety is
what is primarily sought for all motive Systems.
Furthermore, it is not entirely clear how the achievement motive is built on
the variety incentive. The problem lies in how the transition is made from liking
moderate variations in Stimulus situations to liking to produce moderate degrees
of variety, as in liking to do something better. Rats like variations in Stimula-
tion, such as traversing a maze with vertical stripes, but they soon learn to go
where they can get such increases in Stimulation. Does that define an achieve-
ment motive for them? The infants in the study by Kagan et al. liked to see
moderately discrepant visual figures, but they also quickly learned to press a bar
to get such experiences. It is well known that subjeets with a strong achieve-
ment motive prefer moderately difficult tasks (see Chapter 7), just as the little
boy Peter in Chapter 4 was attracted to climbing over the threshold, which in-
volved a slight difficulty for him in getting from one place to the next. But if he
were just interested in producing moderate variety, why did he not stay in the
next room and explore it? Why did he turn back, climb over the threshold
again, and smile at his aecomplishment? One might assume that task difficulty
becomes a regulär form of variety, just as moderate complexity does, so moder-
ately difficult situations become sought in themselves. Or one might assume that
variety has fused with the impact incentive in mastering tasks, since obviously
Peter was also pleased at producing the effect of getting over the obstacle. Re-
search to date has not clarified this point.
THE IMPACT INCENTIVE

Many observers have noted the great importance of exploration, manipulation,
and play in lower animals and the human infant. Robert White (1960) coined
the term effeetance to describe the goal of these activities. To give a lively sense
of what such behavior is like in the young child, White quotes from Gesell and
Ilg (1943):
The child wants to finger the clothespin, to get the feel of it. He puts it in his
mouth, pulls it out, looks at it, rotates it with the twist of the wrist, puts it back
into his mouth, pulls it out, gives it another twist, brings up his free hand, transfers
the clothespin from hand to hand, bangs it on the highchair tray, drops it, recovers
it, retransfers it from hand to hand, drops it out of reach, leans over to retrieve it,
fails, fusses a moment, then bangs the tray with his empty hand.
Even in feeding, which according to drive theory should represent the passive
pleasure of tension reduction, the child becomes active. "Around one year there
is likely to occur what Levy (1955) calls 'the battle of the spoon,' the moment
when 'the baby grabs the spoon from the mother's hand and tries to feed itself.
. . . We can be sure that the child is not motivated at this point by increased
oral gratification. He gets more food by letting mother do it, but by doing it
himself he gets more of another kind of satisfaction—a feeling of efficacy"
(White, 1960).
Later, when the child can walk, "one observes a constant activity of carry-
ing objects about, filling and emptying Containers, tearing things apart and fit-
ting them together, lining up blocks and eventually building with them, digging
and constructing in the sandbox" (White, 1960). White feels there is "one gen-
eral motivational principle" lying behind all these activities: "The word I have
suggested for this motive is effectance because its most characteristic feature is
seen in the production of effects upon the environment." In our terms, the pro-
duction of effects upon the environment is a natural incentive that guides and
directs much of the child's behavior. However, we will refer to it as the impact
incentive, because the term effectance as described by White and understood by
others includes the notions of mastery, competence, and self-determination (see
Deci, 1975), which are probably later derivations of the simple impact incentive.
A noteworthy feature of the impact incentive is that children soon develop
an interest in having big impacts on the environment. They seem to get satisfac-
tion from making louder and louder noises, from knocking things about or pull-
ing them off tables, and from assertive activities of all types. Assertiveness is
particularly likely to occur if the children's ongoing manipulative activity is
blocked or challenged. They react by increased intensity of effort, which quickly
arouses anger-excitement. Such behavior occurs more often in boys than in girls
(see Maccoby & Jacklin, 1974), but it is universally present in all children.
Having big impacts is innately satisfying; it is associated with the positive
emotions in Table 4.2. However, such activities obviously are so dangerous to
children and the environment that caretakers move in very quickly to suppress
them. Thus, so much inhibition develops around being assertive that the satis-
faction in it appears in adults only under very unusual circumstances. Zimbardo
(1970) cites a number of instances in which people seem to enjoy being violent
when normal inhibitions are removed. He notes that students were sometimes
senselessly beaten when they were caught up in "police riots":
The ones who actually got arrested seemed to have gotten caught up among the po-
lice, like a kind of human medicine ball, being shoved and knocked back and forth
from one cop to the next with what was obviously mounting fury. And this was a
phenomenon somewhat unexpected, which we were to observe consistently through-
out the days of violence—that rage seemed to engender rage; the bloodier and more
brutal the cops were, the more their fury increased. (Zimbardo, 1970)
He quotes from an army sergeant who was part of an army intelligence unit
that interrogated (and tortured) Vietcong prisoners: "First you strike to get
mad, then you strike because you are mad, and in the end you strike because of
the sheer pleasure of it" (Zimbardo, 1970). He also mentions a quarrel in which
a man stabbed his girlfriend twenty-five or thirty times—far more than was nec-
essary to kill her: he was carried away by violence, unable to stop. Zimbardo ar-
gues that this kind of pleasure in violence occurs only when natural controls are
somehow removed. For instance, he found that when he "deindividuated" vic-
tims by putting them in bulky hoods, subjects were willing and able to give
them much heavier electric shocks than they were when they could see the vic-
tims' faces. He conducted a field experiment to demonstrate the point more
fully. Zimbardo and some graduate students began attacking an old car with a
sledgehammer to see if others would follow suit. He writes the following:
Several observations are noteworthy. First of all, there is considerable reluctance to

take that first blow, to smash through the windshields and initiate the destruction of
a form. But it feels so good after the first smack that the next one comes more easi-
ly, with more force, and feels even better. Although everyone knew the sequence was
being filmed, the students got carried away temporarily. Once one person had begun
to wield the sledgehammer, it was difficult to get him to stop and pass it to the next
pair of eager hands. Finally they all attacked simultaneously. One Student jumped on
the roof and began stomping it in, two were pulling the door from its hinges, an-
other hammered away at the hood and motor, while the last one broke all the glass
he could find. They later reported that feeling the metal or glass give way under the
force of their blows was stimulating and pleasurable. (Zimbardo, 1970)
It should be clear from these examples that the impact-anger incentive in its
most primitive form does not involve the heavy overlay of learning that nor-
mally accompanies it, inhibits it, and directs it into acceptable Channels. It does
not even imply the intent to hurt another person, which is the way most social
psychologists define aggression.
Distinction Between Impact-Anger and Aggression

What Starts up the anger-aggression sequence? First, there is simple satisfaction
from manipulation—from having impact. Then, at least as psychologists since
McDougall (1908) have argued, threats and interference with manipulation satis-
faction increase the intensity of the response, leading at first to excitement and
then to anger and violence. Violence is not aggression, however, because it does
not necessarily imply intended injury to another person. McClelland and Api-
cella (1945) carried out a simple experiment in which students were repeatedly
criticized (insulted or threatened) by an experimenter as they were trying to sort
some cards as fast as possible. The earliest response to criticism was an in-
creased intensity of responding: the subjects began slamming the cards on the
table. Later this simple assertive reaction differentiated into more instrumental
responses, such as being aggressive toward the experimenter, trying to find satis-
faction in what they had accomplished so far, or giving up. The most primitive
response to threat, however, was undirected assertiveness, excitement, or anger.
Dollard, Doob, Miller, Mowrer, and Sears (1939), the authors of Frustration
and Aggression, attempted to define what elicits the aggressive response by stat-
ing that any interference with a goal-directed activity would create the instiga-
tion to aggression, defined as "injury to an organism or organism Surrogate."
This formulation turned out to be misleading in several ways. In the first place,
children dropping things or students destroying a car or slamming cards on the
table all seem to be enjoying the activity for its own sake; the effect of the activ-
ity in terms of its injury to something is a secondary interpretation of what is
going on. Second, their definition suggests that aggression occurs only in re-
sponse to frustration when in fact, as the examples make clear, pleasure in ag-
gression can occur without prior frustration. It is hard to imagine what frustra-
tion the students were responding to in breaking up the car. Finally, their
definition directed attention away from other reactions to frustration, such as
giving up or trying to find an alternative means of reaching the goal. Often in-
terference with goal-directed activity does not lead to aggression but to other
types of response.
Thus, it seems inaccurate to tie anger too tightly to frustration of just any
goal. On the other hand, a challenge to an impact goal, as in threatening or crit-
icizing someone, does seem more likely to elicit first excitement and then an in-
crease in the intensity of response and the emotion of anger. This is not the
same as blocking any goal sequence. A boy blocked in his desire to get some
candy may simply decide to eat something eise. No excitement need be involved.
However, if he is blocked in his desire to produce an effect—to topple over a
chair, for example—he pushes harder and may get quite excited. It is as if the
salience of the incentive to have an impact has been increased.
Brain Reward and the Impact Incentive

There is a fair amount of circumstantial evidence that the impact incentive is
served by a specific reward System in the brain, if we are correct in assuming
that impact is a milder form of anger-excitement-aggression. This is ceftainly
true in lower animals. For example, electrical Stimulation in different parts of a
cat's hypothalamus will produce either affective attack (hair on end, arched
back, spitting) or "quiet-biting" attack (for example, pouncing on a mouse and
biting it). Furthermore, the cat will press a lever to obtain the brain Stimulation
that produces the quiet-biting attack (Panksepp, 1971). In other words, it can be
inferred that the central Stimulation leading to a quiet-biting attack is pleasur-
able to the cat, in the sense that it will work to get the Stimulation.
The hypothalamus in humans, as well as in cats, is a key portion of the
brain involved in regulating the autonomic nervous System and the hormone re-
lease System, both of which play a central role in affective arousal, particularly
as it involves anger and aggression. The hypothalamus regulates the "fight-or-
flight" response, identified long ago by Cannon (1915) as an organism's way of
mobilizing its resources when frustrated or threatened. The sympathetic nervous
System is aroused, which has far-reaching effects, including increased heart rate,
increased blood pressure, vasodilation in the periphery so more blood can get to
the muscles, and so on. Certain neurohormones, called catecholamines, are the
chemical messengers responsible for neurotransmission in the sympathetic
branch of the autonomic nervous System. Thus, catecholamine release also is
tied to the anger-aggression response regulated by the hypothalamus. The best
known catecholamine is adrenaline, which accompanies anger and aggression,
but its precursors noradrenaline and dopamine also play an important role in
sympathetic activation. As Hamburg, Hamburg, and Barchas (1975) sum it up:
"Studies in animals utilizing a variety of test situations have suggested a very
powerful role of catecholamines in aggressive behavior. Catecholamines are es-
sential for many forms of aggressive behavior, and catecholamine mechanisms
are markedly altered by aggressive behavior."
Of particular interest, therefore, is the fact that electrical brain Stimulation
in the areas related to catecholamine release have been found to be rewarding,
in the sense that animals will work to get such Stimulation. It is also known
that there are catecholamine pathways in the hypothalamus (Sawchenko &
Swanson, 1981; Wise, 1980). This suggests that the release of catecholamines by
electrical Stimulation may be rewarding. The evidence for such a conclusion
comes not only from the fact that hypothalamic Stimulation leads to rewarding
quiet-biting attack in cats, but more directly from evidence that electrical Stimu-
lation in the neighborhood of dopamine fibers, and perhaps noradrenergic fibers,
is rewarding in rats (Wise, 1980) in the sense that they will press a bar to get
the Stimulation. It may be inferred that the rats are pressing a bar to get the
electrical Stimulation that releases dopamine, a catecholamine neurotransmitter,
which then might be conceived of as the physiological System subserving the im-
pact incentive.
Furthermore, drugs that increase catecholamines at the synapse between two
neurons—such as the amphetamines (Axelrod, 1974) or alcohol (Borg, Kvande,
& Sedvall, 1981)—increase both anger-excitement and aggression, as well as feel-
ings of pleasure, in humans (Berlyne, 1967), once again establishing a link be-
tween the catecholamines, anger-excitement, and pleasure. What is more, the sex
difference in assertiveness also is tied to a difference in catecholamine function-
ing. Males are notoriously more aggressive than females, even as small children
(Maccoby & Jacklin, 1974), and the catecholamine System is more responsive to
emotional Stimuli in males than in females. For example, males respond to injec-
tions or the threat of an examination by releasing more noradrenaline and adren-
aline in urine than females do (Frankenhaeuser, Dünne, & Lundberg, 1976).
Finally, physiological arousal, as from exercise or loud sounds, increases ag-
gressiveness (Konecni, 1975). This can be understood as resulting from the fact
that physiological arousal involves activation of the sympathetic nervous System,
which in turn increases catecholamine release—which, according to our argu-
ment, is strongly linked to the anger-excitement-aggression response. This evi-

dence strongly suggests that the natural incentive of having impact is served by
the release of catecholamine neurotransmitters in the brain. Stein (1975) argued
that a particular catecholamine transmitter, norepinephrine, provided the sole
reward mechanism in the brain for all motives. He was on the right track in
general, but his view needs correcting in its details.
Researchers now agree that another catecholamine transmitter, dopamine, is
more important than norepinephrine. They also agree there is no good reason to
assume there is only one reward mechanism. Instead, it seems more likely that
there would be several neurotransmitter reward substances, perhaps one for each
of the primary human incentives and the major motive Systems built upon them.
Certainly the evidence, so far as it goes, links the impact incentive specifically to
catecholamine neurotransmitters. Conclusive proof for such a connection does
not yet exist, but it certainly would not be surprising to discover that a natural
incentive is hooked up with a particular type of reward System in the brain.
Impact Incentive Derivatives

Many authors have been so impressed by the importance of effectance that, like
White, they treat it as a kind of master motive. As children grow they develop a
sense of competence, defined often as the ability to control events. deCharms
(1968) popularized the notion of personal causation and developed a measure of
individual difFerences around whether people feit they caused events (that is, feit
like an Origin) or feit they were a victim of circumstances beyond their control
(that is, feit like a Pawn). Origins are characterized as seeing a goal as a chal-
lenge rather than a threat, setting realistic goals based on their own probability
of success, and understanding the difference between controllable and uncontrol-
lable outcomes (see deCharms & Muir, 1978). Schoolchildren in the inner city
taught to think and act like Origins performed better academically and were
more likely to graduate from high school (deCharms, 1976), as Chapter 14 will
discuss. Feeling like an Origin can be conceived as originating from the cogni-
tive elaboration through training of the natural impact or effectance incentive
until it guides new behaviors in much the same way a motive does. For exam-
ple, it has been shown that people prefer activities they have freely chosen over
those they have been assigned (Perlmuter, Scharff, Karsh, & Monty, 1980), pre-
sumably because they have developed a value around being in control and have
discovered they can satisfy the impact incentive more when they are in control.
This chapter will return to the problem of values, or derived incentives, later.
Much research also has centered on the importance of people's feeling they
can or cannot have an impact on events. Seligman (1975) popularized the notion
of "learned helplessness," which results when an animal or a human being is
presented over and over with a threatening Situation about which nothing can be
done. Thus, dogs who were shocked repeatedly in an enclosure from which they
could not escape eventually became so passive that they failed to escape when
shocked, even when given the opportunity to do so. This would seem to be an
extreme case of extinguishing the natural incentive to have impact through

learning to be helpless. On the other hand, Wortman and Brehm (1975) point
out that the immediate response to a painful Situation is reactance, the attempt
to do something about it, much as we pointed out earlier that blocking the im-
pact incentive seems to increase assertiveness. Only after repeated discovery that
events are uncontrollable does a person or animal learn to be helpless.
The most important motivational derivative of the impact incentive is the
power motive, which will be discussed in Chapter 8. The definition of the power
motive is a "concern for having impact," a concern that presumably develops
out of the many early and late learnings associated with the impact incentive.
THE CONTACT OR SEXUAL INCENTIVES
In his paper on infantile sexuality, Freud (1905/1938) first called attention dra-
matically to the pleasure infants get from rhythmic tactile sensations:
As is shown by the example of thumbsucking, there are predestined erogenous

zones. But the same example also shows that any other region of skin or mucous
membrane may assume the function of an erogenous zone. . . . The thumbsucking
child looks around on his body and selects any portion of it for pleasure-sucking,
and becoming accustomed to this particular part, he then prefers it. If he acciden-
tally strikes upon a predestined region, such as a breast, nipple or genitals, it natu-
rally gets the preference. . . . The rhythmic characters must play some part and this
strongly suggests an analogy to tickling. . . . Pleasure-sucking is often combined with
a rubbing contact with certain sensitive parts of the body, such as the breasts and
external genitals. It is by this path that many children go from thumbsucking to
masturbation.
Freud (1905/1938) called attention to the similarity of this type of pleasure

seeking and sexual gratification: "The pleasure-sucking is connected with a füll
absorption of attention and leads to sleep or even to a motor reaction in the
form of an orgasm. . . . He who sees a satiated child sink back from the moth-
er's breast and fall asleep with reddened cheeks and blissful smile, will have to
admit that this picture remains as typical of the expression of sexual gratifica-
tion in later life."
Freud's observations were later confirmed by Kinsey, Pomeroy, and Martin
(1948), who observed orgasm in boys and girls as young as four months old.
They reported that orgasm in the infant involves "the development of rhythmic
body movements with distinct penis throbs and pelvic thrusts, and obvious
change in sensory capacities, a final tension of muscles, especially of abdomen,
hips and back, a sudden release with contractions—followed by the disappear-
ance of all Symptoms" (Kinsey et al., 1948). Freud, with his usual care, also ob-
served that infants obtained pleasure not only from the mouth and the genitals,
but also from the anal region and from playing with feces. He and others (see
Spitz & Wolf, 1949) also noted the pleasure some infants get from "rocking,"
that is, from getting on their hands and knees and rocking vigorously back and
forth. This too they classed as an expression of sexuality. The key defining char-
acteristic of the sexual incentive for them is rhythmic tactile self-stimulation,
particularly in certain body regions, which rises to a climax and then declines.
More recent studies have confirmed the innate pleasure infants get from this
type of Stimulation when it is provided by others, particularly the mother. Har-
low (1971) and his associates demonstrated this in an ingenious series of experi-
ments in which infant rhesus monkeys were reared with "mother" Surrogates.
The infant could obtain nourishment from all of the Surrogates, but they were
constructed to have a variety of other characteristics. The bodies of the Surro-
gates were made of wire mesh, and some were covered with terry cloth. The
rhesus monkey infants spent much more time clinging to the terry cloth mother
than the wire mesh mother. Furthermore, they reacted less emotionally to a
stränge environment when a cloth mother with which they had been reared was
present (Harlow, 1959). The presence of the wire mother, however, even though
it had been the only source of food, did not reduce the infant's fear in a stränge
environment. In fact, the infants reared by the wire mother showed a much
greater emotional response to a stränge environment than monkeys reared either
normally or by the terry cloth mother. Other coverings for the wire mother,
such as rayon, vinyl, or sandpaper, did not increase the attractiveness of the Sur-
rogate mother for the infant monkeys.
The monkeys' behavior strongly indicates there is some innate affective re-
sponse—joy-happiness-pleasure—released by the soft tactile Stimulation produced
by the terry cloth. Further research showed what other sign Stimuli were part of
the natural contact incentive: The rhesus monkey infants showed a greater pref-
erence for terry cloth mothers that were warm (rather than cold) and that
rocked rather than remained stationary. In other words, they preferred the kind
of rhythmic tactile Stimulation Freud described. They maintained their attach-
ment even though some of the mothers were "abusive"; that is, they blasted
them in the face with air or threw them off.
Human infants also show similar attachments to being with their mothers,
as the next section will describe. They enjoy being tickled—a clear example
of rhythmic tactile Stimulation—particularly in the stomach region, according
to Cicchetti and Sroufe (1976), which is as close as they got to the genital
area. The fact that people cannot tickle themselves successfully Supports the
idea that some such Stimulation must come from another. The importance of
Harlow's work with other primates is the fact that it permits a clearer definition
of the innate natural incentives than is possible with humans, where mothers' at-
titudes and cultural expectations might be expected to teach the infant such
responses.
In what sense was Freud correct in describing such a natural incentive as
sexual? He did so to base his sexual theory of neurosis on what appeared to be
solid biological observation, but he also used the term sexual so broadly that it
is almost synonymous with the word love. In that sense, many would agree that
the kind of contact gratifications he and Harlow observed are the innate bases
for acquired affectional or affiliative motives. However, the sexual motive System
also seems to be involved. Harlow (1971) also found that the contact gratifica-
tions in infancy, either between the mother and infant or between the infant and
peers, were essential to developing normal heterosexual attraction and behavior
later on. Monkeys reared on wire mesh mothers were isolated, could not relate
to other monkeys, and did not form heterosexual attachments.
What still is not clear is how these contact incentives get integrated into the
adult sexual motive, as well as the extent to which they are still part of such
motivation in adults. Harlow's work shows that the physical contact from play-
ing with peers soon replaces the contact gratifications received from the mother.
Eventually, because of the hormonal changes that occur at puberty (see Chapter
9), the contact gratifications sought become more explicitly sexual in nature.
There is also substantial evidence that children and adults continue to get plea-
sure from rhythmic tactual and kinesthetic sensations, since they spend so much
time in amusement parks riding roller coasters and being jostled around in vari-
ous machines. Some of the pleasure may derive from Solomon's Opponent pro-
cess mechanism described in Chapter 4 (see Solomon, 1980). That is, pleasure
may be an innate consequence of the fear aroused in such situations. On the
other hand, much of the jostling and rhythmic movement—swinging on Swings,
for example—seems quite pleasurable even when no fear is involved.
Dutton and Aron (1974) carried out an investigation that suggests that this
type of Stimulation may continue to be related to sexual arousal. They stationed
a female experimenter near the exits from two bridges in a public park. She
stopped men after they had crossed the bridges and asked them if they were
willing to participate in a psychological experiment, which consisted among
other things of telling a brief story about a picture. She then said that if they
wanted to know how the experiment turned out they could call her telephone
number, which she gave them. One bridge was a long Suspension bridge over a
chasm; the bridge tilted, wobbled, and swayed as people walked across it. The
other bridge was quite stable. The experimenter found that the men who had
crossed the Suspension bridge told stories containing more sexual imagery than
the men who had crossed the stable bridge. The men who had been jiggling up
and down also were much more likely to telephone the girl later. In short, they
appeared more sexually aroused by the experience.
Dutton and Aron (1974) and Dienstbier (1979), who conducted other simi-
lar experiments in the laboratory, explain this result as being due to the fear
aroused in crossing the Suspension bridge. That bridge did evoke more fear than
the stable bridge, as determined by actual ratings by other subjects. Thus, it
could be argued that the fear produced physiological arousal, which the men in
the presence of an attractive female attributed to sexuality; when male interview-
ers were used, there was no difference in the sexual content of the stories writ-
ten by men who had crossed either of the two bridges. According to the theory
advanced by Schachter and Singer (1962), all types of physiological arousal are
nearly the same, and whether they will be labeled sexual or fearful depends on
cognitive factors—namely, the understanding of the Situation. According to the
theory, the presence of the female Interviewer suggested to the men that the
greater physiological arousal caused by the Suspension bridge really might be

sexual in nature.
As we shall see in Chapter 12, there are difficulties with this theory, and
there are other possible explanations of the result. One is provided by Solomon's
Opponent process theory, which states that the fear from crossing the Suspension
bridge, much like that involved in sport parachute jumping, leads to the arousal
of an opposing, or pleasurable, emotion. This theory has the advantage of ex-
plaining why the men crossed the Suspension bridge in the first place—namely,
to get to the pleasure that would follow the fear. Emotion attribution theory
really does not explain why people would choose to cross a fearful Suspension
bridge if they could cross a stable one. However, Solomon's theory does not ex-
plain why specifically sexual feelings are aroused to oppose the Suspension
bridge fear rather than, say, the exhilaration feit after parachute jumping. A
much simpler explanation is that the jiggling, joggling, and swaying is part of
the sexual incentive in adults, just as it is in infants. This accounts for the inter-
est in the Suspension bridge, as well as for the increased likelihood of sexual
thoughts in the presence of a woman experimenter. Despite some of Dienstbier's
(1979) later experiments, it seems unlikely that inducing fear without jiggling al-
ways will increase the sexual arousal of men in the presence of women, although
evidence on this point is not conclusive.
Imprinting and Attachment
Just how the contact incentives evolve into social attachments is not known, but
certainly there is a shift from physical to social sources of Stimulation. One ex-
planation Starts with the observation of attachments in lower animals—for ex-
ample, Harlow's study of the way infant rhesus monkeys get attached to their
mothers or terry cloth mother Surrogates and cry when they are left alone. At
an even lower level, ducklings will follow a mother duck or any moving mother
Surrogate to which they are exposed shortly after birth. This process is called
imprinting: the mother normally becomes a natural incentive that is imprinted
on the duckling during a certain critical period in its development.
Some psychologists, particularly Bowlby (1969), have concluded that human
infants appear to be imprinted in a very similar way on their primary caretakers
—usually, but not always, the mother. The infant shows the attachment through
clinging, smiling, following, sucking, crying, and perhaps seeking eye contact.
Which of these behaviors is displayed depends on the context, but the purpose
of all of them is increasing mother proximity. According to attachment theo-
rists, there is nothing particularly primary about clinging and its attendant in-
centive of contact (Ainsworth, 1973; Bowlby, 1969; Sroufe & Waters, 1977).
Therefore, it is probably an oversimplification to derive all the attachment incen-
tives from primary contact pleasures, in the manner suggested by Freud. That
is, in addition to the contact incentive there may be a natural incentive (or one
acquired very early and easily) based on the positive affect associated with being
with others or exchanging responses harmoniously with them. (See Condon's
work discussed in Chapter 9.)
On the other hand, Kagan, Kearsley, and Zelazo (1978) have provided a
somewhat different explanation for the attachment phenomenon, in particular
for the Separation distress infants typically begin to show at around eight
months of age when the primary caretaker leaves. This distress can be taken as
a sign of the baby's attachment to the mother, which parallels the duckling's
distress calling or the infant monkey's crying when the mother leaves. However,
there are difficulties in assuming that crying indicates attachment (Kagan et al.,
1978). Children sometimes cry when a comparative stranger to whom they are
not attached leaves the room. On the other hand, they sometimes cry when
their mother comes into the room, or they do not cry when she leaves in famil-
iär surroundings to carry out her work around the house. 'The major problem
with the attachment explanation of the growth function for Separation anxiety is
the difficulty it has accounting for behavioral differences among children whose
mothers differ in availability" (Kagan et al., 1978). That is, attachment should
be stronger in infants whose mothers are readily accessible, yet children who are
frequently separated from their mothers, as in day care centers, show a normal
development of Separation distress. They are just as likely to be distressed as
children who are with their mothers all the time.
Kagan et al. (1978) believe Separation anxiety is not an attachment phenom-
enon, but one that can be explained in terms of the variety incentive discussed
earlier in this chapter. They believe that around the age of seven months a child
is able to understand for the first time what it means for someone to leave the
room: "During or after maternal departure the child tries to generate a cognitive
structure to explain the mother's absence or a behavior to alter the Situation. If
he cannot do either he is vulnerable to uncertainty" (Kagan et al., 1978). If the
child is uncertain, he or she is likely to be unhappy, as was noted in Table 4.2.
Before the child is seven months old, the mother's leaving the room is not part
of the child's cognitive Schemata for understanding events, so it does not repre-
sent a major discrepancy from what is expected. Later, after about fifteen
months of age, the departure is understood perfectly well, but it no longer repre-
sents a major discrepancy from the child's understanding and therefore no long-
er produces crying in the normal child. In other words, Separation distress can
be understood as an aspect of cognitive development. Such an inference also ex-
plains why the departure of a comparative stranger can produce distress at this
age: children are not upset at the loss of someone to whom they are not at-
tached, but they are upset by the realization of what someone leaving the room
means, since it is disruption of expectations that is unpleasant rather than dis-
ruption of an attachment.
Whereas such theorizing can account for Separation distress, it does not deal
directly with the positive attachment children show for their primary caretakers
or others in their environment. Little Peter in Chapter 4 spent a good deal of
time going up to his mother and snuggling and rubbing himself against her, just
as Harlow's infant monkeys did. The problem is how children get from this
behavior to more developed social attachments. As Michael Lewis (cited in
Henley, 1974) has put it so well, "a major socialization process in terms of
attachment or social behavior, is to move the infant from a proximal mode of
interaction (e.g., touching, rocking, holding) to a distal mode (e.g., smiling and
vocalizing)."
Sexual Arousal from a Distance

Whereas rhythmic contact remains the primary mode of sexual arousal in
adults, there is also some evidence that other kinds of Stimulation are effective.
Several observers have noted that the face and eyes are particularly potent in
arousing the human infant to pleasurable smiling. Sroufe and Waters (1976) feel
that eye-to-eye contact is especially important. As Rubin (1975) reports, people
who are in love spend much more time looking into each other's eyes than do
other people.
In most animal species, nature must arrange for the two sexes to get to-
gether and copulate in order to reproduce the species. In lower animals very
specific sights, sounds, smells, or tactile Stimuli elicit sexual behavior. A male
moth is attracted by a smell given off by the female moth. The peahen is sexu-
ally aroused by the courtship dance of the peacock.
What about humans? Havelock Ellis (1954) has suggested that a human
male can teil that a human female has reached sexual maturity by using his
eyes: he can see that her body contours have changed. On the other hand, a
human female can teil whether a male is sexually mature by using her ears, for
a male's voice changes at puberty. The implication is that the sight of the ma-
ture female figure may have sexually arousing properties for the male, and con-
siderable research evidence Supports this assertion. When males are asked what
arouses them sexually, they list primarily visual cues centering on the female
body (Kinsey et al., 1948). Clark (1952) has shown that exposure to pictures of
nude females will arouse sexual fantasies in men if their inhibitions are removed
by alcohol. Wilson and Lawson (1976a) showed that pornographic films had the
same effects on penile tumescence, again if inhibitions were removed either by
the subjects' drinking or thinking they were drinking.
Kaiin (1972) reported that even the presence of a fully clothed but attrac-
tive female folk singer in a small social gathering could arouse sexual fantasies
in males as shown in written stories, providing the circumstances were not in-
hibiting. Table 5.2 summarizes some of the results of this study. Kaiin counted
the frequency with which stories contained references to physical sex (that is,
kissing and sexual intercourse) when the singer was either present or not pres-
ent, and the experiment was carried out either in an apartment setting or in a
classroom, with or without alcoholic drinks being served. Under inhibiting con-
ditions (that is, either apartment-dry or classroom-wet), the singer did not pro-
duce an increase in sexual imagery in male fantasies. However, under conditions
of disinhibition (when the subjects were drinking in a relaxed social setting in an
apartment), there was a large and significant increase in their sexual fantasies.
Certainly the preference of males for pictures of nude females and for buying
magazines that contain such pictures suggests that such Stimuli have a natural
incentive value for them.
Table 5.2.
MEAN FREQUENCIES OF THEMATIC PHYSICAL SEX IN STORIES WRITTEN IN SIX
CONDITIONS (after Kaiin, 1972)
No Singer Singer S - NS*
1. Apartment-dry 1.18 (17) 0.70 (27) -0.48

2. Apartment-wet 2.48 (27) 4.39 (26) 1.91b
3. Classroom-wet 1.80 (20) 1.31 (32) -0.49
Drinking effect ( 2 — 1 ) 1.30c
Setting effect (3 - 2) -0.68 d -3.08 d
Note: Numbers in parentheses equal number of subjects per condition.

a
Singer — no singer conditions.
b
In apartment-wet conditions, no singer — singer t = 1.96, p < .05 (in the predicted direction).
c
In alcohol experiment (main effect, alcohol), F = 11.69, p < .01.
d
In setting experiment (main effect, setting), F = 7.57, p < .01.
The evidence for the arousing value of the male voice for women is less
clear, although the sexual excitement among women produced by male singers is
a well-known natural phenomenon. It seems to exceed considerably the sexual
arousal potential of the female voice for males, although a systematic study of
this phenomenon has not yet been carried out. One attempt was made by
Beardslee and Fogelson (1958) to see if rhythmic musical Stimulation had more
sexual arousal potential for women than for men. They found no sex differences
in sexual imagery in stories written to control pictures, but when the subjects
wrote stories to musical selections the results were quite different. The males re-
sponded with more overt sexual imagery, probably because at least at this his-
torical period women were more inhibited about writing directly about sexual
matters. So a symbolic sex activity score was developed that was patterned after
the criteria for sexual activity described by Freud in his paper on infantile sexu-
ality. That is, the stories written by the subjects were scored for any references
to motion, rhythm, peak, and penetration—the last category being added as a
metaphorical reference to sexual intercourse. If any two of these scoring catego-
ries appeared, symbolic sexual activity was scored for the story.
As Table 5.3 shows, women provided almost significantly more instances of
symbolic sex activity than men did, even when writing stories to the four neu-
tral musical selections, and the difference was much more marked for the arous-
ing musical selections. The arousing selections contained "a pronounced empha-
sis on rhythm, themes with comparatively large tonal ranges, and a gradual
buildup to a climax. In contradistinction, the neutral music could be character-
ized as having very little rhythmic emphasis and a general evenness in melody
in overall construction" (Beardslee & Fogelson, 1958). Both the men and
Table 5.3.
AVERAGE SYMBOLIC SEX ACTIVITY AFTER TWO TYPES OF MUSICAL
SELECTIONS (after Beardslee & Fogelson, 1958)
Males Females
Four neutral musical selections .61 .93
Four arousing musical selections .20 1.93
Difference, neutral — arousing .59 + 1.00
Difference between differences +41 < .05
women were more sexually aroused by the rhythmic musical selections, but the
women significantly more so. These findings need extension and confirmation,
but so far as they go they suggest women may be more sexually sensitive to cer-
tain types of sound than men are.
Derivatives of the Contact Incentives: Succorance and Nurturance

Besides sexual motivation, what other motives are likely to develop out of con-
tact gratifications? Two possibilities that readily present themselves are what
H. A. Murray (1938) refers to as the needs for succorance and for nurturance,
succorance being based on receiving contact gratifications and nurturance, on
giving them. That is, children may get so much pleasure out of contacts with
the primary caretaker that they develop a strong need for such contact, which
should show up as dependence on the primary caretaker.
Sears, Rau, and Alpert (1965) and Whiting and Whiting (1975) have ob-
served acts of dependency in children in great detail, but they have failed to
make a case for the fact that there is a strong dependency motive. The reason
may be that they defined and measured a wide variety of dependency behaviors
(for example, asking for help, crying when the caretaker leaves, seeking atten-
tion and approval, and so on) and then tried to discover whether dependency
was a behavioral trait by intercorrelating the extent to which children showed
these behaviors. That is, would those who performed dependency act A also
perform dependency acts B and C, and would children who acted dependently
toward parents also act dependently toward peers? In general, they found little
consistency of this type. Even children who were high in the total number of
acts of dependency observed at one time period were not particularly likely to
show more such acts at another time period: the average correlation for depen-
dency scores across time periods was only .36 (Sears et al., 1965).
Whether a child acts dependently or not is more a function of the Situation
(the Status of the two actors, the age, and the setting) than of a general ten-
dency to act dependently. But this conclusion permits no inference about the
presence of a dependency motive—defined as an anticipatory goal State involving
being cared for—since such a goal State might be achieved by nondependent acts
(for example, being a "brave" child when hurt) and since dependent acts can be
instrumental in achieving other goal states. For example, a child may ask for
help to solve a problem while in pursuit of an achievement goal. Furthermore,
much dependent behavior, such as seeking to be with a caretaker, is clearly
guided by the affiliation or attachment incentive. Thus, whether the existence of
another incentive, the need for nurturance, should be postulated can be ques-
tioned. At present there is no adequate measure of the need for nurturance or
any strong empirical evidence that such a need exists.
On the other side of parent-child interaction, many authors have written
about the mother's or father's parental instinct—that is, their desire to care for,
look after, and protect a child. Since the human infant is so helpless for so long,
there obviously would be an evolutionary advantage for survival if parents were
endowed with an innate responsiveness to infant helplessness that would lead to
nurturance. Harlow's (1971) work on rhesus monkeys has demonstrated that
clinging by the baby is a sign Stimulus that evokes cuddling and caressing be-
havior from the mother. If a kitten is substituted for her baby, the mother mon-
key will try to caress it for a time; since it cannot cling, however, the mother
soon stops. If another baby that does cling is substituted, the mother continues
caressing. Clinging can even evoke a normal affiliative response in a "sick" mon-
key who had become an isolate due to prolonged Separation from others (Har-
low, Harlow, & Suomi, 1971).
Among humans there is also widespread evidence that mothers and some fa-
thers (Spelke, Zelazo, Kagan, & Kotelchuck, 1973) naturally like to play with
and look after their infants. Social psychologists have studied in depth the con-
ditions under which people will stop to give assistance to a helpless person or
one who cries out for help. Does a helpless person in obvious need act as a kind
of sign Stimulus that elicits affect and the impulse to help in a way analogous to
the presumed effect of a crying baby on a parent? The evidence does not permit
a clear answer. Certainly some people are Good Samaritans and stop to help
needy people; however, Darley and Batson (1973) showed that even some semi-
nary students on their way to give a talk on the Good Samaritan incident in the
Bible failed to stop to help an obvious victim if they thought they were late and
thus were in a hurry.
Unfortunately, neither this experiment nor the many others on the same
topic shed much light on the incentives or motives involved; they focused on
predicting the action of helping, which is a Joint product of motives, incentives,
skills, values, and opportunities (see Chapter 6). Thus, it is entirely possible that
a victim arouses an impulse to help or nurture that could be discovered if it
were measured directly, but that other factors inhibit the helping act. On the
other hand, a natural incentive to nurture may not exist. It is impossible to
know on the basis of present evidence, despite the large number of studies of
helping behavior.
Little work has been done on measuring a motive like the need for nurtur-
ance, which might arise out of such a natural incentive as the joy-happiness-
pleasure that comes from holding and helping babies. A preliminary approach
has been made by Sara Winter (1969), who studied the fantasies of a mother
while she was nursing her baby compared with her fantasies while her baby was
in the room but not actually being nursed. The stories told after nursing con-
tained more references to positive feelings and orientation to the present time, as
well as fewer instances of instrumental or means-end thinking, than the stories
written when the baby was simply present. Winter concluded that nursing moth-
ers were characterized by a "being" rather than a "doing" orientation. In other
words, mothers and other primary caretakers get contact and other gratifications
out of nursing, touching, and otherwise playing with babies that may in time de-
velop into a social motive that could be described as the need for nurturance.
What is needed to demonstrate this fact is further research of the sort S. Winter
began.
Finally, it is presumed that the affiliative motives—the need for affiliation
and the intimacy motive—develop primarily out of the contact incentives. These
motives will be fully discussed in Chapter 9.
THE CONSISTENCY INCENTIVE
Drive-reduction theorists (see Chapter 3) readily adopted the notion that conflict
could produce tension, which the organism would seek to reduce (see J. S.
Brown, 1961). They reasoned further that variety was a source of conflict or
confusion in the sense that the organism would not know how to behave in a
new Situation. Therefore, if people or animals appear to seek variety, they are
doing so only to reduce tension. The rat's exploratory behavior in a new maze
really is designed to reduce the tension and conflict from uncertainty (Montgom-
ery & Segall, 1955). Thus, some psychologists always have thought organisms
seek consistency in the sense that they wish to avoid conflict. The weakness of
this kind of theorizing, as already noted, is that it does not adequately account
for the positive enjoyment of variety.
McClelland, Atkinson, Clark, and Lowell (1953) dealt with this problem by
postulating that small discrepancies from the adaptation level (or expectancy)
produced positive affect, and large ones produced negative affect. They reported
several studies that showed this to be the case. (See, for example, Figures 5.2,
5.3, and 5.4.) It was a short Step from such a result to the inference that people
would seek to avoid the negative affect resulting from major inconsistencies in
what happened over what was expected. In more general terms, people would
seek consistency in confirming expectations. The chief difficulty with the theory
is that it proved very difficult to define in advance or generalize about what con-
stituted a small or large discrepancy. Does a person who just misses a plane ex-
perience a large or a small discrepancy? It is small in terms of time, but large in
terms of his or her expectancy of getting somewhere. Often multiple expecta-
tions are involved in real life, so it has been hard to pin down the concept of
consistency in practice, when it is defined as avoiding major discrepancies from
expectation.
Nevertheless, a number of independent sources of evidence suggest there is a
consistency, as well as a variety, incentive. It is obviously of great adaptive sig-
nificance to the organism to be able to live in a reasonably stable, coherent envi-
ronment. The perceptual constancies are an example of this phenomenon.

Shapes and colors are perceived as relatively the same even as they undergo sig-
nificant transformations. People seen at a considerable distance are perceived as
being about the same height as they would be close at hand, even though as a
physical Stimulus they present a much smaller image to the eye. George Kelly
(1955) has theorized that people's main goal in life is to form constructs that
will give coherence and order to their conception of the world, in order to en-
able them to understand what is going on: "Each person evolves for his conve-
nience in anticipating events a construction System." Thus, people are able to
maintain order by fitting whatever happens into their personal constructs. This
is a consistency theory in the sense that people try to match up events with ex-
pectations and avoid confusion. The behavior of young children in learning a
language fits here. They seek to match what they say to what they hear and
avoid the discomfort of a mismatch.
At the social level the consistency incentive appears as conformity to social
norms. Automobile drivers learn they are supposed to stop at red lights, so they
stop at red lights. A man knows that he is a Mormon and that Mormons do not
drink, so he does not drink. We drive on the right-hand side of the road, eat
three meals a day, go to school, brush our teeth, and do thousands of things
simply to be consistent—to live in a stable, predictable environment. We develop
many expectations about the way things are and the way we are, and we act in
conformity with these expectations. When we do not know what to expect, we
conform to the expectations created by others. In a well-known experiment
Asch (1951) demonstrated that if subjects hear three or four others say a shorter
line is equal in length to a longer line, most will conform to the group opinion
despite clear visual evidence to the contrary.
It seems unnecessary to search for underlying motives to explain why people
conform to such expectations: there is tremendous economy in doing so, since
conforming reduces confusion, conflict, and uncertainty. The power of the con-
formity incentive has been demonstrated dramatically by Milgram's (1963, 1965)
well-known experiment in which subjects, when told to by an experimenter, con-
tinued to give supposed severe electrical shocks to people despite the fact that
the shocks were known to be dangerous and the subjects in some instances were
crying out in pain. People expect to do what they are told by authority—in this
case, the experimenter. Therefore, they obeyed, despite the härm to their "vic-
tims," because not to do so would have caused even greater discomfort from
disobedience to a strong expectation that they should obey authority.
Social psychologists have studied extensively the variables that affect
whether a person conforms or not. Our interest here is limited to demonstrating
that there is some kind of natural incentive that elicits satisfaction from match-
ing expectations we have about ourselves, or what we perceive, to social or
physical reality.
Cognitive Dissonance
Several experiments have demonstrated that inconsistency—or what has come to
be called cognitive dissonance—motivates people to change their behavior to
make it more consistent. In the original study starting this line of research
(Festinger & Carlsmith, 1959), subjects first spent a long time doing dull and
boring tasks. They were then paid either one dollar or twenty dollars to teil
other subjects that the experiment was really very interesting. In other words,
they were paid either a lot or a little for lying about the nature of their experi-
ence. Later they were all asked by another experimenter to rate how much they
actually had enjoyed doing the experimental tasks. Those who were paid twenty
dollars rated the tasks as slightly more enjoyable than those who had been paid
nothing, who rated them as very unpleasant; those who had been paid only one
dollar, however, rated the tasks as most enjoyable. Why? According to the theo-
ry, lying about the nature of the tasks created cognitive dissonance—a disso-
nance between what the subject is doing (lying) and his or her self-image. Being
paid twenty dollars to lie seemed sufficient justification to explain to the subjects
why they were doing something so unusual. Therefore, they did not change their
evaluations of how pleasant the tasks were. However, subjects who were paid
only one dollar to lie found they were doing something that did not fit their
image of themselves as honest. They were in a State of dissonance or confusion,
so they reduced the inconsistency by concluding that the tasks actually had been
fairly interesting. Thus, they were no longer lying very much. This has led to a
whole series of research projects demonstrating that various types of justification
regularly are used by subjects to make their behavior consistent with their self-
image.
Furthermore, Rokeach (1973) carried out an experiment in which he
showed that demonstrating to subjects an inconsistency in their attitudes would
lead to behavioral change that would reduce the inconsistency. He found that
most Michigan State undergraduates ranked the value of freedom highest,
whereas the value they placed on equality was somewhat lower. Rokeach
pointed out the inconsistency by observing that they believed highly in freedom
for themselves but not for others, like blacks and women, since they had rated
equality somewhat lower. Just pointing out this inconsistency carefully in a lec-
ture was sufficient to motivate the subjects to change their value ranking for
equality, to increase their belief in equal rights for blacks, and to respond posi-
tively more often to a recruiting letter from the local branch of the National As-
sociation for the Advancement of Colored People.
Motives Related to the Consistency Incentive

Some people clearly are more eager to reduce cognitive dissonance than others,
and some try harder to avoid "social" dissonance, or the tension that arises
from not matching their behavior to social expectations. In both cases what mo-
tivates them is a desire to avoid the strong negative affect that for them accom-
panies inconsistency. Not much attention has been given to measuring the
strength of the motive to avoid cognitive inconsistency; more attention has been
given to measuring the strength of the motive to avoid social dissonance by con-
forming to social norms. For example, Crowne and Marlowe (1964) measured
the strength of what they called the approval motive by asking subjects to re-
spond to items such as / never hesitate to go out of my way to help someone in

trouble and / like to gossip at times. It is socially desirable to answer true to the
first Statement and false to the second, although it seems highly unlikely that
people would "never" fail to help someone or not like to gossip occasionally.
People who give many socially desirable responses tend to behave in socially ex-
pected ways. They report to the experimenter that a boring task was really
pleasant and worthwhile, and they go so far as to agree with the judgment of
others that there were four knocks on a tape, for example, when there were
clearly only three. In short, they distort reality to avoid the discomfort of not
behaving in ways that they believe are expected of them.
However, further research has suggested that the motive behind the need for
approval may be fear of rejection by others (see Chapter 10). In fact, as Chapter
10 points out, the current trend is to identify the different types of motives peo-
ple have to avoid the inconsistencies leading to negative affect, depending on the
source of the inconsistency—whether it be failure to meet expectations about
achievement, affiliation (approval of others), or power to have impact on others.
Thus, the consistency incentive, in combination with other incentives, may pro-
duce several types of avoidance motives, just as we found that an avoidance mo-
tive developed out of the distress that arose when the impulse to eat could not
be fulfilled because no food was present. Up to the present, therefore, research
has focused more on difFerent types of avoidance motives than on a Single con-
sistency motive.
INTERACTION OF INCENTIVES
So far we have been treating natural incentives as if they were independent.

However, they also interact. The variety incentive combines with several others.
Animals fed to satiation on one type of food will Start eating again if new food
is presented (arousing the variety incentive) or another animal comes in and
Starts eating (arousing the impact incentive). Animals sexually satiated in regard
to one partner will become sexually active again if a new partner is presented
(see Schein & Haie, 1965). Variety also increases sexual responsiveness in hu-
mans. So, apparently, does the impact-anger incentive. Barclay and Haber
(1965) and Barclay (1969) showed that insulting students and making them
angry resulted in their being sexually aroused, and vice versa: subjects who were
sexually aroused showed an increase in aggressive thoughts or impulses (Barclay,
1971). This cross-linkage was unique to sex and aggression. Anger arousal did
not increase general drive level or other concerns, such as the concern for
achievement or affiliation; other types of arousal, such as anxiety, did not in-
crease sexual and aggressive thoughts. Clark (1955) also found a close link be-
tween sexual and aggressive thoughts in males during sexual arousal.
As mentioned earlier, the impact and variety incentives may combine to
produce the peculiar fascination that leads children and some adults to work at
moderately difficult tasks. They seek the variety of something new but also get
pleasure from producing the new experience themselves.
From the very beginning, natural incentives become embedded through

learning in a complex network of cognitions and expectations. Some of these af-
fect how motive dispositions are satisfied, as will be explained in the next sec-
tion. Others are elaborated cognitively. For example, as has already been noted,
impact experiences lead to people's conceiving of themselves as personally more
or less competent—as an Origin, to use deCharms's terminology, rather than as
a Pawn. That is, children not only have impact experiences; they also develop a
concept or Schema of the seif as Origin by generalizing across these experiences,
which may or may not be conscious. This self-schema may be thought of as a
complex or symbolic incentive that is the product of experience with certain
types of natural incentives.
Similarly, some people may develop a concept of themselves as dependent
on others if they are punished for novelty seeking and too much assertiveness so
that most of their natural gratifications come from contact with a caretaker.
One can only speculate as to how these incentive experiences combine with each
other and cumulate into various outcomes, since no direct research has been
done on the problem. In fact, at this stage there is not even good agreement as
to what the natural incentives are, let alone how they interact with each other.
• THE ROLE OF COGNITION IN THE

DEVELOPMENT OF INCENTIVES
As is clear in the case of the variety incentive, what gives pleasure naturally
changes continuously as a person's understanding grows. What seems complex
and interesting to a child is boring to an adult. The number of dimensions in
which an object or event can differ from expectation increases as understanding
grows. For example, Martindale (1975) has applied the notion of the variety in-
centive to aesthetic appreciation. He has produced considerable evidence sup-
porting the idea that in the art world, works that are moderately different from
what people have been habituated to are most appreciated: "Whatever eise it
must be, a work of art must at some degree be novel, original, or at least differ-
ent from preceding works of art" (Martindale, 1977). But repeated experience
with what is novel for works of art in one period gradually produces an intrinsic
pressure for more novelty in some dimension or another. Martindale has studied
English poetry from the time of Chaucer to the present and found that as he
predicted, there has been a steady increase in a composite, quantitative measure
of variability, as he designed it, in the style of poetry. However, the dimension
in which the novelty occurred shifted from one historical period to another. At
one time the novelty was in word length; at another time, in phrase length; at
another time, in the percentage of words occurring only once in a text; and at
still another time, in the strikingness of the contrasts introduced in the poetry.
Thus, the natural preference for moderate variety persists in poetry as a natural
phenomenon, but there has been an enormous cognitive development of the di-
mensions in which the variety can occur.
The same process affects the impact and consistency incentives. What a per-
son can have impact on as a child often is limited to physical grasping or push-
ing. One day, however, children discover that if they leave the room they can
have impact in the sense that a parent may worry about what has happened to
them. They learn that they can have impact by arguing successfully, by graduat-
ing from College, and by collecting prestigious possessions. Similarly, in the be-
ginning children develop cognitive expectations by observing consistency be-
tween their perceptions and physical reality: they see that a ball rolls off a table
and confirm that it hits the floor. If such a sequence fails to occur after it has
been built into a firm expectation, children are upset emotionally and seek to re-
duce their sadness-distress in some way.
Such cognitive expectations become more and more elaborate and differenti-
ated with experience, so what causes sadness-distress also changes greatly. Very
quickly normative expectations also develop in the social sphere. Children learn
that if they throw their food on the floor, their mother will be angry. At first
they may repeat this response several times simply to cönfirm the fact that what
they expect to happen will indeed happen. From here they go on to develop
very complex expectations of how to behave in the social area, which they nor-
mally confirm by behaving that way, because not to do so is innately distressing.
The need for consistency is the same at various stages of maturity, but how the
consistency is defined becomes enormously elaborated through experience and
cognitive development.
SYMBOLIC INCENTIVES, OR VALUES
Eventually people develop conscious values that guide their attitudes and
behaviors. Sometimes it is easy to see how these values could have developed
through cognitive elaboration from natural incentives, and sometimes it is not
easy. For example, most U.S. Citizens rank freedom as their most important
value (Rokeach, 1973). It can readily be seen how belief in freedom develops out
of the impact incentive if children are allowed a fair amount of freedom in pur-
suing that incentive. That is, having obtained a lot of satisfaction from impact,
they might easily develop the conscious thought that it is "good to be free" so
as to be able to have impact in all sorts of ways. It also can be understood how
people could come to value security consciously if they had experienced much
pain and fear in growing up. But how can people derive a belief in the impor-
tance of equality or salvation? Such values appear to develop out of the general
understanding people of a particular culture develop as to how the world works.
They have no obvious source in natural incentives. In other cases, a value may
have many sources. Money is commonly believed to be an important value or
social incentive, but money means different things to different people. For some
it means security; for others, the opportunity to have impact or to be considered
important; and for still others, a measure of how well they are doing at some
job (see McClelland, 1967).
Psychologists are uncertain about just what values are most important to
study. Rokeach (1973) has done a lot of interesting research on a representative
list of eighteen terminal (end-state) and eighteen instrumental values. Subjects

are asked to rank order each set of eighteen values in the order of importance
for their own lives. As Tables 5.4 and 5.5 show, considerable Variation exists
from one type of subject to another in regard to what is most important to
them. Salvation is ranked fourth in importance by U.S. women but twelfth in
importance by U.S. men. On the other hand, men rank a comfortable life fourth
in importance, whereas it is only thirteenth in importance to women. Physical
seientists regard it as much more important to be self-controlled than do hu-
manities majors, whereas the latter regard forgiving as much more important
than do the former.
The importance of such values also varies across eultures. National security
is second in importance to Israelis but only seventeenth to U.S. Citizens, a not
very surprising finding in view of the threats to IsraePs existence from the Arab
world.
Values have a strong cognitive component. They grow out of people's un-
derstanding of the world and of what their eulture regards as most important.
Values are conscious. Motives, on the other hand, are less cognitively elaborat-
Table 5.4.
TERMINAL VALUE MEDIANS AND COMPOSITE RANK ORDERS FOR U.S.
MEN AND WOMEN (after Rokeach, 1973)
Male Female
Terminal Value N = 665 N = 744 P
A comfortable life 7.8 (4) 10.0 (13) .001
An exciting life 14.6 (18) 15.8 (18) .001
A sense of aecomplishment 8.3 (7) 9.4 (10) .01
A world at peace 3.8 (1) 3.0 (1) .001
A world of beauty 13.6 (15) 13.5 (15) —
Equality 8.9 (9) 8.3 (8) —
Family security 3.8 (2) 3.8 (2) —
Freedom 4.9 (3) 6.1 (3) .01
Happiness 7.9 (5) 7.4 (5) .05
Inner harmony 11.1 (13) 9.8 (12) .001
Mature love 12.6 (14) 12.3 (14) —
National security 9.2 (10) 9.8 (11) —
Pleasure 14.1 (17) 15.0(16) .01
Salvation 9.9 (12) 7.3 (4) .001
Self-respect 8.2 (6) 7.4 (6) .01
Social recognition 13.8 (16) 15.0 (17) .001
True friendship 9.6(11) 9.1 (9) —
Wisdom 8.5 (8) 7.7 (7) .05
Note: Figures shown are median rankings and, in parentheses, composite rank Orders.
Table 5.5.
SIGNIFICANT VALUES DIFFERENTIATING HUMANITIES, SOCIAL SCIENCE, AND PHYSICAL
SCIENCE MAJORS AT FLINDERS UNIVERSITY IN SOUTH AUSTRALIA (after Rokeach, 1973)
Social Physical
Humanities Sciences Sciences
N = 103 N = 162 N = 198 Pa
Terminal values
A comfortable life 14.4 (17) 13.3 (14) 12.2 (13) .001
A world of beauty 10.9 (13) 13.0 (13) 12.9 (15) .001
Instrumental values
Ambitious 10.8 (14) 7.8 (6) 7.8 (6) .01
Capable 10.1 (12) 8.7(11) 8.0 (7) .05
Forgiving 7.8 (6) 8.2 (9) 9.7 (12) .05
Imaginative 8.4 (7) 12.8 (16) 12.4 (16) .001
Intellectual 9.1 (10) 11.4(14) 11.5(15) .05
Self-controlled 10.2 (13) 8.1 (8) 7.7 (5) .05
Note: Figures shown are median rankinj>s and, in parentheses, composite rank orders.
a
Kruskal-Wallis test.
ed, are tied more directly to natural incentives and emotions, and often are un-
conscious. Both motives and values shape behavior, but they should be thought
of as independent determinants of behavior. (This will be discussed further in
Chapters 6, 12, and 13.) Sometimes people speak of values motivating behavior
in the sense that an Israeli may be "motivated" by a concern for national securi-
ty. As has been repeatedly emphasized, however, it is more correct to say that
national security activates or helps determine a response, and to save the term
motivation for situations in which a motive is actually involved rather than some
other determinant of behavior.
Natural incentives may be cognitively elaborated into symbolic incentives, or
values, which guide and direct much of conscious behavior. On the other hand,
the motives built on natural incentives also continue to influence behavior at an-
other level. The point is dramatically illustrated by a physician's description of
the behavior of a patient damaged in the frontal lobe, where cognitive elabora-
tion takes place:
For example, one former patient, whom I will call Mr. Jones, had his right frontal
lobe destroyed by a tumor. Mr. Jones had difficulty with voluntary tasks that were
verbally requested, despite the fact that he understood what was wanted of him.
When asked to pick up his water glass, for example, he made no response. "Do you
understand what I want you to do, Mr. Jones?" I asked. "Yes. You want me to pick
up the glass." "Have you done that?" "No." "Can you?" "Yes." "Go ahead and
pick it up then," I urged him. No response. If he was thirsty, however, he would

pick up the glass and drink from it spontaneously. (Cytowic, 1981)
The man's more primitive, unconscious motivational System, built on the natural
incentive for drinking, was still intact and could activate the necessary response,
but the cognitively elaborated conscious intent to drink no longer functioned.
NOTES AND QUERIES

1. Perry's recurrent dream from In Cold Blood, described in Chapter 2, ended
up with a vision of an oral paradise in which there was abundant wonderful
food and drink. In what way could this be taken to mean that Perry had a
strong hunger motive? Which of the two types of hunger motives outlined
in Figure 5.1 does the dream most probably represent? Why? What types of
childhood experiences surrounding eating would be likely to produce this
motive?
2. If people eat a lot, does it mean they have a strong hunger motive? List
some reasons other than hunger that would lead people to eat excessively. If
people do not eat much, does it mean they have a weak hunger motive? In
the condition called anorexia nervosa, people have great difficulty eating
enough to maintain their weight. Can you imagine a learning scenario in
which they might have developed a strong motive that somehow is satisfied
by not eating?
3. Tying the achievement motive to the variety incentive creates a difficulty
over whether some discrepancies from expectation are more pleasurable than
others. In the Maddi (1961) experiment, if a sentence stem occurred either a
little earlier or a little later than expected, it produced positive affect. How-
ever, in the area of achievement it seems unlikely that an infant or an ani-
mal would like a slight deviation in the less complex direction as much as a
slight deviation in the more complex direction. Would the rats, after being
adapted to a complex environment for some time, spend more time in a
simpler environment? It seems unlikely. In much the same way it seems un-
likely that a child, having become adapted to playing with a complicated
toy, would prefer working with one that was simpler. There seems to be a
built-in bias toward getting pleasure from more complex or difficult tasks
that is not captured by the idea that any slight deviation from expectation is
pleasurable. How would you define the natural incentive, if it indeed in-
volves more than simple variety? Could we think in terms of a complexity
incentive?
4. List some ways in which you enjoy having impact. Do they involve primar-
ily physical activities or relationships with people? How often do they in-
volve aggression in the sense of intent to injure another person? Compare
the lists prepared by men and women. If there are differences, what do they
mean?
5. The eyes are said to be an important source of sign Stimuli in humans. For
example, a certain type of glance is supposed to elicit fear or anger and an-
other type, love and affection. How would you describe the difference be-
tween these two types of glances? Having made the distinction, how would
you determine by experiment whether you were right? Can you think of an
experiment that would indicate whether the emotional responses to these
glances were innate or learned?
6. In what ways could the increased production of sex hormones at puberty af-
fect the way in which natural incentives in the sexual area function? Con-
sider all the aspects of a natural incentive as outlined in Figure 5.1.
7. Do you agree with Freud that love is based on contact incentives, or do you
think it might be based on some other type of natural incentive? Why or
why not? What other type of natural incentive might be involved?
8. From the results of Sara Winter's study of nursing mothers, would you the-
oretically expect women to develop a stronger need for Nurturance (if it
could be measured) than men? Why or why not?
9. Whereas the present trend in motivation research is toward measuring dif-
ferent types of avoidance motives, make a case for the fact that in addition
there may be a natural incentive for cognitive consistency, which might de-
velop into a stronger concern for this goal State in some people than others.
How would you detect its presence?
10. The chapter treats values as independent cognitive derivations from natural
incentives. Does this suggest that important values might be named for the
same natural incentives for which the motives are named? How can you de-
cide what the important values are?
11. Freud (1927/1957) remarks in The Future of an Illusion that "there are two
widely diffused human characteristics which are responsible for the fact that
the organization of culture can be maintained only by a certain measure of
coercion: that is to say, men are not naturally fond of work, and arguments
are to no avail against their passions." After reviewing the evidence in this
chapter, do you agree with Freud's conclusions? How would you specifically
answer his contention that people are not naturally fond of work?
6
Measures of Human
Motive Dispositions
The Motivational Sequence

Demands
Effect of Increasing Demands on Performance
Incentives
Motive Dispositions
Aroused Motivation
Arousing Motives to Detect Their Unique Effects on Behavior

Effects of Arousing the Hunger Motive
Individual Differences in the Strength of the Hunger Motive
Measuring the Strength of Social Motive Dispositions
Coding Thought Content for the Achievement Motive
Coding Thought for the Power and Affiliation Motives
Adequacy of Various Measures of Individual Differences in Motive Strength
Alternative Measures of Motive Strength Evaluated According to the Criteria

for Good Measurement
Sensitivity
Uniqueness
The Effect of Conditions of Test Administration on Imaginative Stories
Reliability
Coding Reliability
Test-Retest Reliability
The Influence of the Set to Be Creative or Consistent
Validity-Utility
A Comparison of the Validity-Utility of Self-report and Fantasy Measures of
the Need for Achievement
172
THE MOTIVATIONAL SEQUENCE
What is needed now is a general model of motivated behavior. Experimen-
talists deal with motivation—with short-term situational influences like food, va-
riety, requests for obedience, or electric shock that arouse approach or avoid-
ance behavior immediately. Personality theorists or clinicians typically think in
terms of motives, that is, stable dispositions that organize or explain much of
what a person says and does. How do these two approaches fit together? What
exactly is a motive disposition, and how should individual differences in its
strength be measured? According to clinicians, some people, like Freud, have a
strong need for fame, recognition, or power. How do we determine how weak or
how strong the power motive is in different individuals? How is a disposition
like the power motive aroused by a Situation, and when aroused, how does it in-
fluence what the person does? These are the questions to which this chapter is
addressed.
Motives are based on emotionally arousing incentives, which were discussed
in the previous chapters. The incentives Start out by being natural in the sense
that they innately give rise to different types of positive or negative emotions. As
we have seen, however, their nature changes rapidly with learning. The moder-
ate degree of variety that produces positive emotional arousal at one time pro-
duces only boredom after exposure to new material. The pleasure children de-
rive from the impact of dropping a spoon wanes as they move on to seeking
bigger impacts.
Eventually a cognitive Schema develops that organizes for children the cate-
gory of situations that evokes different types of positive or negative emotional
arousal. From the subjective point of view, these cognitive Schemata may be
called goals in the sense that the person learns to seek out (or avoid) the situa-
tions involved. Looked at from the point of view of the observer, they are re-
ferred to as either positive or negative incentives, since the person is seen to ap-
proach or avoid the situations involved.
In growing up, some people may develop more elaborate Schemata associ-
ated with a given incentive than other people do. While all children Start out en-
joying having impact, some parents may strongly discourage this activity, so
their child does not develop much pleasure from it or develop a good concept of
how to attain pleasure in this way. Other parents may allow or even encourage
the activity, so their child develops a more elaborate schematic representation of
the many different ways in which he or she can get pleasure from having im-
pact. Some parents may begin to verbalize the desirability of pursuing power
goals by admiring their child for acting assertively or by encouraging him or her
to fight back. In time, as the previous chapter showed, the child may develop a
conscious value for being assertive or influential, although he or she may also
pursue power goals without placing a conscious value on them.
In either case, many more situations get connected through learning with
power goals or incentives. At home, a simple request to turn off the television
may produce a defiant, assertive refusal. At school, playing basketball may
evoke a strong competitive response. Even in a social Situation among friends,
173
children may find ways of calling attention to their importance. When a variety
of cues consistently arouses a class of incentives or goals, we may speak of a
motive having been formed. Obviously, there will be differences in the strength
of such a motive, depending on the experiences of the person. So it will be im-
portant to find a way to measure these differences in strength. The cues or Stim-
ulus situations that give rise to these motives may be referred to as demands,
since they regularly give rise to a motive State.
The way these facets of motivation fit together into a motivational sequence
is illustrated in Figure 6.1. Demands (A), or arousal cues, made in terms of in-
centives (B), will—if they contact an existing motive disposition (C)—lead to
aroused motivation (D), which combines with cognitions, habits or skills to pro-
duce the impulse to act, which combines with opportunities to produce action.
As Chapter 3 noted, some writers (Atkinson & Birch, 1978; Weiner, 1980a) use
the term motivation to refer to the end product of all the personal determinants
of action (including habits and cognitions), but we will restrict the use of the
term here to mean an aroused motive.
Figure 6.1 gives three examples to illustrate how the motivational sequence
operates. Hunger is described in the behaviorist tradition as a primary drive. It
is primary only in the sense that recurrent physiological demands (for exampie,
low blood sugar) get associated very early and very strongly with the incentive
of eating, which involves taste sensations. (See Figure 5.1.) In time, through
learning, even a simple cue like noticing that it is time for supper is enough to
contact or arouse the eating motive complex, which then arouses the intent to
eat. Ordinarily we do not think much about individual differences in the
strength of the eating motive, although the overweight people described in
Chapter 5 spent much more time thinking about eating, preparing to eat, trying
out different kinds of foods, and so forth. Whether the aroused motive to eat
leads to actual eating obviously depends on opportunities (whether food is avail-
able), cognitions (whether the person eats the available food), and habits
(whether the person usually eats at this time).
The second exampie in Figure 6.1 concerns the power motive. Suppose a
girl is about to play a tennis match. Just being in a competitive Situation is
probably enough of a demand to contact the incentive of having impact or act-
ing powerfully, or her coach may actually verbalize the demand by telling her to
"get in there and fight." The coach's mentioning the incentive does not automat-
ically arouse the intent or desire to win. It depends on whether the incentive
used is part of the power motive System the girl has developed. If she has a
greater need for power, then the incentive will contact it and lead to the intent
to win. Whether her aroused power motivation leads to aggressive acts will de-
pend on the Situation, or opportunities (whether she is winning or not); her val-
ues (whether she thinks it is wrong to display temper on the court); and her
habits (whether she is accustomed to throwing her racket away when she
misses).
The third exampie is drawn from the type of material analyzed by Freud
and other clinicans in order to demonstrate that the model works equally well
External, in the Environment Internat, in the Person Cognitions Habits, Opportunities

(Values) Skills I
Stability:
Temporary A. Demands • D. Aroused Motivation • *— Impulse—^ Actions
to Act
1. Physiological or 1. Hunger, desire to eat. 1. Eating.
social cues associated
with eating 2. Desire to win. 2. Aggression.
("It's time to eat").
3. Desire to avoid Nineveh 3. Jonah goes into ship,
2. "Get in there and fight!" and Lord's command. ship's hold, sea, belly
(tennis match). of the whale.
3. Lord telling Jonah

to go preach
in Nineveh.
Stable B. Incentives -*• C. Motive Dispositions
1. Eating 1. Being more concerned

(taste sensations). about eating.
2. Having impact
(anger, recognition).
2. Being more concerned
about having impact
Io
3
(need for power).
3. Danger in Nineveh.
3. Need for security.
Situational
Motivational
Personal
Motivational f
Factors Factors
Figure 6.1.
Elements in the Motivational Sequence Leading to Action.
for such analyses. In the Bible there is a story about Jonah being swallowed by
a whale in a kind of dream sequence. The Lord calls Jonah to preach against
wickedness in the city of Nineveh. Jonah refuses and runs away in a ship, which
gets caught in a storm. He falls asleep in the hold, eventually gets thrown into
the sea to appease the gods, is swallowed by a whale, asks the Lord for forgive-
ness, and is vomited up on the dry land, safe and sound, after three days in the
whale's stomach. Jonah's motive is not hard to identify: he is running away and
going from one hiding place to another—first in the ship, then in its hold, then
in sleep, then in the ocean, and finally in the whale's belly.
The motivational sequence leading to his actions can be conceptualized in
the usual way, as Figure 6.1 illustrates. The demand is that he go preach in
Nineveh, but Nineveh apparently is a negative incentive for him: it represents
danger. This negative incentive contacts his need for security, which arouses the
intent to avoid Nineveh and the Lord's command to go there. The actions
Jonah takes all seem designed to satisfy his need for security. The only thing
unusual about the story is that in real life a whale's stomach would not provide
an opportunity for hiding.
As we shall see, thoughts are good places to look for motivational effects,
because they are not as dependent as actions are on values, skills, and opportu-
nities in order to be expressed. In real life, Jonah's need for security may have
led only to the action of not going to Nineveh; we would have been left trying
to determine why he did not go, when there might have been many reasons. He
might like it where he is; he might have had a girlfriend somewhere eise; he
might not be able to get work in Nineveh. It is only as we get more of his
thoughts represented here by such dream acts as sleeping in the middle of a
storm or being swallowed by a whale that we have an easy time inferring what
his primary motive was.
The Jonah story also illustrates that fear or avoidance motives built on
negative incentives operate in the same way as approach motives. They are
triggered by cues, which evoke a negative incentive (the threat of pain or
discomfort), which contacts a stronger or weaker need to avoid that Situation,
which arouses in turn a stronger or weaker immediate impulse to avoid the
danger.
Figure 6.1 also illustrates how two facets of the motivational sequence can
be thought of as external to the person in the environment, namely, the de-
mands and the incentives. In an experiment, investigators usually manipulate
these two variables: they ask subjects to do something (a demand) for a variety
of possible incentives (to please the experimenter, to earn some money, to show
how good they are, and so on). Presented incentives become goals inside a per-
son only when they contact a relevant motive disposition. Thus, the motive dis-
position and the aroused motive are the two variables that can be thought of as
inside the person. The facets of the motivational sequence may also be classified
in terms of whether they are temporary or stable. Demands and aroused motives
are specific to a particular Situation and therefore are temporary. Incentives like
money, food, or the challenge of moderate risk are relatively stable aspects of
the sequence, as is the motive disposition the subject brings to the experiment.
Measures of Human Motive Dispositions 111
Each of these four aspects of the motivational sequence will now be discussed in
more detail.
Demands
Demands may be classified as biological or social. For a long time psychologists
concentrated on studying biological demands, such as hunger, thirst, and pain
from electric shock, because they appear to be simpler and easier to define and
manipulate systematically. Strictly speaking, these demands are not motives
when they first occur. They may produce restless behavior, as when a baby first
experiences arousal from low blood sugar levels, but the behavior does not sat-
isfy the criteria of being motivated—that is, of being organized or directed
toward an anticipated goal State—until the baby has formed a learned connec-
tion between a demand and what will satisfy it, namely, eating food. A biologi-
cal need caused by a lack of food will, through learning, very quickly and de-
pendably cue off a desire to get food. So it is proper to speak of a hunger
motive having developed out of this regulär biological demand, but it is never-
theless important to realize that the demand and the motive are separate. In ex-
actly the same way a social demand, such as a father's command to a little boy
to drink his milk, does not at the outset automatically lead to the impulse to
drink the milk, or to behavior that could be classified as motivated. The boy
must learn that the command ties in with some incentive-motive System before
he will obey it. Many demands acquire motivating characteristics quite rapidly,
but they do not have them initially.
Social demands can be just as powerful as biological demands. The basic
tool of psychologists, the experiment, would not work unless the force of social
demands was very great. Think of the tens of thousands of undergraduates who
faithfully do even the most bizarre things at the request of psychological experi-
menters year after year! It would be a rare biological demand that could pro-
duce such consistent behavior from human subjects. The ready compliance of
the hypnotized subject to the experimenter's command is well known, but as
Martin Orne (1962) has pointed out, the average subject, even in a nonhypno-
tized condition, also will do almost anything the experimenter asks him to do.
The experimenter has only to say, 'This is an experiment. Go and lie down on
the table." Almost every subject will do so without hesitation.
Remember that none of these demands is effective unless it contacts some
incentive-motive complex. Just because an experimenter asks some College stu-
dents to try to do as well as they can on some tasks given to them does not
mean that all the students will accept the goal of doing well. Whether they ac-
cept it or not depends on the strength of their need for Achievement (see Chap-
ter 7). The reason social demands often are so effective is that they call on the
natural consistency incentive discussed in Chapter 5. People do what they are
asked to do because it is satisfying to do what is expected or what they and oth-
ers have often done before in the Situation. In any case, the effectiveness of de-
mands or instructions for arousing intents must always be evaluated in terms of
the incentives and motive dispositions present in the Situation.
Effect of lncreasing Demands on Performance. In general, stepping up de-

mands increases Performance. This is why parents urge their children to do well
in school or to "hitch their wagon to a star," and why coaches spend time urg-
ing their players to try harder. Psychologists demonstrated long ago that the
more demanding the task, the more effort the subject puts out (Ach, 1910).
More recently, Locke, Shaw, Saari, and Latham (1980) reviewed data showing
that setting higher goals leads to better Performance than setting easy goals in
all sorts of tasks from logging to typing to simple clerical tasks like adding and
card sorting. In other words, people normally do better when greater demands
are made on them to perform.
Does this justify setting impossibly difficult goals for people to get them to
put out more effort? Not really, because the relationship between demand and
Performance is more complicated than such a conclusion implies. In 1908
Yerkes and Dodson concluded that increasing demands lead first to an increase
and then to a decrease in the efficiency of Performance. Animals' problem solv-
ing seems to be best when their drive strength is moderate rather than extreme.
So if demands increase drive strength or aroused motivation, they could have a
negative influence at very high levels.
In a typical study, Birch (1945) observed how chimpanzees solved problems
to get food when their demand for food was increased by depriving them of it
for up to forty-eight hours. For instance, in one task the caged chimpanzees had
to use a short stick to bring in a string attached to a longer stick that could be
used to reach the food. After going without food for only two hours, the chim-
panzees were desultory in their approach to problems like this; most of the time
they failed altogether to solve them in the maximum allotted tirre of one hour.
The average time taken for all tasks (counting one hour for failure to solve the
task) was about thirty-four minutes. As the chimpanzees got hungrier and hun-
grier, failures became fewer, and the average time taken to solve the problems
dropped after twenty-four hours of food deprivation to around eleven minutes.
However, if they were deprived of food for forty-eight hours, their efficiency feil
off, and the average time taken to solve the problems rose to about twenty-one
minutes.
Inefficiency at low and very high demand was qualitatively different, how-
ever. When the animals were not very hungry, they simply did not try very
hard to solve the problems, because they were not very interested in the food.
When they were extremely hungry, on the other hand, they were overexcited
and focused so directly on the goal that they could not take the time to figure
out how to solve the problem. After forty-eight hours of going without food, the
chimpanzee Art persistently reached directly for the food with the short stick, in
one instance making as many as twenty-six consecutive unsuccessful sweeps. The
intensity of the demand prevented him from taking the time to "think out" the
problem. It lowered his efficiency by narrowing the focus of his attention.
People put under great pressure to solve difficult problems make more er-
rors, and their efficiency decreases (see Heckhausen, 1980; Schneider & Kreuz,
1979). This type of observation led Thurstone (1937) to suggest that there may
be an optimal level of demand to get the highest quality of Performance for
tasks of different difficulty. Increasing demand at first facilitates the utilization
of cues relevant to performing the task (Brunei*, Matter, & Papanek, 1955), but
very intense demand may distract people, cause them to become anxious and
not see relevant cues, lead them to make errors (Patrick, 1934), and perhaps
even quit trying altogether. In short, the relationship between increasing demand
and efficiency is curvilinear, assuming an inverted U shape; also, the more com-
plex the task, the lower the optimal level of demand.
Stennett (1957) reported a representative study showing this effect. Subjects
were asked to try to keep noise at a minimum by turning a knob to a "null
point" whenever noise occurred in an auditory tracking task. Their Performance
score was the number of seconds in a one-minute trial in which they managed
to keep "on target" (no noise). Muscle tension and skin resistance (to measure
palmar sweating) were recorded from the unused left arm to get estimates of
physiological activation. Subjects were tested under three types of demand for
good Performance. In the low demand condition, subjects were told from time
to time that they should track as usual, but that the experimenter was simply
calibrating the apparatus and could not record their score. In the optimal de-
mand trials, the experimenter encouraged the subjects to perform better and of-
fered them small sums of money (twenty-five cents) to improve their scores. In
the high demand trials, they were told the best score that had ever been made
with the apparatus and urged to break the record. They were offered five dollars
for each trial in which they equaled or bettered the record, or two dollars plus
avoidance of electric shock to their leg. As Figure 6.2 makes clear, the activa-
tion level as measured by muscle tension increased regularly as the level of de-
mand for better Performance increased, but Performance was best under moder-
ate levels of demand and activation. Other investigators (for example, Duffy,
1962) have shown that a variety of factors enter in to complicate the simple re-
lationship shown here and assumed by the Yerkes-Dodson law. In fact, surpris-
ingly few studies confirm its existence, partly because the investigators have not
conceived of demand as only one aspect of the motivational sequence to be con-
sidered and kept under control.
The distracting effect of too much demand is also illustrated by research on
the effects of induced muscular tension on learning. In these experiments sub-
jects see three-letter nonsense syllables like BMT exposed briefly in a small win-
dow. They are to learn to anticipate and say out loud what letters will be ex-
posed next when they see a given nonsense syllable. They perform this task
under normal conditions and also when squeezing a hand dynamometer at vari-
ous fractions of the maximum squeeze they can make on the instrument. As
Figure 6.3 shows, a mild squeeze on the dynamometer, at about one quarter of
the maximum grip, facilitated learning; however, when the subjects were forced
to squeeze at about three quarters of their maximum grip, learning actually feil
below what it was when no squeezing was required (Courts, 1939). The induced
muscular tension can be thought of as equivalent to increasing demand such as
would be mobilized by stronger, external incentives; if this is so, the expected re-
lationship is obtained; increased demand first increases Performance and then de-
creases it. Very probably it decreases it, because, as in the case of the chimpan-
zee Art, the increased pressure distracts the subject from concentrating on the
Performance itself.
60
Performance
O
!r 50
200 |
T3 <
C
o
f
S? 4 0 150
Activation Shown as Muscle Tension
100 £
Low Optimal High

Demand
Figure 6.2.
Relationship Between Increasing Demand, Activation Level, and Performance. Two values for the
optimal condition have been averaged for simplicity in presentation (after Stennett, 1957).
It is important to remember, however, that demands have an effect on per-

formance only if they contact motives in the individual, as Figure 6.1 makes
clear. Laboratory studies such as those just reviewed, or the field studies sum-
marized by Locke (1968, 1975) show the effects of demands on Performance
only because subjects are willing to commit themselves to trying to do what the
experimenter asks them to do (that is, learn to anticipate nonsense syllables and
squeeze a dynamometer). In real life, if demands are set too high, people may
give up trying altogether or become so anxious that they perform very badly.
Incentives
Incentives are emotionally arousing stable characteristics of the environment or
person-environment interactions that people seek out (positive incentives) or
Dynamometer Tension
Figure 6.3.
Relationship Between Number of Letters Correctly Anticipated and Dynamometer Tension (after
Young, 1961, after Courts, 1939).
avoid (negative incentives). Chapters 4 and 5 discussed at length the various

forms incentives take. They differ from demands in that they are less specific to
a particular time and place. Suppose an experimenter says to a group of sub-
jects, "You are to work on these arithmetic problems, and I will give a prize of
$2.50 to the one of you who solves the most problems in twenty minutes." The
experimenter is making a demand at a particular time and place for perfor-
mance from some particular individuals but is using a generalized incentive—
money.
Atkinson (1958) carried out such an experiment in which money and an-
other incentive were introduced. Ten to twenty undergraduates were assembled
in groups and told they were going to compete for monetary rewards by seeing
who could solve the largest number of arithmetic problems in the time allotted.
Along with a test booklet, students were given an instruction sheet telling them
how large a money prize they might win and what their chances of winning it
might be in terms of how many others they were competing with for it. The
money prize offered was either $2.50 or $1.25 for the best Performance, and the
chances of winning it were for some students one out of twenty, for others one
out of three, for others one out of two, and for others three out of four. That is,
in the last instance the three best performers out of the four competing all
would get the prize.
As Table 6.1 shows, those who were competing for $2.50 performed signifi-
cantly better than those who were working for a possible $1.25, regardless of the
odds under which they were working. Notice also that a higher money reward
maintains higher Performance, particularly when the expectancy of winning is
very low (odds, one in twenty) or very high (odds, three in four). In other
words, money seems to work best when other incentives are not operating—in
this case, the achievement incentive, which is optimal when there is a moderate
chance of winning. As the means at the bottom of Table 6.1 show, Performance
is best when the achievement incentive is optimal—when the person has a fifty-
fifty chance of getting the prize (see Chapter 7). If the odds are very long (one
in twenty), the achievement incentive has less attraction, because the person is
unlikely to win. If the odds are short (three in four), it also has less attraction,
because nearly everyone will win. In these two situations where the achievement
incentive was less important, the larger money reward had the most effect.
When the achievement challenge was greater at moderate odds, the larger
money reward had less influence. The reverse also was true: when the money in-
centive was less important ($1.25), the achievement incentive—as represented by
the difference in Performance at one-in-twenty versus one-in-two odds—had a
stronger effect than when the money incentive was more important ($2.50).
Money as an incentive is misleadingly concrete. Psychologists use it in ex-
periments because it is easier to manipulate than other incentives. However, it
always takes on the meaning of the particular motivational Situation in which it
is employed as an incentive. For example, even in the experimental Situation re-
ported in Table 6.1, it is doubtful that the students were working harder just be-
cause they were offered a somewhat larger money prize. Is it reasonable to be-
lieve that $1.25, even in noninflated money, would mean that much to a College
undergraduate? Instead, the larger prize offered by the experimenter simply may
have meant to the students that a prestige figure placed a higher value on their
performing faster, and their desire to please the experimenter was proportion-
Table 6.1.
EFFECT ON PERFORMANCE OF MONETARY INCENTIVE AND EXPECTANCY
OF WINNING (after Atkinson, 1958)
Expectancy <of Winning

(Difficulty)a
Incentive
Incentive 1/20 1/3 1/2 3/4 Means
$2.50 N = 18 18 18 18
Mean score 50.3 51.8 54.1 51.7 52.0
$1.25 N = 13 13 13 13
Mean score 45.0 50.8 52.0 45.7 48.4
Difference 5.3 1.0 2.1 6.0 4.6
Expectancy means 48.0 51.4 53.2 49.2
a
Fifty is an average score for the group on two tasks—arithmetic and drawing A"s in circles. Incentive dift'erence is
significant at p < .01, difficulty differences at p < .05.
ately increased. In fact, in another experiment Atkinson and Reitman (1956) re-
ported that girls with high affiliative needs were particularly likely to work
harder when a monetary incentive was offered—presumably because they were
precisely the ones who would be more interested in pleasing the experimenter.
Motive Dispositions
Demands are couched in terms of incentives that must contact motive disposi-
tions if motives that ultimately influence action are to be aroused. What exactly
are motive dispositions? We have described them as representing individual dif-
ferences in the strength of anticipatory goal states or the networks of associa-
tions built up around natural incentives. If a boy—through encouragement or on
his own—gets a good deal of pleasure from doing well at moderately challeng-
ing tasks, he will tend to develop a stronger achievement motive in the sense
that more situations will cue off for him, and more frequently, the anticipation
of getting satisfaction from doing such things.
The possibility of satisfaction from such an incentive Situation is seen inter-
nally as a goal. The goal may not be conscious. The boy may not know why he
is attracted to working on challenging tasks. He—or we as observers—may be
aware that he works more energetically at such tasks, is more aware of the pos-
sibilities for satisfaction from this type of Performance, and learns more quickly
to do whatever is necessary to get this type of satisfaction. That is, as in the be-
haviorist tradition, we infer the presence of a central motive State when a per-
son is more active or energetic in pursuing a goal, more sensitive to cues relat-
ing to the goal, and quicker to learn what is needed to get to the goal. In other
words, motives drive, Orient, and select behavior. So we may define a motive
disposition as a recurrent concern about a goal State that drives, Orients, and se-
lects behavior. The word concern in the definition gives away the fact that we
are primarily interested here in human beings; as we shall see, concerns can be
directly measured in the thought processes of humans, but they must be in-
ferred from the actions of animals. Concerns can be inferred from actions, but
it is much more difficult and hazardous to do so than to measure the concerns
directly.
The goal State that defines a motive can be the same as or different from the
incentive present in the demand that is supposed to motivate behavior. Thus, in
Table 6.1, the achievement incentive (defined as moderate challenge) will acti-
vate individuals who have a stronger concern about this incentive—namely,
those we say have a stronger achievement motive—more than those who do not
have such a recurrent concern (see Chapter 7). For the latter the incentive does
not contact the achievement motive disposition, so if the person performs at all,
it must be out of some other motive, such as the desire for social approval or
the motive built on the consistency-conformity incentive. The problem in all of
this is to measure recurrent concerns or motive dispositions, since now we have
moved from the external world to inside the person, and it is harder to know
just what is going on inside a person than what is going on in the environment.
This is the main problem to which we soon will turn our attention.
Aroused Motivation
The final outcome of a motivational sequence is an aroused motive to act in

such a way as to satisfy the motive disposition that has been engaged by de-
mands or arousal cues referring to some incentives relevant to the motive. This
motivation to act combines with cognitions (values), skills, and opportunities to
produce the particular acts that occur.
Often in real life, all these determinants of action are implicit in the form of
a very specific behavioral intent, such as when a man says he cannot find a shirt
and "I want to go to the störe to buy a shirt." From such a stated intent, it is
often safe to predict that he will walk out of the house, get into a car, drive to
the center of town, park, put a dime in the parking meter, walk into the störe,
pick out a shirt, pay for it, bring it home, and put it on. Notice that we can
predict action in this case from the motivational intent because we know there is
a störe where he can buy a shirt, that he can drive a car, that he has the money
to buy the shirt, that he has the time to do it now, that he values wearing a
shirt, and so on.
On the other hand, Statements of general intent, such as "I want to do well
in school," are of little value in predicting action—in this case how well the per-
son will do—because the other determinants of action are not so well known,
such as the person's academic skills, the opportunities provided for studying ver-
sus enjoying life, and so on. Furthermore, even the expression of such an intent
is influenced by factors other than motive dispositions, such as the positive value
people place on academic success, the social desirability of saying one wants to
do well in school, and so on. In short, Statements of conscious intent are gener-
ally not good indicators of an aroused motive, because they are the product of
many factors other than motive strength.
Psychologists interested in understanding intents usually have relied on sim-
ple Statements of wishes or choices and concentrated on the situational factors
that influence them. For example, Atkinson (1957) worked out a System show-
ing that the intent or desire to choose an alternative is a Joint function of motive
strength multiplied by attractiveness multiplied by the difficulty of attaining the
alternative. In commonsense terms, children's desire for a piece of candy is a
Joint function of their hunger, how big and delicious the candy is, and how
much they must pay for it. Atkinson's model will be explained in Chapter 7.
The point to keep clearly in mind here is that this kind of intent or impulse to
act is not identical with the concept of an aroused motive, as shown in Figure
6.1. It may, under certain very specific conditions, indicate the presence of an
aroused motive, but ordinarily such an intent is more complexly determined.
Demands can easily influence aroused motives. In fact, most of the time we
act as if all we have to do to get somebody to do something is ask them to do
it. And so far as many normative situations are concerned, we can indeed rely
on demands' arousing an impulse to act, without worrying about what incen-
tives and motive dispositions are involved. A father teils his daughter to drink
her milk, and she does. A boss asks a worker to get back to work, and he does.
Actually, behind the success of such demands lie various incentives and motives,
but for practical purposes they can be ignored until something goes wrong and a
demand does not produce the expected intent or aroused motive.
Over forty years ago psychologist Kurt Lewin (1935) and his students got
interested in what appeared to be a natural impulse to complete an interrupted
task. They allowed subjects to complete some tasks and stopped them in the
course of performing others; they found that the subjects remembered the in-
completed tasks better than completed tasks and returned to resume working on
them if given the opportunity. What appears to be responsible for the effect
(named the Zeigarnik effect after its discoverer) is the implicit demand, accepted
as a norm by most adults, that people should finish a Job that has been started.
Shifting the demand character of the Situation alters the motive aroused. If the
experimenter suggests that the subjects are to show how well they have done,
they tend to recall more completed tasks (Atkinson, 1955). If subjects are asked
to imagine themselves failing before a large audience in a directed daydream, an-
other implicit demand is implanted—to prove they are not failing—and they re-
call more completed tasks than uncompleted tasks (Russ, cited in Heckhausen,
1967).
Aroused motive states need not be conscious, as Chapter l's review of
Freud's work made clear. They can influence behavior without a person's being
aware that an unconscious wish existed until later evidence reveals its presence.
This fact provides still another reason why conscious motives or stated interests
should not always be taken at face value as indicators of the strength of aroused
motives.
• AROUSING MOTIVES TO DETECT THEIR

UNIQUE EFFECTS ON BEHAVIOR
Motive dispositions are the obvious key to understanding a motivational se-

quence, for without knowing what motives people bring to a Situation, it is im-
possible to know how they will react to a demand or an incentive. Laboratory
experiments on motivation can be carried out, because most subjects bring mo-
tives to the Situation that the experimenter can appeal to in getting them to do
what he or she wants them to do. The fact that subjects sometimes do not do
what they are expected to do calls attention to the fact that individuals differ
even in the motive dispositions they bring to an experiment. What are some of
these motives? How can individual differences in their strength be measured?
One way to detect individual differences in motive strength is to find some
behavior that is the unique effect of the motive and use it as an index of the
strength of the motive. Thus, if we could show that whenever people were angry
their heart beat faster, and the angrier they were the faster their heart beat, we
might try to use heart rate as an index of how angry a person was. However,
the effect has to be unique to the State that produces it. Otherwise we make mis-
takes in inferring backward from effects to their cause. We can infer a person is
angry from an increased heart rate if anger—and anger only—produces such an

eifect, which is clearly not the case. Increased heart rate can also result from
fear or sexual arousal.
This is the difficulty in trying to reason from actions to the motives that
produced them. As we have repeatedly emphasized in discussing Figure 6.1,
many factors other than aroused motives enter into determining an action like
eating or doing well in school. Yet people continuously make the mistake of in-
ferring that because people are doing well in school, they have a strong need for
achievement, or that because others eat a great deal, they are always hungry. In
fact, people may do well in school because they are intelligent (have academic
talent); because they have developed good study habits; or because they are mo-
tivated by a variety of goals unrelated to achievement, such as the need for so-
cial approval or the need to be considered an important person.
Such considerations led psychologists to seek unique effects of aroused mo-
tives in perception or apperception (that is, fantasy). As Chapters 1 and 2
pointed out, Freud and other clinicians repeatedly demonstrated that fantasy—
particularly dreams—provided excellent opportunities for inferring the presence
of different types of motives. This is because fantasies are less influenced by non-
motivational determinants like skills and opportunities than actions are. People
can dream about doing all sorts of things they have neither the skill nor the op-
portunity to do in real life. Thus, it is safer to infer from the fact that a girl
dreams about being president of her class that she has a desire to be important
than it would be if she actually were president of her class: in the latter instance
the behavioral result might be determined primarily by her social skills or the
lack of good alternative candidates rather than by her desire to be important.
In reviewing the behaviorist tradition we found that Skinner had also con-
cluded that operants—that is, "spontaneous" responses—reflect goal-directed
striving more directly than respondents do for the same reason. Respondents, by
definition, are more determined by the external Stimulus Situation, whereas oper-
ants are determined by whatever cues they carry inside themselves. In Skin-
nerian terminology, fantasies are "covert operants." They have the advantage of
all operants in more sensitively reflecting motivation. Furthermore, since they
are covert (that is, thoughts rather than overt acts), they have the further ad-
vantage of not depending on skills or opportunities.
Effects of Arousing the Hunger Motive

Why not arouse motives experimentally and look for their unique effects in fan-
tasy? Atkinson and McClelland (1948) did just that in an early study that com-
pared the content of imaginative stories written after subjects went one, four,
and sixteen hours without food. The objective was to detect precisely how fanta-
sies of hungry young men differed from their fantasies when they were not hun-
gry. The experiment was done in such a way that the men involved did not real-
ize that the effect of hunger on imagination was being investigated; otherwise,
they might have written the kind of stories they thought hungry people should
write. In other words, the experimenters tried to eliminate a nonmotivational or
cognitive determinant of the fantasy responses. This was done by assigning some
submariners to quarters late in the afternoon so they could sleep late and partic-
ipate in some psychological research the next morning, some sixteen hours after
their previous meal. Other subjects wrote stories one hour after they had fin-
ished the noon meal, and again four hours afterward (just before supper). The
content of the stories was searched to identify fantasy elements that were pres-
ent when the hunger motive was strong and not present when it was weak.
The results contained a surprise, as Figure 6.4 illustrates. Hungrier men did
not have fantasies of eating a lot of delicious Steaks and desserts, as one might
have supposed. Conventional wisdom assumes that hungry people dream of
food, so that if people think a good deal about eating, they must be hungry. But
this does not turn out to be true. The category labeled goal activity in Figure
6.4, which represents thoughts about eating, declines in frequency as hunger in-
creases. In contrast, the category labeled instrumental activity, which has to do
with activities or objects associated with getting food, increases with hours of
food deprivation. Stories dealing with food deprivation—particularly with black
market activities to overcome shortages—increased. After the fact, it is easy to
understand how it is more adaptive for hungry people to think about ways of
getting food rather than just passively dream of eating. In another part of the
experiment the subjects were shown a very dim slide, supposedly of three objects
on a table. The hungrier subjects more often reported that they were knives,
forks, spoons, and plates than the less hungry subjects, but they did not more
often mention eatables.
So an aroused motive State—namely, hunger—has some unique and rather
unexpected effects on fantasy. It is doubtful if these effects would have been ob-
tained if the subjects had known they were purposely being deprived of food.
For Sanford (1937) has shown that subjects who fast voluntarily for an experi-
menter do think more often about food. They do what they expect they should
do. Keys (summarized in Sherif, 1948) reported that conscientious objectors in
World War II voluntarily undergoing starvation spent a good deal of time fanta-
sizing about things to eat. However, these fantasies also might have an adaptive
significance, since if the visions of eating had become strong enough, the volun-
teers might have broken their fast. In other words, in this case the subjects' cog-
nitions or expectations were quite different from those of the submariners, who
did not know that hunger was the object of study.
Individual Differences in the Strength of the Hunger Motive

Atkinson and McClelland (1948) also developed an overall "need for Food"
score based on giving one point for the presence of a characteristic in a story
that appeared more often after sixteen hours than one hour of food deprivation
and subtracting one point for a story characteristic that decreased in frequency
from one to sixteen hours of food deprivation. This score averaged 0.74 for sub-
jects who had gone one hour without food, 1.57 for subjects who had gone four
hours without food, and 4.05 for subjects who had gone sixteen hours without
eating. In contrast, when subjects were asked to rate how hungry they were,
100
1 90 — ^ ^ Food Imagery
2 80-
. ^ ^ ^ ^ Instrumental Activity
U
70 -
I 60
c
- Food Thema
o 50 - ^ Deprivation Thema
40 -
^ Goal Activity
3 30 -
CD
20 -
10 -
i i 1
1 Hour 4 Hours 16 Hours
Length of Food Deprivation
Figur'e 6.4.
Percentage of Subjects Showing Selected Food-Related Story Characteristics as a Function of
Increasing Hunger (after Atkinson & McClelland, 1948).
they judged themselves to be significantly more hungry after four hours than
after one hour of going without food, but after sixteen hours they were no hun-
grier than after four hours. Thus, self-ratings of the strength of the hunger mo-
tive provided a less sensitive index of hours of food deprivation than did the
measure based on fantasy.
The next logical step would have been to obtain need for Food scores from
subjects when the hunger motive was not aroused—say, two or three hours after
a meal—and to use them as a measure of individual differences in the strength
of the motive. That is, people who scored very high in the need for Food even
when the motive was not aroused might be assumed to be chronically hungry,
so to speak. At least, they chronically think like hungry people do—namely,
about the absence of food or about getting food. Such people might be further
expected to eat more or eat more often than people for whom the need for Food
score is low.
Such a study has not been carried out, although Schachter and his associ-
ates have extensively investigated overeating, which might be considered a sign
of a chronically aroused hunger motive. They too obtained a somewhat surpris-
ing result (Schachter, 1971b). Overweight subjects are more influenced by exter-
nal cues than underweight or normal subjects are, as was noted in Chapter 5.
They eat more when it is time to eat or when food cues are present than normal
subjects do. But they eat less and find it easier than subjects of normal weight to
skip meals and fast when external cues are not present. Table 6.2 gives a typical
result. In this experiment subjects were given a break after doing some work in
the laboratory and told they could eat as many Sandwiches as they liked during
the break. In one condition, one Sandwich was put out and the subjects were
told there were others in the refrigerator if they wanted more. In the other con-
dition, instructions were the same except three Sandwiches were put out. In the
one-sandwich condition, the overweight subjects ate fewer Sandwiches—about
1.5 on the average—than the normal subjects, presumably because when they
ate up one Sandwich no more food cues were present. However, when three
Sandwiches were offered, the overweight subjects ate significantly more than the
normal subjects and more than they did under the one-sandwich condition.
There are several possible explanations for this result, but one possibility not
investigated by Schachter and his associates is that overweight subjects are
chronically hungrier than normal subjects. If they are, their behavior could be
explained by Atkinson and McClelland's (1948) results. In that study also, hun-
grier subjects did not think more often about food cues when they were absent.
Thus, a hungrier subject might not be expected to imagine more vividly the
Sandwiches in the refrigerator and thus be led to get more of them to eat. On
the other hand, in another part of the Atkinson and McClelland experiment,
subjects were shown some vague objects exposed on a slide, which they were
told were a bowl and a cake. They were asked to judge which was larger, and
the hungrier subjects said the cake was larger more often than did the less hun-
gry subjects (McClelland & Atkinson, 1948). Thus, it would be expected that if
food cues were actually present, hungrier subjects could see them more vividly
and be led by the food incentive to eat more. For a comparable effect of need
on perception, see Bruner & Goodman (1947).
However, without a direct measure of the strength of the hunger motive in
obese and normal weight individuals (as obtained, for example, from the need
for Food score developed by Atkinson and McClelland), it is impossible to
Table 6.2.
EFFECTS OF PRESENT OR ABSENT EATING CUES ON EATING BY
OVERWEIGHT AND NORMAL WEIGHT SUBJECTS (after Schachter, 1971b)
Mean Number of Sandwiches Eaten When Offered

Eating Cues Absent Eating Cues Present
Subjects (One Sandwich Offered) (Three Sandwiches Offered)
Normal weight (N) 1.96 1.88

Overweight (0) 1.48 2.32
Difference (TV - 0) .48 -.44
P <.O5 <.O5
know whether the obese subjects are, in fact, chronically hungrier. All that can
be said is that the finding on the importance of external cues for obese subjects
is consistent with other research on the effects of hunger on perception and fan-
tasy. The Schachter research also illustrates once again the difficulty of using an
action measure to determine the strength of a motive, for overeating could be
caused by many factors other than a strong hunger motive. It might be the re-
sult of a defect in fat metabolism, hypothalamic malfunction such as character-
izes obese rats (see Schachter, 1971a), or eating habits of long Standing.
• MEASURING THE STRENGTH

OF SOCIAL MOTIVE DISPOSITIONS
Coding Thought Content for the Achievement Motive

The procedure adopted in the hunger experiment (Atkinson & McClelland,
1948) has been followed in a number of other studies designed to detect the
unique effects on fantasy of arousing different kinds of social motives. In the
first of these, attempts were made to arouse the achievement motive by telling
some young men that Performance tests they were taking would give an indica-
tion of their general intelligence and leadership capacities, and then manipulat-
ing the amount of success and failure they experienced on those tests (McClel-
land, Atkinson, Clark, & Lowell, 1953). Shortly afterward, subjects were asked
to write brief, five-minute stories to a series of four to six pictures for what was
sometimes described as a test of creative imagination and sometimes as a pic-
ture-story exercise (to deemphasize the testing character of the fantasy collection
technique). Some of the pictures were taken from the Murray Thematic Apper-
ception Test (1938), and some were chosen specifically for this study, such as
the one in Figure 6.5.
Characteristics of stories written after achievement motive arousal were
carefully compared with characteristics of stories written when no attempt to
manipulate motivation was made (called the Neutral condition). The chief differ-
ence was in a characteristic labeled achievement imagery, which was defined as
being present in a story when someone was involved in doing something better.
"Doing better" involved an implicit or explicit Standard of excellence—for ex-
ample, winning a contest; fixing a machine; attaining a unique accomplishment,
such as inventing something; or being concerned over a long period of time with
a Performance goal, such as in pursuing a career to become a doctor. Involve-
ment was indicated by an explicit Statement of a desire or intent to do well; by
affective concern over goal attainment (feeling good after success or bad after
failure); or by extraordinary efforts to achieve a goal, as in "the boy is working
very carefully on his essay."
Once a story was identified as being achievement related because of the
presence of such imagery, it was searched for other characteristics that differen-
tiated the stories written under achievement-oriented versus neutral conditions.
A number of such characteristics were found and carefully defined so that two
Figure 6.5.
Picture Used in the Standard Series for Eliciting Stories to Be Scored for n Achievement (McClel-
land, Atkinson, Clark, & Lowell, 1953).
trained judges could agree almost perfectly on their presence or absence in a

story. They are listed in Table 6.3, which also shows the percentage of stories
containing each characteristic under relaxed and achievement-oriented instruc-
tion conditions. Thus, 33 percent of the stories written under achievement-
oriented conditions contained explicit Statements of the need or desire to
achieve, as compared with only 10 percent of the stories written under relaxed
conditions.
The categories make sense when they are conceptualized as part of a prob-
lem solving sequence, as illustrated in Figure 6.6. In this sequence the person is
seen as trying to achieve a certain goal, defined here in terms of the achieve-
ment imagery definition. In thinking about the goal, the person may express a
specific need for it, as well as anticipate getting it or not getting it and feeling
good if he or she gets it or not good if he or she does not get it. The person
may be pictured as encountering obstacles on the way to getting the goal—
either in himself or herseif or in the environment—and as taking various ac-
tions to reach the goal or overcome the obstacle to the goal. In the course of
taking these actions, the person also may get help. It is as if the arousal of the
achievement motive increases the probability that the person will think about all
aspects of a problem-solving sequence.
To get an overall achievement motive score for an individual, the presence

in a story of any one of the categories in Table 6.3 is counted as + 1 . A story
containing no imagery related to achievement is scored — 1. Originally, working
at a task in a way that is not quite strong enough to be given a + 1 was scored
as 0 (for task imagery), but in later use both unrelated and task imagery were
scored as 0. A given characteristic was scored only one time per story, even
Table 6.3.
EFFECT OF ACHIEVEMENT AROUSAL ON CHARACTERISTICS OF IMAGINATIVE STORIES
(after McClelland, Atkinson, Clark, & Lowell, 1953)
Percentage of Stories Showing Characteristic

Under Two Conditions
Relaxed Achievemen t-orien ted
Story Characteristic and Example (N = 156 Stories) (N = 156 Stories)
Achievement imagery
"He is trying to run faster." 25 54
Need
"He wants to do well." 10 33
Positive anticipatory goal State
"He is thinking how great he will feel if he does." 4 15
Negative anticipatory goal State
"He is thinking how badly he will feel if he fails." 6 14
Positive goal State
"He is happy he succeeded." 2 15
Negative goal State
"He is unhappy he failed." 6 9
Actions
"He has been practicing every day for a week." 3 25
Obstacle in the environment
"His mother thinks he is crazy to spend so
much time running." 6 6
Obstacle in the person
"He worries so much it slows him down." 3 13
Help
"His coach has given him some good advice
on how to improve." 3 10
Achievement themea 18 45
Mean n Achievement scoreb 1.95 (SD = 4.3) 8.77 (SD = 5.3)
a
The whole story deals with an achievement theme. No other significant theme was mentioned.
b
The mean n Achievement score is obtained by adding one point for each characteristic coded per story and subtracting one point for
each story totally unrelated to achievement. All differences are significant except for obstacle in the environment and negative goal State.
Obstacle in the
Environment or
in the Person
Positive Anticipations
The Goal
Negative Anticipations
Positive Goal States

Actions
Negative Goal States
Figure 6.6.
A Problem-solving Model for Conceptualizing Motive Subcategories (Winter, 1973, after McClel-
land, Atkinson, Clark, & Lowell, 1953).
though it occurred several times, to avoid giving extra credit for more verbally
fluent subjects. However, in recent years it has also been the regulär practice to
correct for verbal fluency by adjusting the motive score statistically for the total
number of words in all the stories.
The sum of such characteristics present in all stories (a protocol) written by
a subject is called the "« Achievement score," following H. A. Murray's (1938)
practice of abbreviating need for Achievement to n Achievement. It is obviously
jointly determined by the number of achievement cues in pictures, instructions,
and individuals, as Figure 6.7 illustrates. That is, the n Achievement score per
story increases as a function of how related the pictures are to achievement, in-
creases as instructions get more achievement oriented, and increases as individu-
als differ in their overall n Achievement score. The three determinants combine
in a simple additive fashion such that a person with a low n Achievement score
overall, when responding to a picture containing no achievement cues under re-
laxed conditions, is very unlikely to put any achievement thoughts into a story
written under those conditions. On the other hand, a person with a high
n Achievement score overall, when responding to a high cue picture under
achievement-oriented conditions, is most likely to teil a story saturated with
achievement content. There do not seem to be any interaction effects in the
sense that moderate cue pictures, for example, might differentiate better than
high or low cue pictures between relaxed and achievement arousal conditions.
External events also can influence the content subjects put into their stories.
Figure 6.8 shows how achievement imagery increases in stories written by stu-
dents when they are to take an examination at some later point in time. The
more nearly the picture represents the examination Situation and the closer the
students are to taking the examination, the greater the amount of achievement
imagery written into the stories. Thus, a medium cue picture evokes little
achievement concern a year before the examination, somewhat more three
weeks before the examination, and much more on the day of the examination
itself.
Coding Thought for the Power and Affiliation Motives

The same approach was adopted to develop scoring Systems for the need for Af-
filiation (n Affiliation) and the need for Power (n Power), as will be described in
Measuring Instrument
Increasing Number of Achievement

Cues in Individuais
8
in
Large Number of Cues

in Individuais
1 Small Number of Cues

I in Individuais Increasing Number of Achievement Cues in Pictures
Figure 6.7.
Graphic Representation of Joint Determination of n Achievement Score Obtained from a Single Story by Number of
Achievement Cues in (1) Pictures, (2) Instructions, and (3) Individuais (McClelland, Atkinson, Clark, & Lowell, 1953).
1 Year from Exam 3 Weeks from Exam Same Day as Exam

c
o
U 1.5
S2 1.0
> in
Neutral Low Medium High Neutral Low Medium High Neutral Low Medium High
Degree of Picture Similarity to Exam Situation
Figure 6.8.
Mean Achievement Concern in Stories Written to Pictures of Increasing Similarity to an Examination, Plotted Separately
for Groups of Students for Whom the Examination Was Increasingly Near (Heckhausen, 1967, after Fisch).
subsequent chapters. Thus, any given story is normally scored for several
motives.
Table 6.4 illustrates how stories written to the same picture can have quite
different motivational content. The stories in Table 6.4 are artificially con-
structed to illustrate some of the aspects of coding Systems developed from
studies in which these motives were aroused to define and measure n Achieve-
ment, n Affiliation, and n Power. The story in the left column contains most of
the elements that appeared more often in stories written by young men after the
achievement motive had been aroused. The primary decision about whether to
score a story as achievement oriented depends on whether there is some concern
to do well or to do better expressed in the story. Here George clearly cares
about producing the most practicable drawings so he can build a bridge. Other
aspects of the story mention that he is thinking about how happy he would be if
he achieved the goal, how he must act so as to overcome blocks in the world,
and how unhappy he is when he fails to achieve the goal. It is a nonobvious
finding that arousing achievement motivation just as often led subjects to think
about failure as about success and joy in winning.
Affiliation concerns were aroused by exposing young men to a Situation in
which they were publicly judged by their peers on their popularity and personal
characteristics. Under these conditions their stories reflected a greater concern
for establishing, maintaining, or repairing friendly or affiliative relationships
(Shipley & Veroff, 1952). In the artificial story constructed in the middle column
of Table 6.4 to reflect this type of arousal, George is described as concerned
about maintaining his affiliative relationship with his wife, about personal and
environmental blocks that interfere with this relationship, and about his desire
to maintain it and to take Steps to repair it.
Power motivation has been aroused in a variety of ways (see Winter, 1973).
Table 6.4.
TYPICAL STORIES WRITTEN WHEN ACHIEVEMENT, AFFILIATION, AND POWER MOTIVES
HA VE BEEN AROUSED TO A PICTURE OF A MAN AT A DRAWING BOARD
Achievement Arousal Affiliation Arousal Power Arousal

George is an engineer who George is an engineer who is This is Georgiadis, a
(need, +1) working late. He is (prestige of actor, -[-1)
wants to win a competition in (affiliation imagery, +1) famous architect, who
which the man with worried that his wife will be (need, +1)
(achievement imagery: Standard annoyed with him for neglecting wants to win a competition which
of excellence, +1) her. will establish who is
the most practicable drawing will (block, world, + 1) (power imagery, +1)
be awarded the contract to build a She has been objecting that he the best architect in the world. His
bridge. He is taking a moment to cares more about his work than chief rival, Bulakovsky,
think his wife and family. (block, world, +1)
(goal anticipation, +1) (block, personal, + 1) has stolen his best ideas, and he is
how happy he will be if he wins. He seems unable to satisfy both dreadfully afraid of the
He has been his boss and his wife, (goal anticipation, negative, +1)
(block, world, +1) (need, +1) disgrace of losing. But he comes
baffled by how to make such a but he loves her very much, and up with
long span strong, but remembers (instrumental act, +1) (instrumental act, -f 1)
(instrumental act, -f 1) will do his best to finish up fast a great new idea, which absolutely
to specify a new steel alloy of great and get home to her. (powerful ejfect, -f 1)
strength, submits his entry, but bowls the judges over, and he
does not win and wins!
(goal State, negative, -f V
is very unhappy.
Thema + 1 , Total n Achievement Thema + 1 , Total n Affiliation Total n Power score = -f 7.
score = + 7. score = + 6 .
In one study, Student candidates for office wrote stories while they were waiting
for the election returns to be counted. In another, they wrote stories after they
had been exposed to the influence of an impressive leader (a film of John F.
Kennedy making his inaugural address). Gradually it appeared that the kernel
of the unique effect these different arousal situations were having on thought
content lay in the area of having impact on others or being considered strong,
important, or powerful. Thus, in the right column of Table 6.4, Georgiadis is
described as wanting to win a competition, just as George is in the achievement
story. This time, however, it is clear that winning implies an enhanced reputa-
tion. The concern is not with doing better but with being recognized as the best
architect in the world. Other aspects of this story illustrate the characteristics of
a well worked out motivational sequence involving rivals, fears, and means of
having impact.
Every time an aspect of either the n Achievement, n Affiliation, or n Power
motivational sequence is mentioned in such stories, as indicated by the words in

italics, it is counted as + 1 . An additional point is added for thema if the story
is mostly about a particular subject, and an overall score for the story is ob-
tained by summing the number of different elements present. Few of the charac-
teristics of the overall coding scheine appear in any given story, but people are
asked to write from four to six stories so that they have several opportunities to
include ideas that would be scored under one or the other of the motive-coding
Systems. A score for all stories is called a person's n Achievement, n Affiliation,
or n Power score. Once the motivational codes have been established from the
arousal experiment, individual differences in the strength of a motive are as-
sessed by having people write stories under neutral conditions so that the ob-
server can discover what they spontaneously think about when they are not
under the pressure of any particular motivational demand. What is being mea-
sured is how often thoughts spontaneously turn to achievement, to power, or to
affiliation goal states. Note that subjects do not know what they are supposed to
write in stories of this kind, so conscious values and expectations have less efFect
on what they write; therefore, this writing is a purer measure of their motive
dispositions. The method provides a means of assessing directly the recurrent
concern for a goal State, which is the essence of the definition of a motivational
disposition.
Whereas the motives most commonly studied have been the achievement, af-
filiation, and power motives, the technique of deriving scoring codes for motives
through motivational arousal is completely generalizable, and it has been used to
study other motives as well. For example, some soldiers wrote stories ten hours
before, one-half hour after, and ten hours after, an atomic explosion during
which they were stationed in trenches 4000 yards from the point of the detona-
tion (Walker & Atkinson, 1958; see Chapter 10 of this book). The danger (a
"demand" characteristic) eJicited more fear-related motivation in stories written
before the explosion than in those written when the danger had passed.
In Chapter 1 we discussed the effects of sexual arousal on fantasies, and in
other places we will deal with other concerns discovered to result from motiva-
tional arousal.
Adequacy of Various Measures of Individual Differences in Motive Strength

Coding stories for various types of goal-oriented content is not the most populär
means of measuring individual differences in motive strength. It derived from
H. A. Murray's research reported in Explorations in Personality (1938), but so
also did the method of assessing motive strength by asking subjects how much
they agreed with various Statements related to different motives. Table 6.5 lists
some items Murray also used to evaluate the strength of a person's n Achieve-
ment, n Affiliation, and n Power. One obvious advantage of this approach is
that it is so much cheaper than coding thought content; a subject's replies—say,
on a scale of + 3 to — 3 for each item—can be recorded and can be read by a
machine and a person's total score aatomatically computed without any effort
on the part of the experimenter. So a number of measures of motive strength
Table 6.5.
EXAMPLES OF SELF-REPORT ITEMS USED BY MURRAY TO ESTIMATE
THE STRENGTH OF VARIOUS NEEDS (after H. A. Murray, 1938)
Need for Achievement
1. I set difficult goals for myself, which I attempt to reach.
2. I enjoy relaxation wholeheartedly only when it follows the successful completion of a
substantial piece of work.
3. I work like a slave at everything I undertake until I am satisfied with the result.
4. I enjoy work as much as play.
Need for Affiliation
1. I am in my element when I am with a group of people who enjoy life.
2. I become very attached to my friends.
3. I like to hang about with a group of congenial people and talk about anything that
comes up.
4. I go out of my way just to be with my friends.
Need for Power (Dominance)
1. I enjoy organizing or directing activities of a group—team, club, or committee.
2. I argue with zest for my point of view against others.
3. I usually influence others more than they influence me.
4. I feel that I can dominate a social Situation.
Note: In Murray's questionnaire the Statements for a given need are not presented consecutively, but are
interspersed with others. In later versions of this type of questionnaire, items are included so disagreement also
indicates strength of the need in question, and pairs of Statements equated for social desirability are presented so
the choice is not dictated by the tendency to give desirable responses.
based on such self-reports have been carefully developed over the years, includ-
ing the Edwards Personal Preference Scale (1957), the Jackson Personality Re-
search Form (1966), and the Mehrabian Scales for Achievement and Affiliation
Orientation (1969, 1970). How do these measures of motive strength Stack up
against those obtained from coding thought content? Which is the better method
of assessing motive strength?
Murray believed the best method of measuring motive strength was to in-
clude Information from both these sources, as well as from any other source,
such as Interviews or behavior in games or role-playing situations. In the assess-
ment centers he designed, a relatively small group of persons was studied inten-
sively over a period of days by a number of expert observers. Subjects were in-
terviewed, tested in a large number of different ways, and watched as they
participated in various group activities. Each observer made preliminary esti-
mates of the strength of the subjects' various motives based on all the Informa-
tion available to him or her. Then the observers came together in a diagnostic
Council in which they discussed at length all the Information obtained and esti-
mated the strength of a person's motivation in a particular area. The final judg-
ment by the diagnostic Council of, say, the strength of a person's n Achieve-
ment, was considered to be the best possible way of arriving at the truth about
that motive for that person. How does this method of measuring individual dif-
ferences in motive strength—which we call the clinical method—Stack up
against the self-report method or the method of estimating motive strength from
coding spontaneous thought content? Which is the best method?
• ALTERNATIVE MEASURES OF MOTIVE STRENGTH

EVALUATED ACCORDING TO THE CRITERIA
FOR GOOD MEASUREMENT
To answer such questions, we must review the criteria for a good measure of
any characteristic. Table 6.6 lays out the major criteria and illustrates how each
applies to a physical variable, such as temperature, and a psychological one,
such as motivation. The sensitivity criterion requires that a measure fluctuate or
vary systematically with the presence or absence of, or change in, a characteris-
tic of which it is supposed to be an index. If a lighted match is brought near a
column of mercury in a glass tube and the mercury does not rise, the mercury
is an insensitive indicator of temperature change. If a self-report does not indi-
cate fear when we know the person is about to make a parachute jump (Epstein,
1962), self-report is not a sensitive indicator of a motive State.
The uniqueness criterion requires that the indicator not be influenced by
variables other than the one it is supposed to be measuring: if the mercury rises
as pressure is applied, a high reading can reflect pressure rather than heat. It
was on this ground that we questioned the value of using action as an index of
motive strength, because actions such as doing well in school may result from
intellectual factors that have nothing to do with motivation.
The reliability criterion really has two aspects. In the first place, two ob-
servers reading the same index at the same time should give the same report. In
the second place, two observations of the same characteristic of the same person
under the same conditions should agree.
The validity-Utility criterion is the most important of all, because it is possi-
ble to get a measure that is sensitive, unique, and reliable but that is quite unim-
portant in that it does not teil anything eise about what the person will do in a
variety of situations. Psychology, like all science, is an economizing game, the
goal of which is to measure the least number of variables that will give the most
information about how a person will behave in a variety of situations. In prac-
tice, the validity-utility of a measure of human motive strength is judged by the
network of correlations (Cronbach & Meehl, 1955) established between it and
other things the person does. The network also must make theoretical sense.
That is, to be sure we have a measure of motive strength and not some other
variable, it must be shown that people who score high on the measure behave
as if they were more motivated in this area; for example, if they are high in
Table 6.6.
CRITERIA FOR A GOOD MEASURE OF TEMPERATURE AND MOTIVATION
Application to
Criteria ofa
Good Measure Meaning Temperature Motivation
Sensitivity Index reflects underlying Column of mercury rises Behavioral index reflects
variable sensitively. regularly as heat is applied. induced motivational states.
Uniqueness Index reflects changes in Column of mercury does not Behavioral index is relatively
that variable only. rise as pressure is applied. uninfluenced by habits,
values, and so on.
Reliability Index gives same reading Mercury rises to same point Behavioral index gives same
under same conditions. if same heat is applied. value for an individual on
repeated testing.
Validity-utility Index reflects a variable of Temperature as measured by Motive index shows
importance. a thermometer is theoretical properties of a
theoretically and practically motive and is useful in
useful in predicting what predicting behavior under
will happen to substances various conditions.
under various conditions.
n Achievement, they must be more energetic in achievement situations, they

must be more sensitive to achievement cues, and they must learn material re-
lated to reaching the achievement incentive faster.
Let us now see how each of the three methods of measuring individual differ-
ences in motive strength Stacks up against the four criteria of a good measure.
Sensitivity
The thought content codes for measuring motive strength pass the sensitivity
criterion for a good measure, because they were derived by comparing and con-
trasting differences in fantasy under unaroused and aroused conditions. The seif-
report measures pass it less adequately. As mentioned earlier, the thought mea-
sure of need for Food (n Food) increased monotonically with hours of food
deprivation, but self-ratings of hunger did not increase appreciably between four
and sixteen hours of food deprivation. Parachute jumpers do not report being as
afraid as physiological indexes suggest they are, probably because they are sup-
pressing their feelings of fear in order to go through with the experience. Simi-
larly, self-reported fatigue ratings do not increase in a simple fashion with hours
of going without sleep, whereas thought codes do (see, for example, E. J. Mur-
ray, 1965).
Personality theorists working with questionnaires usually distinguish be-

tween a trait measure (asking people what they want usually) and a State mea-
sure (asking people what they want right now; see Spielberger, Gorsuch, &
Lushene, 1970). This corresponds to the distinction between a motive disposition
and an aroused motive. To satisfy the sensitivity criterion, theorists should show
that an aroused motive (for example, anxiety in response to a threat) increases
the State anxiety score and that this somehow relates to the trait anxiety score
the person also reports. So far little research has been carried out on whether
motive self-reports are sensitive to motivational arousal; what little there is sug-
gests that they are not likely to be sensitive, since they are strongly influenced
by other variables, for example, the conscious value placed on particular motive
states such as appearing not to be very anxious.
The clinical method of estimating motive strength can hardly be evaluated
against the sensitivity criterion, because by definition it includes all sources of
information, including the fantasy codes, which do pass the sensitivity criterion.
However, since the final clinical judgment also includes self-report measures,
which are relatively insensitive to motivational changes, it might be inferred that
they are not as sensitive to motive changes as the fantasy measures are by
definition.
The sensitivity criterion also implies that the effects of variations in motive
strength on thought content will be the same for all people everywhere. Heat
applied to a column of mercury will cause it to rise in Peoria and Timbuktu in
exactly the same way. But is this true in psychology? Does achievement motive
arousal have the same effects on fantasy in different cultures? Perhaps in some
cultures it will lead to thoughts about achieving cooperatively, and in others it
will lead to thinking about doing things individualistically, as Maehr and
Kleiber (1981) have argued. In The Achievement Motive (McClelland et al.,
1953) it was reported that Navaho males responded to achievement arousal in
the same way other males did. That is, they wrote more stories containing
achievement imagery (as defined from earlier arousal experiments on people in
the United States) than did Navaho males under relaxed testing conditons. Since
then, the basic achievement arousal experiment has been replicated in Germany
(Heckhausen, 1963), in Brazil (Angelini, 1959) and in Japan (Hayashi & Habu,
1962). The thought characteristics associated with achievement arousal seem to
be much the same across different cultures.
On the other hand, some early results obtained using women subjects raised
some doubts as to the generality of this conclusion. Neither the achievement-
arousing instructions nor the experiences of success and failure that worked for
men served to increase the n Achievement scores of women exposed to the same
conditions. For example, in one study the mean n Achievement score for College
males was —1.46 under relaxed conditions and 3.46 after failure, whereas for
College females the mean scores were 1.24 and 0.93, respectively (McClelland et
al., 1953). Apparently either the aroused achievement motive affects women's
thought processes differently than it affects men's, or the achievement-arousing
instructions that were used failed to arouse the motive for women.
The first possibility raises some very serious questions about this whole ap-
proach to motive measurement, because it would mean that a whole different

scoring System would have to be developed to assess achievement motive
strength in women. Fortunately, this did not prove to be necessary for two rea-
sons. In the first place, it was found that n Achievement scores for women
(using the scoring criteria developed from the arousal experiments with men)
correlated with female Performance in the same way as for men. That is to say,
women who scored high in n Achievement behaved in the same way as men
who scored high in n Achievement, indicating that the measures were as valid
for women as for men. In the second place, it was found that the achievement-
oriented instructions used to arouse men simply were failing to arouse the
achievement motive in some groups of women. In later arousal studies (for ex-
ample, French & Lesser, 1964), references to the importance of intellectual,
career-oriented achievement did increase achievement motivation in women (see
Stewart & ehester, 1982). Furthermore, in Germany, Brazil, and Japan, the
male achievement-orienting instructions produced the same effects on women's
as on men's fantasy. In addition, it was found that judging popularity and per-
sonal attributes in public served as achievement incentives for some U.S. women
in the sense that under these conditions, women's stories showed the same rise
in achievement concerns as the men's stories did when their intelligence and
leadership capacity were the incentives used (McClelland et al., 1953).
Thus, methods of arousing the achievement motive (demands plus incen-
tives) may differ in various subgroups of the human population, but so far no
one has reported any Special fantasy effects of achievement motive arousal that
are not included in the coding System originally developed for young men in the
United States. Unfortunately, not enough research has been done in other cul-
tures on arousing motives to be sure the same conclusion can be drawn about
thought content related to the power and affiliation motives.
Uniqueness
How do the various methods of assessing motive strength Stack up against the
uniqueness criterion? The clinical method fails it signally. Suppose a number of
clinicians pool their judgments of a patient's motivation for treatment, as in a
Menninger Clinic study (Luborsky & Sargent, 1956). They all use a number of
clues based on their knowledge of a patient's test scores, social behavior, or con-
duet in a therapeutic hour. Each clinician makes a numerical estimate of how
motivated for treatment a particular patient is, and by and large they agree. Is
this not the best possible estimate of a patient's motivation? But note in the dia-
gram in Figure 6.9 that the judges' ratings of the level of psychosexual develop-
ment and ego strength correlate highly with their ratings of motivation for treat-
ment. Furthermore, levels of psychosexual development and ego strength are
themselves highly correlated. In other words, the clinicians could not really dis-
tinguish very well among these three variables. Because of their common train-
ing in the psychoanalytic tradition, they found it difficult to imagine a person at
a low level of psychosexual or ego development also being highly motivated for
treatment. So it is not clear what clues the judges were actually using in rating
Measures ofHuman Motive Dispositions 203
the patient's motivation. They may have been responding primarily to his or her
level of psychosexual or ego development. If so, the estimate of motivation is
really not a measure of motivation at all, but of other things that, in the judges'
minds, imply that the patient is motivated to a greater or lesser degree. In short,
if the judges are allowed to use many clues in arriving at their assessment of
motive strength, the measure is "impure," a hodgepodge of assessments of many
different response tendencies the judges believe are related to motive strength.
Self-reports also do not pass the uniqueness criterion very well, in part be-
cause they are subject to various types of response biases. For example, some
subjects are "yea sayers"; others are "nay sayers" (Couch & Keniston, 1960).
The yea sayers tend to agree more with all sorts of Statements and the nay say-
ers to disagree more. A high score on the first four items in Table 6.5 may indi-
cate not only a higher n Achievement, but also a greater tendency to agree with
any and all Statements. Schedules can be corrected for this bias by equalizing
the number of Statements that have to be agreed or disagreed with in order to
get a high score.
Another type of bias comes from the fact that most people tend to agree
more with socially desirable items. This bias can be corrected by equating items
for their social desirability and forcing the subject to choose between them (Ed-
wards, 1957; Jackson, 1966).
However, the main difficulty with self-ratings so far as the uniqueness crite-
rion is concerned is that they reflect the person's self-image and values. Thus,
most U.S. Citizens feel they should have a high achievement motive and will
agree with Statements indicating that they do. On the other hand, most U.S. cit-
izens feel they should not be high in the power motive and will disagree with
items indicating that they have such a need. Once in a discussion section for a
class in motivation, a Student vehemently argued for nearly the whole hour that
his n Power score obtained from the picture-story method was too high. By the
end of the time, everyone in the class was convinced that the student's score
was exactly right.
The main contribution of the Freudian revolution to personality theory was
Level of Psychosexual Development
.85
Motivation for
Treatment
84
Ego Strength
Figure 6.9.
Intercorrelations of Personality Variables as Rated by Judges at the Menninger Clinic (Luborsky &
Sargent, 1956).
the demonstration that people's conscious values and unconscious motives are
not necessarily the same. This book will adopt the Convention of referring to
self-report measures of motive strength as values (abbreviated to v) to avoid
confusion on this point.
People's values as reflected in self-ratings have a number of determinants
that are not motivational in character. For example, Lynn (1969) developed a
self-report measure of v Achievement in which entrepreneurs scored higher than
students. He inferred from this fact that he had a good measure of achievement
motive strength, since entrepreneurs should score higher than students. What he
failed to take into account, however, was that entrepreneurs know they have
been more successful; if asked, they will say they set difficult goals, prefer work
to play, and so forth, because that readily explains to them why they have been
more successful. What is taken as an indication of their achievement motive
strength is really simply a part of their overall self-image.
Do the thought measures pass the uniqueness criterion? To begin with, in
general they are uninfluenced by values or response sets. People who respond to
questionnaires showing that they highly value achievement, affiliation, or power
do not usually score significantly higher on the corresponding motives as scored
in fantasy (see Child, Frank, & Storm, 1956; deCharms, Morrison, Reitman, &
McClelland, 1955). Those who value social approval, as measured by the
Crowne-Marlowe measure (1964), also do not score higher on any of the three
social motives, as checked in a sample of seventy-six adults studied by McClel-
land, Constantian, Pilon, and Stone (1982).
Can a person fake a high motive score by writing stories with the necessary
characteristics? If so, a high score might be obtained even though the motive
level was low. Asking subjects to fake a high n Achievement score did not pro-
duce a significant increase in the mean n Achievement score (McClelland et al.,
1953), but obviously anyone who has learned the scoring Systems as outlined in
Table 6.4 can fake a high score. Curiously enough, however, if after learning to
fake a high score people tend to think that way much of the time, the motive in
fact would appear to have greater salience for them. As we shall see in Chapter
14, one of the ways of increasing motive strength is to teach people to write sto-
ries saturated with the thoughts characteristic of people with a strong motive.
The Effect of Conditions of Test Administration on Imaginative Stories. A

more serious problem arises from the fact that story content is very easily influ-
enced by the conditions of test administration, as the arousal experiments dem-
onstrate. This means that the score in an individual's stories may not be the re-
sult of the person's own motivation so much as the result of other factors in the
testing conditions (see Figure 6.7). Ever since Murray's time it has been stressed
that if stories are to reflect personal characteristics, they must be obtained under
relaxed conditions in which the subject feels natural and spontaneous. McClel-
land et al. (1953) reported that ego involvement tended to inhibit the ability of
the subjects to fantasize, making their stories unable to be scored.
Lundy (1981a) has systematically investigated the problem by comparing
the results obtained when stories were written under the usual relaxed or infor-
mal conditions with those written after the introduction of three forms of "pres-
sure" that he thought would evoke stereotyped stories not based on personal
motives. He told one group of subjects that the story writing was a test that
would reveal "hidden personality tendencies" including "imperfections or minor
defects in the personality" (Lundy, 1981a). He stressed to another group of sub-
jects that they should be very careful in writing their stories because he wanted
to get them under conditions "which are standardized and carefully controlled"
(Lundy, 1981a) so that he could compare the results for normal subjects with
those written by people who were psychotic or brain damaged. Before asking a
third group to write stories, Lundy simply administered a personality test deal-
ing with affiliative matters, which was supposed to teil them whether they were
the kind of person who was always at ease or quite insecure in groups. He feit
that all the Special instructions would make the subjects more cautious and
more likely to write stereotyped stories rather than ones reflecting their personal
motives. He developed a Special coding device, called the ego control score, to
see if in fact this occurred. The score measured the amount of incoherence in a
story—that is, as shown by jumping from one thought to another—and the
amount of transcendence as shown by the subjects' bringing in characters and
objects not in the picture.
All three types of Special instructions designed to increase pressure de-
creased transcendence and incoherence; this significantly increased the ego con-
trol score, showing that the stories were more stereotyped. What is even more
important is that the motive scores obtained when stories were written under
Special instructions lost their validity. That is, they no longer correlated with be-
haviors characteristic of the motives in question, as they had when obtained
under relaxed conditions.
Lundy also had obtained a number of other measures of the usual correlates
of n Achievement, n Affiliation, and n Power, such as the kinds of television
shows the students typically watched. He found, as expected, that the subjects
higher in n Achievement were more likely to watch information shows on televi-
sion, those high in n Affiliation were more likely to watch soap operas and com-
edies, and those high in n Power were more likely to watch sports on television
than students low in any of these motives. This was true, however, only when
the motive scores were obtained under relaxed conditions. When the students
feit threatened in any way because of the instructions, the validity coefficients
—that is, the correlations with activities associated with the motives—were re-
duced significantly and, for all practical purposes, they disappeared. Figure 6.10
summarizes the results in a very striking way. It shows that the classes that
took the test under relaxed conditions—referred to as the personal Schemata
group—all scored low in ego control and high in the validity estimates of the
motive scores obtained. The reverse was true for the three types of Special in-
struction groups: they scored higher in ego control and lower in the validity of
the motive measures obtained.
In another study, Lundy (1980) reviewed all reports between 1966 and 1980
that included a measure of n Achievement, n Affiliation, or n Power. He found,
for example, that of the 195 reports that employed a measure of n Achievement,
.70
.60
.50
.40
.30
«„ .20
c
;§ .10
00
l
£-.10
I-.20
-.30
-.40
-.50
-.60
1 1 1 _|_ J I 1 l I I
-1.6 -1.2 -0.8 -0.4 0.0 0.4 0.8 1.2 1.6 2.2
Ego Controlled Ego Relaxed
Mean Ego Control Scale Score
a
Fisher z-transformations to normalize the distribution of coefficients.
Figure 6.10.
Scatterplot of Validity Coefficients and Mean Ego Control Scale Scores for All Classes (Units of
Analysis) in the Study. p = personal Schemata group; n = normative group; s = structured group;
e = ego-threatened group (after Lundy, 1981a).
124 found significant differences using this measure, and fifty-one reported insig-
nificant differences. He then went on to look for factors that might be responsi-
ble for the fact that some investigators found the measure to be valid and oth-
ers, invalid. Among the factors that were important was the Status of the test
administrator. When the administrator was an authority figure—which could be
interpreted as introducing some kind of pressure on the subjects—the validity
coefficients of the motive measures were very much less than when the adminis-
trator was a fellow Student or behaved in a casual manner. Lundy also found
that manipulations of instructions strongly affected the validity of the scores ob-
tained by the test. Even simply calling the story-writing procedure a personality
test rather than a picture-story exercise adversely affected the validity of the
scores obtained from the stories.
In short, motive scores obtained from imaginative stories do not uniquely
reflect motive differences under all conditions of test administration. It is ex-
tremely important that the conditions of administration be relaxing for the re-
sults to satisfy the uniqueness criterion. What is particularly valuable about
Lundy's research is that he has developed a measure of ego control that can
serve as a check for experimenters as to whether or not they have succeeded in
relaxing the subjects taking the test. If the average ego control score is too high,
there is evidence that other factors will influence the motive scores and that
they will no longer be good indicators of motive strength.
Reliability
Fantasy measures of human motives pass the sensitivity and the uniqueness
criteria for a good measure, but they have difficulty satisfying the reliability cri-
terion. In fact, fantasy makes many psychologists nervous, because it has a
"now you see it, now you don't" quality that is the very antithesis of good sci-
entific measurement. One analyst sees one motive pattern in a dream, while an-
other sees something eise. The problem of getting high observer agreement has
been solved for measures of the sort illustrated in Table 6.4. The criteria for the
presence or absence of achievement, power, or affiliation imagery and their sub-
categories are so objective that it is possible to train observers to agree almost
perfectly on the scores they obtain for a given set of stories.
Coding Reliability. Two methods of testing coding reliability have been used.
The first involves the percentage of agreement on the presence of a category.
Suppose two judges, A and B, score six stories for the presence or absence of
achievement imagery. Suppose Judge A scores achievement imagery in Stories 1,
3, and 5, and Judge B scores it in Stories 1, 3, 5, and 6. The percentage of
agreement is computed by multiplying the number of stories in which they
agreed the category was present (namely, three) by two ( 3 x 2 = 6) and divid-
ing by the sum of the number of times each judge scored for the presence of the
category (3 + 4 = 7). This gives a ratio of 6/7, or 86 percent, which is a rep-
resentative figure for the amount of agreement usually found for different cate-
gories in the various motive scoring Systems.
Note that in this ratio it is necessary to have a denominator that Covers
how frequently a judge scores a given category, since this tends to vary some-
what; if we only looked at the percentage of agreement, without taking the level
of scoring into account, it would be technically possible to get fairly high agree-
ment if one judge scored every story as containing the category.
The percentage of agreement is a conservative reliability estimate, because it
does not include agreement on the absence of a category. The reason it does
not is that some of the categories are quite infrequent, so including agreement
on absence could overestimate the reliability of coding. For example, suppose
Judge A scores "help" in the n Achievement scoring System in Story 1 and
Judge B, in Story 3. There is perfect disagreement on presence; however, they
do agree that the category is absent in four of the stories, which would yield a
percentage of agreement score of (2 X 4)/12, or 75 percent, which seems like
an overestimate.
Another way of checking coding reliability is to correlate the total scores

for an individual obtained by two different coders. These correlations between
well-trained scorers run between .85 and .95, indicating that two different judges
arrive at about the same score for the same person (see McClelland et al., 1953;
Winter, 1973). The only difficulty with this measure is that it is still technically
possible to get high correlations between two sets of scores while one judge is
scoring more than the other judge overall, so the mean scores for a group of
subjects could be higher for Judge A than for Judge B even though their agree-
ment on the rank order of the two sets of subjects was nearly perfect. Thus, the
same judge must be used in scoring all subjects to be compared, or the two
judges must be calibrated very carefully so as to arrive at the same level of
scoring.
Test-Retest Reliability. The second meaning of reliability involves whether

testing individuals on repeated occasions yields the same motive score. One tra-
ditional way of checking for this kind of reliability is to divide the test into two
equal halves and see if the person scores the same on both halves. This is some-
times referred to as the split-half method of estimating reliability. For example,
if a Student gets an n Achievement score of 4 on Pictures A, C, and E, he or
she should get a score of 3 to 5 on Pictures B, D, and F rather than, say, a
score of — 1, which would indicate great inconsistency in the two halves of the
test. The correlation from sets of n Achievement scores obtained from two
groups of three pictures each was reported to be as high as .64 in The Achieve-
ment Motive (McClelland et al., 1953), but a later review of research suggested
that these internal consistency correlations generally run from .30 to .40, indi-
cating unsatisfactory reliability by traditional psychometric Standards (Entwisle,
1972). In contrast, internal reliability correlations for self-report measures are
much higher, from .70 to .90 (Entwisle, 1972; McClelland, 1980).
However, Atkinson, Bongort, and Price (1977) have argued that traditional
psychometric theory is wrong in assuming that the response to any test item
(here, a story written to any picture) is an isolated and completely independent
test of the strength of some underlying characteristic such as n Achievement.
Instead, they have developed a theory that Stresses the continuity of the thought
stream. In their view, response tendencies having to do with achievement, affilia-
tion, or power are continuously competing for expression in thought. If the
achievement tendency is expressed in connection with the first picture, its ex-
pression "uses up" the tendency and reduces it in strength so that the next
strongest tendency—say, for affiliation—will appear in the next story. Thus,
there is no reason to expect consistency in response tendencies from one item to
the next. In fact, a well-known phenomenon in psychology, called associative re-
fractory phase (see Telford, 1931), states that subjects are unlikely to repeat an
association they have just given.
Atkinson and Birch (1978) took their analysis further and developed a Com-
puter program that, with certain built-in assumptions, generated curves like
those shown in Figure 6.11. In this Simulation there are three tendencies com-
peting for expression. It is further assumed that the subject is telling four sto-
ries, which divide the time line into equal quarters. Tendency X is for achieve-
ment. If we sum the time spent expressing X in the four stories (when the ten-
dency to write about X was stronger than the tendencies to write about Y and
Z), we obtain the values of 7, 3, 4, and 5. If we compare the achievement ten-
dencies expressed for Stories 1 and 4 with those for Stories 2 and 3 on two al-
ternative halves of the test, we get values of 12 and 7, indicating a major dis-
crepancy, or low internal inconsistency.
Using their program, Atkinson, Bongort, and Price (1977) generated twenty-
five Computer simulations; in each, subjects were arbitrarily classified as high,
middle, or low in n Achievement. They then measured the percentage of time
that each subject would spend thinking about achievement, according to their
Computer Simulation with different assumptions about the internal consistency
measures from one time period to the next. Their measure of the validity of the
total time spent thinking about X measure was the extent to which it agreed
with the n Achievement level assigned to an individual in advance in the
Computer program. That is, if a person was assigned to the top third of the
n Achievement scoring distribution in advance, he or she should be in the top
third of those spending time thinking about achievement in the Computer
Simulation.
In Figure 6.12 they plotted construct validity (the percentage of subjects
Time
Figure 6.11.
Response Tendencies Competing for Expression During the Telling of Four Stories (after Atkinson, 1980).
correctly placed in various thirds of the n Achievement distribution by the time-

spent measure) against a measure of internal consistency (here, Cronbach's
alpha, a more generalized measure of internal consistency than the usual split-
half correlations). Notice that the overall time spent thinking about achievement
is an excellent index of "true" n Achievement scores, even when the internal
consistency measure is very low or even, in one case, negative. Their conclusion
is that "the construct validity of thematic apperceptive measures does not re-
quire internal consistency reliability as supposed by traditional test theory" (At-
kinson et al., 1977).
Not everyone accepts all the assumptions involved in the Computer model
generated by Atkinson and Birch. Others will remain perplexed by the practical
problem of deciding which score to use if, in fact, they are different on different
halves of the test. But at the least this approach suggests that the need for inter-
nal consistency in measures of this sort must be carefully thought through.
Another way to look at the fantasy measures of motive strength is in terms
of a multiple regression model in which the pictures represent independent esti-
mates of the true score. In multiple regression it is advantageous for several de-
0.20 0.10 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 0.80 0.90 1.00
Cronbach's Alpha (a)
Figure 6.12.
Relationship of Construct Validity and Internal Consistency (a) in Twenty-five Simulations of TATs
Varying Design and Parameter Values. Construct validity refers to the percentage of subjects
correctly placed by total time in achievement activity into rank-ordered thirds of true motive
strength defined by Computer input (after Atkinson & Birch, 1978, after Atkinson, Bongort, &
Price, 1977).
terminants of an overall score to be uncorrelated with each other, because their

contribution to the total score will be greater if each is not influenced by its as-
sociation with another determinant. McClelland (1980) puts it this way: Motives
in psychology are used to explain "inconsistencies in behavior just as habits ex-
plain consistencies. When a hungry dog is trying to get out of a cage to get to
food, it will try a wide variety of different responses. Why should we expect it
to provide internally consistent responses? If it is whining, it may not be
scratching; if it is trying to push through the slats, it is not pawing at the latch.
One would not expect high correlations among these responses any more than
one would expect signs of achievement motivation in different stories to be
highly correlated." It is the total sum of all these different responses that should
be the best indicator of motive strength.
The Influence of the Set to Be Creative or Consistent. But here again a prob-
lem arises. The correlation of the total n Achievement or n Power score from a
set of four to six stories obtained on one occasion is not very close to the same
scores obtained from the same individual on a second occasion. The correlations
typically run from .20 to .40 (Entwisle, 1972; McClelland et al., 1953; Winter,
1973), again suggesting a reliability that is too low for the measures to be taken
very seriously. Part of the reason for this variability lies in the fact that the
chief requirement for a true reliability test is not met, for the subjects are not in
the same Situation the second time they take the test. They are set to respond
differently the second time. They have seen the pictures before and wonder why
they are being asked to take the test again. The instructions teil them to "be
creative," which they interpret to mean that they should teil "another story" to
the same picture. They are set to be creative. If they teil different stories to the
same pictures, they are unlikely to get similar motive scores. If different pictures
are used, the test-retest correlations are scarcely any higher, perhaps because the
subjects are still under the impression that they should be creative and teil dif-
ferent stories, but more probably because the subjects find the experience stränge
and wonder what is going on.
As Lundy's results discussed in the previous section show, if subjects feel in
any way threatened by the procedure and not completely relaxed, the scores ob-
tained seem to be the product of normative or stereotyped, rather than personal,
Schemata. There is still the problem of which score to use for a given individual,
and in general, research has shown that the scores obtained from the first testing
are more valid than those obtained on second testings, probably because the sec-
ond testing is not understood by the subjects (see Atkinson, 1980).
Winter and Stewart tested the possiblity that it was the "be creative" set
that lowered test-retest correlations. They instructed the subjects as follows: "Do
not worry about whether your stories are similar or different from the stories
you wrote before. Write whatever stories you wish" (Winter & Stewart, 1977).
Under these instructions the test-retest correlation for the n Power score rose to
.58, as contrasted to correlations in the .20s when the second test was adminis-
tered with the usual "be creative" instruction. In testing some high-school stu-
dents a year later, Lundy (1981b) also told them they were free to teil stories
that were the same as or different from those they told last time. He obtained a
test-retest correlation of .56, which compares favorably with the stability of mea-
sures obtained from questionnaires after one year.
In a sense, the reliability issue is the obverse of the sensitivity criterion. Mo-
tive measures based on coding spontaneous thought patterns are unusually sensi-
tive to situational disturbances, so replicating experiments and getting the same
scores for individuals on two different occasions represent more difficulties than
are common in psychological research. Nevertheless, they do yield reliable re-
sults if care is taken to keep the subjects relaxed and unusual response sets are
removed.
Klinger (1968) understood this problem but attempted to solve it in the
wrong way. He tested subjects alone in a seven-by-seven-foot soundproofed
room, eliminating all personal contact with the experimenter. Instructions were
given to the subject by means of signs, audio tapes, slides, and films. They were
asked to sit quietly for twenty-five minutes in order to reduce such things as as-
sociative variability. This did not increase the reliability of the measures ob-
tained, but as Winter (1973) points out, the Situation for the subject was a very
peculiar one; it was not at all patterned after H. A. Murray's early insistence
that subjects must be kept relaxed and spontaneous, the conditions which have
been shown ever since to yield the most valid and relaible motive scores.
Test-retest and internal reliability correlations for self-report and clinical es-
timates of motive strength are much higher—in the ränge of .70 to .90. So far
as the clinical judgments are concerned, the high correlations do not mean
much, because—as the last section pointed out—the judges are responding to
some overall picture of the subjects rated, which remains stable, rather than to
the specific motivational aspects of the subjecf s personality.
The reported consistency of self-reports represents an exaggerated estimate
of "true" consistency for several reasons. The most important is that subjects
answering a questionnaire are set by the instructions to respond consistently,
since they are told to be honest and frank. If they answer a question differently
on a second occasion, they might easily feel either dishonest on one occasion or
the other, or at the very least, foolish in not knowing themselves very well. The
tendency toward consistency is abetted also by response sets like the tendency to
admit or deny unfavorable characteristics or to agree or disagree with most
items.
Furthermore, the subjects are typically asked more or less the same question
in a dozen different ways (see Table 6.5). If people say in one part of the ques-
tionnaire that they are very attached to their friends, it might seem like a reflec-
tion on their honesty or intelligence to say in another part of the questionnaire
that they do not go out of their way to be with their friends. If they had said
that they set difficult goals for themselves, it just seems reasonable that they
should also agree that they work like a slave.
Worst of all, many motivational inventories contain items referring to the
past, which ought to evoke the same response every time the subject answers the
question. If subjects say they can remember "playing sick" to get out of some-
thing, they should say that every time they are confronted with the item. A dif-
ficulty with items like these is that they have to continue to be scored as indicat-
ing the presence of a particular need even though a person might have changed.
In this case, sensitivity is sacrificed to reliability. Despite these difficulties, most
self-report motive inventories have satisfactory, though somewhat inflated,
reliability.
Validity-Utility
From the behaviorist tradition covered in Chapter 3, we learned that motives
drive or energize, Orient, and select behavior. A more motivated animal is more
energetic, it directs its attention toward cues relevant to the motive more readi-
ly, and it learns more quickly to take the acts that lead it to the goal it is seek-
ing. So presumed measure of motive strength in people should satisfy these
criteria in order for the measures to be considered valid indicators of differences
in motive strength. That is, people high in n Achievement, as compared with
people low in n Achievement, should be more energetic in producing achievement-
related acts, should be more sensitive to achievement-related cues, and should
learn material related to n Achievement incentive more quickly. The motive
measures based on operant thought content have been tested against these
criteria, whereas self-report and clinical judgments for the most part have not.
Scattered throughout the chapters on the different motive dispositions are many
studies indicating the validity-utility of the measures, so it is only necessary here
to give a brief summary of the type of study that demonstrates validity accord-
ing to these criteria. So far as the energizing criterion is concerned, studies have
shown that the n Achievement score is associated with more frequent entrepre-
neurial acts (Andrews, 1967), the n Power score with more frequently getting
into arguments (McClelland, 1975), and the n Affiliation score with more fre-
quent affiliative acts (Constantian, 1981; see Chapter 13 of this book).
So far as the orienting or sensitizing function of motives is concerned, an
early study (McClelland & Liberman, 1949) demonstrated that subjects high in
n Achievement showed faster recognition of achievement-related words rapidly
presented in a tachistoscope if the words were positive (like success) rather than
negative (like failure). It was as if they were "preset" to see such words so that
they recognized them sooner than other subjects. The same effect has been dem-
onstrated even more dramatically for subjects high in n Power as they are ex-
posed to pictures representing power or neutral cues.
Through electrodes attached to their scalps over the occipital area of the
brain, which receives visual sensations, it has been shown that subjects high in
n Power respond with a greater electrical potential change to pictures represent-
ing power than to neutral pictures as early as 0.1 to 0.2 second after the picture
has been exposed. Those low in n Power show no difference in visual responsive-
ness to power and neutral pictures (McClelland, Davidson, & Saron, 1979; see
Chapter 8 of this book). What is particularly important about this Unding is that
it demonstrates that compared with subjects low in n Power, the brains of sub-
jects high in n Power are preset to respond more strongly to power cues in the
occipital area, where visual Information is first projected. The difference is in the
receiving System, and not just in the verbal or motor response to motive-related
cues.
So far as the criterion of facilitating learning is concerned, high n Achieve-
ment was early shown to lead to improvement in Performance where the task
was difficult enough for improvements to occur. Figure 6.13 shows the relevant
findings. Subjects were asked to unscramble letters like WTSE to make a word
and continue doing so for twenty minutes (McClelland et al., 1953). Subjects
high and low in n Achivement started out at about the same level of perfor-
mance in the first four minutes. By the last four minutes, however, the subjects
high in n Achievement were unscrambhng somewhat more than five words more
in the four-minute period than they had at the beginning, whereas those low in
n Achievement were performing pretty much the same at the end as they had
been at the beginning.
The difference in gains is highly significant. Similar studies have been car-
ried out for the other motives. Subjects high in n Power learned associations
faster between pictures and words if they are power related than did subjects
low in n Power, whereas there was no difference in learning speed if the picture-
word pairs were unrelated to power (McClelland, Davidson, Saron, & Floor,
1980). Subjects high in n Affiliation learned complex networks of social relation-
ships more quickly than did subjects low in n Affiliation (see Chapter 9).
25
High n Achievement
24
•3
I
I 23
CJ
™ 22
21
20
19
2 3
Four-Minute Periods
Figure 6.13.
Mean Output of Scrambled Words per Four-Minute Period for Subjects with High and Low
n Achievement Scores (McClelland, Atkinson, Clark, & Lowell, 1953).
A Comparison of the Validity-Utility of Self-report and Fantasy Measures of

the Need for Achievement. In one study, the ability of a motive score based
either on fantasy coding or on self-report to satisfy the validity-utility criterion
for a measure of motive strength was directly compared. The results are shown
in Table 6.7. Here the orienting function of a motive is assessed by the propor-
tion of subjects taking moderate risks in a ringtoss game. The expectation is that
subjects high in n Achievement should focus their attention more on what has
been defined as the achievement incentive—that is, a moderately challenging
task. Subjects classified as high in n Achievement by the fantasy measure do sig-
nificantly more often focus attention on the moderately challenging task, as ex-
pected according to theory.
But if subjects are classified as high in n Achievement by the self-report
measure—here obtained from the Edwards Personal Preference Schedule
(1957)—they do not behave according to expectation but even focus on the
achievement incentive less often than those low in n Achievement according to
the self-report measure. The energizing function of motivation was measured in
this study by the percentage of subjects persisting longer than average when
they were taking a final examination. Once again the fantasy measure correctly
predicts who will persist longer, and the self-report measure does not.
Table 6.7.
VALIDITY-UTILITY OF FANTASY AND SELF-REPORT MEASURES OF
n ACHIEVEMENT IN PREDICTING BEHAVIOR (after Atkinson & Litwin, 1960)
Motive Function Fantasy Measure Self-report Measure

A. Orienting Percentage taking moderate risks in a
ringtoss game
High n Achievement 61 36
Low n Achievement 36 62
p = .04 p = .02
B. Driving Percentage persisting longer than
average in a final examination
p = .03 n.s.
C. Selecting Percentage above average in final
examination grade
p = .02 n.s.
So far as the selecting function of motivation is concerned, subjects classi-

fied as high in n Achievement by the fantasy measure performed better on the
final examination, suggesting either that the motive had led them throughout the
course to learn the material better or that they dealt with the questions more ef-
ficiently at the time of the test itself. And once again, the self-report method of
classifying subjects as high or low in n Achievement did not successfully predict
which ones would perform better on the final examination. This last result is
somewhat atypical for reasons that will be discussed later, since subjects high in
n Achievement do not usually get higher grades. However, the results are in-
cluded here because they compare the validity of the two ways of measuring
motive strength against the three validity criteria for a motive in the compass of
a Single study.
The fact is that those who use self-report or clinical assessment methods of
measuring motive strength have relied heavily on the "face validity" of their
measures. It is enough to observe that items like those in Table 6.5 obviously
deal with the content of various motive Systems. Why worry if people classified
as high or low by this method actually behave as they ought to behave if they
were more or less motivated on the dimension in question? Yet as we have seen,
the self-reports may reflect other variables, like values, that are not motives, and
it is only by using the functional criteria for the presence of a motive that we
can distinguish between motives and values.
Values do influence behavior. For example, deCharms et al. (1955) report
that students with high self-reported achievement drives were more likely to be
influenced by a professor's opinions on the quality of a painting than students
low on this measure. So those who say that they work hard and strive for
achievement clearly value achievement in that they tend to look for Standards of
excellence in judging the worthwhileness of something. Conscious values can
clearly affect judgments without leading to behavior that can be regarded as
motivated.
A number of self-report measures of differences in motive strength have
been developed, because they are cheaper and easier to use (Atkinson & O'Con-
nor, 1966; Edwards, 1957; Hurley, 1955; Jackson, 1966), but they either have
not been tested against the validity criteria or have not shown good validity-
utility. The one exception is a carefully designed achievement orientation test de-
veloped by Mehrabian (1969). The items on the test were taken from achieve-
ment motivation theory and do not deal as openly with achievement values as
those in the early Murray schedule and its later derivatives. For example,
Mehrabian included items such as "I think more of the future than the present
or the past" or "I would rather work on a task where I alone am responsible
for the final product than one in which many people contribute to the final
product." He found that a thirty-four-item scale of this type correlated signifi-
cantly with the n Achievement score obtained from fantasy and also had some
validity in the sense that people who scored high on it correctly solved more
Problems than those who scored low on it. However, the correlation with the
fantasy n Achievement score is not high, and the Mehrabian measure has the
disadvantage of correlating significantly with measures of social desirability, indi-

cating that it is not a pure measure of motive strength. Thus, there is little justi-
fication for using it as a substitute for the fantasy n Achievement score, which
has been extensively validated.
In summary, it is clear that the measures of motive strength based on cod-
ing thought content are better than the measures obtained from self-reports or
clinical judgments for a variety of reasons. The fantasy measures were more
carefully derived by actual comparison of how people think when a motive is
aroused as compared with how they think when it is not aroused. The fantasy
measures provide a more unique index of motive strength, since they are rela-
tively uninfluenced by values, response biases, and other determinants of behav-
ior. They are probably as reliable as the self-report measures when the concept
of reliability is properly understood, and they certainly have more proved valid-
ity by the criteria generally established in behavior theory for determining
whether behavior is motivated or not. The major difficulty with the fantasy
measures is that they are very sensitive to situational influences, and great care
must be taken when stories are obtained to be sure subjects are relaxed and
spontaneous. Self-report measures of motive strength are reliable and inexpensive
to obtain, and they provide a good index of a person's value orientations. As
measures of motive strength, however, they have shown little validity-utility.
Whereas clinical judgments of motive strength may seem to have more face va-
lidity, since they take everything into account, they are practically worthless as
independent measures of motive strength alone, precisely because they take so
much into account. That is, it is difficult to know how much a rating reflects
motive strength or any number of other variables.
For these reasons, the emphasis in this book is on research done using the
fantasy measures of motive strength, because they deal more narrowly with
Problems of motivation per se. However, the influence of information obtained
in self-reports will be discussed again in Chapter 12, on cognitive influences on
motivation, and in Chapters 13 and 14, which deal with the interaction of mo-
tives and values in determining response strength and in changing behavior.
NOTES AND QUERIES
1. In the muscle tension experiment, it is usually assumed that increased ten-

sion somehow directly affects the learning. But the result obtained might
also be explained in motivational terms. Try to make such an analysis. Re-
member that the subject is trying to please the experimenter by doing two
things at once—squeeze a dynamometer and learn some nonsense syllables.
How could increasing emphasis on the first task at first facilitate and then
inhibit performing the second? Would such an analysis suggest that the in-
verted U effect might be obtained by requiring the subject to perform any
two tasks, even if one did not involve direct physiological arousal?
2. The text suggests several reasons why very heavy demands may make per-
formance less efficient. What are they? Has your Performance ever suffered
when either you yourself or others expected too much of you? Why?
3. The text states that the money incentive means different things, depending
on the motives of the people involved. Give some examples. For instance,
what might the money incentive mean for poor people, for people with a
strong achievement drive, or for people concerned about fair pay?
4. Under what conditions is an intent or desire for something most likely to
indicate the presence of an aroused motive? According to Figure 6.1, such a
Situation shold occur when the other determinants of an intent are mini-
mized or held constant. Describe some situations in which this happens.
5. If different incentives arouse the achievement motive for different groups of
people, does that mean we have to think in terms of different kinds of
achievement motivation? For example, some women's achievement motive is
not aroused by references to their leadership capacity but by references to
their social skill; offering money rewards increases n Achievement more in
working-class than in middle-class people (see Douvan, 1956). What exactly
does this mean? Is the motive itself different in different groups, or is the
way it is expressed likely to be different?
6. Several organizations have been trying to teach people "success" moti-
vation. Is there such a thing as success motivation? How does it relate to
n Achievement or other motives psychologists measure?
7. Why do you think arousal instructions that increased n Achievement in
men in the United States also tended to arouse the achievement motive in
women in other countries more consistently than in the United States?
8. Blood pressure varies greatly depending on what a person is doing, yet phy-
sicians believe their measures of blood pressure are reliable enough to be de-
pended on. What precautions do they take to insure greater reliability of
their measures? Is what they do analogous to what should be done to insure
the reliability of measures of motive strength obtained from fantasy?
9. Validity-utility is sometimes defined in terms of whether something is mea-
suring what it is supposed to be measuring—that is, is it giving a true esti-
mate of the characteristic in question? But how do you know what provides
a true measure of a motive? If a person says "I have a strong need to
achieve" over and over again, is that not valid evidence that he or she truly
has a strong need for achievement? Why or why not?
10. Suppose an experimenter (a faculty member) says to a Student, "I want you
to perform as well as you can on this experiment, in which you will get a
mild shock for mistakes." Make as complete a descrition as you can of all
the incentives and motives that might be involved for the Student. If your
picture of the motivational Situation turns out to be complex, might this ex-
plain why the results of such experiments are often confusing?
11. If measures of anxiety State go up under the threat of electric shock, do you
think that means self-reported anxiety would provide a good estimate of the
anxiety avoidance motive? For example, suppose people exposed to the

threat of shock say they are "nervous" more often. Does that mean people
who say they are nervous in general, even when they are not threatened by
shock, are more likely to have a stronger anxiety avoidance motive? Why or
why not? In other words, might the response pass the sensitivity, but not
the uniqueness, criterion of a good measure?
12. The fantasy measure of motives employs written daydreams, but Freud
based his analyses on night dreams. What would be the advantages and dis-
advantages of asking subjects for dreams they had had and scoring them?
(For a reference, see LeVine, 1966.)
7
The Achievement Motive
Measuring the Need for Achievement
Evidence That the Need for Achievement Score Measures a Motive
What Is the Incentive for the Achievement Motive?

Moderate Task Difficulty as an Achievement Incentive
Atkinson's Model for Predicting Moderate Risk Preference
A Cognitive Explanation of the Preference for Moderate Risk
How High Need for Achievement Affects Performance
Influence of Variations in the Challenge a Task Presents
Responding to Moderate Challenges in Everyday Life
Persistence
Effects of Too Much Motivation on Performance
Other Characteristics of People with a Strong Need to Achieve

Personal Responsibility for Performance
Need for Performance Feedback
Innovativeness
Social Consequences of a Strong Need to Achieve

Occupational Success
Entrepreneurial Success
Relationship of the Achievement Motive to the Protestant Ethic and the Rise
of Capitalism
Protestant Reform, Early Independence Training, and the Achievement Motive
Other Influences Affecting the Achievement Motive Early in Life
Parental Emphasis on Learning to Control Autonomie Functions
223
I n Chapter 6 we examined in detail how individual differences in motive
strength can best be measured, because in science, measurement is of central im-
portance. Without it we would still be in the position of McDougall, speculating
about what motives there are and how they affect behavior. This chapter and
the next chapters in Part 3 attempt to summarize what has been found out
about several social motives that have been measured by the method recom-
mended in Chapter 6—by coding fantasy or spontaneous thought patterns. For
each motive we will Start by explaining how the method of measuring it in fan-
tasy was developed, then turn to the evidence indicating it really is a measure of
a motive using the validity criteria established in Chapter 6, and finally summa-
rize what is known about how people behave who score high in the motive mea-
sured in this way. Naturally, there has been curiosity about how people develop
a strong motive of one type or another, so at the end of each chapter is a sec-
tion summarizing what is known about how people acquire such a motive.
The emphasis on measurement may seem boring or unnecessary, but the
fact is that progress has been made in the field of motivation only as some stan-
dardization in measurement occurred. Before that, investigators kept Coming up
with different or contradictory results or insisting that what was really meant by
some motive, such as the need to achieve, was quite different from what some-
body eise had claimed. For example, do you think Jesus had a strong need to
achieve? He certainly was ambitious in his efforts to save the world. In fact, he
was so ambitious that he gave up his life for the cause. The whole point of this
chapter is that once the method it describes of measuring the need to achieve
has been established and accepted, the question about Jesus can be empirically
answered. His sayings can be scored for achievement imagery, and the frequency
of this imagery can be compared with its frequency in the sayings of others. If
Jesus scores high, then by definition he has a high n Achievement. Otherwise,
he does not. There can be no further argument on the point.
Thus, it is obviously of primary importance to establish how the motive is
measured in the first place. The scientific meaning of a concept like the need to
achieve is defined by how it is measured once that has been established. Of
course, the method of measurement can be questioned, or another method pro-
posed, but whatever the method, it must pass the criteria for a good measure
discussed in Chapter 6. The measures described in this and subsequent chapters
for the most part do pass those criteria and have led to the accumulation of
some very new and exciting information about the way certain key social mo-
tives govern our behavior both individually and socially.
• MEASURING THE NEED FOR ACHIEVEMENT

The strength of the achievement motive in individuals is best measured by the
n Achievement score derived from coding the thought content of imaginative
stories, as explained in Chapter 6. The n Achievement score is usually obtained
from stories written to pictures, but it also can be validly obtained from stories
written to sentence stems such as "A father and his son looking out at a field"
(McClelland, 1977a) or from the French (1958a) Test of Insight, in which the
subject is asked to explain why a man behaves as he does in a brief written de-
224
The Achievement Motive 225
scription, such as "Joe is always willing to listen" or "Ray works much harder
than most people." Heckhausen (1963) used the picture-story format but devel-
oped a slightly different coding System for n Achievement that yielded four sep-
arate measures: Hope of Success (HS), Fear of Failure (FF), the difference be-
tween the two (HS — FF, or the Net Hope score), and the sum of HS -f FF.
Atkinson (1958) and his students also were interested in the negative
aspect of achievement motivation, or fear of failure, and developed a Resultant
Achievement Motivation (RAM) score, which was obtained by cross-classifying
n Achievement scores with scores on a Test Anxiety questionnaire developed by
Mandler and Sarason (1952). Subjects above the median in n Achievement and
below the median in Test Anxiety are considered to be most clearly oriented to-
ward approaching achievement, whereas those below the median in n Achieve-
ment and above the median in Test Anxiety are considered to be most oriented
toward avoiding achievement (or fearing failure; see Atkinson & Litwin, 1960).
Most of the results Atkinson and his students report are obtained by contrasting
the Performance of these two groups.
However, it is not entirely clear—either on theoretical or empirical grounds
—that the fear of failure is the direct opposite of the need to achieve (as as-
sumed by this procedure). Therefore, in this chapter we will consider so far
as possible only the effects of the achievement motive by itself on behavior,
leaving the treatment of the fear of failure and other avoidance motives to
Chapter 10.
One other method of measuring n Achievement levels has been used with
some success. It is based on Aronson's (1958) finding that individuals high in
n Achievement in fantasy spontaneously draw different shapes—that is, doodle
differently—as compared with individuals low in n Achievement. The former
draw more discrete S shapes and diagonals, whereas the latter produce more
"fuzzy" multiple waves: in short, people low in n Achievement "scribble" more,
make less effort to introduce variety, and are satisfied with simple repetitive mo-
tions. A simple test of this difference can be obtained by flashing very briefly a
couple of slides with some barely visible markings on them and asking the sub-
ject to reproduce them. Since nothing much can be seen, what the subjects pro-
duce represents their spontaneous drawing tendencies, which can be coded to
yield an n Achievement index, scoring + 1 for diagonals and S shapes, and — 1
for multiple waves and a few other such characteristics.
This index has the advantage that it is apparently even less influenced than
the picture-story exercise by the subject's sets introduced by test administration
conditions, as reviewed by Lundy (1981a). McClelland (1961) found it to corre-
late better than the verbal measure with theoretical behavioral correlates of
n Achievement in a cross-national study involving different test administrators
in different countries. It has also been used to estimate n Achievement levels in
ancient civilizations—particularly those with no written records—since it can be
applied with minor changes to pottery designs (see Chapter 11). However, since
it lacks "face validity"—that is, it does not have obvious meaning in itself—it
has not been widely used.
As Chapter 6 mentioned, there have been many attempts to develop "objec-
tive" measures of the strength of the achievement motive by employing various
questionnaires. None of these measures has proved to be a consistently valid in-

dicator of the strength of the achievement motive in the sense that it passes the
criteria for a good measure described in Chapter 6 as well as the n Achievement
score does.
Typical of such attempts is Schmalt's Achievement Motivation Grid (1976).
Following a procedure adopted some years earlier in the Iowa Picture Interpre-
tation Test (Johnston, 1957), Schmält asks subjects to look at pictures and pick
Statements that describe what the characters in the pictures are thinking. The
various Statements written below the pictures represent aspects of the n Achieve-
ment scoring System. Achievement motive scores obtained in this way do have
some predictive value for goal-setting behavior, but not for other well-known
characteristics of achievement motivation measured by coding fantasy. The re-
sult is typical of such research, because a conscious cognitive, evaluative compo-
nent is much more strongly present in picking choices among alternatives in a
questionnaire than in generating spontaneous responses to a picture or a sen-
tence stem. This cognitive, evaluative component is not part of the motive sys-
tem per se, although it influences it in many ways (see Chapter 12). In everyday
language it represents values (that is, the extent to which a person values
achievement) rather than motives (that is, the extent to which a person gets sat-
isfaction from thinking about or doing things better; see also Chapter 5).
• EVIDENCE THAT THE NEED FOR ACHIEVEMENT

SCORE MEASURES A MOTIVE
As we concluded from studying the behaviorist tradition in Chapter 3, motives

drive, Orient, and select behavior. Chapter 6 presented some preliminary data
showing that individuals who score high on various fantasy-based motive coding
Systems generally behave differently in these three respects from those who score
low on the same motive-scoring Systems. As compared with subjects low in
n Achievement, those who score high in n Achievement show various signs of
a higher level of physiological activation: They show more muscle tension
when concentrating on a task (Mücher & Heckhausen, 1962); a stronger gal-
vanic skin response, indicating more imperceptible sweating due to activation of
the sympathetic nervous System (Raphelson, 1957; Vogel, Baker, & Lazarus,
1958); and a higher critical flicker fusion frequency (Wendt, 1955). That is,
if two adjacent lights are flicked on and off in rapid succession, they will
eventually fuse into perceived motion from one to the other, if the time between
the two lights is short enough. A measure of alertness or vigilance can be
obtained by determining how short the interval between the two lights must
be before the person reports apparent fusion or apparent motion. Subjects high
in n Achievement report fusion only after a much shorter period of time has
elapsed between the two light flashes, indicating greater vigilance than subjects
with low n Achievement.
The fact that subjects low in n Achievement sometimes show signs of
greater physiological arousal (see Raphelson, 1957; Vogel et al., 1958) appears to
be due to the fact that very often subjects low in n Achievement are high in an-
other motive—namely, the fear of failure, which, since it is anxiety producing,

leads to physiological arousal.
Subjects high in n Achievement are also more oriented toward or more sen-
sitive to certain Stimuli. As compared with those low in n Achievement, they
perceive achievement-related words more quickly when they are exposed very
rapidly in a tachistoscope (McClelland & Liberman, 1949; Moulton, Raphelson,
Kristofferson, & Atkinson, 1958). That is, they pay more attention to cues re-
lated to Performance improvement. Bartmann (1965) showed that they gained
more in learning how to do a task from a period of programmed instruction
than subjects low in n Achievement. They tend to recall more interrupted tasks
in a context in which interruption may be interpreted as not doing very well
(Atkinson, cited in McClelland et al., 1953).
This greater attention to the Performance aspects of a Situation also pro-
motes better learning. As compared with subjects low in n Achievement, those
who score high perform better at anagrams and scrambled word tasks (see
Chapter 6), mental arithmetic (Wendt, 1955), and conceptual problems requiring
insight (see Heckhausen, 1967). What is common to these tasks is that they are
at least moderately challenging. Tasks that are too easy or routine—such as can-
celing e's and o's—do not always elicit better Performance from those higher in
n Achievement, nor do impossibly difficult tasks, as we shall see in a moment.
In any case, there is ample evidence that n Achievement as measured in this
way leads to learning or Performance improvement, thus satisfying the third cri-
terion for the existence of a motive—namely, whether it serves to select out and
make more likely certain responses after they are rewarded.
Much energy has gone into trying to determine whether there is a relation-
ship between n Achievement and grades in school. Usually such a relationship is
not found to be significant; this has led a number of investigators (see Entwisle,
1972) to conclude that the n Achievement score cannot be valid because it does
not predict scholastic achievement. However, there is no theoretical reason for
predicting that high n Achievement should lead to better Performance in the
classroom under all conditions, any more than there is reason to believe, as ex-
plained earlier, that high n Achievement should always lead to better perfor-
mance regardless of the incentives present (see Table 7.1 in the next section). If
there is strong achievement pressure (see McKeachie, 1961; Wendt, 1955) or if
extrinsic incentives of any kind are introduced, there should be no relationship
between high n Achievement and grades. Nor should there be a relationship if
the work is very easy or much too hard. In other words, the relationship should
exist only when an achievement incentive or challenge exists. To expect other-
wise is like contending that hunger as a drive should lead to better Performance
even if its incentive of food is not present (see McClelland, 1980).
• WHAT IS THE INCENTIVE

FOR THE ACHIEVEMENT MOTIVE?
If food is the reward or incentive for the hunger drive, what is rewarding for
the achievement motive? We have repeatedly emphasized that "doing something
better" is the natural incentive for the achievement motive. It now is time to be-
come more precise, for people can do better for all sorts of reasons—to please
the teacher, to avoid criticism, to gain the approval of a loved one, or simply to
get some time off from work. What should be involved in the achievement mo-
tive is doing something better for its own sake, for the intrinsic satisfaction of
doing something better.
The point is nicely illustrated by an experiment conducted by French
(1955). Having first obtained n Achievement scores from story completions for
the Test of Insight under Standard conditions, French varied the incentives pro-
vided for performing better at a digit-symbol Substitution task on a second occa-
sion. Her subjects were officer candidates in school at a U.S. Air Force base.
Using a code that showed which digits went with which letters, the subjects
were to put the appropriate digits under each of a long series of letters typed
out below the code. To one group of subjects she said in a casual, friendly way,
"We are just experimenting today and we appreciate your cooperation very
much. We want to find out what kind of scores people make on these tests."
Table 7.1 lays out the motivational sequence in the terms used to describe it in
Figure 6.1. The demand to do the task is stated here in terms of an incentive to
please the experimenter, which should be appropriate to a motive disposition to
affiliate with others (n Affiliation; see Chapter 9), meaning that the aroused mo-
tivation to do well should be highest in subjects for whom the need for affilia-
tion is aroused.
A second group of subjects were told in a formal and serious manner by an
airman that the tests they were taking measured "a critical ability—the ability
to deal quickly and accurately with unfamiliar material. It is related to general
intelligence and will be related to your future career. Each man should try to
perform as well as possible." In other words, the incentive was for individuals
to show how capable they were, which should be related to the need to
achieve.
Finally, a third group was told that "the five men who make the best
scores in five minutes will be allowed to leave right away—the others will have
more practice periods and more tests." Here the incentive is to leave early and
get out of work, which should appeal particularly to those who dislike work or
need rest.
French also measured the degree to which the affiliation and achievement
motives were aroused by administering a second form of the Test of Insight just
after she gave the instructions mentioning different incentives to the various
groups. She found, as expected, that mentioning the achievement incentive had
the greatest effect on arousing achievement motivation for those high in disposi-
tional n Achievement on the first test (mean level = 6.27). The level of achieve-
ment motivation aroused for these subjects from the other incentives was lower
(mean = 4.20 for the Relaxed condition and 5.40 for the Time-off condition).
The level of achievement motivation aroused for those low in dispositional
n Achievement was significantly lower across all incentive conditions, as expect-
ed. For those high in n Achievement, the picture is quite consistent and easily
understandable: the demand to do the task couched in terms of achievement in-
centives contacted their high achievement motive disposition and aroused more
achievement motivation, which expressed itself in performing better.
As Table 7.1 shows, introducing an achievement incentive for the subjects
—to show how capable they were—produced the highest average number of
digit-symbol substitutions, but its effect was most marked on those who had
scored high in dispositional n Achievement on a previous occasion. Under this
incentive the subjects high in n Achievement produced almost twice as many
substitutions as those low in n Achievement. On the other hand, the two other
incentives—involving affiliation or time off from work—did not elicit better per-
formance from those high in n Achievement as compared with those low in
n Achievement.
For Group 1, mentioning an affiliative incentive for doing the task ("to
please me") meant the men for whom this instruction had aroused the affiliation
motive performed significantly better. In this group the correlation between
aroused affiliation motivation and subsequent Performance was .48, p < .01, as
compared with similar correlations of —.02 and —.13 in the other incentive
groups. In other words, any affiliation motivation aroused in the other condi-
tions did not lead to better Performance, because better Performance was not
tied in these conditions to the affiliative incentive. To put it another way,
aroused affiliation motivation will lead to better Performance if better perfor-
mance is cognitively understood to be the way to satisfy the affiliation motive
(that is, to please the experimenter). This makes the point once again that
better Performance does not always signify more aroused achievement motiva-
tion. Better Performance can be undertaken to satisfy many different
motives.
What motivation was aroused in the group promised time off cannot be de-
termined very well, because only achievement and affiliation motivation were
scored. Those low in dispositional n Achievement had the highest level of
aroused achievement motivation in this condition, but it is possible that it was
avoidance motivation (which could have shown up in one of the measures of
fear of failure; see Chapter 10) that led them to work harder to escape from
work. It is also possible that for such people the motivation to achieve is
aroused primarily by external pressures, or that some other type of motivation
was aroused that was not scored.
The main point is that those high in n Achievement actually performed
somewhat less well than those low in n Achievement when the incentive was
time off from work. So subjects high in n Achievement do not always perform
better than those low in n Achievement. They apparently perform better only
when an achievement incentive is present in the Situation. As these results sug-
gest, an achievement incentive is one in which a person gets satisfaction from
doing something better for its own sake, or to show that he or she is more capa-
ble of doing something. As Deci (1975) has shown in a number of experiments,
if extrinsic incentives are provided for doing something, the intrinsic satisfaction
from doing it well tends to be lost, and subjects high in n Achievement do not
perform better under such conditions.
Evidence for the importance of intrinsic achievement satisfaction to subjects
Table 7.1.
INFLUENCE ON PERFORMANCE OF VARIOUS INCENTIVES IN COMBINATION WITH n ACHIEVEMENT (after French, 1955)
Work Output: Average

A. Demand • # Incentive • C Motive Disposition >*D. Aroused Motivation Digit-Symbol Substitutions
Group 1 Do this task To please me Need for approval or To work to get approval 16.5:
affiliation High n Achievement = 17.7
Low n Achievement = 15.4
Group 2 Do this task To show your intelligence Need for Achievement To work to do better 23.5:
and leadership ability High n Achievement = 29.8
(how capable you are)
/'difference < -01
Group 3 Do this task Best five can leave early Need for rest, escape To work to escape 20.3:
and get time off from from work Situation High n Achievement = 18.2
work
high in n Achievement comes from a number of sources. For example, Wendt

(1955) found that when subjects were performing complex mental arithmetic
tasks, those high in n Achievement did significantly better than those low in
n Achievement so long as the experimenter left them alone. However, if the ex-
perimenter scheduled the work by constant reminders of what the subjects
should be doing, the better Performance of those high in n Achievement tended
to disappear. In an entirely different Situation, McKeachie (1961) studied the
Joint effect of teaching style and students' motive levels on grades obtained in
College. He found that when external achievement cues in the classroom were
judged to be high, students high in n Achievement tended to do less well in the
course than those low in n Achievement. To put it the other way around, if
achievement cues were low—that is, if the instructor did not make a point of
encouraging competition or setting high Standards of achievement—those high
in n Achievement tended to do better than those low in n Achievement. That is,
they readily find intrinsic satisfaction in achieving, but external promptings tend
either to distract them or to encourage the students with low n Achievement to
do better.
Moderate Task Difficulty as an Achievement Incentive

In the French experiment summarized in Table 7.1, the nature of the achieve-
ment incentive was spelled out for the subjects by the experimenter. However, in
most of the work in this area, the achievement incentive is defined by the intrin-
sic difficulty or challenge of the task. If the incentive is to "do better," neither a
very easy task nor a very difficult one provides an opportunity to do better. If
the task is easy there is no question of doing it better, since anyone can do it; if
it is very difficult there is also no question of doing it better, because everyone is
likely to fail in attempting it.
It follows that moderately difficult tasks should provide people the best op-
portunity of proving they can do better. A large number of empirical studies
have demonstrated that subjects high in n Achievement prefer working on mod-
erately difficult tasks in which the probability of success lies somewhere between
.30 and .50. The reason is that they should prefer working under conditions
where the achievement incentive is maximal. Subjects low in n Achievement
show quite a different preference curve for tasks differing in difficulty.
Some typical results are shown in Figures 7.1 and 7.2, as reported by
deCharms and Carpenter (1968). The subjects were over two hundred children
from the fifth and seventh grades in inner city schools. After being given the
Standard test for n Achievement, they were familiarized with a set of words to
be spelled and arithmetic problems to be solved that were carefully graded into
levels of difficulty: "After all children had attempted all items, the various levels
of difficulty were explained to them and their own ability at each level was im-
pressed upon them by pointing out how many they had completed correctly"
(deCharms & Carpenter, 1968). It was then explained to them that they would
take a test individually in which they could choose the level of difficulty of each
item they would attempt. They would get more points for performing an item
correctly at a more difficult level. The subjects also knew what their probability
of success was at the level of difficulty they chose, because they could look back
at their previous Performance at that difficulty level. That is, there were ten
items at each difficulty level, and if they had done two correctly at that diffi-
culty level previously, their probability of success at that level of difficulty would
be two out of ten, or .20.
Figures 7.1 and 7.2 show that subjects high in n Achievement regularly
chose to work at difficulty levels where their probability of success was moderate
(from .30 to .50), whereas subjects low in n Achievement showed no such
preference. If anything, they biased their choices toward very difficult items
where their probability of success was quite low (from 0 to .20). In the case
15
I
-4 X\ \
\A
o
U
10 -
/
>
J \\\>
V yv Low n Achievement
c
i
i
High n Achievement
i
i
i 1 1 I 1 1 1 1 1 1 1
10 20 30 40 50 60 70 80 90 100
Probability of Success
Figure 7.1.
Percentage of Tasks Chosen at Each Level of Probability of Success on the Arithmetic Risk-taking
Task by Subjects Above and Below the Mediän n Achievement Score (after deCharms & Carpenter,
1968).
of spelling, there was a trend for the subjects low in n Achievement to choose
very easy items, which has also been found in other studies (see McClelland,
1958a).
Atkinson's Model for Predicting Moderate Risk Preference

In a classic paper, Atkinson (1957) presented a formal model that explained
these results. He assumed that the strength of the tendency to achieve on a
task (or the strength of preference for various tasks) is a Joint function of the
motive to achieve, the expectancy or probability of success (Ps), and the incen-
tive value of success (/5), where incentive value is defined as one minus the
15
10
Low n Achievement
o
U
High n Achievement
J_ _L I I J_ I J_ I I I I
0 10 20 30 40 50 60 70 80 90 100
Figure 7.2.
Percentage of Tasks Chosen at Each Level of Probability of Success on the Spelling Risk-taking
Task by Subjects Above and Below the Mediän n Achievement Score (after deCharms & Carpenter,
1968).
probability of success (1 — Ps). In other words, it is assumed that the value

of success is directly proportional to its difficulty. As Table 7.2 illustrates, if
the probability of success is moderate, or fifty-fifty, the product of Ps X Is is
maximal, explaining why moderately difficult tasks have the largest incentive
value. As Table 7.2 also makes clear, if the motive to achieve (Ms) is higher, as
in high n Achievement, the preference for moderately difficult tasks over either
easy or difficult tasks should be even greater (.75 to .27 versus .25 to .09).
The model also suggests that even those low in n Achievement should show
some preference for moderately difficult tasks, which is not the case, as Figures
7.1 and 7.2 show. However, Atkinson handled this problem by assuming that
subjects low in n Achievement are actually afraid of failure and tend to avoid
moderately difficult tasks for this reason, thus canceling out this preference (see
Chapter 10).
Atkinson's model introduced a confusion in terminology, as already men-
tioned, which must be made explicit to avoid trouble. He sometimes referred to
the product of the three variables in Table 7.2 as the aroused motivation to
achieve, which confuses the aroused achievement motivation with what we have
called the impulse to act (or to choose one task over another)—which is a prod-
uct of motivational and nonmotivational variables and which Hüll called excit-
atory potential (see Chapter 3). The level of achievement motivation aroused in
a Situation can be determined by giving a TAT under conditions of arousal (and
comparing the n Achievement level obtained with its level under normal testing
conditions), but Atkinson did not mean (and has never shown) that the aroused
achievement motivation level measured in this way is a Joint function of the
three determinants in Table 7.2. Rather, he meant to predict with his formula
the tendency to approach or work hard at various tasks, which should be distin-
guished from the achievement motivation actually aroused, as shown in the mo-
tivation sequence outlined in Figure 6.1.
Table 7.2.
AROUSED TENDENCY TO ACHIEVE AS A JOINT FUNCTION OF MOTIVE TO
ACHIEVE (Mv), EXPECTANCY OR PROBABILITY OF SUCCESS (Pv), AND
INCENTIVE VALUE OF SUCCESS (7V), WHERE 7V = 1 - Ps (after Atkinson, 1957)
Low n Achievement: High n Achievement:

Task Ms X Ps X Is = Approach Ms X Ps X I5 = Approach
Task A (hard) 1 X 0.10 X 0.90 = 0.09 3 X 0.10 X 0.90 = 0.27
Task B 1 X 0.50 X 0.50 - 0.25 3 X 0.50 X 0.50 - 0.75
Task C (easy) 1 X 0.90 x 0.10 = 0.09 3 X 0.90 X 0.10 - 0.27
Atkinson's model explains quite well the actual distribution of choices for
tasks of different difficulty, as Table 7.3 demonstrates. To test it, Litwin (1958)
used a ringtoss game in which choosing to stand at different distances from the
peg represented choosing tasks differing in difficulty. He had determined the in-
centive value of a ringer from different distances by asking a number of students
how much money up to one dollar should be awarded as a prize for a ringer
from each distance. The second column in Table 7.3 gives the answers he re-
ceived. He also asked the subjects to estimate what the chances in one hundred
were of an average person throwing a ringer from the same distances. The re-
sults are recorded as perceived success probability (Ps) in the fourth column
in Table 7.3. Litwin observed that the money reward column matched closely
the perceived difficulty of the task, as estimated from the reciprocal of the prob-
ability of success (1 — Ps). Choices or the approach tendency then should be a
Joint function of expectancy of success (Ps) times the value of success at that
level of difficulty figured either in terms of money or difficulty (1 — Ps). He
multiplied the two variables, Ps X (1 — Ps), and compared the predicted
attractiveness of various distances with the distribution of distances from which
students actually chose to throw. The curves match fairly well. The value times
probability of success equations predict that most choices will fall between
Lines 10 and 15, with fewer at shorter and longer distances, and this is what
happened.
Other psychologists have worried about the best way to estimate the incen-
tive value of tasks of different difficulty. Calculating it on the basis of 0.9 — Ps
Table 7.3.
THEORETICAL AND ACTUAL ATTRACTIVENESS OF TASKS OF DIFFERENT DIFFICULTY IN
A RINGTOSS GAME (after Litwin, 1958)
Incentive Value
(Predicted
Success Attractiveness, Value
Value Probability X Expectancy) Proportion
Lines Distant Cents for Difficulty Ps Difficulty Actual Choice
from Pega Ringer (1 - Ps) Perceived Actual Cents X Ps X Ps Out of 400 Times
1-3 $0.04 .01 .99 .04 .01 .00
4-6 0.14 .12 .88 .70 .12 .11 .05
7-9 0.29 .30 .70 .31 .20 .21 .09
10-12 0.50 .45 .55 .20 .27 .25 .33
13-15 0.62 .62 .38 .15 .24 .24 .34
16-18 0.81 .75 .25 .14 .20 .19 .18
a
Lines started ten inches from peg, and subsequent lines were ten inches apart.
instead of 1 — Ps gives results in this instance that more closely approximate

where people actually stood, for they apparently tended to underestimate in
practice the real difficulty of what they were attempting. That is, as a compari-
son of Columns 4 and 5 in Table 7.3 illustrates, they perceived the probability
of success as greater than the actual probability of success. Thus, they stood far-
ther away more often than they should have, according to the model. This is
also true of the results shown in Figures 7.1 and 7.2. That is, the actual peak
preference for task difficulty is in the .30 to .40 ränge rather than .50 as pre-
dicted by the model (see Schneider, 1978).
Teevan, Burdick, and Stoddard (1976) have even argued that incentive value
is not always inversely related to task difficulty. They found that whereas it was
true that subjects tended to associate larger incentives with greater task diffi-
culty (lower probability of success), this seemed not to be the case when subjects
were asked in advance how important it was to them to succeed at tasks of va-
rying difficulty. Under these conditions, subjects said it was most important to
them to succeed at the easiest and not the hardest tasks, in an apparent reversal
of the Atkinson assumption. However, this may only have meant that they
feared looking ridiculous for failing at very easy tasks, and not that they valued
success at the easy tasks more in the Atkinson sense.
Atkinson and Litwin (1960) also classified subjects in this experiment ac-
cording to whether they were high or low in n Achievement. Those with high
n Achievement tended to choose moderately difficult tasks (throwing from Lines
10 to 15) significantly more often than those with low n Achievement. In other
words, they showed the same trend as the whole group, only more so. This is as
it should be according to the model outlined in Table 7.2. A further example
will make the point even clearer:
Resultant
Motive Approach
Strength X Ps X 75, or 1 - Ps = Tendency
4 X .50 X .50 - 1.00
4 X .70 X .30 0.84
Difference = 0.16
1 X .50 X .50 0.25
1 X .70 X .30 = 0.21
Difference = 0.04
That is, the difference in the attractiveness of a moderately difficult task (Ps =
.50) over an easy task (Ps — .70) is greater (.16) for those high in n Achieve-
ment than for those low in n Achievement, where it equals .04. The greater
preference of subjects with high n Achievement for moderately difficult tasks has
been confirmed in a number of different studies (see Figures 7.1 and 7.2). For
example, Heckhausen (1963) found that they typically chose to work on mazes
that were only slightly more difficult (not easier or much harder) than mazes
they had succeeded in tracing earlier.
A Cognitive Explanation of the Preference for Moderate Risk

Weiner (1980a) has argued that the reason subjects with high n Achievement
choose moderately difficult tasks is that such tasks are more diagnostic of how
well they are doing. If the task is easy they will not know whether success was
due to their efforts, because everyone can do it, and if the task is difficult they
will also not be able to teil what their efforts produced, because they will fail.
Thus, they seek moderately difficult tasks to get information on the impact of
their efforts on Performance. In other words, they prefer these tasks not because
they get more pleasure out of working with them in Atkinson's sense, but be-
cause they can find out better from performing such tasks whether they can at-
tribute their success to their own efforts—which is the way Weiner defines
n Achievement (see Chapter 12).
Trope (1975) created a Situation in which the diagnostic value of a task var-
ied independently in relationship to its difficulty level. Diagnosticity was varied
by telling the subjects that 90 percent of subjects with high ability can do the
task as compared with only 60 percent of subjects with low ability. For another
task the comparable figures were 52 percent success for the high ability person
and 48 percent success for the low ability person. Under these conditions sub-
jects with a presumably strong achievement motive more often chose to work on
the easier task with high diagnostic value (90 percent versus 60 percent) rather
than on the more moderately difficult task with low diagnostic value (52 percent
versus 48 percent).
There is some evidence that subjects high in n Achievement prefer moder-
ately difficult tasks because they have more feedback value. W. U. Meyer (1975)
carried out the following experiment on a group of police trainees in Germany:
Each subject was asked to imagine a Situation where he had to shoot at a target
with 10 rings from 9 distances. For each distance, difficulty was conveyed by social
norms indicating what percentage of prospective policemen hit the "10" on the tar-
get (2, 6, 13, 32, 52, 73, 81, 85, and 98% respectively). The subjects then had to
imagine that they were going to shoot from each distance once and they would max-
imally concentrate on one shot. Given these conditions, the subjects estimated their
probability of hitting the 10 on the target from each distance. After indicating the
specific probability levels, the subjects were asked: "Suppose that immediately after
shooting from each distance the target disappears. You have no feedback whether
you were hitting or missing the 10. Suppose further that you can get feedback about
hitting or missing the 10 on the target for one distance (assume that there is no
other feedback possible). At what distance would you like to know if you were hit-
ting or missing the "10"?"
Figure 7.3 is from a reanalysis of Meyer's data by Heckhausen (1975). It

shows that the subjects higher in n Achievement chose much more often to get
feedback on how well they were doing at moderate levels of difficulty for them.
Subjects low in n Achievement showed no particular trend, or if anything, pre-
ferred getting feedback for low probability of success. This suggests, but does
not prove, that the subjects high in n Achievement were making their choices in
70
60
X>
3
in 50
o Positive Net
00
Hope Score
40
30
•5
20
10
Low Intermediate High

(0-3) (4-7) (8-11)
Subjective Probability of Success
Figure 7.3.
Frequency of Information Choice Within Low, Intermediate, and High Ps Range for Subjects with
Positive and Negative Net Hope n Achievement Scores (after Heckhausen, 1974).
order to get feedback, for it is possible that the tendency to choose moderate
levels of difficulty here was simply a generalization of their tendency always to
do this rather than because the instructions told them they would get feedback
at this level. Unfortunately, Weiner's attempts to show that subjects high in
n Achievement prefer more diagnostic tasks are inconclusive, because he used an
objective test measure of v Achievement rather than the usual n Achievement
score based on spontaneous thought content.
HOW HIGH NEED FOR ACHIEVEMENT

AFFECTS PERFORMANCE
Influence of Variations in the Challenge a Task Presents
So far, research findings indicate that subjects high in n Achievement prefer
working at levels of moderate risk where the probability of success is somewhere
in the ränge of .30 to .50. They are more attracted to such tasks, but do they
actually perform better when the challenge is moderate? Many studies have
shown that they do. A typical result is shown in Figure 7.4. In this study sub-
jects were asked to try to solve sixty anagrams placed on a page (Raynor &
Entin, 1982b). They were told that in order to be successful and move on to the
next task, they had to complete either fifty-four, thirty, or six out of the sixty
anagrams in the time allotted. If they did not meet the "moving-on" criterion,
they had to sit quietly while others continued to work. As Figure 7.5 shows,
when the moving-on criterion was of moderate difficulty (requiring half of the
anagrams to be solved), those high in n Achievement performed significantly
better than when the moving-on criterion was either very difficult or very easy.
Furthermore, those low in n Achievement did not show this pattern at all.
Karabenick and Yousseff (1968) also showed that subjects high in n Achieve-
32 -
31
30
\
; Wortced Coi-rectiy
\ High RAM
29
28
\
27 \
E
03
ob 26
03
c
- \
o 25
Numbei
24
23- ^•^^ Low RAM
22-
21-
i 54 of 60
1
30 of 60
1
6of60
Moving-on Criterion
Figure 7.4.
Number of Anagrams Worked Correctly as a Function of Resultant Achievement Motivation
(RAM) and Moving-on Criterion (after Raynor & Entin, 1982b).
ment performed better than subjects low in n Achievement only when learning
moderately difficult paired associates as compared with very easy or very diffi-
cult ones.
Atkinson (1958) had reported a similar result earlier in an experiment, some
of the results of which were reported in Table 6.1 in Chapter 6. In this study
subjects understood that they had different probabilities of success because they
were competing either with a larger or smaller number of other students. As
Figure 7.5 shows, the subjects high in n Achievement performed significantly
better at middle levels of probability of winning (one in three or one in two) as
compared with a very slight chance of winning (one in twenty) or an almost
certain win (the best three out of four students getting the money award). In
this case, a probability of success of fifty-fifty produced better Performance even
for those low in n Achievement, as ought to be the case according to the Atkin-
son model presented in Table 7.2. The effect of the intrinsic challenge of tasks of
different difficulties is accentuated in Figure 7.5 because the extrinsic, monetary
incentive was relatively low. Even so, it may have accounted for the extra effort
from those low in n Achievement at Ps = .50, as contrasted with the result in
High rt Achievement
53 {Ns = 5 t o 8 ) / ^ ^ ^
52 -
51 _ / / \ \
o / / \ \
50
/ /
8H 49
c c
es o /
48 _ /
a o \
47 \ \
- y* Low n Achievement
c 46 (Ns = 4 to8) \
45 -
\
44 -
i 1 1 1
20 3 2 4
Expectancy of Winning (Probability)
Figure 7.5.
Performance as a Function of n Achievement and Expectancy of Winning When the Monetary
Incentive Is Low (after Atkinson, 1958).
Figure 7.4, because they thought they had a better chance of getting the money
reward.
One difficulty in generalizing about the behavior of those low in n Achieve-
ment is that they are working in various Performance situations for a variety of
incentives induced by the instructions and the setting other than the achieve-
ment incentive of primary interest. Also, the definition of low n Achievement
varies from study to study: in the Atkinson study it is simply those who are
below the median in n Achievement, whereas Raynor and Entin (1982b) fol-
lowed the later Atkinson procedure of combining the n Achievement score with
a test anxiety score so that a person low in resultant achievement motivation
(RAM) is defined as someone low in n Achievement and high in anxiety, indi-
cating fear of failure. Whatever the results for those low in n Achievement, the
findings do agree that those high in n Achievement perform better under condi-
tions of moderate challenge.
At first glance the results for the easiest competition, where chances of suc-
cess are three out of four, or .75, would seem to present a problem for the At-
kinson model outlined in Table 7.2, for under this condition the subjects low in
n Achievement perform, if anything, somewhat better than those high in
n Achievement. This result cannot be predicted by a model that allows only for
motive strength to combine multiplicatively with Ps, since according to the
model, those high in n Achievement will always do better at all tasks, no matter
how easy or hard they are.
What is happening here is that those with high n Achievement are perceiv-
ing the easy task as even easier for them, so it has little incentive value to com-
bine with their high motive strength. See the following numerical example:
Approach
Ms X Ps X /5, or 1 - Ps = Tendency
Stated = .75
3 X Perceived = .95 X .05 = .14
1 X Perceived — .75 X .25 = .19
Difference in favor of low n Achievement = .05
In the example it is assumed that the stated probability of success of .75 is per-
ceived by those high in n Achievement as Ps = .95, which then yields an ap-
proach tendency by multiplication that is actually less than that for the subjects
low in n Achievement despite their lower motive strength. In general, people
with high n Achievement estimate tasks to be easier for them at the outset
(McClelland et al, 1953; Pottharst, 1955), so they are attracted to, and work
harder at, more difficult tasks and less hard at easy tasks than people with low
n Achievement.
Responding to Moderate Challenges in Everyday Life

Hoyos (1965) applied Atkinson's risk-taking model to driving. He predicted that
individuals with high n Achievement should be better drivers because they
would avoid extreme risks; in fact, he found they were less often involved in ac-
cidents and traffic violations. However, they took more calculated risks, such as
overloading, driving without a license, or committing parking violations.
O'Connor, Atkinson, and Horner (1966) used the risk-taking model to ex-
plain the effects of ability grouping on Performance in the classroom. They rea-
soned that in a normal classroom where children of very different abilities are
present, those with high ability perceive the probability of success to be high,
whereas those of low ability perceive it to be low. In such classes students with
high n Achievement, in comparing their Performance with that of other stu-
dents, would not be maximally challenged, for they would see that they could
easily do better than the really poor students and that they did not stand a
chance of doing better than the very good students. If they were grouped with
students of similar ability in homogeneous classes, however, they should be more
often presented comparisons in which they would stand around a fifty-fifty
chance of doing better than the other person; this corresponds to the optimal
achievement incentive value according to Atkinson's model. The results of com-
paring the Performance of students high and low in n Achievement when in ho-
mogeneous or heterogeneous classes are shown in Table 7.4. Clearly, at all abil-
ity levels, subjects high in n Achievement do much better than those low in
Table 7.4.
MOTIVATIONAL EFFECTS OF ABILITY GROUPING IN SIXTH GRADE: PERCENTAGE OF
STUDENTS SHOWING ABOVE-MEDIAN GROWTH FOR THEIR INTELLIGENCE LEVEL IN
READING AND ARITHMETIC ON CALIFORNIA ACHIEVEMENT TEST (after Atkinson & Birch,
1978, after O'Connor, Atkinson, & Horner, 1966)
, A , . w . . Homogeneous Classes Heterogeneous Classes

Resultant Achievement Motivation
(n Achievement-Test Anxiety) N Percentage Above Mediän N Percentage Above Mediän
IQ 125 and above
Both Areas Both Areas
High 24 71 37 46
Low 10 50 27 37
IQ 113 to 124
Both or One Area Both or One Area
High 11 90 17 41
Low 17 65 19 58
IQ 112 and below

High 8 88 8 38
Low 52 14 36
Note: When all levels of intelligence are combined, the difference between motivation groups within the homogeneous classes and the
difference between classes for the high motivation group are statistically significant.
n Achievement in homogeneous classes, whereas they do not do significantly

better in heterogeneous classes. This result is a nice confirmation of the predic-
tive value of Atkinson's model.
Persistence
The risk-taking model has also been applied to explaining the tendency to persist
in working at tasks of different difficulty levels. At first it was thought that sub-
jects high in n Achievement would persist longer in working at any task. For
example, French and Thomas (1958) found that 47 percent of the subjects high
in n Achievement persisted up to the time limit in working at an insoluble task,
as compared with only 2 percent of those low in n Achievement.
However, Feather (1961) reasoned that persistence ought to depend on
probability of success according to Atkinson's model. That is, subjects high in
n Achievement should persist longer when they begin to fail at an easy task
than when they fail at a very difficult task. He in fact found this to be the case,
as Figure 7.6 illustrates. The figure also suggests why this difference should be
obtained: As a subject begins to fail at an easy task that is supposed to have a
.70 probability of success, the perceived probability of success begins to get less
and moves into the area of maximal attraction for subjects high in n Achieve-
ment—namely, the area of moderate probability of success. On the other hand,
a hard task with a probability of success of only .05 should have little attraction
for subjects high in n Achievement; as they begin to fail, it should have even
less attraction, and persistence should be low. Thus, whether subjects high in
n Achievement persist when they begin to fail at a task depends very much on
the perceived difficulty of the task.
Feather predicted and found the reverse result for subjects low in
n Achievement: they persisted longer after failure at a difficult than at an easy
task (see also Feather, 1963). This result is predicted from Atkinson's model, be-
cause subjects low in n Achievement are perceived as fearing failure and avoid-
ing tasks of moderate difficulty (see Chapter 10). It follows that success at easy
and difficult tasks would have the opposite effect. Since success at an easy task
(Ps = .70) would increase the probability of success even more, the attraction of
continuing to do the task would be less for those high in n Achievement, and
they would tend to stop doing it. On the other hand, success at a very difficult
task might move the probability of success into the region of greater attractive-
ness for those high in n Achievement.
Effects of Too Much Motivation on Performance

As noted in Chapters 3 and 6, psychologists have found evidence that too much
motivation can interfere with Performance. So far no one has demonstrated a
curvilinear, or inverted U-shaped, relationship between n Achievement and per-
formance, indicating that too high a level of n Achievement leads to somewhat
lower Performance than a moderate level of n Achievement. However, several
studies have shown that if n Achievement is associated with a high n Affiliation,
100
High n Achievement
H o
80 Low n Achievement
I-
-o S
40 -
=1
•sl
20 -
1
.05 .20 .40 .60 .70 .80 1.00
Stated Probability of Success

Hard Task Difficulty Easy
Figure 7.6.
Relationship of High and Low n Achievement to Persistence at Easy and Difficult Tasks After
Failure (after Feather, 1961).
and incentives for both motives are present in the Situation, Performance can
suffer. For example, Entin (1968) asked junior high school boys to solve arith-
metic problems under either one of two types of feedback conditions. In one in-
stance they were told privately how well they had done—which was assumed to
be relevant to the achievement motive—and in the other condition their scores
were listed publicly where all could see them—an incentive relevant to the affili-
ation motive, because subjects high in n Affiliation tend to want social approval
(Chapter 9).
In a reanalysis of Entin's results, Atkinson and Birch (1978) assumed that
subjects who were low in n Achievement and n Affiliation and who received pri-
vate feedback would have the lowest level of motivation to perform, whereas
those high in both motives and receiving public feedback would have the highest
level of motivation. Subjects with other combinations of motive strength and
feedback were considered to be moderately motivated. The data in Figure 7.7
from this study show the inverted U-type curve supposedly characterizing the
relationship between increasing motivation and Performance. Neither the sub-
jects with the lowest or highest motivation performed as well as those with a
more moderate combination of motives and incentives to perform.
Low Moderate High

Motivation to Perform
Figure 7.7.
Effect of Increasing Motivation to Perform on Efficiency (after Atkinson & Birch, 1978, after Entin,
1968).
Similar results were obtained in a study by Sorrentino (1974). Instead of

using the incentive of private versus public feedback, he employed the incentive
of contingent versus noncontingent paths to going on to the next task, as ex-
plained in connection with the moving-on criterion used with the experiment
summarized in Figure 7.4. Under contingent conditions subjects were under
pressure to achieve a certain level of Performance before they could continue
with other parts of the experiment. Figure 7.8 displays the results for the
subjects high in n Achievement when they were also either low or high in
n Affiliation and working under contingent or noncontingent conditions. The
motivation to perform in this experiment can be conceptualized as a product
of three motivational factors: high n Achievement, high n Affiliation, and a
contingency incentive. For the first point in the curve, only one of the motiva-
tional factors (high n Achievement) is present. For the second two points in
the Performance curve, two out of the three motivational factors are present
(high n Achievement and either high n Affiliation or the contingency incentive).
For the fourth point in the curve, all three motivational factors are present.
Once again, the total strength of motivation shows an inverted U-type relation-
ship to Performance. The students solved more arithmetic problems under mod-
erate motivational pressure than under either low or very high motivational
pressure.
16 -
NC C NC C
Low n Affiliation High n Affiliation
Figur'e 7.8.
Performance Scores for High n Achievement-Low Test Anxiety Subjects in the Noncontingent (NC)
and Contingent (C) Conditions Who Are Also Classified as High and Low in n Affiliation (after
Sorrentino, 1974).
OTHER CHARACTERISTICS OF PEOPLE

WITH A STRONG NEED TO ACHIEVE
Personal Responsibility for Performance

On theoretical grounds, it has always been assumed that subjects high in
n Achievement would prefer being personally responsible for a Performance re-
sult, because only under such conditions could they feel satisfaction from doing
something better. Horowitz (1961) carried out a simple empirical test of this
theoretical expectation. He designed a board game in which the players as sales-
people were supposedly crossing the United States and trying to make as much
money as possible in encounters they were described as having in various cities.
In each encounter the players were asked to choose between two alternatives,
each of which gave them a moderate chance (Ps = .33) to make a certain
amount of money. In one alternative they could do some work—that is, take
personal responsibility for the outcome. They could choose to work on an arith-
metic problem, the solution to which, they were told, approximately one in
three students could reach in the time allotted. If they succeeded in solving the
problem in the time allotted, they would be considered to have made the sale
and its accompanying financial reward. Or they could choose another Option,
represented, for example, by the opportunity to invest in digging an oil well,
where the chances were also one in three of finding oil. To determine whether
or not they won on this alternative, they rolled a die with the understanding
that if either of two out of the six numbers came up, they would win. In other
words, people who played the game were repeatedly confronted with the oppor-
tunity either to take a chance through the roll of a die or to take an equal
chance in which they had to work to obtain the result. Under these conditions
the subjects high in n Achievement chose to take personal responsibility in situa-
tions of moderate risk significantly more often than those low in n Achievement.
A number of studies have shown that the Atkinson risk-taking model does
not apply when the risks involve pure chance rather than personal responsibility.
That is, subjects high in n Achievement are not more attracted to moderate
risks in games of chance such as poker dice (Hancock & Teevan, 1964; Littig,
1963; Raynor & Smith, 1966). However, a study by McClelland and Watson
(1973) shows that this is not always the case. They found that in a game of rou-
lette, subjects with high n Achievement tended to prefer moderate risks over
very short or very long odds, just as if the outcome depended on their perfor-
mance (see Figure 8.5). Perhaps the tendency to choose moderate risks is so
strong in subjects high in n Achievement that it generalizes to some situations
in which chance only is involved.
The focus on one's own ability to produce a Performance result may explain
why subjects high in n Achievement show less interpersonal sensitivity in the
achievement area (Berlew & Williams, 1964). In an experiment, subjects had an
opportunity to get to know each other by working on a common problem in
five-person groups. At the end of the Session the subjects rated themselves on
the extent to which they possessed certain characteristics such as being keen,
practical, warm, or influential. Each subject was also asked to judge how others
in the group would rate themselves. Those high in n Achievement, particularly
when the experimenters focused their attention on how important the work was
for determining leadership ability and intelligence, proved much less able to
judge accurately how the others would rate themselves on achievement traits
like being keen and practical. In short, they focused so much on doing well
themselves that they paid no attention to how achievement oriented the other
members of the group were.
Need for Performance Feedback

Theoretically, subjects high in n Achievement should prefer working in situa-
tions where they get feedback on how well they are doing. Otherwise they have
no way of knowing whether they are doing better than others or not. Several
different types of studies have confirmed the importance of Performance feed-
back to them. In an early study, French (1958b) introduced two different kinds
of feedback to two groups of subjects working on a common task. For one
group she drew attention to various things members of the group had been
doing that contributed to improving Performance on the task. For the other
group she drew attention to aflfiliative behaviors that contributed to the group's
working well together. She found that as contrasted with the subjects low in
n Achievement, those high in n Achievement worked subsequently more effi-
ciently after Performance feedback than after affiliative feedback. They wanted
to know how well they were doing in solving the problems worked on rather
than how well they were getting along with other members of the group.
In quite a different confirmation of this point, Kagan and Moss (1962) dem-
onstrated in a longitudinal study that boys with high n Achievement were rated
more interested in and better at mechanical activities such as carpentry or con-
structing model vehicles than those low in n Achievement. The likely explana-
tion for this finding is that construction activities give very concrete Performance
feedback. The boys knew how well they had done in terms of whether the
model they had constructed worked or whether the pieces of the airplane they
were making fitted together evenly. In contrast, social behaviors, or even study-
ing in school, do not give people such immediate direct Performance feedback
on how well they have done. And children higher in n Achievement do not tend
to perform better socially or in school than those lower in n Achievement.
Another type of very concrete Performance feedback is provided by pro-
grammed learning, in which a teaching machine gives immediate feedback on
the correctness of each step taken to learn something. Bartmann (1965) has
shown that subjects high in n Achievement profit more under such feedback
conditions than subjects low in n Achievement.
Money is a common way of giving people feedback on how well they are
doing. There is evidence that whereas money is not an incentive for subjects
high in n Achievement, they do use it as information on how successful their
Performance has been. For example, in the Atkinson (1958) experiment, some of
the results of which are presented in Figure 7.5, the subjects high in n Achieve-
ment performed better across all levels of expectancy of winning than subjects
low in n Achievement so long as the money incentive was $1.25. However, if
the money incentive was raised to $2.50, the curves for the high and low
n Achievement subjects across various expectancies of winning were nearly iden-
tical, and there was no difference in overall Performance. Furthermore, the sub-
jects high in n Achievement did not perform significantly better for a $2.50 in-
centive across all conditions than for a $1.25 incentive. In contrast, the subjects
low in n Achievement performed significantly better for the larger monetary
incentive than the smaller one. This and several other studies (for example,
Douvan, 1956) show that offering money rewards to subjects intrinsically high
in n Achievement does not increase their striving. In fact, as noted earlier, intro-
ducing extrinsic incentives to people who are intrinsically motivated for achieve-
ment may cause them to work less hard.
In the ringtoss experiment described in Table 7.3, however, the cents to be
awarded for a ringer from different distances rises at a more rapid rate for those
high in n Achievement than for those low in n Achievement (see Figure 7.9).
This would appear to mean that subjects with a strong achievement motive
want a more difficult achievement to be recognized by a larger monetary reward
than subjects with a weaker achievement motive. That is, subjects higher in
n Achievement tend to see money as a measure of success, rather than as an in-
centive to do better.
$1.00
0.90 High n Achievement /

Subjects (N= 10) /
Q 0.80
g> 0.70
£ 0.60
o
| 0.50
g 0.40
o
1 0.30 Low n Achievement

£ 0.20 Subjects (N= 10)
0.10
30 60 90 120 150 180

Distance of Lines from the Peg (Inches)
Figure 7.9.
Average Money Reward Assigned for Hits from Different Distances in a Ringtoss Game by Subjects
with High and Low n Achievement (after Litwin, 1958).
Innovativeness
Doing something better often implies doing it differently from before. It may in-
volve finding a different, shorter, or more efficient path to a goal. It leads to
cost-benefit calculations, such as "How can I get the same result with less
work?" or "How can the same amount of work produce a bigger result?" In
fact, it would be more accurate to call the achievement motive the efficiency mo-
tive, since the notion of doing things better involves efficiency calculations,
whereas achievement is a more generic term that can be applied to achieving the
goals for any motive.
It follows that individuals high in n Achievement should be more restless
and avoid routine. They should be more likely to seek out information to find
better ways of doing things. They should be more innovative. The very fact that
they are always seeking moderately challenging tasks means that they tend al-
ways to be moving on from what they had been doing to doing something a lit-
tle more challenging. Or consider the results shown in Figure 7.6 again. If peo-
ple high in n Achievement find a task unusually difficult, they will not persist in
working on it. If they succeed at a moderately difficult task, it will become eas-
ier and therefore less attractive to them, so they will move on to doing some-
thing eise.
A variety of studies support these expectations. When Indian farmers
viewed a telecast about a new irrigation technique, those high in n Achievement
retained more information from the program than those low in n Achievement
(Sinha & Mehta, 1972). In view of this, it is scarcely surprising to learn that
farmers in countries as different as India and Colombia adopt more innovative
agricultural practices if they are high rather than low in n Achievement (E. M.
Rogers & Neill, 1966; E. M. Rogers & Svenning, 1969). In a totally different oc-
cupational category, Sheppard and Belitsky (1966) also showed that individuals
high in n Achievement were more active in seeking out new information than
those low in n Achievement. They found that when blue-collar workers were
laid off, those with higher n Achievement started looking for work sooner, and
many more of them used at least five of eight different job-seeking techniques,
such as checking with a larger number of different companies, going out of town
to seek employment, checking newspapers from other towns, and so on. Those
with low n Achievement typically sat around waiting for their Company to call
them back to work.
Using quite a different measure of innovativeness, deCharms and Moeller
(1962) showed that the rise and fall of n Achievement scored in populär litera-
ture in the United States (see Chapter 11) was closely paralleled by a similar
rise and fall in patents issued per million inhabitants (see Figure 11.10 in Chap-
ter 11). If it can be assumed that the amount of achievement imagery in populär
literature reflects in a general way the number of persons with high n Achieve-
ment in the country, it could be taken as evidence that the larger the number of
such people, the more likely they are to come up with innovative ideas for doing
something better that get recorded in the patent index.
Seeking short cuts can also lead to cheating, and there is evidence that sub-
jects high in n Achievement are more likely to cheat. Mischel and Gilligan
(1964) conducted a study in which children were asked to keep track of their
own score in using a ray gun machine while the experimenter was out of the
room. Unknown to the children, the machine also kept accurate track of their
scores. The children high in n Achievement were more likely than those low in
n Achievement to report that they had made more hits than they actually had.
McClelland (1961) has argued that this is why entrepreneurial groups high in
n Achievement tend to acquire the reputation of being dishonest or tricky; it ap-
pears they are so fixated on finding a short cut to the goal that they may not be
too particular about the means they use to reach it.
Other studies have shown that individuals high in n Achievement are more
restless. They tend to travel more and are more likely to migrate (Kolp, 1965).
The tendency toward change shows up even in graphic expression. As noted
earlier, Aronson (1958) found that those high in n Achievement produced
Single, discrete, and different S shapes and diagonals, whereas those low in
n Achievement tended to scribble over and over in the same space, producing
fuzzy lines and multiple waves. What the graphic designs characteristic of indi-
viduals with high n Achievement would seem to reflect is the constant search
for variety or new ways of doing things.
• SOCIAL CONSEQUENCES
OF A STRONG NEED TO ACHIEVE
Occupational Success
Psychologists often seem to expend most of their energy doing research on how
College sophomores perform various tasks in the laboratory. The early research
on n Achievement was no exception to this rule. Out of it came the generaliza-
tions that people high in n Achievement tend to seek out and do better at mod-
erately challenging tasks, take personal responsibility for their Performance, seek
Performance feedback on how well they are doing, and try new and more effi-
cient ways of doing things. Then it occurred to investigators that these charac-
teristics should have important effects on the way people behave in "real life,"
in the social world. So a whole line of research was undertaken that led ulti-
mately to the conclusion that the achievement motive is a key factor in eco-
nomic growth. That story is told in the rest of this chapter and in Chapter 11.
For once, psychologists were able to demonstrate that what began as a purely
academic or theoretical study of how the achievement motive affects perfor-
mance had applications in the real world that included an improved understand-
ing of how a basic motive influences the economic well-being of nations.
Very early it was realized that if individuals high in n Achievement had the
characteristics just reviewed, they ought to behave in ways that, under certain
circumstances, would lead to greater success in the real world (McClelland,
1955). For example, Mahone (1960) argued that they should be more realistic
in setting occupational aspirations for themselves. That is, they should neither
underaspire (prefer low-level occupations relative to their ability) nor overaspire
(choose occupations way beyond their ability). Rather, in accordance with At-
kinson's risk-taking model, they should aspire to occupations that they stood a
moderate chance of succeeding at relative to their ability. Mahone tested this
hypothesis by having judges rate the career aspirations of a number of under-
graduates at the University of Michigan as to whether they were underaspiring,
realistic, or overaspiring relative to the students' intelligence test scores, grade
point averages, and major field. The results were striking. Of the vocational
choices made by those high in n Achievement, 81 percent were confidently
agreed to be realistic versus unrealistic, as compared with only 52 percent of the
choices of those low in n Achievement, a difference that is highly significant sta-
tistically. This should give subjects high in n Achievement a better Start toward
occupational success, since they would be more likely to pursue occupations that
are realistic in terms of their abilities and Performance to date.
Crockett (1962) checked this possibility with a different source of data.
Using motive scores obtained from a national sample survey (Veroff, Atkinson,
Feld, & Gurin, 1960), he found that individuals with high n Achievement had
shown greater upward occupational mobility than those with low n Achieve-
ment. Figure 7.10 summarizes the results. He determined whether the subjects'
occupation had higher, the same, or lower prestige than their father's occupa-
tion. Those with high n Achievement had more often moved up on the occupa-
tional scale, as Figure 7.10 shows. Furthermore, the result cannot be the result
of some general "motivational" factor, since no differences in n Affiliation or
n Power scores were obtained for those who were occupationally upwardly mo-
bile. Duncan, Featherman, and Duncan (1972) confirmed Crockett's findings
after carefully examining and recoding the occupational classifications.
As would be expected from such findings, another national sample survey
conducted in 1976 found that U.S. men high in n Achievement are oriented pos-
itively toward work (Veroff, 1982). They report more job satisfaction, evaluate
their Jobs as interesting, do not see work as interfering with family, and prefer
work to leisure. Perhaps as a result of their good work adjustment, they report
few Symptoms of ill health, rate their overall happiness as high, attend church
frequently, and do not take drugs to relieve tension. In the same survey, high
n Achievement was not associated with work satisfaction among women, per-
haps because women's Jobs are not intrinsically as interesting as men's or be-
cause, as Veroff suggested, having been brought up before the women's libera-
tion movement, they were still oriented toward doing well in more traditional
i 100
JHigh
Je 80 -
|LOW
1
2 60 -
•g «
li 20 -
N=117 N=116 N=128

n Achievement n Affiliation n Power
Motivational Level
Figure 7.10.
Upward Occupational Mobility of Respondents High and Low in n Achievement, n Affiliation, and
n Power (after Crockett, 1962).
women's role activities. For instance, women high in n Achievement more often
reported participating in challenging leisure activities, perceived their marriages
as involving much interaction with their husbands, and worried about whether
they were doing well as mothers. Like the men high in n Achievement, they
also reported a higher sense of well-being, or zest, indicating that they were
finding achievement satisfactions in life, if not in work.
Perhaps as a result of the tendency to move upward or to perform better
occupationally, men from the middle class generally have higher n Achievement
scores than those from the lower class (Nuttall, 1964; B. C. Rosen, 1956, 1959;
Veroffet al., 1960). At the very highest socioeconomic class levels, n Achieve-
ment scores tend to drop off somewhat (B. C. Rosen, 1959), perhaps because
achievement challenges are less at this high level of prosperity and success. An
interesting exception to this trend, which tends to confirm such an interpreta-
tion, is provided by U.S. blacks. Their n Achievement score is higher the farther
they have risen up the socioeconomic class ladder (B. C. Rosen, 1959). There is
no dropping off in n Achievement among blacks classified as belonging to the
two highest classes, as there is among most ethnic groups. But this is quite con-
sistent with the notion that achievement opportunities are still very great even at
this socioeconomic class level for an oppressed minority like U.S. blacks. In fact,
blacks even at the highest occupational level are not matched in opportunity
with others at this level. At the other end of the scale, U.S. blacks of lower
socioeconomic Status have the lowest average n Achievement score, which
McClelland (1961) attributes partly to their history of dependency in a slave-
holding society and partly to their lack of achievement opportunities.
Several studies have been carried out to investigate the success of individu-
als high in n Achievement in real life. Kaltenbach and McClelland (1958) found
that individuals perceived as leaders in small towns had higher n Achievement
scores than those not so perceived. J. D. W. Andrews (1967) followed the ca-
reers of executives in a large Company over a three-year period and found that
those with high n Achievement scores received a significantly larger number of
promotions than those with low n Achievement scores.
However, this was true only in an achievement-oriented firm in which those
who did well were given an opportunity to advance. In another firm ruled
in an authoritarian manner by the owner, there was no relationship between
n Achievement and promotion. Later research demonstrated that n Achievement
was associated with promotions in a large firm only for managers who did not
manage large numbers of other people but who made individual contributions
on their own. McClelland and Boyatzis (1982) found that n Achievement in
managers at the time of entry into the American Telephone and Telegraph
(AT&T) Company was associated with promotion up to Level 3 in the Company
after sixteen years, but not above that point. They reason that at Level 4 and
above in the Company, the managerial job demands more influencing of other
people, and hence a higher n Power, rather than doing one's own work particu-
larly well, which is what characterizes individuals high in n Achievement. (See
Chapter 8.)
Entrepreneurial Success
According to McClelland (1961), high n Achievement should make people par-

ticularly likely to be interested in and able to do well at business, for business
requires that people take moderate risks, assume personal responsibility for their
own Performance, pay close attention to feedback in terms of costs and profits,
and find new or innovative ways to make a new product or provide a new Ser-
vice. These are precisely the characteristics that laboratory research had shown
belong more to the person high than low in n Achievement. McClelland (1961)
showed that generally speaking, young men higher in n Achievement are in fact
more attracted to business occupations in several countries. Tessler, O'Barr, and
Spain (1973) report that this is true even in a non-Western culture, the Kanuri
of Nigeria, as it becomes industrialized and its young people learn about busi-
ness occupations.
Andrews (1966) found that even in College, students with high n Achieve-
ment are more entrepreneurial than those low in n Achievement. They do not
get higher grades (Entwisle, 1972), but they more often investigate a course's re-
quirements before signing up for it, talk to a teacher about an examination be-
fore it is given, contact the teacher about the way an examination was graded,
and so forth.
McClelland reported in The Achieving Society (1961) that an associa-
tion between n Achievement and entrepreneurship occurred across cultures.
Among preliterate tribes, 75 percent of those with high n Achievement in
their folktales were characterized as having at least some full-time entrepre-
neurs, as contrasted with only 38 percent of the tribes with lesser amounts
of achievement imagery in their folktales. He also reported that business exe-
cutives in three countries (the United States, Italy, and Poland) had higher
average n Achievement scores than other Professionals of the same educa-
tional background.
A further examination of this difference indicated that it was particu-
larly the salespeople from these countries who scored significantly higher in
n Achievement than other managers. The explanation for this finding is the
same as the one given for the failure of n Achievement scores to predict promo-
tions into Jobs where the primary responsibility is managing others rather than
one's own Performance. Salespeople manage their own Performance, whereas
general managers manage other people's Performance. That is, salespeople must
make decisions about what prospects to call on, choose moderate risks, take per-
sonal responsibility for making calls, carefully monitor their success and failure
in making sales, and find innovative or new ways of persuading people to buy.
A strong achievement motive should interest them in doing all of these things
and therefore help them to like selling more and to do better at it. General
managers, on the other hand, are more likely to spend time trying to influence
other people rather than do things themselves (see Chapter 8). As might be ex-
pected from these findings, individuals with a strong need to achieve are more
interested in investing than in saving money (Regelmann, cited in Heckhausen,
1973) and make better managers of investment portfolios than individuals with
low n Achievement scores (Wormley, 1976).
Most of these studies are not longitudinal. Therefore, despite the strong
theoretical expectation that the achievement motive could be responsible for
entrepreneurial success, it is possible that people develop high n Achievement
because of the entrepreneurial requirements of the Jobs in which they find
themselves.
However, empirical findings indicate that there may be a causal relationship
between n Achievement scores and entrepreneurial success. One is the longitudi-
nal study already reported at AT&T in which n Achievement at entrance into
the Company was related to promotion up to Level 3, sixteen years later. An-
other study involved a fourteen-year follow-up of College students varying in
n Achievement to see what occupations they entered (McClelland, 1965b). As
contrasted with those low in n Achievement, those who scored high in College
were significantly more often found to be in business—most of them in small
business—some fourteen years after they took the test. And Wainer and Rubin
(1969) showed that the heads of more successful small research and develop-
ment companies had higher n Achievement than heads of similar, less successful
companies.
The most definitive data come from a study by Kock (1965) in which he ex-
amined the Performance of a number of comparable small knitwear firms in Fin-
land. He found that the n Achievement score of the owner-manager (or, in the
case of somewhat larger firms, of the three or four top executives) significantly
predicted increases in the number of workers for the firm over time, increases in
the gross value of Output, and increases in the gross amount of investment in ex-
panding the firm (McClelland & Winter, 1969/1971). Furthermore, training in
n Achievement improves the Performance of small business entrepreneurs
(Chapter 14).
• RELATIONSHIP OF THE ACHIEVEMENT MOTIVE

TO THE PROTESTANT ETHIC AND THE
RISE OF CAPITALISM
After studying the entrepreneurial characteristics of individuals high in

n Achievement, McClelland wondered if the achievement motive might not be
the basis for the association between the Protestant Reformation and the rise of
capitalism described so eloquently by the German sociologist Max Weber
(1904/1930). Weber had argued not that Protestants had invented capitalism,
but that Protestant businesspeople and workers were more energetic and entre-
preneurial than Roman Catholics. He feit that the Protestants' increased drive
was to prove through their own Performance that they were not damned, but
among the elect. Having broken with papal and church authority, they were
more on their own, responsible for their own Performance, and less dependent
on church authority to grant them indulgences for misbehaviors.
McClelland (1961) observed that the characteristics of Protestant entrepre-

neurs as described by Weber seemed to be similar to those of individuals high in
n Achievement, and also that the Protestant reform should have led Protestant
parents to stress self-reliance and monitoring one's own behavior more carefully.
This, in turn (according to data then available), should have promoted the de-
velopment of n Achievement particularly among Protestant children. Figure 7.11
summarizes these presumed relationships. The connection Weber outlined be-
tween Protestant reform and the dynamic spirit of modern capitalism was to be
explained, according to McClelland, by greater independence and mastery train-
ing by Protestant parents, which promoted the development of the achievement
motive, which led to the production of more active entrepreneurs, which should
lead to more rapid economic growth. McClelland set about systematically test-
ing the presumed causal links in this chain of events. The relationship between
D and E in Figure 7.11, or between high n Achievement and becoming a more
active entrepreneur, clearly exists, as has been discussed.
Protestant Reform, Early Independence Training, and the Achievement Motive

What about the connection between C and D in Figure 7.11—between indepen-
dence and mastery training by parents and n Achievement scores in sons? An
early study by Winterbottom (1958) showed that mothers of high n Achieve-
ment eight- to eleven-year-old boys reported, in response to a questionnaire, hav-
ing set high Performance Standards for them at an earlier age than mothers of
children with low n Achievement scores. The data are summarized in Figure
7.12. Mothers of boys with high n Achievement scores say they insisted more
often that the boys by the age of five or six be active and energetic, try hard
Weber's Hypothesis
A. Protestant Reform ^ B. Spirit of Modern Capitalism

(Self-reliance Values) (More Rapid Economic Growth)
E. More Active Entrepreneurs
C. Independence and Mastery D. n Achievement

Training by Parents *" in Sons
Figure 7.11.
Hypothesized Eflfect of the Protestant Reformation on n Achievement and Economic Growth (after
McClelland, 1961).
100
90
o
2 80 High n Achievement
Children
T3
70
-o
C
cd
60
50
40
30
/ Low n Achievement
eö 20
/ Children
10
/
1 1 1 1
4 5 6 7 9 10
Age of Child
Figure 7.12.
Cumulative Curves Showing the Proportion of Total Demands Made Up to Each Age Level as
Reported by Mothers of Children Scoring High and Low on n Achievement (McClelland, Atkinson,
Clark, & Lowell, 1953, after Winterbottom, 1958).
things for themselves, do well in school, look after their own possessions, and so
on.
Later investigations showed that this finding was limited to achievement de-
mands and did not cover caretaking demands, such as being able to undress and
go to bed by themselves, and that making achievement demands too early, as
well as too late, could be associated with lower n Achievement scores (McClel-
land, 1961). For example, Hayashi and Yamaushi (1964) reported that in Japan
among very young children, low achievement motivation is associated with high
demands on the mother's part, although this relationship reverses itself among
older children. It looks as if the mothers who are promoting the achievement
motive tailor their demands to just slightly above what the child can do easily.
In other words, they are maximizing the achievement incentives for the children
by setting tasks in which the probability of success is neither too low nor too
high; the tasks are in the moderate ränge of difficulty.
McClelland (1961) reports that in Brazil, where on the average mothers
make achievement demands on sons early, the n Achievement score of sons is
positively associated with the age at which their mothers make demands:
the mothers who set later ages for the achievements have sons with higher
n Achievement scores. The reverse is true in Germany, where the average age
set for the same achievements is almost a year later than in Brazil. In Germany,
the mothers who expect the achievements at an earlier age have sons with
higher n Achievement scores. In other words, there is an optimal age for moth-
ers to expect the achievements listed in the questionnaire if they wish to develop
the achievement motive. If the mother demands the achievement too early
(probability of success too low) or too late (probability of success too high), the
child is unlikely to get pleasure out of performing and therefore does not de-
velop high n Achievement.
Rosen and D'Andrade (1959) examined the behavior of parents as they ob-
served their sons working on a task. The boys, for example, were asked to Stack
blocks as high as they could with their left hand when blindfolded while their
parents watched. The parents were asked how well they thought their son could
do on the task, and then, as the child tried to build the tower, the comments of
the parents were recorded and coded. The results are shown in Figure 7.13,
with separate curves for the mothers and fathers. Both mothers and fathers
of the sons high in n Achievement tended to set somewhat higher levels of
aspiration for their son's Performance than did the parents of sons with low
n Achievement scores. Furthermore, they expressed more warmth and affection
for their sons than the parents of sons low in n Achievement. Finally, whereas
the mothers of the "highs" were more authoritarian, giving their sons more spe-
cific directions than the mothers of the "lows," the reverse was true of the fa-
thers. The fathers of the highs were much less authoritarian than the fathers of
Parent of "Highs" Is
Parent Behavior Lower Hjgher
Variables Standard Score Units: -0.5 0 +0.5

1. Level of Aspiration
Block Stacking(l), Height
Block Stacking (2 and 3), Height
Block Patterns to Be Copied, Difficulty
Ringtoss Distance
2. Warmth (Positive Affect)
3. Authoritarianism
a
Rejection (Irritation) ^ ^
a
Specific Directions 1
a
Pushing (Urging On) Father 1
a
Dominance (Decision Making) /
''Patents of'hiehs" predicled to be lower in \ariahle. permittinu more independence.
Figure 7.13.
Mean Differences in the Behavior of Parents of Sons with Low and High n Achievement Working
in Task Situations (after McClelland, 1961, after B. C. Rosen & D'Andrade, 1959).
the lows. What is important about this finding is that domineering behavior on
the part of the mother does not interfere with n Achievement, but it does if it
comes from the father. In terms of the hypothesis being checked, independence
training—that is, allowing the son to proceed on his own—does characterize the
fathers of sons with high n Achievement scores, but not the mothers.
Other results reported by McClelland (1961) confirm the strong connection
between paternal authoritarianism and low n Achievement scores. In Turkey,
for example, the only businessmen with high n Achievement scores were those
who had escaped paternal authoritarianism at an early age.
To return to Figure 7.11, what about the connection between A and C—be-
tween Protestantism and independence and mastery training? McClelland,
Rindiisbacher, and deCharms (1955) reported significantly later ages at which
independence and mastery were expected for Italian and Irish Catholic families
as compared with Protestant and Jewish families, and B. C. Rosen (1959) re-
ported similar differences for French Canadian and Italian Catholics versus
Protestants and Jews. More careful research demonstrated that these differences
tended to disappear in certain groups and may be related to social class differ-
ences. McClelland summarizes the research by saying that whereas some tradi-
tional Roman Catholic groups have the authoritarian attitudes associated with
low n Achievement, other Catholic groups have moved toward more general
norms for independence and mastery training, so one would not expect an over-
all Protestant-Catholic difference in n Achievement scores at present. In fact,
such a difference did not exist in the United States in the 1950s (Veroff et al.,
1960) or 1970s (Veroff, Depner, Kulka, & Dowan, 1980).
What seems more crucial is the spirit of reform or renewal that originally
characterized Protestants in the sixteenth Century and that many of the Protes-
tant churches lost as they became more established. For the spirit of reform is
likely to translate itself into adherents of the new faith's believing that they are
better than adherents of the old way of doing things. This emphasis on being
better translates itself into higher Standards of Performance for themselves and
for their children, as well as into higher levels of n Achievement. McClelland
(1961) found that preliterate cultures tended to have higher levels of n Achieve-
ment in folktales, which stressed independence training (McClelland & Fried-
man, 1952), and direct individual contact with the divine rather than contact
through established church authority. If people had to set Standards for their be-
havior based on their personal revelations rather than on church authority, they
tended to be more achievement oriented and more interested in personally figur-
ing out what was best for them to do.
Furthermore, McClelland (1961) observed what happened in two compara-
ble Mexican villages when one of them had been recently converted to a radical
form of Protestant Christianity—a modern replay of the Protestant Reformation.
The newly converted Protestants thought of themselves as superior in every way
to the "ignorant" traditional folk Catholics in the other village. In order to read
the Bible, they had become literate in their own language; they did not worship
"idols"; they did not drink and showed more self-control in various ways; and
they had become interested in better health and agricultural practices with the
help of Protestant missionaries. In other words, they had set for themselves
much higher achievement goals than had the people in the traditional nearby
village. This achievement orientation had clearly affected the children in the vil-
lages, since those from the Protestant village showed evidence of significantly
higher n Achievement. Thus, Protestant reform from this point of view is simply
a Special case of any type of ideological movement that helps people set higher
achievement goals for themselves. So Weber's hypothesis, or the relation be-
tween A and B in Figure 7.11, translates into a relationship between achieve-
ment ideology and more rapid economic growth. Further tests of this hypothesis
are discussed in Chapter 11.
However, the findings reported there suggest that reformist ideology is
translated so rapidly into the achievement motive, more energetic entre-
preneurship, and rapid economic growth that it is highly unlikely the change
comes about through the effect of reform on child-rearing attitudes. Thus,
in terms of Figure 7.11, A (the spirit of reform) must affect B (economic
growth) directly, as well as C (child-rearing) indirectly, which in a gener-
ation affects D (n Achievement), E (entrepreneurship), and B (economic
growth).
The chief value of this investigation lies in the way it was able to translate
Weber's argument into measurable variables in terms of which his hypotheses
could be scientifically tested. What was concluded in the course of making such
tests was that the Protestant Reformation was a Special case of an increased so-
cial emphasis on excellence or achievement, which, as in other such cases, fos-
tered the achievement motive, entrepreneurship, and economic growth, and may
have indirectly influenced child rearing for a time—until the reform became in-
stitutionalized. Not all social scientists accept this interpretation of Weber's hy-
pothesis—some because they do not like psychological explanations for complex
social phenomena and some because they disapprove of measuring variables and
testing their relationships in a scientific way.
• OTHER INFLUENCES AFFECTING

THE ACHIEVEMENT MOTIVE EARLY IN LIFE
Several investigators have tried to find the origins of achievement motivation in
the earliest years of a child's life. Cortes has suggested that there may be a he-
reditary factor involved, since he found that mesomorphic boys (with more mus-
cle relative to fat and skin) have significantly higher n Achievement scores (and
also n Power scores) than other boys (Cortes & Gatti, 1972). Their better mus-
cular development should enable them to perform the many motor activities of
early childhood with more pleasure and success, building the positive affective
associations that make up n Achievement through improved Performance.
From observing children's success and failure in working at various tasks,
Heckhausen (1967) concluded that achievement motivation first appears around
the age of three to three and a half, when "the success or failure of one's activ-
ity directs the pleasure or disappointment no longer only at the outcome of the
activity as such but rather at the seif, so that with success the child experiences
pleasure about his competence, and with failure experiences shame about his
incompetence".
Veroff (1969) agrees that the evaluation of Performance is an essential part
of achievement motivation and distinguishes it from simple joy in exercising
some skill, or what German psychologists had earlier called function pleasure.
But he sets the time at which this occurs even earlier, somewhere between the
age of one and a half to two and a half years, when language develops and chil-
dren can in effect say to themselves, I did it. He concludes that "autonomous
achievement goals generate from repeated experiences of pleasure in the child's
new found capacity to do what he was previously unable to do" (Veroff, 1969).
He measured the tendency to set such goals by the extent to which children
from kindergarten on would choose tasks that were moderately difficult for
them, as contrasted with very easy or very difficult tasks. He found that the
concept of achievement in comparison to what other children can do develops
somewhat later, and that both this concept and feelings of pleasure over perfor-
mance finally integrate into the n Achievement score measured in thought pro-
cesses, which he found to increase regularly from kindergarten through sixth
grade.
There is other evidence that the achievement motive is associated with the
development of maturity. Lasker (1978) demonstrated that n Achievement
scores were strongly associated with stages of ego development, as defined by a
measure introduced by Loevinger (1976), in samples from the island of Curagao
and from Chicago. Subjects at Loevinger's earliest, or impulsive and seif-
protective, stages showed almost no n Achievement; those at the middle, or
self-protective and conformist, stages had moderate n Achievement, whereas
those at the highest, or conscientious, level of ego development had high average
n Achievement scores.
This suggests that only subjects who have matured fully can have really
high n Achievement scores. Certainly the age trend reported by Veroff Supports
the notion that n Achievement scores increase with maturity. Lasker further
demonstrates that this result is partly due to the fact that certain of the scoring
characteristics for the n Achievement System appear only or primarily in the
records of subjects at Loevinger's higher levels of maturity. Lasker divides scor-
ing categories into five sets as follows, beginning with those that appear at the
lowest levels of ego development: (A) "imagery" and "theme"; (B) categories in
the form A + action and success anticipation; (C) categories in the form B +
need and world block; (D) categories in the form C + personal block and help;
and (E) categories in the form D + positive and negative feelings and failure
anticipation.
Studies by others (for example, Nakamura, 1981) move a few of these cate-
gories from one set to the next, but on the whole they confirm Lasker's results.
Figure 7.14 shows how scoring categories of the most elementary type begin to
appear at the lowest levels of ego development, whereas those of Types B and C
appear only at the conformist stages and categories of Types D and E appear
only at the highest stages. The explanation for this fact is that thinking about
30
27
24
21
18
15
12
A B C D E A B C D E A B C D E
il. ih
A B C D E A B C D E A B C D E
Ego Stage: Impulsive Self-protective Self-protective/ Conformist Seif-a wäre Conscientious
Conformist
Achievement Story Form
Figure 7.14.
Histogram of Percentage of Achievement Stories of Different Forms by Ego Stage As Defined by Loevinger (1976) (after
Lasker, 1978).
possible obstacles in the way of achievement represents a higher level of ego de-
velopment, and a person's thinking about his or her own characteristics as possi-
ble blocks represents an even higher stage of self-awareness. A person's stating
how he or she would feel in relation to an achievement event, or even anticipat-
ing failure, would represent a still further differentiation of the seif from an on-
going story sequence. But whatever the interpretation, the results strongly sug-
gest that the highest levels of n Achievement are dependent on overall maturity.
Parental Emphasis on Learning to Control Autonomie Functions

Is there anything that triggers the development of the achievement motive more
in some individuals than others? A study by McClelland and Pilon (1983) sug-
gests that parents may have an influence much earlier than was suggested by the
studies on achievement goal setting for children by mothers in middle child-
hood. They administered the Standard picture-story test for n Achievement to

adults at the age of thirty-one whose mothers had been extensively interviewed
as to their child-rearing practices twenty-six years earlier by Sears, Maccoby,
and Lewin (1957) when their children were five years old. McClelland and Pilon
correlated the ratings on some sixty-eight child-rearing practices, as reported by
the mothers, with the adult n Achievement scores of their children twenty-six
years later. Some of the results of this investigation are shown in Table 7.5.
At the top of the table are listed a number of early child-rearing practices
that might be supposed to relate to adult n Achievement, according to theory or
earlier research on middle childhood. None of these practices related consis-
tently to adult n Achievement. The only one that fitted with earlier results is
that mothers who feit it was very important for their child to do well in school
tended to have children—particularly if they were sons and from white-collar
families—who had higher n Achievement. This is consistent with earlier studies,
most of which employed middle-class boys as subjects.
The results for three new child-rearing practices are also shown in Table
7.5, two of which showed significant relationships to adult n Achievement across
several subsamples of the total population. That is, since these results were not
expected and might have been due to chance in the large number of correlations
tested, it is necessary to see if they hold up or cross-validate in different samples
of subjects. Scheduling of feeding and severity of toilet training are both associ-
ated with high n Achievement, not only in white-collar and blue-collar families
and for males and females, but also for two randomly generated sets of subjects
out of the total of seventy-eight. What is curious about this result is that both
practices refer to a very early period in the child's life; scheduling feeding nor-
mally begins at birth, and toilet training in this sample during the second year
of life. This places the origins of the achievement motive earlier than was sug-
gested by Heckhausen or Veroff, although their reasoning that parental emphasis
on mastery is important would apply to these areas of behavior as well. Both
practices relate to the earliest tasks set for children by the parent—in one case,
to regulate their hunger to correspond to the times when they are fed, and in
the other, to regulate their sphincters so they are released only in appropriate
places.
Two characteristics appear to differentiate what the child learns in these
areas as compared with other areas. First, the learning involves control of auto-
nomic functions before the child is capable of much cognitive processing of what
is going on, of having much sense of seif, or of understanding that " I " did it.
These later cognitive additions to affective associations centered around early
Performance in these areas may have more to do with the value consciously
placed on achievement.
Second, the pleasurable experiences in these areas derive from progressive
mastery of or progressive improvement in timing hunger or eliminating the
body's wastes. We hypothesized in Chapter 5 that the need for achievement
might derive from the natural variety incentive, from the pleasure people feel in
being able to produce a response that confirms a moderate expectation of suc-
cess. Having many affective experiences of this type based on different expecta-
Table 7.5.
CORRELATIONS OF EARLY CHILD-REARING PRACTICES WITH SOCIAL CLASS AND ADULT
n ACHIEVEMENT (after McClelland & Pilon, 1983)
Correlations with n Achievement for

White- Blue-
Social All Collar Collar
Child-rearing practice Class" Subjects Familiesb Familiesc Males Females
(Scale: 1-8 or 1-9) (N = 78) (N = 78) (N = 46) (N = 32) (N = 38) (N = 40)
Early tasks -.18 -.10 -.05 -.13 -.05 -.15
Mother punishes dependency .30** -.12 -.05 -.28 -.01 -.24
Important child do well in school -.13 .16 .30* .01 .24 .10
Regulär System for earning money .05 -.12 -.04 -.28 .05 -.37*
Use of reasoning .24 -.06 -.09 -.07 -.21 .14
Scheduling of feeding .06 .33** .25t .45** .29t .38*
Severity of toilet training -.37*** .41*** .57*** .28| .43* .38*
Standards for neatness -.05 .23* .39** .03 .36* .10
Note: N varies slightly because of missing child-rearing ratings. The n Achievement score has been corrected for regression on protocol
length. Correlation of n Achievement with the social class of the family of origin = .06.
a
Based on the prestige of the father's occupation and income on a scale of 1 to 9 (Sears, Maccoby, & Lewin, 1957, p. 424; direction here is
reversed).
b
Social class scaled 6 through 9.
c
Social class scaled 1 through 5.
t/? < 10 in the predicted direction.
*p < .05.
**p < .01.
***p < .001.
tions should develop a variety motive or a play motive, but not an achievement
motive. What characterizes high n Achievement is getting pleasure from progres-
sive improvement in a particular line of endeavor, not from getting pleasure from
several unrelated lines of activity. What parents who stress progressive mastery
of autonomic control appear to be doing is placing a constraint on the variety
incentive by focusing the pleasure derived from it on increasing probabilities of
success in a given area.
We might question whether it is really these early child-rearing practices in
themselves that promote the achievement motive. Perhaps they are part of the
mother's general attitudes, which continue to shape the child's achievement mo-
tivation over the ensuing years. It is true that both of these practices are part of
larger strictness-permissiveness syndromes, but neither of these overall syn-
dromes, nor their other components, are related in any way to adult n Achieve-
ment scores. It is scheduling feeding and toilet training only that appear to be
associated with later n Achievement (McClelland & Pilon, 1983).
Such a result makes it easier to understand why the presumed early affec-
tive learning that shapes the n Achievement score is not adequately represented
in conscious choices of achievement values, since this learning occurs before
later cognitive developments represented by achievement values. What seems
most likely is that an affective core for n Achievement is strongly or weakly es-
tablished very early from experiences of progressive mastery in learning about
feeding and toilet training, and that this affective core gets elaborated—that is,
either reinforced or extinguished by later experiences and elaborated by the nat-
ural course of cognitive development. Certainly there is no reason not to believe
that the achievement goals set by parents in middle childhood cannot also
strengthen or diminish the achievement motive core formed earlier in life. Thus,
specifically designed educational experiences in adulthood also can strengthen
the achievement or any other motive (see Chapter 14). The extent to which a
child is taught or learns later to consciously value achievement is seen in this
model to be quite independent of the achievement motive based on the pleasure
and pain experienced in relation to improving Performance in early life.
NOTES AND QUERIES
1. Try scoring the Sermon on the Mount or other sayings of Jesus accord-
ing to the scoring definition given in Chapter 6 to see if Jesus had high
n Achievement.
2. Several characteristics of people who have high n Achievement are de-
scribed in the chapter. Try to explain the way they doodle in terms of these
characteristics.
3. Would you expect people who work hard and conscientiously to have high
n Achievement? Why or why not?
4. Generally, students high in n Achievement do not get better grades in
school. Why do you think this is true? What motives, in your opinion,
should lead people to do better in school?
5. Some people just seem to be "born losers"; they keep getting into situations
where they have to lose out, or they seem to have a great deal of bad luck.
Is there information in this chapter that might suggest why they have so
much bad luck and what kind of motivation they might have?
6. Score the imaginative stories you wrote for Chapter 1 for n Achievement.
Make a checklist of the characteristics a person high in n Achievement
should show, and decide whether you have shown those characteristic on a
number of occasions (recall them specifically). Does the amount of achieve-
ment imagery in your stories agree with the number of achievement-related
characteristics you have shown? Do you believe you have high or low
n Achievement? Were there any surprises in the result? Is the value you
place on achievement higher or lower than your estimated n Achievement
score?
7. In Table 7.4 it appears that subjects high in n Achievement more often are
high in intelligence also, at least in homogeneous classes. Could it be that
some students do better in homogeneous classes because they have higher
intelligence, and not because they have higher n Achievement?
8. Economists often argue that people in general will do what it is in their self-
interest to do; that is, they will be guided rationally by the probability of
success of various alternatives. For example, if there have been several crop
failures in a row, farmers on the whole will give up farming and move to
the city, because the probability of success in farming is so low. Explain, in
terms of Atkinson's model of the relationship of n Achievement and the
probability of success, who will be likely to move or not move after crop
failures, and under what conditions.
9. Which type of person would be most likely to be happy alone on a desert
island—a person high in n Achievement, high in n Power, or high in
n Affiliation? Why? What would that person be likely to be found doing?
10. It has been said that whereas money is not everything, it will buy every-
thing in first place. Is this true for a person high in n Achievement?
11. Do you think creative writers or inventors would be more likely to be high
in n Achievement? Why?
12. If any reform movement should promote the achievement motive, should
not countries newly converted to Communism be higher in n Achievement?
(See Chapter 11 for evidence on this point.)
13. In terms of the description of how the achievement motive functions and
how it is developed, why do you think U.S. blacks of lower-class back-
ground scored low in n Achievement in the 1950s?
14. It has been argued that it is absurd to conclude that lower-class blacks have
lower n Achievement when they obviously want to do many things well, al-
though they may not be the same things middle-class whites want to do
well. Comment on this argument. Would the argument have the same force
if the achievement motive were called the efficiency motive?
15. Why should paternal authoritarianism lower n Achievement levels in sons
more than maternal authoritarianism?
16. Psychologists have been intrigued by the possibility that cultural "accidents"
might have unintended consequences for the development of human motives
like the achievement motive. McClelland (1961) speculated that wealthy
families in ancient Greece or in the U.S. South might have unwittingly low-
ered the n Achievement levels of their children by acquiring household
slaves who spoiled the children, undermining the emphasis on independence
and mastery needed to develop high n Achievement. Wendt (1961, 1974)
has suggested that the tendency for eminence to be associated with birth at
certain times of the year might be due to higher levels of n Achievement ac-
cidentally acquired by eminent people because they learned to walk in the
summer months, when there were greater opportunities to get pleasure from
independently mastering such an important skill.
Now that it appears that the achievement motive is developed by paren-

tal emphasis on strict toilet training, speculate on the possible accidental
side effects on n Achievement levels of the introduction of easily disposable
diapers. How could your speculation be tested?
17. Under what circumstances could Catholicism be associated with a greater
emphasis on Performance improvement and higher levels of n Achievement?
8
The Power Motive
• Measuring the Need for Power
• Evidence that the n Power Score Measures a Motive

Energizing Function of the Power Motive
Orienting Function of the Power Motive
Selective Function of the Power Motive
• Outlets of the Need for Power

Aggressiveness
Negative Self-image
Entry into Influential Occupations
Search for Prestige
Acting So As to Be Recognized in Small Groups
Risk Taking
Application of the Atkinson Risk-taking Model to n Power
• Catharsis
• Role of the Power Motive in Drinking
Drinking Conceived as a Means of Reducing Anxiety
Effect of Social Drinking on Thoughts About Power and Inhibition
Relationship of the Power Motive and Inhibition to Problem Drinking and Other
Activities
• How Maturity Modifies the Expression of the Power Motive

Deriving a Measure of Social-Emotional Maturity
Alternative Manifestations of the Power Motive as a Function of Maturity
Sex Differences in the Outlets for the Power Motive at Different Stages of Maturity
• Controlled and Impulsive Assertiveness in Organizational Behavior

Leadership Motive Syndrome Among Managers
Motive Profiles of U.S. Presidents
Motive Profile of Radicals
• Inhibited Power Motive Syndrome and Susceptibility to Illness
Cardiovascular Disease
Impaired Immune Function and Illness
• Origins of the Power Motive

Parental Permissiveness About Sex and Aggression
Sources of the Leadership Motive and Don Juan Syndromes
Loss of Status and the Development of the Need for Power
268
MEASURING THE NEED FOR POWER
Practically every Student of personality, from Freud to McDougall to Murray

and Cattell, has found that human beings are characterized by a need for power,
aggression, or domination. Anthropologists, biologists, and philosophers like
Nietzsche have all been impressed by human beings' aggressive urges.
The story of how a measure for the motive involved here was finally derived
is considerably more complicated than it was for the achievement motive. An
early measure patterned after the scoring System for n Achievement was discov-
ered to have some limitations and was eventually modified, expanded, and re-
fined into the measure generally used today. The account of its development il-
lustrates how science proceeds by trial and correction and, once again, how
crucial precise measurement is in science.
The original coding System for n Power was derived by Veroff (1957) from
examining the content of stories written by Student candidates for office while
they were waiting for election returns to be counted. The idea was that students
seeking office would be more likely than others to want power and that their
power need would be more apt to be aroused while they were waiting to see if
they would get power than under more neutral conditions. The stories written
when the power motive was aroused contained many more themes about control
of the means of influence than stories written by comparable subjects under con-
trol conditions.
So for scoring purposes, Veroff defined the goal of the power motive as ex-
erting influence, and he looked for increases in the subcategories already defined
as part of goal-seeking behavior for the achievement motive (see Table 6.4 in
Chapter 6). He found such categories as instrumental activity aimed at a power
goal, or blocks interfering with attaining it, to be more frequent in the stories
written by those waiting for election returns than by others. However, he did
not specifically look for new categories that might be unique to the arousal of
power motivation.
The n Power scoring System derived in this way was validated by examining
how subjects who scored high in it under nonarousal conditions behaved in a
variety of circumstances. For example, Veroff (1957) reported that those who
scored high in n Power were rated by their instructors as trying more frequently
to convince others than subjects who scored low in n Power. However, as subse-
quent studies showed, n Power scored in this way sometimes was associated
with assertive behavior and sometimes with unassertive behavior. For example,
Terhune (1968a, 1968b) found that teams consisting of individuals high in
n Power were more assertive, less cooperative, and more deceitful in a multi-
nation Simulation game than other teams. But Berlew (1959) had previously re-
ported that subjects high in n Power in small discussion groups were less asser-
tive, tending to give opinions and suggestions less often than other subjects. Fur-
thermore, sometimes people without social power in society scored high in
n Power, and sometimes those with social power scored high.
Does that mean that power motivation measured in this way is aroused by
having or not having social Status or power? In a national sample, Veroff, At-
269
kinson, Feld, and Gurin (1960) found that men with less education scored
higher in n Power than those with more education, whereas the reverse was true
for women: those with more education scored higher in n Power than those
with less education. Veroff and Veroff (1972) explained such apparent inconsis-
tencies by reasoning that the n Power score was really measuring fear of appear-
ing weak or fear of being deprived of power. Thus, both less educated men and
more educated women would be likely to feel somewhat oppressed or relatively
weaker in the competition for power in U.S. society. Hence, their high n Power
scores would represent a kind of defensive reaction to compensate for their rela-
tive deprivation or weakness.
The same line of reasoning could explain the inconsistency between behavior
in small groups and in the multination Simulation game, for in the latter the
participants never confronted each other directly. Rather, they wrote notes to
each other as diplomats would, and they could appear assertive without the dan-
ger of being personally shown up in front of others. In contrast, in Berlew's
discussion groups, the individuals were reacting to other people face to face
and might have avoided assertive behaviors for fear of being confronted and
shown up.
Such an interpretation of the Veroff n Power score suggested that it might
contain more of an element of anxiety than would be ideally desirable. Certainly
the conditions under which power motivation had been aroused included a
strong anxiety component: the Student politicians wrote the stories while they
were waiting for the votes to be counted and were worrying whether or not they
were going to win. This suggested that power motivation ought to be aroused in
individuals who had more certainty about the outcome of their power impulses.
So Uleman (1966) aroused power motivation by appointing students to be "ex-
perimenters" in a two-person gambling game for small stakes. The experiment-
ers were told that the purpose of the game was to study the effects of frustra-
tion, and they were taught how they could always win the gambling game so
that they could frustrate the subjects. As Orne (1962) has pointed out, the role
of being an experimenter is an extremely powerful one. Therefore, it was ex-
pected that power motivation would be aroused in the students assigned the ex-
perimenter role, since they knew how to defeat their subjects. This turned out to
be the case. Stories written by the Student experimenters were found by Uleman
to contain many more themes relating to successful, interpersonal influence than
stories written under control conditions. In contrast to Veroflf, Uleman looked
for new scoring categories not specifically related to the problem-solving se-
quence involved in the n Achievement scoring System and found several new
ones, such as more frequent references to the prestige of actors in the experi-
menters' stories than in those written by others. He developed a coding System
called n Influence, which has shown some promising correlations with a confi-
dent style of interpersonal influence (see Uleman, 1972).
Faced with these somewhat conflicting results, Winter (1967, 1973) under-
took a thorough revision of the n Power scoring System that includes categories
from both the Veroff and Uleman Systems and adds some new ones. Winter de-
rived this System by identifying shifts in story content under two new types of
power motivation arousal and then checking to see if the same shifts occurred
The Power Motive 271
under the Veroff and Uleman conditions for arousing power motivation. In the
new arousal experiments, subjects observed a powerful figure in action. In one
case, the power figure was a hypnotist behaving in a confident, successful man-
ner and ordering hypnotized subjects to do all sorts of things in front of a class,
after which the class members wrote their stories. In the other, the power figure
was John F. Kennedy giving his inaugural address as president shown on film to
a class of business school students.
In contrast to stories written under control conditions, those written after
exposure to the power figures contained more references to a person or persons
concerned about having "impact, control, or influence over another person,
group, or the world at large" (Winter, 1973). This became the new definition of
whether power imagery was present in a story or not. A further explanation
of the definition is given in Table 8.1, along with the subcategories Winter
found were more frequently present under all or most of the power motivation
arousal conditions. Note that he also distinguishes between Hope of Power—
an approach component of the motive—and Fear of Power—an avoidance
component.
The Winter revised scoring System for n Power is the one now in widest
use, and it has been employed in most of the studies to be reported in this chap-
ter. It correlates around .44 with the Veroff and Uleman scoring Systems (Win-
ter, 1973). Thus, it is not really measuring a different variable, but it represents
a refinement and expansion of earlier scoring Systems. The kinds of behavior
with which it is associated suggest that it is not as related to the fear of weak-
ness as the Veroff n Power score is. For example, in national sample surveys
conducted in 1957 and 1976, female College graduates scored high in n Power
less than 50 percent of the time according to the Winter System, in contrast to
more than 50 percent of the time according to the Veroff System; this is why
Veroff argued that their n Power was a reaction to their relative deprivation.
Women do score higher in n Power, according to the Winter System, in the pro-
fessions and sales, which are occupations that ought to require more power as-
sertiveness (Veroff, Depner, Kulka, & Douvan, 1980). Furthermore, in the 1976
survey, men high in fear of weakness (the Veroff System) had lower self-esteem
(if College educated), drank and took drugs to relieve tension, and attended
church less; however, they saw fatherhood as very important and feit good
about work competence, which Veroff (1982) interprets as compensation for
their underlying sense of weakness. Those high in n Power (the Winter System)
in the same survey showed none of these characteristics, except the tendency to
drink more. (See the discussion of drinking later in the chapter.)
• EVIDENCE THAT THE n POWER SCORE

MEASURES A MOTIVE
Energizing Function of the Power Motive
According to theory, motives drive, Orient, and select behavior. What is the evi-
dence that the power motive as measured this way drives, or energizes, the or-
Table 8.1.
BRIEF VERSION OF THE n POWER SCORING SYSTEM (after
Winter & Stewart, 1978)
Power Imagery: Scored if some person or group of persons in the story is concerned
about establishing, maintaining, or restoring power—that is, impact, control, or influence
over another person, group, or the world at large. Examples: (1) Someone shows power
concern through actions that in themselves express power. (2) Someone does something
that arouses strong positive or negative emotions in others. (3) Someone is described as
having a concern for reputation or position.
Subcategories, to Be Scored Only If Power Imagery is Scored:
Prestige: The characters are described in ways that increase or decrease their prestige.
Settings, titles, adjectives of Status, reputation, and alliance with some prestigeful person
or institution are all examples of prestige.
Stated Need for Power: An explicit Statement that the character wants to attain a power
goal. Not inferrable from mere instrumental activity.
Instrumental Act: Overt or mental activity by a character indicating that he or she is
doing something about attaining a power goal.
Block in the World: An explicit obstacle or disruption to the attempt to reach a power
goal.
Goal Anticipation: Some character is thinking about the power goal, with either positive
or negative anticipations.
Goal States: Affective or feeling states associated with attaining or not attaining the
power goal.
Effect: A distinct response by someone to the power actions of someone eise in the story,
or indication of widespread effect on the world at large.
Hope/Fear Distinction:
All stories are scored Hope of Power unless one or more of the following occurs in the
story, in which case it is scored Fear of Power: (1) The power goal is for the direct or
indirect benefit of someone eise. (2) The actor has doubt about his or her ability to
influence, control, or impress others. (3) The writer of the story suggests that power is
deceptive or has a flaw, as by the use of contrast, irony, or explicit Statement. Included
are cases in which characters feel happy after power failures and sad after power
successes.
Note: This brief version is intended for illustrative purposes only. It is not adequate for actual scoring
purposes. A complete version of the n Power manual, together with practice stories and procedures for
learning the System, is available in Winter (1973, Appendix 1). Scoring System Copyright © David G.
Winter, 1973.
ganism? Steele (1973, 1977) conducted an experiment that provides information

on this point. He aroused power motivation in Student subjects by having them
listen to a tape of an actor giving excerpts from famous inspirational Speeches,
such as Winston ChurchiH's Speech at Dunkirk or Henry v's Speeches in Shake-
speare's Life of King Henry v. In the control condition, subjects listened to

some tape-recorded travel descriptions. All students then wrote imaginative sto-
ries to pictures. Steele also attempted to arouse achievement motivation before
the subjects wrote their stories by having them work on a series of tasks—tasks
they were told measured their intelligence and alertness.
He found, as expected, that the stories written after the inspirational
Speeches contained significantly higher n Power scores than those written after
the travel tapes. On the other hand, the n Achievement scores obtained after the
ego-involving instructions were not significantly higher than the n Achievement
scores obtained after a more relaxed introduction to the tasks, probably because
the mean n Achievement score under the "relaxed" instructions was not as low
as could be expected from comparison with findings in other groups.
Steele measured activation in the subjects either through changes in their
moods, as determined from the Thayer Adjective Checklist (1967), or by shifts
in the concentrations of epinephrine (adrenaline) in the urine of the subjects
from before to after presentation of the Speeches. He looked for shifts in epi-
nephrine Output because emotional or physiological arousal is regularly associ-
ated with activation of the sympathetic nervous System, which increases the out-
put of certain neurohormones—for example, the catecholamines epinephrine and
norepinephrine—which are excreted in urine (see Frankenhaeuser, 1978). Table
8.2 presents the results for gains in epinephrine excretion. The biggest average
gain occurs from before to after listening to the inspirational Speeches. The gains
for the other three experimental conditions are lower and about the same. Con-
spicuously, there is no greater gain in epinephrine excretion for the achievement
arousal condition than for its control, which could mean either that the achieve-
ment arousal instructions failed or that the arousal of achievement motivation is
not associated with catecholamine shifts as closely as power motivation is.
The most striking result in Table 8.2 is that the amount of power imagery
present in the stories written after the inspirational Speeches correlates highly
with gains in epinephrine excretion. In other words, the subjects who showed
the greatest signs of physiological arousal were those who injected the highest
level of power concern into their stories. Gains in norepinephrine excretion were
also highly and significantly correlated with n Power scores after the inspira-
tional Speeches (r = .66; p < .01). Note also that shifts in signs of physiologi-
cal arousal are not significantly correlated with the motivational content of sto-
ries under any of the other conditions. It disappointed Steele particularly that
gains in epinephrine excretion under conditions of achievement arousal were not
significantly associated with n Achievement scores after arousal, since he
thought that any type of motivational arousal should have the same physiologi-
cal consequences. The results are not conclusive, because the correlation is in
the same direction although insignificant and because Steele may not have suc-
ceeded in arousing the achievement motive very strongly. His findings, however,
also raise the possibility that catecholamine shifts are more associated with
power than with achievement motivation.
An indirect indication that the power motive may be particularly associated
with catecholamine function comes from an occasion in which I determined that
of those who sat consistently to the right of me in class (from the students'
Table 8.2.
RELATIONSHIP OF EPINEPHRINE INCREASE TO MOTIVATIONAL
CONTENT AFTER DIFFERENT TYPES OF EXPERIMENTAL AROUSAL
(after McClelland, 1976, after Steele, 1973)
Mean Gains in
Correlations of
Epinephrine Epinephrine Change with
Excretion Motive Scores
(Nanograms
Condition N per Minute) n Power n Achievement
Power arousal (stirring tapes) 16 + 1.51 .71* .04
Power control (travel tapes) 14 + 0.27 .38 .06
Achievement arousal (ego-involving 19 + 0.35 -.02 .25
tasks)
Achievement control (relaxing tasks) 14 + 0.47 .10 .36
*p < .05.
point of view), 66 percent were high in n Power as compared with only 25 per-
cent of those who sat to the left of me (p < .05). Sitting on the right means
students must look more often to the left to observe the instructor. Looking to
the left is associated with dominance of the right hemisphere of the brain (see P.
Bakan, 1978), and there is evidence that norepinephrine function is concentrated
more on the right side of the brain (Oke, Keller, Mefford, & Adams, 1978; Rob-
inson, 1979).
The connection of the power motive and right hemisphere function is fur-
ther supported by the finding that there is a right hemisphere advantage in rec-
ognizing emotional Stimuli (Ley & Bryden, 1982) and the finding that n Power
scores are significantly correlated (/* = .34; N = 72; and p < .01) with recog-
nizing better the emotional tone of content-filtered speech in the taped version
of Rosenthal's test of nonverbal communication (Rosenthal, 1979). In this test,
subjects hear very brief excerpts of a person saying something to someone eise
that have been electronically filtered so they cannot recognize what is being said,
although they still can recognize the speaker's emotional tone. They might have
to decide, for instance, whether the sounds they hear are of a mother scolding a
child or a man making love to a woman. Some people are much better than
others at picking the correct response. This should indicate greater right hemi-
sphere dominance, and such people do tend to be higher in n Power. So we
might infer that those high in n Power tend to dominance of the right hemi-
sphere, which is also more associated with norepinephrine Output than the left
hemisphere.
In a second study, Steele (1977) focused on the adjective checklist measure
of activation. In particular, he employed Thayer's measure of General Activa-
tion, which covers the extent to which subjects describe themselves as feeling
lively, vigorous, energetic, and füll of pep. Steele measured the General Activa-
tion level and n Power before and after some travel tapes and inspirational
Speeches. As Figure 8.1 demonstrates, the inspirational Speeches produced
a marked increase both in the n Power score and the General Activation
level score. Furthermore, he found that for those who heard the inspirational
Speeches first, the amount of an individual's gain in General Activation level
was significantly correlated (r = .79; p < .01) with his or her n Power score in
stories written subsequently. This result exactly parallels that shown in Table 8.2
using the epinephrine measure of general activation. Steele also obtained a simi-
lar result when he gave the TAT first. That is, those who scored highest in
n Power on the TAT also showed the largest gain in General Activation when
they listened to the inspirational Speeches (r = .65; p < .01). In other words,
physiological activation to power Stimuli is closely associated with n Power in
thought content.
14
[ ] n Power
12
\ General Activation
g 5.00 10
o
5 4.00
6 b
c
o
2.00 4 ^
1.00
Before Travel After Travel Before Inspirational After Inspirational

Speech Speech Speech Speech
Figure 8.1.
Effects of Inspirational Speeches on n Power Scores and General Activation (data from Steele,
1977).
Whereas subjects high in n Power are more activated by power-related Stim-

uli, they tend to be Sensation avoiders rather than Sensation seekers, perhaps for
that very reason. In Steele's (1973) study, those whose n Power scores tended to
be highest in response to the power Stimuli of the inspirational Speeches de-
scribed themselves as Sensation avoiders. That is, they scored low on Zucker-
man's Sensation Seeking Scale (1974; see Figure 5.7 in Chapter 5 of this book).
The correlation between n Power scores under these arousal conditions and the
Sensation Seeking Scale was —.57, p < .05, whereas there was no correlation
under any of the other conditions. A similar result has been reported by David-
son, Saron, and McClelland (1980), who used event-related scalp electrical po-
tentials as a measure of the subjecf s response to Stimulation. In this experiment,
as sounds increased in intensity, subjects low in n Power reacted with electrical
Potentials of larger and larger amplitude, as recorded from their scalps. The
trend for those high in n Power was the reverse. They responded more to a
weak sound than those low in n Power, but as the sound increased in intensity,
their evoked event potentials declined. They were less responsive to stronger
Stimulation, perhaps because they are already so activated that they react by re-
ducing the response to Stimulation from without.
Orienting Function of the Power Motive

Individuais high in n Power are more sensitive to power-related Stimuli than to
neutral Stimuli, as demonstrated by findings assessing the electrical responsivity
of the brain to various Stimuli. In this type of experiment, electrodes are at-
tached to the scalps of human subjects over various areas of the brain to pick
up very small changes that occur in electrical potentials due to the response of
that portion of the brain to Stimulation. In one experiment (McClelland, David-
son, & Saron, 1979) the subjects were shown power-related pictures (for exam»
ple, a boxer or a car accident) several times interspersed with neutral pictures
(for example, an old man Standing in a boat or a boy loafing on a bed). The
measure of the electrical response to the power and neutral picture exposures
was the sum of the amplitude of the first negative deflection (Nx) and the second
positive deflection (P2) in electrical potential. (See Figure 8.2 for an example of
how the deflections are identified.)
If there is any difference in the way subjects high and low in n Power re-
spond to power or neutral Stimuli from electrodes on the scalp, it would be ex-
pected to occur first over the occipital area at the back of the head, because that
scalp region lies over the portion of the brain where visual sensations are pro-
jected. That is exactly what happened. The subjects high in n Power, as com-
pared with those low in n Power, showed a much larger N{ -f Pi amplitude re-
sponse (shown by the two dots in Figure 8.2) to power than to neutral pictures
over the occipital area, but not over the frontal area. Note that the difference in
responsivity to power over neutral Stimuli occurs very early—within the first
quarter of a second after the Stimulus has been exposed. This is presumably be-
fore the subjects could have processed the sensations from the pictures very far
and reacted more to them perhaps because they evoked power associations.
Rather, those high in n Power seem already "preset" to react more sensitively
to power cues as soon as they occur.
Figure 8.2 illustrates the difference in electrical scalp potential responsive-
ness to power Stimuli of a subject high in n Power and a subject low in n Pow-
er. For this graph, the A^ -f P2 amplitude of the electrical scalp response to
power words (not pictures) is being recorded in the frontal area (Davidson,
Saron, & McClelland, 1980). It is greater for the subject high in n Power than
for the subject low in n Power. The difference occurs about 100 to 150 millisec-
onds after Stimulus onset. This response is in the frontal rather than the occipi-
tal area, as in the previous study: word meaning is processed more in the
frontal area, whereas visual Stimuli are processed first in the occipital area. The
Nx -h P2 responsiveness was significantly different for subjects high and low in
n Power in response to power, but not neutral, words when Stimulus intensity
was low. When the intensity of the spoken words was raised, however, the Sen-
sation avoidance characteristic of subjects high in n Power appeared.
Selective Function of the Power Motive

Mc Adams (1982b) has reported that subjects high in n Power recall more
"peak" experiences that are described in power terms. A peak experience is one
a person describes as having had great emotional importance. For instance, a
male Student, when asked for such experiences, may recall how emotionally
Subject Characterized by
High n Power
Low n Power
Amplitude ~
of 10 Microvolts
Wave
100 Milliseconds
Time
Figure 8.2.
Event Electrical Response Potentials from Scalp Leads in the Frontal Area in Response to Power-
Related Words Spoken at a Moderate Intensity Level for a Subject High and a Subject Low in
n Power. The electrical response potentials are based on a mean of twenty word Stimuli per
average. Stimulus onset coincides with the beginning of the wave form. The dots correspond to the
first negative (/V,) and second positive (P,) peaks. This peak-to-peak amplitude was the dependent
measure employed (after Davidson, Saron, & McClelland, 1980).
charged he was when he played a part in a play that was loudly applauded by
the audience. This would be considered a power peak experience, because he
was obviously moved by having impact on others.
Subjects high in n Power had more power-related peak experiences to begin
with. A more controlled experiment was carried out by McAdams and McClel-
land (1983) in which the subjects listened to a taped recording of someone tell-
ing a story about a picture they were looking at. The story contained thirty
facts relating to power, thirty related to intimacy, and fifteen neutral facts. At
the conclusion of the story, the subjects filled out a Mood Adjective Checklist
and then unexpectedly were asked to recall the story in as much detail as possi-
ble. To control for ability in recall, the number of power facts remembered was
compared with the number of neutral facts remembered. The subjects high in
n Power recalled significantly more power facts relative to neutral facts than
those low in n Power. This difference was shown to persist at least twenty min-
utes later in another experiment in which recall was delayed and the intervening
time filled with other activities.
Selective recall should lead to faster learning of power-related materials by
subjects high in n Power. An experiment by McClelland, Davidson, Saron, and
Floor (1980) reported such a trend. Subjects who participated in the evoked po-
tential experiment previously described were asked to learn to associate either a
power or a neutral word to a power or a neutral picture Stimulus. That is, when
the Stimulus appeared, they were given two seconds to say the word before it
appeared. After a number of presentations of the twenty picture-word pairs,
most of the subjects had learned most of the associations. To test the relative
speed with which subjects high and low in n Power were learning the different
types of associations, the number of correct responses made on Trials 3 and 4
was compared. As compared with the subjects low in n Power, those high in
n Power gave a significantly higher number of correct responses on Trials 3 and
4 to the pairs in which one of the components was power related—that is, either
a power-neutral or neutral-power pair. There was not a significant difference in
the number of correct responses given by the two motive groups to the power-
power pairs—perhaps because these were the easiest pairs for all subjects to
learn—or to the neutral-neutral pairs. The faster learning of power words to
neutral pictures by subjects high in n Power is illustrated in Figure 8.3.
This study also demonstrated a link between n Power and catecholamine
function, as indexed by the increase in 3-methoxy-4-hydroxyphenylglycol
(MHPG) in urine collected before and after the experiment. MHPG is largely a
metabolite of brain, rather than peripheral, norepinephrine turnover (Maas, Hat-
tox, Greene, & Landis, 1979). In this experiment there was a significantly
greater power pair learning advantage for subjects both high in n Power and
brain norepinephrine turnover (an increase in MHPG from before to after the
experiment) as compared with those low in both measures. Furthermore, the re-
call of more power than neutral facts from the story in the experiment previ-
ously cited is also associated significantly with being high in n Power and high
in brain norepinephrine turnover during the course of the experiment, as con-
trasted with being low in both variables (McClelland & Maddocks, 1984). The
fact that high brain norepinephrine turnover is not in either experiment connect-
ed—in combination with high n Affiliation—with better recall of affiliative facts
suggests that greater general arousal is not involved, but an arousal specific to
power motivational content.
Since the injection of norepinephrine into rat brains increases seif-
Stimulation (see Chapter 4), and since exciting drugs like amphetamines (or
"speed") increase norepinephrine at the synapse, it might be inferred that brain
norepinephrine represents a physiological reward System—not, as Stein (1975)
has argued, for all kinds of drive Systems, but specifically for the power motive.
To put it in everyday language, suppose brain norepinephrine turnover (as in-
dexed by MHPG in urine) represents the extent to which power Stimuli "turn
on" subjects, or make them "feel good." Clearly, the pictures used in this exper-
iment would have more turn-on value for some subjects than others. So these re-
sults mean that subjects who are high in n Power will learn power-related mate-
5.0
4.0
Subjects
I 3.0 Low in
n Power
0* •=
2.0
I
Z 1.0
Learning Trial
Figure 8.3.
Average Number of Correct Power Word Anticipations to Neutral Pictures on Successive Learning
Trials (McClelland, Davidson, Saron, & Floor, 1980).
rials faster if the materials have turn-on value for them than if they do not; in
particular, they will learn the power-related materials faster than subjects low in
n Power, for whom the Stimuli have no turn-on value.
The Situation is analogous to putting hungry rats in a maze with food at the
end that they either like or do not like. The rats who are hungry and like the
food will learn fastest, and those who are not hungry and dislike the food will
learn slowest. From this point of view, brain norepinephrine turnover can be
conceived of as the index of the reward or incentive value of power-related Stim-
uli. For a related finding at the animal level, see Weiss, Stone, & Harreil (1970).
Obviously, such an interpretation is speculative, based as it is at present on two
experiments involving relatively few subjects. It is consistent, however, with indi-
rect evidence and theory that link n Power to catecholamine function.
OUTLETS OF THE NEED FOR POWER

Aggressiveness
Just as in the case of the achievement motive, much research has focused on
how people with a strong power motive behave under various circumstances. In
particular, are they more aggressive? That is a question of Special interest to
those who tend to define motives in terms of characteristic acts. As it became
apparent that high n Power did not always lead to aggression, which is tightly
controlled and regulated in modern society, the attention of researchers shifted
to more subtle ways in which the impulse to assertiveness could express itself.
It might be expected that individuals high in n Power would be more com-
petitive and aggressive. Some early results suggested that this was so. Winter
(1973) found that College students who had received varsity letters for playing
competitive sports scored significantly higher in n Power than other men. Com-
petitive Sports were defined as those in which there was direct competition be-
tween one person and another or one team against another, such as football,
tennis, or hockey. McClelland (1975) later confirmed this finding for a sample of
older men no longer in school. Those higher in n Power voluntarily participated
in more competitive sports than those low in n Power. Furthermore, in the sam-
ple of older men, the n Power score was significantly correlated with the fre-
quency with which the men reported they got into arguments.
However, none of these relationships hold for women, which reminds us
once again that a motive is only one of the determinants of action. Values and
habits or skills also determine whether the power motive will erupt into asser-
tive action. To be openly assertive has been valued for men but not for women,
so even if a woman high in n Power has the impulse to argue, she might inhibit
it because she accepts the sex-role value of being friendly and cooperative.
Nowhere is this clearer than in aggression. Both men and women high in
n Power confessed to having more impulses to being aggressive that they control
at a certain level of maturity than men and women low in n Power (see Table
8.6 later in the Chapter). That is, they were first asked whether they had ever
feit like carrying out a number of impulsive aggressive acts like yelling at some-
one in traffic, throwing things around the room, or taking towels from a hotel
or motel; then they were asked whether they had ever carried out any of these
actions. More men and women high in n Power at a certain level of maturity
confessed to being angry and to having thought about doing such things, but
they did not in fact do them more often than individuals low in n Power
(McClelland, 1975).
Whether a man actually carries out aggressive actions depends very much
on social class values, as Winter (1973) demonstrated. He found that high
n Power was significantly associated with carrying out more impulsive aggressive
acts among working-class men, but not among middle-class men. Thus, there is
no simple, one-to-one relationship between n Power and aggressive actions.
Whether such actions will occur for those high in n Power depends on the other
determinants of action, namely, the person's skills or habits and values.
Veroff (1982) also pointed out that the power motive is particularly Situa-
tion dependent, because its goal is to "make a splash" or create excitement, and
the way to do this varies markedly with the Situation. So what has gradually
emerged out of research on the power motive is a picture of a person striving to
be assertive in ways that either are appropriate to the Situation or that are not
and therefore produce feelings of guilt and anxiety over the aggressive impulses.
What is not in question is the general conclusion that people high in n Power
strive to be assertive.
Negative Self-image
The disposition to be aggressive or assertive leads people to view themselves
negatively for having what are generally considered to be antisocial tendencies.
A sample of adult men and women aged around thirty-one years were asked to
fill out the Gough Adjective Checklist (Gough & Heilbrun, 1975). The task is
simply to mark any of three hundred adjectives that people think describe them-
selves. Table 8.3 lists the adjectives checked more often by men and women
high in n Power as compared with those low in n Power. The men with a
strong power motive recognized that they feit more assertive—rebellious, resent-
ful, and sulky—and obviously judged this characteristic rather negatively, as we
shall see. The self-picture of women high in n Power was similar, although the
adjectives used were slightly different. They feit more often cynical, bitter, and
resentful.
What is particularly striking about both sets of the self-descriptive adjectives
chosen is that they are so negative. There were many opportunities to pick posi-
tive adjectives like active, adaptable, adventurous, cooperative, or courageous,
which could signify an assertive disposition. It is as if the individuals high in
n Power recognize their aggressive impulses; judge them negatively, just as soci-
ety would judge them; and have a rather negative self-image in consequence.
Aggressive-assertive behavior is punished beginning in early childhood because it
is often antisocial, and people learn that it is bad to be aggressive in many—
although not all—situations. Hermann (1980) has confirmed these trends in cod-
Table 8.3.
CORRELATIONS BETWEEN FREQUENCY WITH WHICH SELF-DESCRIPTIVE
ADJECTIVE IS CHOSEN AND n POWER SCORE FOR MEN (N = 38) AND
WOMEN (N = 40)
Men Women
Correlations with Correlations with
Adjectives n Power Score Adjectives n Power Score
Coarse 47** Cynical .35*
Complicated .42* Complicated .35*
Disorderly .49** Disorderly .29t
High strung .43* Bitter .40*
Rebellious .43* Self-pitying .44*
Resentful .41* Resentful .39*
Sulky .52** Moody .35*
Spunky .39* Unaffected .35*
Hurried .38*
*p < .05.
**p < .01.
ing for n Power Speeches by members of the governing body of the Soviet
Union, the Politbüro. She found that those high in n Power did not favor
detente with the United States (r = —.34; p < .05)—that is, they favored
a more assertive foreign policy—and had a more negative self-image (r = .51;
p < .01).
National sample survey data from the United States also confirm the fact
people high in n Power view themselves more negatively (Veroff et al., 1980).
Both men and women with strong power motives report themselves as feeling
more inadequate or dissatisfied with various aspects of their lives and having
Problems with drinking or drug taking to relieve tension. (The relation of power
motivation to drinking will be taken up later.) What is clear is that people high
in n Power have more emotional problems, a fact confirmed in yet another sam-
ple of adults in which high n Power among men was correlated with trouble in
sleeping and among women, with more unpleasant dreams (McClelland, 1975).
Entry into Influential Occupations

One of the ways individuals can exercise influence in more socialized ways is
through the occupations they choose to enter. Winter (1973) found that male
students planning to have a career in teaching, psychology, the ministry, busi-
ness, or journalism had significantly higher n Power scores than students plan-
ning to enter other occupations, including law and medicine. These results were
for the most part confirmed in longitudinal studies that showed that men who
actually entered the first set of occupations had higher n Power scores years be-
fore as undergraduates than those who did not enter these occupations. The re-
sults are summarized in Table 8.4, and they seem to hold up not only in the
United States, but also for a small sample of University of Oxford graduates in
England.
Winter argues that these results are consistent with theory because the occu-
pations chosen and entered by the men higher in n Power all provide more
scope for power and influence than the other occupations. More specifically,
teachers and the clergy are in a public position trying to influence an audience.
Psychologists and journalists have more inside information: they know what
other people do not know and are in a better position to influence people
through this Special knowledge. Sonnenfeld (1975) confirmed the tendency of
people high in n Power to seek positions of public influence in a study of stu-
dents who volunteered to work for a College radio Station. He found that the
correlation between n Power scores and the time devoted to the Station was .53
(TV = 30;/? < .01). In contrast, n Achievement correlated —.65 (p < .01)
with time spent at the radio Station, presumably because there was no tangible,
lasting product or achievement outcome to display as a result of time spent. In
similar fashion, Mueller (1975) discovered that women who perform publicly as
musicians score higher in n Power than music teachers.
Managers are also in a position to exercise power, although, as we shall see
later in this chapter, more is known about the specific kind of power motive
syndrome that is related to success in different types of managerial positions. In
Table 8.4.
MEAN n POWER SCORES OF COLLEGE MEN ASPIRING TO CERTAIN OCCUPATIONS AND
MEN IN THOSE OCCUPATIONS (after Winter, 1973)
Mean n Power Scores

Aspirants: Occupants: Alumni:
Wesleyan Harvard Oxford
Occupation T-Scores (N) Raw Scores (N) Raw Scores (N)
Teaching 53.7^ (48) 4.85a (27)
Teaching and clergy 10.13a (15)
Psychology and clergy 54.1a (31) 4.18a (17)
Business and journalism 52.4a (34) 6.60a (15) 13.86a ( 7 )
Law 49.0 (27) 3.59 (29)
Medicine 46.4 (14) 3.35 (20)
Other 47.0 (26) 3.06 (65) 5.44 (16)
a
Significantly higher than all others.
contrast, doctors and lawyers may have influence through their skills rather
than their powers of persuasion, although this obviously varies with the type of
lawyer. Courtroom lawyers arguing cases should have higher n Power than
lawyers involved in settling estates, but Winter's data do not permit such a
comparison.
In general, national sample surveys from 1957 and 1976 confirm these re-
sults (Veroff et al., 1980). Male Professionals (including teachers), as well as la-
borers and operatives score higher in n Power; farmers and omce-secretarial
workers score lower in n Power. The results for laborers and operatives suggest
that high n Power might lead people to occupations where they have impact on
things as well as people, although Winter has argued that the power motive is
an interpersonal motive. In fact, the scoring definitions include only interper-
sonal impact. Still, if the power motive develops out of the impact incentive, as
suggested in Chapter 4, it might be expected that it would lead to pleasure in
having impact over things as well as people. Unfortunately, the results for farm-
ers do not particularly fit in with this line of argument: farmers might also be
thought to have wide scope for having power over things, but they score low in
n Power. However, they may feel relatively helpless in the face of nature's capri-
cious whims, and they have little scope for interpersonal influence, which could
explain why those low in n Power end up in farming. The whole matter de-
serves more careful study.
Search for Prestige

One way individuals high in n Power can appear powerful in a socially accept-
able way is to collect Symbols of power, or what Winter (1973) called prestige
possessions. Both at Harvard and Wesleyan Universities he found that n Power
scores were significantly correlated with the number of such prestige possessions
owned by undergraduates, such as cars, wine glasses, College banners, a tape re-
corder, wall hangings, an electric typewriter, and so on. When asked to give
their preference in cars, those high in n Power more often chose a foreign car
and one that was more maneuverable than did subjects low in n Power. Similar
results were obtained for a group of fifty adult males around thirty-three years
of age when a different set of prestige items was used that included a color tele-
vision set, a rifle or pistol, and a convertible car. The n Power score among
these men was significantly correlated with the number of such items they
owned (McClelland, 1975).
Credit cards are another symbol of prestige. They represent the power to
purchase important Services, often from prestigious hoteis, restaurants, and
Stores. Winter (1973) reported a significant correlation between the number of
credit cards in their wallets and n Power scores among a group of executives in
a large manufacturing Company. A similar significant correlation was found
among men of the lower middle class and working class (McClelland, Davis,
Kaiin, & Wanner, 1972). McClelland (1975) has reported that the n Power
score is significantly correlated with the number of credit cards owned among
115 women, though not among the men in this study. It seems clear that the
power motive leads people to collect whatever Symbols of prestige are appropri-
ate for the position they occupy in life.
Acting So As to Be Recognized in Small Groups

Grades in school are an important way of gaining recognition for young people,
so it is not surprising to discover that a significant correlation has been found
(Costa & McClelland, 1971) between n Power scores in eighth grade and
twelfth-grade rank in class (r = —.27; N = 64; p < .05). The students with a
stronger n Power worked harder to get higher grades and therefore ended up
ranking higher when they graduated from high school. The result gains added
significance from the fact that n Power is not at all correlated with IQ, the
other major predictor of rank in class; from the fact that n Achievement score
in this sample, as in many others (see Chapter 6), is not significantly correlated
with grades earned in school; and from the fact that the relationship is predic-
tive. That is, one might expect people with better grades to be proud of them-
selves and perhaps develop a higher n Power, but here the higher n Power pre-
dicts better grades four years later. The finding is interesting and fits theory, but
the same energy should be spent on checking it that has gone into showing
there is not a relationship between n Achievement level and school success.
Winter (1973) reasoned that in small groups, individuals high in n Power
should work to become known and to build alliances. He asked the members of
a psychology class to list four friends. He then made up a master list of all
these names and submitted it to the members of the class, asking them to check
which of the people on the list they knew. A recognition score for each name on
the list was obtained by giving it a point for each time the person was recog-
nized by anybody in the class. Winter found that the higher a person's n Power
score, the lower the average recognition score of the friends he or she had listed.
In other words, people high in n Power (mostly male in this sample) choose as
friends individuals who are not particularly well known or in a position to com-
pete with them for prestige. Thus, they seem to Surround themselves with lesser
known people who can be led.
A more direct confirmation of this characteristic had been obtained by
Fodor and Farrow (1979), who arranged a Situation in the laboratory in which
business school students thought they were supervising three workers in another
room. They were told that they could say anything they liked through a two-
way communication System to foster the productivity of the workers in making
objects out of Tinker Toys. In return, they heard comments from the workers,
which in order to standardize the conditions were actually tape-recorded State-
ments. In one condition Worker C made comments aimed at ingratiating the Su-
pervisor, such as "You know, I really like your approach. You're going to be a
good Supervisor." In another condition his comments were neutral. Fodor and
Farrow found that subjects high in n Power, as compared with those low in
n Power, evaluated more favorably the ingratiating worker in terms of his ability
to work, his worth to the Company, and their willingness to rehire or promote
him. The same difference did not appear for other workers or for Worker C
when he made neutral comments. People high in n Power like to have people
around them who respect them and are loyal supporters.
Furthermore, people high in n Power tend to act in a group so as to call at-
tention to themselves. Participants in an achievement motivation training course
in India (see Chapter 14) were asked to list the names of those they had known
before the course and those they had gotten to know during the course. The
gain in the number of times a person was mentioned from before to after the
course—the gain in his recognition by others—was significantly correlated with
his n Power score. Apparently those high in n Power behaved in ways that
made them more visible to other members of the group (Winter, 1973).
McAdams, Healey, and Krause (1982) have clarified the nature of this be-
havior further in a study of the nature of friendship episodes engaged in by Col-
lege students. They found that men high in n Power, as compared with men low
in n Power, reported that they were more often in the Company of a group or
gang of four or more friends. This would give them more opportunity to stand
out in the crowd in some way than if they were in a dyadic relationship. Fur-
thermore, students of both sexes higher in n Power more often described their
role in the friendship episodes as involving "agentic" striving rather than listen-
ing and sharing: "The high-power individual tends to experience friendships in
an agentic manner, understanding them in terms of opportunities to take on
dominant, Controlling, organizational roles. . . . In the agentic mode, relation-
ships are apprehended in power terms. Seif and other are understood as sepa-
rate. . . . The friends take advantage of various opportunities that may arise for
self-display and self-expansion within the bounds of the relationship" (McAd-
ams, Healey, & Krause, 1982).
Schnackers and Kleinbeck (1975) demonstrated how exploitations of others
occur by having three subjects participate in a con game, one of whom was high
in n Power. The object of the game was to see which player could get to the
goal first by throwing dice and using "power cards" of different values that
could be multiplied by the score on the dice. Coalitions could also be formed be-
tween two players that, if maintained, would always defeat the third player. But
it was also possible to break a coalition at any time after it had been formed. As
expected, Schnackers and Kleinbeck found that as compared with the subjects
low in n Power, those high in n Power entered into more coalitions and also
broke up more coalitions in order to take advantage of their partners. They also
won more points. The results are summarized in Figure 8.4. The correlation be-
tween n Power scores and points scored in the third round was .45, p < .005.
Obviously, the subjects high in n Power exploited the others in the Situation in
every possible way in a game that encouraged this type of behavior. Notice that
people high in n Power, while not more aggressive in all situations, tend to be
so when the values governing a social Situation favor it.
In a study of small groups of strangers interacting for a while, D. F. Jones
(1969) found that those high in n Power talked more and were judged to have
"most influenced the other participants." However, they were not best liked nor
judged as having contributed most "to get the job done and come to a good
conclusion" (Jones, 1969). Furthermore, Watson (1974) has reported that men
_ | High n Power _
30 - 140
j J Low n Power
24 — — —i -- 120
18 - —i - 100 £
I
'S
0,
12 - 80
—
6 —i -__ 60
1
Goes It Makes an Accepts 1 2 3
Alone Offer to an Offer Round
a Third from a
Person Third
Person
Figure 8.4.
Behavior of Subjects High and Low in n Power in a Game. (Lefl) Frequency of three ways of
breaking a coalition. (Right) Points won in three consecutive rounds (after Heckhausen, 1980, after
Schnackers & Kleinbeck, 1975).
high in n Power (using its hope of Power aspect, as described in Table 8.1) eval-
uate other members of the group more often negatively than men low in n Pow-
er (Winter & Stewart, 1978). Thus, it is not surprising to learn that they are
also judged not to be particularly helpful in small group discussions (Kolb &
Boyatzis, 1970). In fact, people low in n Power have been found to be more
helpful in promoting group decision making based on the facts in the case
(Fodor & Smith, 1982). Those high in n Power are judged to have been more
influential by the others in the group, but they are apparently too assertive to be
good at bringing other people out.
Other characteristics in combination with high n Power are necessary to
make a good leader. For example, Constantian (1978) studied one group en-
gaged in studying its own processes in which the behavior of the participants
was monitored week by week over a whole Semester in terms of the three di-
mensions Bales has used to classify small group behavior (Bales & Cohen,
1979). Roughly speaking, Bales' upward-downward dimension reflects the
amount of dominance behavior a person shows in the group; his forward-back-

ward dimension, the amount of task-related behavior shown; and his positive-
negative dimension, the amount of friendly behavior a person displays. At the
end of the term, Constantian collected the number of nominations each person
in the group received for the position of being "best leader." She then at-
tempted to predict the variance in the best leader nominations from seven deter-
minants: scores on the three dimensions of behavior coded by Bales; the motive
scores for n Achievement, n Affiliation, and n Power; plus a measure of inhibi-
tion also obtained from the TAT (see the later discussion on the leadership mo-
tive syndrome in this chapter). She found that she could predict over 90 per-
cent of the Variation of the leadership nomination score from these variables,
but that only three variables contributed significantly by themselves to the re-
sult. These were the n Power score, the positivity of behavior score, and the
task orientedness score. In other words, subjects high in n Power, in positivity,
and in task orientation were most likely to be nominated as best leaders.
What is especially significant about this finding is that the general negativity
of subjects high in n Power we have mentioned is here canceled out by the re-
quirement that the person behave positively and in a task-oriented manner in
the group. In other words, high n Power leads to successful leadership if it is
combined with the behavior traits of positivity and task orientation. Once again
we are reminded that behavior must take into account not only motives, but
also values and skills—in this case, two skills (positivity and task orientation)
specifically rated by the Bales procedure.
However, when Constantian attempted to cross-validate these findings in an-
other self-analytic group, she obtained quite different results. The same variables
no longer predicted the best leader nominations. The explanation lies in the
different makeup of the group. Leadership depends not only on personal char-
acteristics, but also on the characteristics of the other members of the group—
the environmental determinants in Lewin's well-known equation of behavior
as a Joint product of the person and the environment. For example, if a group
contains a strong abrasive personality, it may not be possible for anyone—no
matter how favorable his or her personal characteristics—to assume effective
leadership.
Risk Taking
Outward Bound is a training program designed to increase courage and self-
reliance by exposing participants to a number of physical dangers they must face
and overcome. They have to scale walls, walk narrow planks some distance
above the ground, escape from being thrown in the water tied up, and survive
for three days alone in the woods with almost no resources. Fersch (1971) found
that when adolescents of high-school age, mostly from poverty backgrounds,
were sent to this program, a number found the training so rough that they
dropped out. But if they were high in n Power, they were significantly less likely
to drop out. In other words, those with a stronger power motive were more
willing to endure the physical risks and dangers. Furthermore, those whose
n Power score remained high after the program, as reflected in a retest at the
end of the summer, also showed signs of increased social competence in school a
year later.
McClelland and Watson (1973) studied risk taking among seventy-two Stu-
dent volunteers in situations in which the outcome depended either on effort or
on chance. In the first task, the subjects were asked to pick the level of diffi-
culty of a problem they wanted to work on after having had some experience in
trying to solve similar problems. As expected, the subjects high in n Achieve-
ment, as contrasted with those low in n Achievement, preferred to work on
tasks that were moderately difficult for them judged in terms of their prior per-
formance. However, the subjects high in n Power were not moved by the
achievement incentive to choose moderately difficult problems, nor were they
more likely to choose extremely difficult problems to solve.
The second task involved deciding what bets to place in roulette, in which
the odds of winning vary all the way from fifty-fifty (placing a chip on red or
black) up to one in thirty-five (betting on a Single number out of the thirty-six
available). After the game was fully explained to the subjects, they were told
they would be allowed to place fifteen bets and that they should mark on the
sheet how they wanted to distribute them. These choices were made in private,
without anyone knowing what they were. Then the subjects actually played the
game at the roulette table, choosing any ten of the bets they had previously de-
cided to make. In this condition the bets were made publicly, and wins and
losses could be observed by the others in the group. When the subjects were de-
ciding what bets to make privately, there was a tendency for those high in
n Power to choose more extreme bets, but it was not significant. When they ac-
tually played the game publicly, however, the correlation between n Power
scores and the riskiness of the bets placed was significant. Furthermore, as con-
trasted with those low in n Power, the individuals high in n Power used more
of the extreme bets (seventeen or thirty-five to one) they had chosen in the
private condition. Sixty-one percent of them used all, or all but one, of the
extreme bets, as contrasted with only 34 percent of those low in n Power
(p < .05).
Figure 8.5 shows the distribution of bets according to probability of success
for those high in n Power, high in n Achievement, or high in n Affiliation. The
biggest difference is for the riskiest bet, which only the subjects high in n Power
chose most often, in contrast to the other subjects, who tended to choose more
moderately risky bets. The results for the high n Achievement subjects are unex-
pected, since according to theory they should have chosen options with moder-
ate probability of success only when they had a chance of influencing the out-
come. However, this tendency may be so engrained in them that they continued
to choose moderate probability of success options even in a chance Situation.
The curve for the subjects high in n Affiliation only tends to be skewed toward
the safest bets, which suggests that they are not particularly attracted by this
kind of competitive game.
60
Bets by Subjects with High

50
n Achievement Only (N= 130) /^
/
40
Bets by Subjects with High
n Affiliation Only (/V=30)
30
£ 20-
10 Bets by Subjects with High

n Power Only (N= 150)
T
_L J_ _L
.60 .50 .40 .30 • .20 .10* .00
t
Vi:\ o r 1:1
t t t
2:1 o r 5 : l 8:1 o r 11:1 17:1 o r 3 5 : l
Figure 8.5.
Proportions of Roulette Bets Placed at Various Odds by All Subjects and Subjects High in n Power,
n Achievement, and n Affiliation Only (after McClelland & Watson, 1973).
Application of the Atkinson Risk-taking Model to n Power

At first, it would appear that subjects high in n Power are not behaving as they
should according to Atkinson's model (see Table 7.2), which holds that people
will be most attracted to options where the probability of success is moderate
(Ps = .30 to .50). Whereas Atkinson deals only with the tendency to approach
achievement alternatives, the model is stated in terms general enough to raise
the question of whether it applies to any kind of motivational choice.
McClelland and Teague (1975) explored the extent to which the model pre-
dicted choices of subjects high and low in n Power when the alternatives in-
volved exercising power rather than performing better. They asked students to
imagine they were a member of a board that had to make a decision on some
public issues. Each issue was described in some detail (in one case it had to do
with whether to build a coal-fueled electric power plant despite its polluting ef-
fect), and four positions the board could take were outlined (for example, build
the plant with exacting control specifications, delay its building, build it irnmedi-
ately to promote the economy, or do not ever build it). To provide some control
over the probability of success variable, in each case it was indicated roughly
how populär each alternative was with the board members. Subjects were then
asked to indicate which of the four options they would argue and vote for. They
also indicated on a scale of 1 to 10 how likely it was that each position would
be adopted by the board (the probability of success variable) and how good they
would feel if each of the four positions was adopted (the incentive value of each
Option for them).
In this instance, the Atkinson model, which states that approach is a func-
tion of probability of success times incentive value, did not predict the choice of
various alternatives at all well. It failed particularly in predicting the choices of
those high in n Power, since in their case increased probability of success did
not make it more likely that they would choose an alternative. In fact, the re-
verse was true; the more unlikely they judged an alternative was to be adopted,
the more they tended to choose it. They were attracted to extreme risks, just as
they were in playing roulette. It is possible that they were choosing unpopulär
positions in order to stand out more.
This experiment can be criticized, however, on the grounds that the proba-
bility of success variable does not represent the probability of the subjects' suc-
cess in a Situation in which the outcome depends on a choice they make. So
McClelland and Teague (1975) also conducted an arm-wrestling experiment in
which all participants were asked whether they preferred to wrestle with some-
one stronger or weaker than they or about equal in strength. After choosing, the
subjects were asked to estimate on a scale of 0 to 10 what their chances of win-
ning were (the probability of success variable), as well as how good they would
feel if they won (the incentive value of winning). Unexpectedly, there was no
correlation between expectancy of winning (or Ps) and the rated incentive value
of winning either among the subjects high or low in n Power. This is surprising,
because Atkinson assumes in his model that there is a perfect negative correla-
tion between these two variables, that is, that Is = 1 — Ps. It might be ex-
pected that the less the probability of winning or the stronger the Opponent, the
greater the satisfaction would be from beating him or her, but this does not turn
out to be the case here.
Furthermore, as might be expected from this result, the probability of beat-
ing the Opponent seemed to have little effect among those high in n Power in
determining their choice of how strong an Opponent to fight. Their choices were
almost perfectly predicted by the incentive value they assigned to winning over
the Opponent they chose. Including their estimate of their chances of beating the
Opponent, as in the Atkinson model (Ps X / 5 ), actually predicted their choices
less well.
One possible interpretation of the roulette experiment and these experiments
is that the meaning of the probability of success may not be the same for sub-
jects high in n Power as it is for subjects high in n Achievement: it may not
refer to the probability of success or winning through people's own efforts or
Performance, but to the probability of being considered important in some way.
Thus, it might be possible that Atkinson's model would apply to power choices
if it were assumed that in such cases 1 — Ps = Ps, since the more difficult the
task undertaken, the more likely people would be to call attention to themselves
for attempting it. Then we might argue that their choices could best be pre-
dicted by 1 — (Ps X / 5 ), although Is cannot be defined, as it is in the achieve-
ment area, by 1 — Ps. The whole matter needs more careful study, but at the
very least we have to be very cautious about applying Atkinson's risk-taking
model to choices involving motives other than the need for achievement.
The meaning of the attitude of people high in n Power toward risk taking
may be made clearer by an example from everyday life. Politicians, if they ac-
tively seek office, are generally high in n Power (see Winter & Stewart, 1978).
They thus are often good at calling attention to themselves, at getting their
names in newspapers or on television, and at creating media events that will
lead to name recognition. They may identify with a particular issue or take an
extreme stand (a low probability of success) in order to get better recognized.
Suppose that a number of them in Congress decide to do something about
the problem of alcoholism, which is costing the country billions of dollars annu-
ally in terms of highway accidents, treatment facilities, and lost worker produc-
tivity. They decide to do something really important and visible—to create a
government institute to combat alcoholism in every possible way, for which they
succeed in getting $500 million appropriated annually. The probability of success
that interests them is whether they can be recognized as doing something big
and visible about the alcohol problem. They have a lesser interest in whether the
institute so created will actually succeed in decreasing alcoholism. That would
be of more interest to individuals high in n Achievement. When Congress ap-
propriates the money, those high in n Power believe they have done something
big about a big problem. Their choice is determined more by the incentive value
of having created a big institute than by its probability of success in achieving
the goal of decreasing alcoholism. In fact, if later evaluation shows that the in-
stitute is not lessening alcoholism, they can only lose prestige, whereas if evalua-
tion shows the institute is being helpful, it only confirms what they said would
be true all along. So they tend not to be as interested in Performance feedback
as individuals high in n Achievement are.
• CATHARSIS
It has long been supposed that expressing a motive or an emotion reduces its
strength or intensity. The ancient Greeks called this catharsis and used the term
to refer to the emotional release that occurs from watching terrifying and ag-
gressive scenes in a play. The idea gained further support from the observation
that for homeostatic physiological drives like hunger, consummatory behavior
like eating reduces the desire to eat. And Freud conceived of instincts or drives
as pressures to act in certain ways that could be reduced through achieving their
primary aim—or at least they had to be partially satisfied by the achievement of
secondary or substitute aims, a process that explains Symptom formation or
neurosis.
The idea of catharsis has been very appealing to psychologists, but they
have had difficulty pinning down exactly in what sense it occurs. At the most
general level, it can be taken to mean that any response, once it occurs, is less
likely to occur again right away. This built-in refractory phase to repeating a re-
sponse occurs (see, for example, Telford, 1931) and has adaptive significance,
since it prevents the organism from repeatedly making the same wrong response.
The principle is considered so basic by Atkinson and Birch (1978) in their
model of the dynamics of action that they have stated as a basic assumption
that "the consummatory force (C), which reduces the strength of an action ten-
dency (T), comes from expression of that tendency in an activity, from the oc-
currence of the activity itself."
Most students of catharsis, however, have used the term catharsis in the
much narrower sense of release or reduction in a motive or emotion, particu-
larly aggression, through some kind of action or action substitute such as fanta-
sy. In particular, people have wanted to know whether watching violence on
television increases or decreases aggression. If catharsis occurs from watching
people chase and kill each other on television, should not children who watch
many such television programs be less aggressive afterward than they were be-
fore? Should they not be less aggressive than other children equally aggressive at
the outset, who have watched other programs? Or more generally, in terms of
the concepts used in this chapter, does finding an outlet for the power motive (a
more generalized form of the aggressive motive) decrease power motive arousal
at the time or perhaps, if it is continued over time, decrease dispositional
n Power?
A great deal of research has been carried out in an attempt to answer such
questions, but unfortunately, much of it does not bear directly on the question
of how expressing a motive affects motive strength in itself. Most studies have
found, for example, that children who watch violent television shows often are
more aggressive in behavior afterward than other children (Feshbach, 1970; Ru-
binstein & Sprafkin, 1982). At first this would seem to mean that catharsis does
not occur: Experiencing violence through watching a drama does not reduce ag-
gressivity, as the Greeks thought. It increases it. But a closer analysis reveals
that such studies do not yield any easily interpretable information about what
has happened to the aggressive motive as a result of watching violent television
shows. This is because motives—as we have repeatedly emphasized—whether
dispositional or aroused, are only one of the determinants of the tendency to re-
spond in a certain way (for example, aggressively). Other determinants include
knowledge or skills and values.
In a typical study (Feshbach, 1970), one group of children was asked to
watch violent television shows for some weeks while another group watched
other kinds of shows. When their aggressiveness was measured afterward, it was
found that those who had watched the violent shows behaved more often in an
aggressive manner. This may have had nothing to do with what happened to
their aggressive or power motives, since they had obviously learned much more
than the other children on how to be aggressive: their aggressive knowledge and
skill had increased. Furthermore, the fact that some authority figures had told
them to watch violent television shows may easily have suggested to the children
that such behavior was all right; the value they placed on behaving violently
may have been subtly altered. So the increase in aggressiveness may have had
nothing to do with changes in the aggressive motive, which could have stayed
the same or even decreased while other determinants of the aggressive response
increased.
Any number of studies have shown that whether or not aggression occurs
and whether one type of aggressive response reduces the likelihood of another
type depends on the subjects' cognitive understanding of who is frustrating them
and why and whether being aggressive in one way or another helps them to
achieve their goals (see, for example, Zumkley, 1978). There are so many non-
motivational determinants of whether a person is aggressive or not that in most
studies it is impossible to determine how the motive has been affected just from
observing changes in aggressive behavior.
A few studies have focused more narrowly on whether catharsis occurs for
motives, as they are defined and measured in this book. The classic investigation
was carried out by Feshbach (1955). He arranged to insult a number of stu-
dents, that is, to arouse their power motivation. Some of them then were asked
to write imaginative stories to pictures, whereas others performed different tasks
supposedly measuring their ability. All students then filled out a sentence com-
pletion test and answered questions about their attitude toward the research and
the experimenter who had insulted them. As compared with a control group of
students who had not been insulted, those who had been insulted included more
aggressive themes in the sentence completions and reported themselves to be
more angry at the experimenter, so the power motive clearly had been aroused.
But the insulted students who had been allowed to fantasize about their anger
by writing stories included a significantly lower level of aggressive themes in
their sentence completions than the insulted students who had worked at other
tasks. This suggested that engaging in aggressive fantasies had had a cathartic
effect for subjects whose power motivation had been aroused. However, two crit-
icisms of this interpretation can be made. Perhaps those who wrote aggressive
stories feit guilty for doing so, so that they inhibited further thoughts of aggres-
sion in their sentence completions. And perhaps writing stories is a more relax-
ing or interesting thing to do than taking tests, so it subtly improved the sub-
jects' attitude toward the research and made them less angry. That is, the
reduction in their aggressiveness might be due not to having expressed it, but to
a change in the instigation to aggression.
Feshbach (1961) conducted another experiment in which students who had
been insulted either watched a fight film or a neutral film, after which their as-
sociations were obtained and scored for aggressiveness. He found that those who
had seen the fight film had lower power motivation afterward (gave fewer ag-
gressive associations) than those who had seen the neutral film. It is difficult to
see how differences in what the subjects were doing could explain the result, as
in the previous experiment, so there was some catharsis for those whose power
motivation had been aroused.
Feshbach, however, also found that for subjects who had not been insulted
(whose power motivation had not been aroused), the fight film actually increased
their power motivation, as reflected in more aggressive associations, as compared

with the power motivation aroused after the neutral film. In other words,
whereas the fight film reduced power motivation for those in whom it had been
aroused, it increased power motivation for those in whom it had not been
aroused.
These results have been confirmed for dispositional, as distinguished from
aroused, power motivation (McClelland & Maddocks, 1983). Subjects whose
n Power had been assessed previously in the usual way saw either a power-
arousing film about the successes of the Nazis in Germany and their persecution
of the Jews, or a neutral film about gardening. Their n Power scores then were
obtained again from stories written after seeing the films. Just as in Feshbach's
power motive arousal study, those who were dispositionally high in n Power
showed a lower level of n Power after seeing the Nazi film than the neutral
film, indicating a catharsis effect for them. Those initially low in n Power, how-
ever, showed a much higher level of n Power after the Nazi than the neutral
film: their power motive had been aroused. The interaction effect was highly sig-
nificant. So catharsis does not occur when the motive or emotion is not present
or aroused. Quite the opposite occurs. If it is not present, cues that normally
arouse it do arouse it. Thus, other things being controlled for, violent television
shows for children might arouse power motivation in those who had little to
Start with, even though it might decrease it for those high in n Power.
Attempts have also been made to determine the effects on power motivation
of participating in overtly violent activities. Stone (1950, summarized in McClel-
land, 1951) scored aggression in stories written before and after the football sea-
son by football players and other matched male students. He found that whereas
both groups had the same average level of concern about aggression before the
season, the football players scored significantly lower in thoughts relating to ag-
gression after the season. Apparently, daily participation in sanctioned physical
aggression for several months had lowered the level of concern for at least this
aspect of power motivation. Unfortunately, other aspects of the power motive
System were not scored, so it cannot be determined whether the practice of this
type of activity merely shifted power motivation into other Channels. As noted
earlier, Fersch (1971) also found a decrease in people's overall n Power scores
after a summer of participating in physically demanding, risk-taking activities in
Outward Bound. Unfortunately, however, he did not have a control group, so it
cannot be determined whether any group of high-school students would decrease
in n Power over the summer vacation.
Thus, there is some evidence that outlets for the power motive reduce its
strength for those in whom the motive is strong or aroused. The conclusion is
not nearly as firm as it should be, however, because most research has focused
on testing for cathartic effects on aggressiveness, which has important nonmo-
tivational determinants. Also, very little work has been done on cathartic effects
for other motives. For example, does succeeding at a task lower n Achievement
scores? Atkinson's model in Table 7.2 assumes that to the extent that success
changes the probability of success (and, thereby, achievement incentives), it can
alter the tendency to perform an act. But what about the effect of success on the
motive itself? McClelland, Atkinson, Clark, and Lowell (1953) found that suc-
cess at tasks was followed by an n Achievement level that was somewhat, but
not markedly, lower than that obtained after failure at the same tasks. This de-
serves more careful study to distinguish the effects of catharsis on motives from
its effects on behavior.
ROLE OF THE POWER MOTIVE IN DRINKING
Drinking Conceived as a Means of Reducing Anxiety

From 1940 to 1960, most psychiatrists and psychologists believed that the mo-
tive behind drinking alcohol was to reduce tension and anxiety. There were sev-
eral reasons for this belief. Alcohol is a central nervous System depressant.
Therefore, it was reasoned, people would drink it only to depress something that
was upsetting them, such as anxiety. Also at this time, the behaviorist model of
drive as tension and reward as tension reduction (see Chapter 3) was populär,
and alcohol could be conceived as rewarding because it reduced tension. Fur-
thermore, people often say that they drink to "forget their troubles," although
this is more common in the United States than in other countries. Finally, a
number of clinical studies suggested that alcoholics were anxious because of
frustrated dependency needs and drank to feel more comfortable and relaxed
(see McClelland, Davis, Kaiin, & Wanner, 1972).
Kaiin, Kahn, and McClelland (1965) argued that the predominant view of
the motive for drinking was biased by the notion prevailing among psychologists
at the time that reward was tension reducing, as well as by the view in the gen-
eral public (which had culminated during Prohibition) that alcohol was bad for
people so that the only reason it would be sought was to get rid of something
worse. Whereas it is true that excessive consumption of alcohol can lead to
deadening thoughts and feelings, most people all over the world drink alcohol in
small amounts. Why should most people take only about two drinks at a time?
It did not seem reasonable to suppose that such a small amount of alcohol was
significantly reducing anxiety.
Furthermore, Kaiin et al. (1965) argued that everyone was inferring what
was going on in the minds of people drinking alcohol rather than directly study-
ing what was going on in their minds. Finally, they feit that the settings used to
study the effects of alcohol were biased against finding any positive effects. Usu-
ally the alcohol was consumed in a formal laboratory setting while an experi-
menter, an authority figure in a white coat, administered alcohol in predeter-
mined doses. To remove the effects of Suggestion, the taste of alcohol was often
disguised so that subjects would not know what they were drinking. In what
was considered an ideally controlled experiment, both the experimenter and the
subjects were "blind" as to what the subject was getting, because either alcohol
or saline solution was infused intravenously into the subjects. Under these condi-
tions it was scarcely surprising that very few subjects ever reported that they feit
good or lively. Such procedures do not "remove Suggestion," as was often
claimed, but in fact introduce very powerful suggestions derived from the experi-
menter as authority figure, uncertainty as to what one is drinking, being forced
to consume predetermined amounts of alcohol, and the lack of the normal social
Supports that go with drinking (see McClelland et al., 1972).
Therefore, Kaiin et al. (1965) decided that since alcohol has to be consumed
in some kind of setting, it would be best to use the settings under which alcohol
is normally consumed. Then at least the experimenter would be in a better posi-
tion to discover the motives for normal social drinking. They ran cocktail par-
ties of the type men usually attended in all-male Colleges at which individuals
were allowed to drink as they normally would choose to drink rather than as
they were told to drink by an experimenter. From time to time during the party
the proceedings were interrupted and the participants wrote TAT stories to find
out what was on their minds at the time. Other control parties were also run
during which stories were collected when good food and nonalcoholic beverages
were served. On other occasions alcoholic or nonalcoholic beverages were served
as they normally would be during an evening discussion in someone's apart-
ment. To avoid tapping subjects' preconceptions about the effects alcohol is sup-
posed to have, the participants in these experiments were simply told that the
researchers were interested in the effects of a party or evening discussion on
imaginative thought processes (see McClelland et al., 1972).
Effect of Social Drinking on Thoughts About Power and Inhibition

The early results of these studies showed that as drinking progressed, sexual and
aggressive thoughts increased significantly among men over what they were at a
comparable time period in the control parties or over what the thoughts had
been before the drinking started. Then it was realized that these types of imag-
ery, along with some other types that increased in frequency with drinking, ac-
tually represented subtypes of power imagery as it was defined by Winter
(1967). A reexamination of the data showed that n Power scores increased as a
function of drinking alcohol. It looked as if men were drinking to feel more
powerful rather than to reduce their anxiety, at least to judge from their
thoughts at the time.
Davis (in McClelland et al., 1972) checked this finding by creating an expe-
rience in which men feit they were being required to act more powerfully but
were unsure whether they could. He accomplished this by having them lead a
blindfolded man through the crowded streets of a city. Davis found that after-
ward, when the pairs were relaxing in a bar, those who had been under power
stress from acting as guides in this somewhat embarrassing way drank more
heavily than those who had been blindfolded. In other words, the men who feit
somewhat inadequate about the assertiveness and responsibility required of them
drank more to feel more powerful.
It was later discovered that women responded differently to alcohol: Their
n Power score did not rise while drinking at parties with both sexes present (see
Wilsnack, 1974). Instead, drinking made them feel warmer and more feminine.
In short, alcohol leads men to feel more powerful (that is, stronger) and women
to feel more friendly. At present it is not known whether this difference is due
to physiological patterns relating to the interaction of alcohol with different hor-
monal patterns in men and women, or whether it is due to different social role
expectations for the two sexes. There are also ethnic differences in the effects of
alcohol (Ewing, Rouse, & Pellizari, 1974).
In the same study, after men had at least four or five drinks, signs of inhibi-
tion also decreased significantly, giving some support for the common notion
that alcohol releases inhibitions (McClelland et al., 1972). Two such signs were
measured. One was called Time Concern and was scored whenever a story men-
tioned time in any way, such as how long it took to do something or when a
person did something. There was a significant decrease in the number of individ-
uals mentioning time from the beginning to the middle to the end of a cocktail
party. The other measure was called Activity Inhibition; it was simply a fre-
quency count of the number of times the word not appeared in the protocols.
The percentage of protocols showing at least two nots declined significantly
from the beginning to the end of the party (McClelland et al., 1972).
Thus, drinking had two effects on men: it increased power concerns and de-
creased inhibitory thoughts. Furthermore, the quality of the power thoughts var-
ied significantly, depending on whether the subjects showed high or low signs of
inhibition. On the one hand, if a man was low in Activity Inhibition, his power
thoughts much more often focused on personal dominance—on winning at
somebody eise's expense, as in a zero-sum game ("If I win, you lose"). On the
other hand, if a man scored high in Activity Inhibition, power imagery was
much more often stated in terms of doing good for others, for humanity, or for
some good cause. Furthermore, the power outcome was described in more un-
certain terms or with irony.
The contrast is illustrated by the following two plot summaries of stories
written to a picture of a boxer:
1. "He is fighting the champ—a chance to win a big purse, retire to a beach in
Tahiti."
2. "He is fighting the champ—a chance to win a big purse. His kid is in the
hospital and needs an expensive Operation."
The first type of power imagery is called personalized power (p Power), because
the goal is exclusively personal. The second type of imagery is called socialized
power (s Power), because the power drive seems to be socialized in the Service
of others.
Figure 8.6 shows the effects of different amounts of alcohol consumption on
the frequency of p Power and s Power imagery in stories written by participants
in the cocktail parties. The curves show the "two-drink effect" the investigators
were originally looking for. After about two drinks, the s Power curve reaches
its highest point and then declines with further drinking. In contrast, the p Pow-
er curve increases steadily with more drinking. The result is not surprising; it is
foreshadowed by the fact that the measures of inhibition decline with heavier
drinking, and p Power imagery is associated with lower inhibition, just as
s Power imagery is associated with higher inhibition. Thus, small amounts of

drinking in men leads to socialized thoughts of being more powerful—of being
big, strong, and important on the Job or in the family—but further drinking
leads to thoughts of personal dominance—of beating other people in competition
or perhaps even beating them up.
Relationship of the Power Motive and Inhibition

to Problem Drinking and Other Activities
If it is feelings of power that men get out of drinking, perhaps men who are
high in n Power to Start with drink more. More specifically, those who are high
in p Power, or high in n Power and low in Activity Inhibition, should drink the
most. This, in fact, turned out to be the case. In two separate samples of men, a
p — s Power index correlated significantly with a quantity/frequency index of
60
A
I \
50 - -
/ \ p Power^-»
\ ^ ^
1
1//
5
£ o* \
40 - - \
c u
\
O rt
\
co o
1/ \
8 o 30 -
O —
/ f \
o g
//
//
s Power \
\
\
\
i
10 -
J— i 1 i
2 4 6 10
Light Moderate Heavy
Alcohol Consumption (Ounces ot 86 Proof Alcohol)
Figure 8.6.
Proportions of Subjects Writing Stories Containing Personal or Socialized Power Imagery with
Increasing Alcohol Consumption (after McClelland, Davis, Kaiin, & Wanner, 1972).
reported alcohol consumption (r = .32; TV = 50; p < .05 and r = .24;

N = 108; p < .05).
Figure 8.7 shows the same results in a slightly different format. Subjects in
the two samples of men were classified according to whether they were (1) high
in n Power and low or high in Activity Inhibition or (2) low in n Power and
low or high in Activity Inhibition. Clearly, in both samples the men who were
high in n Power and low in Activity Inhibition were much more likely to be
classified as heavy drinkers by self-report than subjects in the other three catego-
ries. The overall level of drinking reported in the second sample is lower than in
the first, but the relative percentages of heavy drinkers in the various categories
is similar. It is particularly worth noting that a high n Power results in heavy
drinking only if Activity Inhibition is low. If Activity Inhibition is high, the
Proportion of these individuals reporting heavy drinking is very low.
These studies also demonstrated that a classification by power and inhibition
affected activities other than the amount of drinking. Figure 8.8 shows the re-
sults of a factor analysis of a number of activities the 108 men in one of these
samples said they engaged in. As Chapter 2 explained, factor analysis involves
correlating the extent to which people report they engage in pairs of activities
(as examples from this study, consuming much wine and also lots of beer, or
smoking many cigarettes and spending a great deal of time working). From all
70
o j First Sample of Adult Males
60
I Second Sample of Adult Males
II 50
>i 40
X8 H
1
2 2 30
| | 20
o 10
Low
Inhibition
mHigh
Inhibition
Low
Inhibition
n
High
Inhibition
High n Power Low n Power
Figure 8.7.
Percentage of Heavy Drinkers Among Men High and Low in n Power and High and Low in
Activity Inhibition (after McClelland, Davis, Kaiin, & Wanner, 1972).
the intercorrelations of pairs of activities, factors that best account for most of
the relationships obtained are extracted by a statistical procedure.
In this case, two independent factors—one characterized by high n Power
and the other by Activity Inhibition—did a good job of accounting for the inter-
correlations of all the variables. For example, beer and wine consumed are not
highly related: people who drink a large amount of wine do not necessarily also
drink a large amount of beer. Wine-drinking scores load fairly high on the As-
sertiveness (power) factor and on the Restraint factor, whereas beer drinking
scores low on assertiveness and fairly high on lack of Restraint. This way of
sorting activities began to make sense of some of the earlier findings on the cor-
relates of n Power scores. For example, office holding, as well as wine drinking,
loads high on the Assertiveness, or n Power, dimension and fairly high on the
Restraint, or Activity Inhibition, dimension, whereas the amount of liquor con-
Factor I: Assertiveness
'ou • Sports Participation

Speeding*
— • Office Holding
- ß0 • Time with Women
Liquor Consumed — "Sporty" Maga^ines

# Violent TV
40
• Wine Consumed
n
Travel • Power
Reckless Driving
.20
Time Hanging Beer Consumed . . . . . . . . . Time Working
Around • Activity Inhibition #
1 • 1 1 1 1 1 I I 1 1 #l 1 1 1 Factor II: Restraint
-.80 -.60 -.40 -.20 .20 .40 .60
-.20
Cigarettes Smoked
Time Watching TV
— .40
•Age
- -.60
Figure 8.8.
Plot of Drinking and Other Activities Variables on Assertiveness and Restraint Factor Dimensions
(after McClelland, Davis, Kaiin, & Wanner, 1972).
sumed loads high on the Assertiveness factor and low on the Restraint factor.
Similarly, sports participation goes with high n Power and somewhat with Re-
straint but not with lack of Restraint. Driving too fast and getting caught for
speeding goes with Assertiveness and lack of Restraint, as would be expected. It
became clear from this analysis that inhibition is a very important moderator
variable for n Power: it determines whether the power motive expresses itself in
more socialized, controlled ways or in more reckless attempts to show off as
being powerful.
HOW MATURITY MODULATES THE

EXPRESSION OF THE POWER MOTIVE
Deriving a Measure of Social-Emotional Maturity

Even greater clarity in the understanding of the outlets for the power motive
was introduced through the discovery that they varied depending on the
social-emotional maturity of the person. As Chapter 2 described, Erikson (1963)
elaborated Freud's theory of psychosexual stages into a full-blown scheme of
stages of psychosocial, or social-emotional, development.
Stewart (1973) decided that if progress were to be made, some method of
measuring the strength of various stage orientations had to be developed. She
decided to use the approach that had been successful in producing measures of
human motives like n Achievement and n Power. She selected groups of College
freshmen, each representing a "pure" stage orientation, in order to compare and
contrast the content of the imaginative stories they wrote. That is, she picked
out the six freshmen who should be most clearly oriented toward the oral stage,
as defined by the criteria of psychoanalysis. These were the students who ate,
drank, and smoked much more than other freshmen and who did not score high
on any of the behavioral criteria for the other psychosexual stages. She then de-
veloped a formal coding System for the content of their stories that differenti-
ated them from stories written by freshmen not clearly in one stage or another.
In similar fashion, she picked out the six freshmen most clearly in the anal stage
as indicated by the number of self-regulating rituals they performed before going
to bed at night, such as taking a shower, emptying their pockets, brushing their
teeth, urinating, setting the alarm, opening a window, and so forth. The phallic
stage was represented by young men who said they had dated many different
girls in high school and who cited "reputation enhancement" and "sex" as the
most important motives for dating. Freshmen males were considered to repre-
sent the genital stage if they had a steady girlfriend to whom they were faithful
and if they reported that on dates they were most likely to study and make love.
In this way they satisfied Freud's judgment that "love and work" represent
maturity.
Stewart found that the stories of these "pure types" differed from other sto-
ries in ways that could readily be predicted by psychoanalytic theory, and this
result encouraged her to think that perhaps she had found a way of measuring
at least the earliest of Erikson's stages (see Table 2.4). Some of her results are
summarized in Table 8.5. The stages are organized in this table according to
two dimensions: whether the source of power is outside the seif or in the seif,
and whether the target of power is the seif or another.
This classification System generates the four psychosexual stages. In the
first, or oral, stage the source of power is another—usually the mother—and the
target of power is the seif; this defines the dominant modality, which is intake,
as when a mother nurses a baby. In the second stage the source of power shifts
to the seif, but the target remains the seif, as when children attempt to gain
control over themselves in what Erikson calls the stage of autonomy. In the
third, or phallic, stage the source of power remains the seif, but it is now di-
rected toward others, as when a child attempts to influence others. The modality
is assertion. Finally, in the fourth, or genital, stage the source of power shifts
from the seif to some higher outside, institutional authority such as the family,
the church, or the State, and the individual acts under the influence of higher
authority to influence others. As an example of how this highest stage is at-
tained, Erikson (1963) pointed out that after individuals have gone through the
"egocentric" Stages II and III of adolescence, they may get married. If they do,
they sooner or later discover they are involved in a bond or an "institution"
that is bigger than themselves as individuals. The marriage partners do things
on behalf of the marriage that perhaps they would not do on their own. The
point is even clearer when they become parents and find themselves doing some-
thing for the child as parents that perhaps they would not do on their own.
Table 8.5 also briefly lists the themes in stories that uniquely characterized
individuals classified as belonging to each of these stages. In the authority
area, for example, those in Stage I more often pictured authority as benevolent
and providing things for people, whereas those in Stage II saw authority as criti-
cal and requiring individuals to control themselves. In Stage III, individuals de-
scribed rebellion against authority, which should be characteristic of the asser-
tive stage. In Stage IV, authority was described as no longer personal—the
direct attempt of one person to influence another—but was characteristically ei-
ther vague, impersonal, or institutional.
The other thinking characteristics found in the stories of individuals at each
stage fitted fairly well with expectations about what people at those stages
should be feeling and thinking, according to psychoanalytic theory. Thus, the
subjects in the oral stage were pictured as waiting, just as Freud had to wait in
his oral dream described in Chapter 2. Those in the anal stage described people
Clearing disorder, which corresponds to compulsiveness in the Erikson scheme
(see Table 2.4 in Chapter 2). Those at the phallic stage described people taking
initiative, but also showing guilt from failure and the desire to escape, which
can lead to inhibition and hysterical Symptoms, according to Table 2.4.
To fill out the picture of each stage, Table 8.5 also lists values that suppos-
edly go with each level of development, as they have been defined by McClel-
land, Constantian, Pilon, and Stone (1982). Thus, parents would stress the im-
portance of respect for authority in Stage I, the importance of self-reliance in
Stage II, the importance of developing skills in Stage III, and the importance of
Table 8.5.
STAGES OF PSYCHOSOCIAL MATURITY (after McClelland, 1975) I
Target of Power Source (if Power I
Other Seif I
I. Psychosexual Stage: Oral II. Psychosexual Stage: Anal 5'
Modality: Intake Autonomy

Formula: "It strengthens me" "I control myself'
Thinking Authority: Benevolent authority Critical authority
Characteristics3 Relations to things or people: Gets what he or she wants Does not get what he or she wants
Seif Feelings: Loss, despair Incompetence
Actions: Is passive Clears disorder
Values Respect for authority, Self-reliance, willpower,
decency, obedience independence, courage
To be inspired, healed To know
IV. Psychosexual Stage: Genital III. Psychosexual Stage: Phallic

Modality: Generativity or mutuality Assertion
Formula: "It moves me to serve" "I have impact"
Thinking Authority: Impersonal authority Rebellion against authority
characteristicsa Relations to things or people: Differentiates among people Escapes
Other Feelings: Mixtures of joy and sorrow Hostility, anger
Actions: Schedules work Action leads to failure
Values Understanding of others, Doing well in school
tolerance, serving and on the job, developing
common good skills, having influence
To show compassion To have power
a
These characteristics briefly describe the coding System developed by Stewart (1973) to score for various stages of social-emotional matunty. The values are those presumed to be
consciously important for people at that stage.
understanding and serving others in Stage IV. From the child's point of view,
children would be learning to value inspiration and healing in Stage I, knowl-
edge in Stage II, assertiveness in Stage III, and compassion in Stage IV. McClel-
land (1975) has attempted to show that powerful figures such as gurus or great
spiritual leaders like Jesus have typically been described as having expressed
power in all of these valued ways. They healed, showed unusual knowledge, in-
fluenced others effectively, and were particularly compassionate.
Alternative Manifestations of the Power Motive as a Function of Maturity

On theoretical grounds, the power motive would be expected to express itself in
different ways, depending on a person's level of maturity. If a person is primar-
ily oriented toward Stage I, the power motive should express itself in the intake
modality, and if toward Stage III, in an assertive modality.
However, there is no reason to think a person oriented toward Stage III
would show behaviors characteristic of other stages. Thus, to sum or average
the scores for all possible outlets for the need for power across all stages for a
person would be meaningless. If this were done, each person would get a high
score for the outlets characteristic of his or her dominant stage and low scores
for behaviors characteristic of other stages, and everyone should end up with
about the same total or average score. Thus, not much correlation would be ex-
pected between n Power and the sum total of all possible types of outlets for the
power motive characteristic of all stages. Rather, a significant correlation be-
tween the n Power score and a higher score on any one of the behaviors charac-
teristic of the different stages would be expected.
To test this theoretical expectation, McClelland (1975) chose activities that
were theoretically and empirically characteristic of each stage as follows:
1. Stage I (the intake modality): power-oriented reading.

2. Stage II (the autonomy modality): indications of self-control or having many
aggressive impulses but not expressing them.
3. Stage III (the assertion modality): frequently expressed anger to people.
4. Stage IV (the mutuality modality): frequent memberships in voluntary
organizations.
In a sample of eighty-five adult males, none of these possible outlets for the
power motive correlated significantly with the n Power score. This is as it
should be according to theory, because individuals in Stage I should not join
more organizations, and those in Stage IV should not engage in more power-
oriented reading. So in a miscellaneous group of men representing all four
stages, one would not expect any one of these outlets to be strongly correlated
with the n Power score. Also, their sum should not correlate highly with
n Power, because a high score for activities characteristic of the primary-stage
orientation would be canceled out by low scores for outlets of other stages. In
fact, in this sample of adult males, the correlation of n Power with the average
expression of these four alternatives was only .19, which is not significant.
However, the correlation of n Power with the maximum expression of any

of the four alternatives for each man was .32, p < .05. That is, scores on each
alternative were standardized to make them comparable, and a man was as-
signed the highest standardized score he got on any of the four activities. Men
high in n Power generally had a higher standardized score on some one of these
outlets than men low in n Power. This confirms the fact that high n Power
leads to greater participation in some power-related activity, although which ac-
tivity a man expresses depends on the presence of a moderator variable, namely,
the level of his maturity.
Clinicians have often observed that motives express themselves in alternative
ways. Sometimes if a Symptom is cured, another one takes its place; this is be-
cause the motivational problem remains the same, and if one outlet is blocked it
seeks another outlet. In Perry's dream from In Cold Blood, discussed in Chapter
2, he first expressed his power motive in terms of the assertive modality—
reaching out to grab something valuable. When this outlet was blocked by the
snake's swallowing him, he retreated to thoughts characteristic of Stage II,
which centered on the problem of self-control. When he was punished for lack
of self-control, he retreated to the oral, intake stage and dreamed of an endless
supply of delicious food and drink.
The statistical evidence simply confirms what clinicians have long been say-
ing: motives seek alternative outlets depending on the stage of maturity and on
whether a particular outlet is blocked or not. In short, motives show alternative
manifestations. Traits, skills, or habits, however, tend to be quite consistent. A
person tends to be consistently shy or late or good at playing the piano. So if
someone is cured of shyness, for example, by behavior therapy (see Chapter 12),
there is no evidence that he or she will develop other Symptoms. Motives, not
traits, seek alternative manifestations.
The same results were not obtained for women, perhaps because the outlets
most characteristic of each stage were not as appropriate for women as for men
(McClelland, 1975). The reason is that values, as well as stages of maturity, af-
fect the way the need for power expresses itself. McClelland found that women's
values differed from men's values. Regardless of their n Power score, the adult
women in his sample were more sociocentric than the men, as other studies
have reported. They said they had been taught more sociocentric virtues by
their parents than men did. They belonged to a larger number of voluntary or-
ganizations. They had loaned material possessions more frequently, and they
more often said they would volunteer to look after children because they liked
to do it.
As contrasted with women low on n Power, those high in n Power dieted
more often, had a larger number of credit cards, reported a higher average daily
fluid intake (of juice, coffee, and other nonalcoholic beverages), and stated they
were more willing to donate parts of their body to others after death. This sug-
gests that since the women in this sample valued helping others more than the
men did, the power motive in them operates to make them want to have re-
sources to give to others. If their body is disciplined through dieting and kept
filled, and they have more credit cards, they should be in a stronger position to
help others. They even think of using their body as a resource for others after
death. In contrast, because the male role places a value on assertiveness, men
high in n Power are more assertive or argumentative rather than interested in
accumulating more resources to share with others.
Sex Differences in the Outlets for the Power Motive

at Different Stages of Maturity
Winter (1982) has made a convincing case for the fact that, in general, the
power motive functions in the same way for women as it does for men. It is
aroused to the same degree in women and men by observing a hypnosis demon-
stration or by inspirational Speeches. Many of its action correlates are the same
for College women and College men. People of both sexes high in n Power are
more likely to join and hold office in voluntary organizations and to pursue ca-
reers like teaching and journalism, where the opportunity for influencing others
is great.
The chief difference in male and female behaviors associated with high
n Power scores lies in the area of what Winter (1982) calls the profligate expan-
sive impulse. Men high in n Power are more apt to fight, drink, gamble, and try
to exploit women sexually. (See the section later in the chapter on the Don Juan
syndrome, as well as Table 8.9.) The same is not true for women, because, Win-
ter argues, their power motives are expressed in more socially responsible ways.
As evidence, he draws on the cross-cultural studies of Whiting and Whiting
(1975), which show that older siblings generally show more responsible behavior
because they look after younger siblings and that girls are more often assigned
this task, so they grow up to be more socially responsible or sociocentric, as was
just noted. Winter has further validated this hypothesis by showing that among
members of either sex who have had no younger siblings to look after, high
n Power is associated with the profligate expansive impulse, whereas among
those with younger siblings, it expresses itself in more socially responsible ways,
such as office holding. Since women in our culture generally, although not al-
ways, are taught sociocentric values, their power motive expresses itself less
often in profligate, and more often in socially responsible, ways. This line of rea-
soning illustrates once again the great importance of considering values in trying
to predict what activities a motive will be associated with.
Because of differences in sex role values, the outlets for the power motive
differ for men and women, as shown in Table 8.6, which has been prepared to
summarize how the power motive expresses itself at different stages of psycho-
social maturity (McClelland, 1975). The correlations in the table were computed
in the following way. Both the n Power score and the score for a particular
stage obtained from the Stewart coding System were converted to Standard
scores with a mean of 50 and a Standard deviation of 10. Then they were
summed, since they then had equal weights, and the sum was correlated with a
particular characteristic, such as the number of power-oriented magazines (for
example, Playboy or Sports Illustrated) a person read. Since there were a large
number of correlations in this study, and some could have arisen by chance, all
the correlations listed in this table were checked in two independent random
samples of the total to make sure they were approximately the same. If the r's
differed markedly from each other in the two samples, they have been excluded.
Thus, the first correlation in the upper left corner of Table 8.6 means that men
who scored high in n Power and also high in Stage I reported reading more
power-oriented magazines than those who scored low on both of these variables.
In another part of the questionnaire, subjects were asked how often they
talked to their relatives or friends about matters that are often kept secret, such
as health problems, pressures at work, or sexual matters. Men high in n Power
with a strong Stage I orientation share such information with others. They also
tend to be more "intraceptive" in the sense of stating that they feel fantasies are
a very important part of their lives, that it is important to understand the un-
derlying motives of other people, and that they sometimes think of natural ob-
jects as possessing human qualities. The trend is not significant, but it is theoret-
ically very important, because it fits with the next finding that among the men
who can be clearly classified as identifying with their mothers or fathers, those
in Stage I more often identify with their mothers. They say they resemble their
mothers more than their fathers and prefer metaphors for death like "a compas-
sionate mother," which suggest merging with the divine ground.
This complex is referred to as pre-oedipal identification with the mother and
fits psychoanalytic theory, since people in the oral intake stage should still be
dosest to their mothers and should be absorbed in the "omnipotence of thought
in which the psychic rather than the material world is the ultimate reality"
(McClelland, 1975). It is as if they have not yet quite emerged from the early
symbiotic relationship with their mother and are oriented toward maintaining a
feeling of unity or oneness with the world, which also expresses itself in terms of
sharing more information about themselves.
In Stage II the men with high n Power also behave as expected. They re-
port Controlling more aggressive impulses, rejecting institutional demands on
them, and no longer asking for personal help from their parents. They are ex-
pressing their autonomy or independence.
In Stage III they show the typical assertive characteristics Winter (1973)
has associated with the legend of Don Juan, the great seducer. They prefer to be
free to love several women at once, they lie more, and they drink more and for
the wrong reasons (to forget their troubles). They collect valuable objects sym-
bolic of their importance. Not surprisingly, they have little interest in child care.
The picture of how men high in n Power behave in Stage IV is not very
clear, probably because male values in our culture do not encourage the sharing
activities that should characterize people at the Stage IV level of maturity. Nev-
ertheless, they do report sharing more potentially secret information with their
wives and list fewer possessions, which should characterize people who have
gotten beyond the more egocentric Stages II and III. As for the theoretically
most important characteristic, they show only a very weak tendency to join
more voluntary organizations, although in other samples this correlation is
much stronger.
In view of the sketchiness of the picture of men in Stage IV, it is reassuring
Table 8.6.
OUTLETS FOR n POWER AT DIFFERENT STAGES OF PSYCHOSOCIAL MATURITY FOR ADULT MALES CV = 85) AND
FEMALES (N = 115) (after McClelland, 1975)
Correlations with n Power Score Plus Stage Score

Stage I: Intake r Stage II: Autonomy r
Men Reads power-oriented material .23* Controls anger .22*
Shares more secrets .31** Rejects institutional responsibility .22*
Is more "intraceptive" (psychic minded) .17 Does not seek personal help from parents .19*
Has pre-oedipal identification with mother .35*
Women Has stronger male inspiration than female inspiration .22* Controls anger .17*
Would invest a $10,000 gift .24* Wants freedom from restrictions on love .20*
Has more physical Symptoms .19* Shares less with mother .16t
Has oedipal identification with father .29*
Stage IV: Generativity/Mutuality Stage III: Assertion

Men Has many organizational memberships .12 Wants freedom to love several women .25*
Has fewer possessions .19t Lies more .22*
Shares more with wife .20t Collects valuable objects .22*
Reports heavier alcohol consumption .19
Drinks for wrong reasons .27*
Dislikes child care .25*
Women Has many organizational memberships .30** Expresses anger to people .17t
Shares more with husband .20* Likes to travel .19*
Tries new foods .25**
Keeps sex life secret .20*
Is more "intraceptive" (psychic minded) .20*
Feels "oceanic oneness" with the world .24* 3»
ip < .10.
*p < .05.
**p < .01.
to know that those who scored higher in Stage IV upon entrance to a large cor-
poration were significantly more likely to be promoted to higher levels of mana-
gerial responsibility in the Company over the next sixteen years. Among noncol-
lege graduates scoring high in Stage IV, 52 percent had been promoted to Level
3 or above in sixteen years in one Company, as contrasted with only 27 percent
ofthose low in Stage IV (p < .05; McClelland & Boyatzis, 1982). In other
words, their stronger orientation toward serving an institution rather than them-
selves (as in Stages II and III) had led to more rapid promotion in the Compa-
ny. The difference was not significant for College graduates, because in this Com-
pany College graduates tended to get promoted automatically because of their
background rather than because of their personal characteristics.
The outlets of the power motive for women are quite diflferent. In Stage I,
women high in n Power are more oriented toward men, just as men in Stage I
are more oriented toward women. In McClelland's study (1975), the subjects
were asked to list the initials of people who were particularly inspirational to
them as they were growing up, to rate the amount of inspiration they received
from each one, and then to identify whether the person was male or female.
Women high in n Power in Stage I report receiving more inspiration (intake)
from men and also identified more with their fathers by the same measure used
to classify the men as identifying more with their mothers. That is, women high
in n Power and Stage I say they resented their fathers more and prefer meta-
phors that describe death in violent terms rather than in terms of union with
the divine ground. If given $10,000, these women would prefer to invest it. They
list having more physical Symptoms. These facts suggest that women high in
n Power in Stage I are particularly oriented toward thinking of themselves as
needing resources to give: they want to have more money to share and are more
bothered when their bodies are not functioning perfectly.
In Stage II they also show typical signs of independence. Like the men in
this stage, women who score high in Stage II tend to have more aggressive im-
pulses, which they control, than those who score low in Stage II. In love rela-
tionships they prefer to go steady with someone but be free to go out with
someone eise if they desire. They share less with their mothers. In both in-
stances, they seem to be expressing the spirit of independence characteristic of
Stage II.
In Stage III, women high in n Power express more of their anger openly,
travel more, like to try new foods, and keep their sex life a secret. Clearly, they
are more assertive, just as the men are in this stage. But what are we to make of
the fact that they are also more intraceptive, or psychic minded? If we are right
in assuming that psychic mindedness for the men in Stage I reflects a fusion
with the maternal source of life, we might infer that women in Stage III are un-
consciously beginning to identify with their own sex and to share that feeling of
"oceanic oneness" with the world. If Stage III for men high in n Power means
behaving in an assertive, Don Juan style, the parallel myth for women is surely
Diana the Huntress running through the forest with her nymphs and hunting
game.
The picture of women high in n Power in Stage IV is much clearer than it
is for men. They join more voluntary organizations than women who score low
in Stage IV, as they should according to theory. They share more secrets with
their husbands. Perhaps the expected relationships are stronger for women be-
cause values in U.S. culture stress helping others more for women than for men.
It is interesting to note that each sex Starts out oriented toward the opposite
sex in Stage I, then breaks this dependency tie in Stage II, becomes assertive
and identifies with its own sex in Stage III, and then is in a position to return
and share with the opposite sex in Stage IV, having firmly established its own
identity in Stage III. The picture is very much what one would expect based on
psychoanalytic theory. Little boys Start out attached to their mothers; they must
break this tie (Stage II) and form a firm masculine identity in Stage III, and
then they are ready to share, at the most mature stage, on the basis of equality.
Women Start out dependent on their fathers, break that tie, establish their femi-
nine identity, and return to share on an equal basis with male partners in mar-
riage. Cultural values of the United States tend to emphasize the importance of
Stages II and III.
Psychologists, psychiatrists, and school teachers are very concerned about
the dependency crisis: they emphasize the crucial importance of people's break-
ing their ties with their parents and becoming self-reliant and independent. They
are also concerned about the crisis involved in going from Stage II to III, be-
cause they worry about individuals who stay alone; they insist people must be-
come more extroverted and assertive in the real world. However, there is rela-
tively little emphasis in U.S. culture, except in religious circles, on going from
Stage III to Stage IV. In fact, scales developed by other psychologists to mea-
sure maturity (Kohlberg, 1969; Loevinger, 1966) tend to stop at Stage III, in
which the füll actualization of the seif is viewed as the highest level of develop-
ment. Yet in psychoanalytic theory and in all the world's religions, there is a
still higher stage in which people sacrifice some of their egotistic concerns for
Service to others.
How does n Power relate to attaining various levels of maturity? It facili-
tates attaining Stage IV for women and inhibits it for men. The way this was
discovered was by correlating either the Stage IV score alone or the Stage IV
score plus the n Power score with the maximum expression of any one of a
cluster of outlets characteristic of that stage. For men, the Stage IV score alone
correlated .23 with the outlets characteristic of it, but adding the n Power score
to the stage score decreased the correlation to an insignificant .14. The reverse
was true for women. Adding the n Power score to their Stage IV scores in-
creased the correlation with the outlets characteristic of that stage from .24 to a
highly significant .39. "If Stage IV represents moving towards mutuality and
equality, as the theory maintains, then it is easy to see that a high n Power,
tending to accentuate the assertive male role, would make it more difficult for
men to show behavior characteristic of Stage IV. Just the opposite is true of
women: if they are to move towards mutuality, they must behave in a more as-
sertive way than is characteristic of traditional women" (McClelland, 1975).
Women move out of the stage of being dependent on men and are more able to
share with them on an equal basis if they have a high n Power.
CONTROLLED AND IMPULSIVE ASSERTIVENESS

IN ORGANIZATIONAL BEHAVIOR
Leadership Motive Syndrome Among Managers

Considerable attention has focused on men whose assertiveness at Stage III is ei-
ther impulsive (Stage lila) or disciplined and controlled (Stage Illb), for the re-
search on drinking has shown that inhibition is an important moderator of how
the power motive expresses itself. Theoretically, men in Stage lila should be im-
pulsively assertive and should not make particularly good partners or managers,
whereas men in Stage Illb should be better officers of organizations or manag-
ers, since they discipline their power impulses. In a general way, this has turned
out to be true. Among Service managers in one large organization, 73 percent of
those classified as Illb (high Stage III, high Activity Inhibition) had been pro-
moted to higher levels in the organization, as compared with only 40 percent of
those classified as lila (high Stage III, low Activity Inhibition). However, the
reverse was true of line and sales managers; 88 percent of those classified as lila
were promoted to higher levels in the organization, as compared with 63 percent
of those classified as Illb. The diflference between these differences is significant
(McClelland & Boyatzis, 1982). In other words, it depends somewhat on the de-
mands of the managerial job. Service managers need to be more disciplined in
managing consumer requests, whereas line managers may need to be more
openly assertive.
The more common way of identifying men at the controlled assertive stage
is not by using the Stage III score itself, but by using a simpler substitute scor-
ing System that correlates highly with it. Men whose n Power score is high—
and higher than their n Affiliation score—and who are high in Activity Inhibi-
tion also tend to be high in Stage Illb, (that is, high in Stage III and high in
Activity Inhibition). The correlation between these two sets of scores is .58,
N = 3U,p < .001 (McClelland & Boyatzis, 1982). The high n Power, low
n Affiliation, and high Activity Inhibition syndrome is called either the imperial
power motive syndrome or the leadership motive syndrome.
The equivalent to the lila pattern (high n Power, low n Affiliation, and low
Activity Inhibition) is referred to as the Don Juan or conquistador syndrome
(McClelland, 1975). Men characterized by this syndrome tend to make poor
husbands, as judged by combined ratings of marital satisfaction and ability of
husband and wife to work together (McClelland, Colman, Finn & Winter,
1978). This finding is supported by others showing that high n Power alone in
men is associated with poor dating relationships (Stewart & Rubin, 1976) and
with being married to wives who are not allowed to compete with them in terms
of careers (Winter, Stewart, & McClelland, 1977). On the other hand, husbands
who are either at Stage IV or characterized by the leadership motive syndrome
(the equivalent of Stage Illb) tend to have more satisfactory marriages (McClel-
land et al., 1978).
In a longitudinal study of College students, Winter, McClelland, and Stewart
(1982) found that those who showed the leadership motive pattern in school be-
haved more responsibly in life ten years afterward. They had joined more volun-
tary organizations, more often held office in them, and participated in more po-
litical activities. This was true for both men and women, but only for those with
children. Thus, it appears that being required to behave more responsibly, as in
rearing a child, helps direct the power motive—even in this controlled form—
away from a self-expressive mode toward a socially responsible one. This may
explain the fact that women high in n Power do not have the same difficulties as
men in dealing with interpersonal relationships like dating and marriage, be-
cause women tend to be more socialized to be responsible for younger children
(Winter, 1982). Thus, they are not as likely as men to use interpersonal relation-
ships to satisfy their power needs.
Several studies (Boyatzis, 1982) have traced the effects of these power mo-
tive syndromes on managerial behavior. McClelland and Burnham (1976)
showed that male sales managers with the leadership motive syndrome had sub-
ordinates who rated the climate of the office higher on such dimensions as the
amount of organizational clarity or team spirit. These higher morale scores were
significantly associated with more sales, thus indicating that men with the lead-
ership motive profile were more effective sales managers. In contrast, salesmen
working for male managers with the Don Juan syndrome gave the office climate
lower scores for organizational clarity and feit less sense of personal responsibili-
ty, indicating their managers were less effective. The leadership motive syndrome
was also found to be associated with greater success as a senior officer in the
U.S. Navy. Division and executive officers with the syndrome were rated as per-
forming better than those without it. (See Table 13.2.)
The most definitive study of this sort was carried out on male managers at
AT&T whose motive scores were obtained upon entrance to the Company
(McClelland & Boyatzis, 1982). The careers of the managers were tracked over
a sixteen-year period, and the level to which they had been promoted was deter-
mined at that time. As the results in Figure 8.9 show, men characterized by the
leadership motive syndrome at entrance were more likely to be promoted to
higher and higher levels of management in the Company over time. Very few of
them had remained at Management Level 1, at which they had entered the Com-
pany, and nearly half of those at Management Level 4 and above were charac-
terized by the leadership motive profile.
By way of contrast, those with high n Achievement peaked in their careers
at Management Level 3. The nonlinear trend is significant. The explanation
seems to lie in the fact that individuals high in n Achievement are used to doing
things by themselves and for themselves, as explained in Chapter 7. They are
able to advance in the Company so long as their job involves the individual con-
tributions they make. However, at higher levels the focus of the job shifts to in-
fluencing others. The greater success of those with the leadership motive syn-
drome at this level can be explained on the grounds that they are interested in
influencing others (the high n Power score), they are not unduly concerned
about whether they are liked or not (the low n Affiliation score), and they are
self-controlled (the high Activity Inhibition score). If men are high in n Affilia-
tion, they regularly make poorer managers except in very Special kinds of posi-
t 60
High
c
n Achievement
550
Z 40
30
/ Leadership
/ Motive
/ Syndrome
B
c
£ 10
1 2 3 4
and Above
Management Level After 16 Years
Figure 8.9.
Management Levels Attained After Sixteen Years by Men High in n Achievement and with the
Leadership Motive Syndrome (after McClelland & Boyatzis, 1982).
tions, such as employee relations managers (McClelland & Burnham, 1976).

This is because they are so interested in other people and in maintaining good
relationships with them that they find it difficult to make hard decisions that
might hurt other people's feelings. Yet success in management depends on ap-
plying the same Standards of judgment to all people. Managers cannot make too
many exceptions in terms of individual needs without making people feel they
are unfair, and it is precisely those high in n Affiliation who are likely to be
most swayed by the needs of particular individuals.
Furthermore, people with the leadership motive syndrome also have other
characteristics that should make them good managers. Table 8.7 summarizes
these characteristics under four main themes: respect for institutional authority,
discipline and self-control, caring for others, and concern for just reward
(McClelland, 1975). The correlations were obtained using a slightly different
way of identifying which subjects were characterized by the leadership motive or
the Don Juan syndromes. The subjects were divided into those that scored ei-
ther high or low in Activity Inhibition, and then within each group the extent
to which a person was higher in n Power than n Affiliation was correlated with
the characteristic in question. Thus, the definitions of the two syndromes are the
same. In the first numerical column, those who score higher in n Power than in
n Affiliation and are also high in Activity Inhibition (which defines the leader-
ship motive syndrome) show the characteristics listed. For example, they tend to
join more organizations. In contrast, those in the second numerical column who
score low in Activity Inhibition and higher in n Power than n Affiliation (the
Don Juan syndrome) join organizations significantly less often, and so on
throughout the table. More information on those with the Don Juan syndrome
will be provided in Table 8.9 later in the chapter.
In general, individuals with the leadership motive pattern—whether men or
women—show more respect for institutional authority, and they favor discipline
and self-control. In fact, they like to work. This is particularly interesting, be-
cause common sense might suggest that subjects high in n Achievement would
like to work, which is not really the case. As pointed out in Chapter 7, individ-
uals high in n Achievement really like to get out of work by being more effi-
cient. In contrast, those with the leadership motive syndrome, also called the
controlled power motive syndrome, appear to enjoy work, because it involves
being in control of things. They also show more public concern for others. They
say they would be more likely to share some of a $10,000 gift with others. And
there is some indication that they are more concerned about a just reward in the
sense that they tend to feel that the most appropriate metaphors for death sig-
nify murder. That is, if they respect institutional authority, discipline themselves,
and care for others, they may be led to think that it is quite reasonable that
they should get a just reward. In this context, death may seem particularly un-
just. As McClelland (1975) points out, institutional religion that strongly favors
these themes gets around the problem of "death as injustice" by promising re-
wards in the afterlife (Christianity) or in terms of being reborn at a higher level
(Hinduism or Buddhism). But in the present context it is clear why a belief in
centralized authority, hard work, serving others, and justice should make effec-
tive managers. In contrast, those with the Don Juan syndrome either do not
subscribe to these virtues or behave in ways directly opposed to them, which
should mean they would not make very effective managers.
The person high in n Power at Stage Illb is a socially responsible person
who manages things well and often assumes a leadership role in organizations or
the Community. In what sense is such a person not completely mature? The crit-
ical difference between Stage Illb and Stage IV lies in the reasons why such
people serve others. Are they doing it on behalf of themselves or truly out of
commitment to a higher good? Those in Stage Illb see the power and authority
Coming from themselves, and they are serving others to extend their own influ-
ence. Thus, they have not achieved the highest level of maturity in which they
have become true selfless instruments of higher authority. Jesus characterized
such people as praying or performing charitable acts in public so that they
could be recognized for what they had done. People truly in Stage IV would not
care whether what they had done was recognized or not.
Motive Profiles of U.S. Presidents

As a further check on these Undings, Winter (1973) reviewed the managerial ef-
fectiveness of U.S. presidents in terms of their motivational profiles. Table 8.8
summarizes the scores for each president since Theodore Roosevelt on n Power,
n Achievement, and n Affiliation as obtained by coding the motivational imag-
Table 8.7.
PERSONAL CHARACTERISTICS ASSOCIATED WITH LEADERSHIP MOTIVE PATTERN:
HIGH n POWER* > n AFFILIATION, HIGH ACTIVITY INHIBITION* (after McClelland, 1975)
I
Correlations of Correlations of
High n Powera High n Powera
— n Affiliation — n Affiliation
When Activity When Activity
Inhibition Is Inhibition Is
Highb Low Highb Low
Characteristics Among Men (N = 29) (N = 42) Characteristics Among Women (N = 28) (N = 46)
Respect for institutional authority Respect for institutional authority
Joins more organizations 25t -.35* Is elected to more Offices 33* .05
Prefers psychiatrist for personal help 29t -.11 Accepts institutional responsibility 37* -.11
Discipline and self-control Discipline and self-control
Feels work is enjoyable 41* .03 Feels work is not boring 40** .01
Keeps feelings to seif 31t -.12 Teils no lies 40** .04
Caring for others, altruism Caring for others, altruism
Shares $10,000 gift with charities 23 -.32* Shares $10,000 gift with friends 36* .07
Taught sociocentric virtues by parentsc 30t -.24 Lists people as among precious possessions 48** .02
Concern for just reward Concern for just reward
Thinks metaphors for death as murder are .27t .01 Thinks metaphors for death as murder are .30t .20
more appropriate more appropriate
a
Score of 3 or more.
^Score of 2 or more.
c
Sociocentric virtues include being kind and friendly to others.
t/> < -10.
*p < .05.
**p < .01.
ery in their inaugural addresses (Winter & Stewart, 1978). It might be supposed
that the content of these Speeches might reflect either the interests of the speech
writer or particular issues of importance at the time more than the personal
characteristics of the presidents. However, Winter and Stewart's data strongly
suggest that this is not so. They found that the general prestige of a president,
as well as his strength in action or assertiveness versus passivity as indepen-
dently rated by Maranell (1970), were strongly positively related to the presi-
dent's n Power score and strongly negatively related to his n Affiliation score.
Just by looking at Table 8.8 we can observe that the presidents generally re-
garded as strong—namely, both Roosevelts, Wilson, Truman, Kennedy, and
Johnson—all had n Power scores that were relatively high and higher than their
n Affiliation scores, which is the pattern associated with more effective manage-
ment if it is also combined with a high Activity Inhibition score (which Winter
did not measure). The relatively less effective presidents were either low in
Table 8.8.
MOTIVE SCORES FOR THE TWENTIETH-CENTURY PRESIDENTS USING
INAUGURAL ADDRESSES (after Winter & Stewart, 1978)
Standard-scored Motives
(Mean -= 50; Standard
Deviation = 10)
n Power —
President n Achievement n Affiliation n Power n Affiliation
T. Roosevelt 56 45 63 + 18
Taft 35 39 33 - 6
Wilson 43 41 49 + 8
Harding 40 48 41 - 7
Coolidge 38 42 38 - 4
Hoover 47 45 38 - 7
F. D. Roosevelt 52 40 54 + 14
Truman 47 41 59 + 8
Eisenhower 41 55 43 -12
Kennedy 58 58 63 + 5
Johnson 61 45 56 + 11
Nixon 64 58 48 -10
Ford 40 80 46 -34
Carter 63 53 51 - 2
Reagana 62 50 67 + 17
Mean score per 1000 words 4.80 3.33 5.53
Standard deviation 2.53 2.41 2.08
a
Scores provided later by David Winter.
n Power, like Coolidge, or higher in n Affiliation relative to n Power, like Eisen-

hower, Harding, and Ford.
It is also worth noting that the three presidents who were higher in
n Achievement than the other two motives and for whom n Power was not
greater than n Affiliation—namely, Hoover, Nixon, and Carter—were all re-
garded as having difficulty getting things done. They had the characteristics of
individuals with high n Achievement in that they thought things through on
their own and came to good Solutions, but they were less effective in managing
others and in getting their good ideas implemented than presidents with a
higher need for Power. The fact that Nixon was forced to resign for using illegal
shortcuts to reach his goals is dramatic confirmation of the tendency of individ-
uals high in n Achievement to behave dishonestly if necessary to achieve their
goals, as reported in Chapter 7. Winter also showed that the higher a president's
n Power score (especially relative to his n Affiliation score), the more likely it
was that the country would go to war during his administration and the less
likely that he would favor Strategie arms limitation. The same trend is reported
for the relation of n Power to n Affiliation scores in populär literature in
Chapter 11.
Finally, a word about Ford is in order, because his n Affiliation score is so
much higher than that of any other president in the table. Both his n Power
and n Achievement scores are well below average. One might well wonder how
he got to be president with this motive profile. The answer, of course, is that he
is the only president on the list who did not actively seek the office and who
was not directly elected to it. When he pardoned Nixon after sueeeeding him in
office, he made a remark dramatically characteristic of the approach of individu-
als high in n Affiliation. He said, 'The man has suffered enough," which was
undoubtedly true. From the more universalistic point of view, however, it was
also true of many other people who had worked for Nixon and gone to jail for
partieipating in similar activities. This is typical of managers high in n Affilia-
tion: they make exceptions to rules in terms of the particular needs of individu-
als, a style of managing often seen as unjust by others.
Ford became president at a time when the country had suffered a number of
setbacks—in the war in Vietnam abroad and the Watergate scandal at home. He
responded to the crisis by emphasizing putting the country back together again
—stressing the importance of affiliation, as a man high in n Affiliation is likely
to do. A president with a different motive profile—Franklin Delano Roosevelt,
for example—could be expected to respond quite differently to a crisis, just as
Roosevelt responded differently to the crisis of the Great Depression in the
1930s with themes of achievement and power, which eventually led the country
into World War n.
Motive Profile of Radicals

Rothman and Lichter (1978; Lichter & Rothman, 1982) have carried out exten-
sive studies of the personality characteristics of Student and adult New Left rad-
icals as they surfaced in the 1960s. Those who protested against society's op-
pression of blacks, women, and the poor were generally pictured at the time as
being liberated from the power concerns that supposedly activated those who
ran the social System. They were described as "flower children" who believed in
self-actualization, equality, intimacy, and Community; more sympathetically,
they were thought to be dedicated to the cause of bringing about a change in
the way society was governed so as to provide greater opportunities for op-
pressed women and minorities. Such concerns would be consistent with having a
strong socialized power drive.
However accurate these images may have been in general, they did not
characterize radical leaders. Lichter and Rothman (1982) reported that as con-
trasted with nonradical students, active Student radicals scored significantly
higher in n Power and Stage III assertiveness, as well as significantly lower in
n Affiliation and the intimacy motive. Non-Jewish radicals scored significantly
lower in Activity Inhibition (see also Rothman et al., 1977) than all other Stu-
dent groups. In other words, they showed the Don Juan motive syndrome char-
acterized by impulsive aggressive behavior and a rejection of institutional re-
sponsibility. (See the summary of the behavioral correlates of the Don Juan
syndrome found in a sample of adult men in Table 8.9.)
Lichter and Rothman also found that adult radicals showed the same pat-
tern, including also a lower Stage IV orientation, indicating an inability to pro-
gress beyond Stage III to a more sharing, integrative level of maturity. In other
words, if the motive profile of the most successful presidents represents social-
ized power motivation, which leads to effectiveness in managing the establish-
ment, the motive profile of those most active in opposing the esHblishment rep-
resents the rebellious type of personalized power characteristic of Stage lila.
Even here, though, it is important to remember that the action a motive
profile leads to depends on other variables in the person-environment interac-
tion. Greene and Winter (1971) found that black Student leaders were signifi-
cantly higher in n Power than other blacks, and that among black students
reared in the North, a high n Power was associated with being directly active in
the black Community and unwilling to work within the white power System. In
other words, high n Power in these students led to radical protest. However,
among black students reared in the South, a high n Power was significantly cor-
related with being pragmatic or willing to work within the System. Greene and
Winter explained the difference as being due to diflferent estimates of the proba-
bility of success of alternative strategies depending on past experience with racial
discrimination. Northern black students had grown up in a System that was
technically integrated and open equally to all, yet they had often feit discrimi-
nated against because of their color. So working within the System to them
would mean accepting the Status quo and little probability of further success in
winning their rights. In contrast, black students who grew up in the South
would have been exposed to an overtly discriminatory System, so to them a for-
mally integrated, nondiscriminatory System represented a Step forward. In both
cases high n Power led to greater activity, but of diflferent types, depending on
their evaluation of how successful one strategy or another would be in advanc-
ing their rights.
Table 8.9.
PERSONAL CHARACTERISTICS ASSOCIATED WITH THE DON JUAN
MOTIVATIONAL PATTERN IN MEN: HIGH n POWER MINUS
n AFFILIATION, LOW ACTIVITY INHIBITION (after McClelland, 1975)
Correlations of High
n Power — n Affiliation
Among Men Low in
Activity Inhibition
Characteristics (N = 42)
Phallic assertiveness
Reports more frequent physical fights .231
Has higher maximum consumption of any alcoholic drink .27*
Talks more about sex life .25 f
Does not recall distress for which he was responsible .28*
Scores higher on Stage III (Phallic) score on Maturity Scale .33**
Scores higher on any alternative in male Stage III cluster of .35**
actions (promiscuity, drinking, lying, collecting valuable
objects)
Rejection of institutional responsibility
Joins fewer organizations .35
Rejects institutional responsibility .33
Would not share $10,000 gift with charity .32
fp < .10 in the predicted direction.

*p < .05 in the predicted direction.
**p < .025 in the predicted direction.
INHIBITED POWER MOTIVE SYNDROME

AND SUSCEPTIBILITY TO ILLNESS
Cardiovascular Disease
McClelland (1976) observed that individuals high in n Power, particularly those
with the leadership motive syndrome, behaved in many ways like victims of
heart attacks. Friedman and Rosenman (1974) had conducted extensive studies
showing that individuals with coronary artery disease were much more likely
than normal individuals to be irritable, always in a hurry, hard driving, and
tense, often because of repressed anger. They labeled these characteristics Type
A behavior and contrasted it with Type B behavior, which was typical of people
who are much more relaxed and easygoing, willing to take things as they come,
and unlikely to have heart attacks. They reasoned that the "driven" quality of
Type A individuals meant that they had more chronically active sympathetic
nervous Systems, which in time would put a strain on the cardiovascular System,
since chronic sympathetic activation increases heart rate and has other effects,
such as releasing epinephrine, that could be damaging to the cardiovascular sys-
tem. Type A behavior, particularly since it involved suppressed assertiveness,
suggested that inhibited power motivation, as represented by the leadership mo-
tive syndrome, might lie behind Type A behavior. The implication of the sympa-
thetic nervous System and catecholamine release also suggested that n Power
might be involved.
The potential importance of the inhibition variable is indicated by the fact
that high n Power by itself is not related to Type A behavior, nor is high
n Achievement or n Affiliation (Matthews & Saal, 1978). McClelland (1979b)
found in three different samples that individuals with the inhibited power or
leadership motive syndrome had significantly higher blood pressure, indicating a
strain on the cardiovascular System. The most convincing of these studies was
longitudinal. Motive scores were coded from TATs provided by adult males
some ten years after they had graduated from College. Their blood pressures
were checked twenty years after that, when they were in their early fifties. The
results are shown in Figure 8.10. Those with the leadership motive syndrome—
that is, those with n Power higher than n Affiliation and those high in Activity
Inhibition or self-control—had significantly higher diastolic blood pressure
twenty years later than men with other motive combinations. If a diastolic blood
pressure of ninety millimeters of mercury is considered indicative of clinically
significant hypertension, 61 percent of the men who had the inhibited power
motive syndrome in their early thirties showed signs of hypertension twenty
years later, as contrasted with only 23 percent of the men in the other three mo-
tive classifications (p < .01).
Impaired Immune Function and Illness

McClelland, Davidson, Floor, and Saron (1980) reasoned that chronic sympa-
thetic nervous System activation might lead to more susceptibility to infectious
disease, because more epinephrine release could damage immune function and
thus reduce resistance to infection. In a small sample of male subjects, they
found that those with the inhibited power motive syndrome excreted a higher
level of epinephrine in urine averaged over two occasions than other subjects, as
well as that a higher level of epinephrine excretion was significantly associated
with lower concentrations of immunoglobulin A in saliva (S-IgA), a measure of
immune function. They further found, as expected, that lower levels of S-IgA
were associated with reports of more illness in the past year, particularly from
upper respiratory infections or colds. This finding makes good sense, because
S-IgA is the body's first line of defense against cold viruses that enter through
the mouth and nose.
McClelland and Jemmott (1980) added the important fact that the amount
of stress to which the person was exposed combines with the inhibited power
motive syndrome to make the person more likely to get sick. Their findings are
High Control
90
Low Control
ii
80
n Power > rc Affiliation n Affiliation > n Power

(Age 30)
Figure 8.10.
Mean Diastolic Blood Pressure in Men Aged Fifty-one to Fifty-three Classified by their Power
Motive Type at Age Thirty (after McClelland, 1979b).
summarized in Table 8.10. Students were asked to list any illnesses they had
had in the past year and to rate them for severity on a scale of 1 to 100, "where
100 means you were very very sick with a high fever, nearly died, etc.; 50
corresponds to having a case of the flu that kept you out of circulation for
more than three days; and 1 means you hardly noticed the problem" (McClel-
land & Jemmott, 1980). The illness reports of the subjects were classified ac-
cording to their motive dispositions (whether n Power was higher or lower than
n Affiliation), their Activity Inhibition score (whether high or low), and the
amount of power stress they reported in terms of life events in the past year
that could be considered power or achievement related, such as failing an exam-
ination, involvement in a major sports event, or a major worsening of financial
State. As the data in Table 8.10 indicate, all three of these variables contributed
to how sick the students reported they had been. The most sick were those who
were high in the inhibited power motive syndrome and also high in power
stress (mean illness score = 195.5), whereas those subjects who had none of
these characteristics had the lowest average illness score (mean = 47.4). The
comparisons at the bottom of Table 8.10 show that students with any two of
the characteristics reported themselves more sick on the average than the re-
maining students.
McClelland and Jemmott also found that other motive-stress combinations
were not associated with more illness. For example, those high in n Power and
Table 8.10.
MEAN SEVERITY OF REPORTED ILLNESSES AMONG GROUPS DIFFERING IN n POWER,
POWER STRESS, AND INHIBITION (after McClelland & Jemmott, 1980)
High Activity Inhibition0 Low Activity Inhibition

Subject Groups N Mean SD N Mean SD
6
n Power > n Affiliation
High power stressc 10 195.5 195.1 8 79.9 56.5
Low power stress 13 103.7 95.9 10 59.3 49.4
n Affiliation > n Power41
High power stress 5 66.4 74.8 11 84.5 64.2
Low power stress 10 51.7 40.9 8 47.4 23.8
Sources of variance among versus within groups: df = 7/67, F = 2.32, p < .04.
Mean Severity
Planned Comparisons N of Illness F Values
High n Power, high power stress, high Activity Inhibition 10 195.5
Versus low on all three variables 8 47.4 10.1, p<.0\
Versus all other subjects 65 72.9 13.9, /X.001
When Activity Inhibition unknown
High on n Power and power stress 18 144.1
Versus all other subjects 57 71.9 6.5, p<.02
When power stress unknown
High on n Power and Activity Inhibition 23 143.6 11.5, p<.0\
Versus all other subjects 52 65.2
a
Score of 2 or more.
^T-score n Power > 45 and > T-score n Affiliation.
c
Above the median or 4 or more power-achievement life events checked for last year.
d
T-score n Power < 50 and < T-score n Affiliation.
in affiliative stress (as occurs in falling out of love or the death of a family mem-
ber) did not report themselves to have been more sick, nor did those high in
n Affiliation with high affiliative stress. It is the inhibited and stressed power
motive that is particularly bad for a person's health.
Some further evidence of this relationship is provided in Figure 8.11, which
presents data obtained from male prisoners (McClelland, Alexander, & Marks,
1982). The men higher in n Power and reporting an above-median number of
power stresses reported more severe illnesses on the average than other subjects
(including those low in n Power and high in power stress). Furthermore, those
higher in n Power and in power stress showed signs of impaired immune func-
tion, indicated by a lower mean concentration of S-IgA. Thus, the difference
could not be due to a tendency of some subjects to complain more about every-
thing, including both illness and stress. Finally, 58 percent of those below the
median in S-IgA concentration reported an above-median severity of illness, as
contrasted with only 31 percent of those with above-median concentrations of
S-IgA (p < .05). In other words, the presumption is that stressed high power
motivation is damaging immune function, which makes the individuals more
susceptible to disease (McClelland, 1982). Another study has shown that the in-
hibited power motive syndrome is associated not only with lower S-IgA concen-
trations, but also with another indicator of immune function, natural killer cell
activity (McClelland, Locke, Williams, & Hurst, 1982).
One word of caution is in order, however: all the studies involving physio-
logical functions (epinephrine excretion and S-IgA concentrations) have been
done on men, and there is at least some preliminary indication that the relation-
ships may not hold in the same way for women. In a further study that showed
the usual correlation between high n Power and high power stress and lower
~~~[ Severity of Illness
KJ* I in t «i-IoA rnnrpntratinn

>-< 1 1U 1 \
"5
cu x>
o
| 100 - -25 o
|
nesses
5 80 — - 20
All
o
£ 60 - 15
-
u
1 40 - 10
poi
o
u U
c
8 20 - 5 2
N=18 N=\6 N = 6S N=60

n Power > n Affiliation All Other Subjects
High Power Stress Subjects
Figure 8.11.
Mean Severity of Illness and Concentrations of Immunoglobulin A in Saliva (S-IgA) Among Male
Prisoners High in n Power and Reported Stress Compared with All Other Prisoners (data from
McClelland, Alexander, & Marks, 1982).
concentrations of S-IgA in men, no such relationship was found for women

(McClelland & Kirshnit, 1982).
• ORIGINS OF THE POWER MOTIVE
Parental Permissiveness About Sex and Aggression

McClelland and Pilon (1983) also examined the child-training antecedents of the
adult power motive in the study described in Chapter 7 that investigated the an-
tecedents of the achievement motive. They found that mothers who said when
their children were five years old that they were permissive about sex and ag-
gression had children who developed into adults with higher n Power scores.
The results are summarized in Table 8.11.
The score on permissiveness for sex and aggression is derived from a factor
analysis performed years earlier by Sears, Maccoby, and Levin (1957), who con-
ducted the interviews with mothers about child-training practices. Some of the
training practices included in the overall factor score are listed in Table 8.11 to
better describe what was actually involved. The results for the total factor score
are consistent in random subsamples of the total population, as well as for males
Table 8.11.
CORRELATION OF EARLY CHILD-REARING PRACTICES WITH ADULT n POWER SCORES
(after McClelland & Pilon, 1983)
All Subjects Random Subsamples Males Females

Child-rearing Practices (N = 78) (N = 38) (N = 40) (N = 38) (N = 40)
Permissiveness for sex and aggressiona .31*** .33* .30| .28f .32*
For masturbation .31** .39* .20 .33* .27t
For sex play .28* .22 .35 .37* .23
For aggression to siblings .24* .26 .24 .32* .17
For inappropriate aggression to children .12 .04 .18 .06 .14
For aggression to parents .30** .39* .20 .18 .37*
Demands for child to be aggressive -.05 -.10 .03 -.15 -.01
Encouragement of child to fight back .16 .08 .25 -.08 .44*
Praise for good behavior at table -.32** -.26 -.40* -.47* -.23
Note: The n Power scores have been corrected for regression in protocol length. Correlation of social class with n Power = .30,* and
correlation of Social class with Permissiveness for sex and aggression = .59.**
a
Factor score from Sears, Maccoby, & Levin (1957).
tp < .05 in the predicted direction.
*p < .05.
**p < .01.
***p < .001.
and females. They are somewhat less consistent for components of the score, but
there is considerable consistency for permissiveness about masturbation and
about aggression toward brothers and sisters and parents. Mothers who are re-
laxed about sexual and aggressive assertiveness are most likely to have children
who grow up high in n Power.
McClelland and Pilon also found that it is not permissiveness in general
that leads to high n Power, for Sears et al. also rated permissiveness about feed-
ing, toilet training, bed time, noise around the house, physical mobility, and de-
pendency. Permissiveness in none of these areas was significantly related to adult
n Power scores.
Why should permissiveness about sex and aggression be especially impor-
tant? We cannot be certain, but the discussion in Chapter 5 of the natural incen-
tive of "having impact," which presumably underlies the power motive, suggests
that the impact incentive may more immediately and regularly be present and
result in sympathetic activation in the areas of sex and aggression. If the power
motive is served by the catecholamine System involved in sympathetic arousal,
as suggested throughout this chapter, then finding that early experiences in the
areas of sex and aggression, which also involve sympathetic arousal, lie at the
basis of the need for power is entirely consistent with such a viewpoint.
The fact that both sex and aggression are involved is consistent with a great
deal of physiological and psychological evidence that these two types of activi-
ties are related (Barclay, 1971; Moyer, 1976). The findings suggest that children
who are allowed to develop associations around the pleasures from such sexual
and aggressive impacts develop a strong n Power. If they are punished for these
behaviors, they are likely to develop negative associations to having impact in
the sex and aggression areas, which would generalize to avoiding impact in all
types of situations. The fact that praise for good behavior at the table is nega-
tively associated with adult n Power scores suggests that some kinds of inhibi-
tion of impulsive-assertive behavior can interfere with the development of the
power motive. Furthermore, the use of physical punishment for boys is signifi-
cantly negatively associated with adult n Power scores (r — —.36; p < .05);
this is not true for girls, however, perhaps because physical punishment is less
common for girls. It is also noteworthy that positive encouragement for girls to
fight back increases their n Power scores. In other words, society expects girls to
be less assertive and aggressive so that their pleasure from "impact" normally
would be more often inhibited, and it takes positive encouragement on the part
of the parents for them to develop the associative network that makes up the
need for power.
Sources of the Leadership Motive and Don Juan Syndromes

Of particular interest also are the parental training practices associated with the
leadership or imperial power motive syndrome and the Don Juan syndrome. As
has been noted, the leadership or inhibited power motive syndrome is associated
with effective managerial Performance, belief in centralized authority, a liking
for work, impulse control, and a strong sense of justice. McClelland (1975) has
argued that this syndrome in the West is most closely associated with patriar-
chal authority Systems. Boys who grow up to have this syndrome should be
reared by strong, self-controlled fathers with whom the boys can eventually
identify as adults. In McClelland and Pilon's (1983) sample there is some evi-
dence that this is what happened. In one part of the interview the mothers were
asked whether they or the fathers were primarily responsible for the child's up-
bringing. If the father was reported to have been the most important child-
rearing agent, the son was more likely to be characterized by the leadership
or imperial power motive syndrome as an adult (r = .35; p < .05). The
same was not true of daughters, nor was high n Power by itself significantly as-
sociated with the father's being more important in child rearing. Father-reared
sons are more likely to develop the leadership or socialized power motive syn-
drome.
There is also evidence that having a younger sibling promotes the kind of
socialized power activities characteristic of those with the leadership motive
syndrome (Winter, 1982). The reason, as explained earlier, is that having a
younger sibling to look after promotes a sense of responsibility for others,
which socializes the expression of the power motive. In contrast, those who are
only or youngest children are more likely to show the profligate expansive im-
pulse that characterizes the Don Juan syndrome (high n Power and low Activ-
ity Inhibition). Unfortunately, no direct test has been made to see whether indi-
viduals with the leadership motive syndrome are more likely to have younger
siblings.
In discussing the Don Juan legend, which we have associated with high
n Power and low Activity Inhibition, Winter (1973) suggests that Don Juan
may have been reared particularly by his mother and reports evidence that
strong mothers, as indicated by the fact that they were teachers, tended to have
sons with higher n Power scores than other mothers. There is also considerable
evidence (reviewed in McClelland, 1975) that the absence of the father as a
child-rearing agent leads to a son's identifying with his mother and then devel-
oping a kind of defensive masculinity in later life to cover up his underlying
feminine identification (see Whiting, 1965). The defensive masculinity involves
the same kind of impulsiveness-aggressiveness, sexual promiscuity, display, and
antiinstitutional activities that characterize the conquistador or Don Juan syn-
drome (see Table 8.9). This would suggest that mother dominance in child rear-
ing should be associated with a Don Juan syndrome in adulthood. Such is the
case in these data. Sons whose mothers said they were the more important
child-rearing agent tended more often to grow up with the Don Juan or impul-
sive power motive syndrome (r = .46, p < .05). Again, the same was not true
of girls, nor was mother dominance in child rearing associated with higher
n Power by itself in their sons.
Some caution is indicated in interpreting these findings, because the num-
bers of people with the two Special power motive syndromes was small in each
case, as well as because the results would probably hold only for societies in
which families are primarily patriarchal. For example, Ross and Glaser (1970)
found the expected relationship between impulsive-aggressive behavior and fa-
ther absence among Chicano or Spanish American families in the Los Angeles
area. They compared the family backgrounds of well-behaved and delinquent
boys from the same low-income neighborhoods. Among Chicanos they found
that 59 percent of "good" boys had fathers who were the primary decision mak-
ers, as contrasted with only 25 percent of the "bad boys," for whom either the
mother was the primary decision maker or there was no father present in the
home. The same was not true of black families however, where an equal propor-
tion of the good and bad boys came from families with no fathers present.
Among blacks, the crucial question seemed to be whether there was a strong
mother figure. Seventy-three percent of the good boys had mothers as primary
decision makers, as compared with only 39 percent of the bad boys. In other
words, impulsive-aggressive behavior of the Don Juan type in boys seems to be
associated with the absence of the culturally normal primary authority figure,
which is usually, but not always, the father.
Loss of Status and the Development of the Need for Power

To locate child-rearing practices associated with the adult power motive only
raises a further question. What would lead parents to behave in those particular
ways? In the case of the achievement motive, parents influenced by ideological
reform movements seem more likely to emphasize that their children should do
things better in contrast to the old ways of doing things (see Chapter 7).
What would lead parents to stress assertiveness or, at least, to not inhibit it?
Some indirect scraps of evidence suggest that a loss of or threat to Status is as-
sociated with a compensatory increase in n Power. Men who have lost a parent
through death and black men score higher in n Power than other men (Veroff et
al., 1980). In both cases it can be argued that there has been some threat to the
men's sense of competency and power. Unemployment should be an even
greater threat in the case of men. No direct evidence exists on the n Power
scores of unemployed men as compared with others, but McClelland (1976) did
find a significant correlation between high levels of unemployment in a given de-
cade in U.S. history and high n Power scores, as coded in populär literature for
that decade (see Chapter 11).
More convincing data comes from a study conducted by Heller (1979) on
children of the parents involved in the Jewish Holocaust in Europe. He located
a sample of Jewish students, all of whose parents had been forced to migrate to
the United States from Europe because of oppression by Nazis in Germany and
elsewhere. He divided the sample into those whose relatives and immediate fam-
ily members had died in concentration camps and those whose families had not
suffered to nearly the same degree. Many of the parents in what he called the
"high-stress" group were themselves concentration camp survivors. He reasoned
that those who had personally suffered and whose relatives had died would have
been subjected to much greater loss of Status than the parents who had not suf-
fered to the same degree. He found that 67 percent of the children of parents
who had undergone greater stress scored high in n Power, as compared with 44
percent of children of parents who had undergone lesser stress (p < .05).
Furthermore, there was other evidence that the higher n Power scores were
directly related to the highly stressed parents' placing more emphasis on the sur-
vival of Jews as a group. The children of the highly stressed parents were not
more religious in belief or practice, but they reported that they had been re-
quired to learn more about Jewish history and customs, and they had more fac-
tual knowledge about such things. Furthermore, many more of them disap-
proved of intermarriage with non-Jews than did the children whose parents had
been forced to emigrate from Europe but had been less threatened. This was
particularly true of women. Eighty percent of the daughters of highly stressed
parents disapproved of intermarriage, as contrasted with 44 percent of the
daughters of less stressed parents (p < .01). The finding is especially meaning-
ful because in Jewish culture it is the woman who transmits Jewishness to her
children.
In this case we have no direct knowledge of how permissive the parents
were in early childhood toward sex and aggression, but there is plenty of evi-
dence that the parents who had feit the most threatened by the possibility of
Jews being wiped out were those who insisted that their children be assertive in
preserving Jewishness, which led to higher n Power. Heller also obtained infor-
mation directly from the parents and found that those who had been under
greater stress also reported that they had placed greater demands on their chil-
dren to learn about Jewish culture than the parents who had been under less
stress.
Incidentally, the children of the more highly stressed parents also had lower
n Affiliation scores than the children of the less stressed parents: they cared less
about liking or being liked by others (see Chapter 9). It is ironic that Germans
with the high n Power-low n Affiliation syndrome (see Table 11.6 in Chapter
11) persecuted Jews in Europe and evoked a similar power motive syndrome in
children of those persecuted Jews in the United States. With such a motive pat-
tern, one could certainly expect Jews to fight back, and it is interesting to ob-
serve that the leaders in the radical protest movements in the 1960s were over-
whelmingly Jewish (see Lichter & Rothman, 1982; Rothman, 1981; Rothman &
Lichter, 1978) and were characterized by this particular motive syndrome.
Heller did not measure inhibition directly, but he did employ Winter's hope
of Power and fear of Power subscores. Power stories classified as indicating
hope of Power center on themes of personal dominance in zero-sum games (see
Table 8.1) and are significantly associated more often with high n Power and
low inhibition (Winter, 1973). Heller found that the children of concentration
camp survivors, who had parents under high stress, scored significantly higher
in hope of Power than children of Jewish parents under lesser stress. In con-
trast, there was no significant difference in the scores of the two groups of chil-
dren on fear of Power, which is associated with themes of expressing power on
behalf of others, and with high n Power and high inhibition. Thus, oppression
may lead to the direct encouragement of a desire to retaliate or strike back in
an impulsive-aggressive way rather than to a more socialized type of power mo-
tivation. More research is needed on other types of deprivation of Status, but

preliminary evidence suggests that it leads parents to emphasize or release as-
sertiveness in their children, which allows them to develop strong n Power, and
perhaps low n Affiliation and impulsive or retaliatory assertiveness.
NOTES AND QUERIES
1. Think through the implications of the finding that individuals high in

n Power seem to be "preset" to register a bigger response to power Stimuli.
Is this a sign of what Hüll meant by the anticipatory goal response? After
all, it does not surprise us in terms of learning theory that a hungrier per-
son will salivate more at the sight of a steak than at the sight of a stone.
But is an electrical response in a portion of the brain the same as a salivary
response? Must greater electrical responsivity be the result of prior learning,
or could it reflect some other process?
2. If the power motive is associated with greater right brain dominance, what
other behaviors would you expect individuals high in n Power to show? Ex-
amine the handedness literature, for instance (Gur & Gur, 1974).
3. Why do you think men and women high in n Power consider themselves to
be complicated and disorderly?
4. Score the imaginative stories you wrote earlier for n Power. Then make a
list of each of the common outlets for n Power for one of your sex and
class background, and check whether or not you generally behave in those
ways. Is there agreement between your n Power score and the number of
behavioral signs of n Power you show? Explain any discrepancies, if you
can, in terms of other determinants of your behavior, such as values, skills,
or opportunities.
5. Design an experiment that would test the hypothesis that the Atkinson risk-
taking model applies to people high in n Power if probability of success
means the probability of being noticed rather than the probability of
succeeding.
6. Only the men high in n Power chose to arm wrestle with opponents who
were stronger than they were in the McClelland and Teague (1975) experi-
ment. How would you explain this result either in terms of the Atkinson
model of risk taking or any other model?
7. Do you believe men think more about sex, aggression, and power after
drinking because they know they are drinking and expect that drinking
should affect them that way? In other words, are the effects of social drink-
ing on power thoughts the result of learning, of Suggestion, or both? Assem-
ble arguments on both sides of this issue.
8. It is argued in the text that the attempt to conceal from the subjects what
they are drinking so distorts the Situation that whatever results are obtained
are atypical or meaningless. Is there any other way to distinguish between

the effects of alcohol in itself and the effects of the expectations that go with
drinking? What about comparing the effects of actually drinking alcohol
with the effects when people believe they are drinking alcohol but actually
are not? (See, for example, Briddell, Rimm, Caddy, Krawitz, Sholis, &
Wunderlin, 1978; Lansky & Wilson, 1981).
9. In Perry's dream described in Chapter 2, when he was frustrated he tended
to regress to an earlier stage, whereas in Freud's dream, frustration led to
progression to a later stage. Stewart (1978) has confirmed that major life
events like marriage or going to a new school tend to lead to regression in
stage thinking. Do you think the power motive might modify this tendency
in any way?
10. A major unsettled theoretical point is whether the Freud-Erikson stages of
maturity refer primarily to stages in development of the power motive or,
more generally, to stages in development that could affect how all motives
express themselves. Which do you think is more likely to be the case? How
could you determine which point of view is correct? (Consult McClelland,
1975.)
11. Have you noticed any evolution in the way your n Power has expressed
itself over the years that might correspond to changes in your level of
maturity?
12. What does it mean that great spiritual leaders, such as Black Elk (see Nei-
hardt, 1932/1972), are described as having powers involving healing, secret
knowledge, control over the forces of nature, and bringing people together
into an organized nation? How are these characteristics related to expressing
the power motive at various stages of maturity? (For a discussion, see
McClelland, 1975.)
13. In what type of management positions would you think a higher n Affilia-
tion relative to n Power might be more important? If you cannot imagine,
read Chapter 9.
14. Can you estimate from the figures in Table 8.10 whether stress or inhibition
has a greater effect in producing illness for those higher in n Power than
n Affiliation?
15. In view of the various explanations given for the development of either the
Don Juan or leadership motive syndromes, what social events would be
likely to increase the appearance of each characterological syndrome?
16. Design an experiment that would test whether high n Power is associated
with the desire to have impact on things as well as on people.
17. Philosophers, psychologists, and sociologists have often assumed that people
are inherently selfish—that in the end they are motivated to do whatever is
to their advantage. A typical Statement of this position is "Man is selfish.
Individuais will try to maximize their outcomes" (Walster, Walster, &
Berscheid, 1978). Analyze this Statement in motivational terms. What could
it mean? In what sense, for example, are individuals high in n Power trying
to maximize their outcomes by taking extreme risks? Could it be said that
individuals high in n Achievement are trying to maximize their outcomes
more than those high in n Power? What outcomes?
18. An advertisement for Fortune magazine said: "To make it in business . . .
you need all the smarts, guts, and ambition you can muster. Ambition? You
bet. It's finally out of the closet. At last you can be frank about your drive
for success. That's what the fast track is all about." What motive is being
appealed to here?
19. There is some evidence from the study of identical and fraternal twins that
assertiveness, aggressiveness, or dominance is determined in part by heredity
(see Loehlin & Nichols, 1976). According to the view of the power motive
and its biological correlates presented in this chapter, formulate some hy-
potheses explaining how inherited biological characteristics might make it
likely that a stronger n Power would develop in some people than others.
Review Sheldon's (1942) evidence that mesomorphs with strong muscular
development are more assertive, as well as the Suggestion in Cortes and
Gatti (1972) that mesomorphs may be higher in n Power.
9
The Affiliative Motives
• The Meaning of Love
• The Sexual Motive

The Biological Basis of Sexuality in Males
The Biological Basis of Female Sexuality
Is the Sexual Response a Motive?
Homosexuality
• Measuring the Sexual Motive in Fantasy

May's Scoring System for Sex Differences in Fantasy Patterns
Evidence That Deprivation-Enhancement Scores Measure the Sexual Motive
• The Need for Affiliation

Methods of Measuring the Need for Affiliation
Evidence That the Need for Affiliation Score Measures a Motive
• Characteristics of People with a Strong Need for Affiliation
Performing Better When Affiliative Incentives Are Present
Maintaining Interpersonal Networks
Cooperation, Conformity, and Conflict
Managerial Behavior
Fear of Rejection
• The Intimacy Motive

Methods of Measuring the Intimacy Motive
Contrast Between the Affiliative and Intimacy Motives
Evidence That the Intimacy Score Measures a Motive
• The Affiliative Motives and Health
• Origins of the Affiliative Motives
• Relationship of Sexuality to Affiliation and Intimacy
333
THE MEANING OF LOVE
People appear to have a basic need or desire to be with other people, just as
most animals prefer to be with other members of their species. Part of the need
is sexual in origin and biologically adaptive, because the two sexes must get to-
gether in order to reproduce the species. The need to affiliate with others in-
cludes sexual contacts, but it is much broader, including various types of emo-
tional interpersonal attachments that may grow out of natural contact incentives
as outlined in Chapter 4. What has always Struck observers about this need is
how important it is to life and health, how pervasive it is, and how it appears in
many different forms. The word love is commonly used to describe various types
of affiliative ties, and everyone agrees that it is important to satisfy the love
need, yet no one is quite sure, in the words of a populär song, "What is this
thing called love?" Before we review modern psychology's attempt to answer
this question, it will be helpful to turn first to an ancient treatment of the topic
in Plato's Symposium. The Speakers at this banquet, as reported by Plato, man-
aged to mention most of the important themes that have characterized discus-
sions of the psychology of love ever since.
In the Symposium some men get together in ancient Athens for an evening
of eating, drinking, and entertainment. Socrates joins them, and they look for-
ward to a lively time, for he is known for his ability to challenge their ideas and
make them think. The medical man present, Eryximachus, proposes that instead
of getting drunk, they spend the evening talking about love, singing its praises,
and trying to define and understand it. He contributes to the discussion by argu-
ing that the practice of medicine involves "the knowledge of the principles of
love at work in the body," as well as that the good practitioner must be able
"to bring elements in the body . . . into mutual affection and love and know
how to create love and harmony among the different elements in the body."
He sounds like another doctor, Sigmund Freud, who over two thousand
years later also argued persuasively that properly functioning love (or libido, in
his terminology) was essential to mental and physical health. Both men believed
that harmony was an essential aspect of love and health. For Freud, disharmony
and disease resulted from the failure of the love instinct to develop normally.
For Eryximachus, love was the result of different elements in the mind or body
functioning together the way different notes in a chord produce a pleasing sound
or different beats on a drum create rhythm. For this reason he thought of music
as contributing to creating harmony: "Music, by implanting mutual love and
sympathy, causes agreement between these elements, just as medicine does in its
different sphere, and music in its turn may be called a knowledge of the princi-
ples of love in the realm of harmony and rhythm." We will run into the notion
of rhythm and mutual reciprocity again and again in other treatments of love.
Remember also that Freud argued that rhythmic repetitive self-stimulation was
the cue that naturally elicited the sexual response.
Eryximachus distinguishes between sacred and profane love, or heavenly
and vulgär love, for he also realizes that love can cause all kinds of trouble.
Love must be disciplined lest it lead to debauchery. Love must be expressed in
an orderly manner, for "when inordinate love gets the upper hand . . . it causes
The Affiliative Motives 335
widespread destruction and injury." He concludes by arguing that love in gen-

eral "exercises a multifarious and great, or, to speak more accurately, an om-
nipotent sway, but it is the love whose object is good and whose fulfillment is
attended by sobriety and virtue . . . that possesses the greatest power, and is the
author of all our happiness and makes it possible for us to live in harmony and
concord with our fellow creatures." Few modern psychologists would disagree
with this conclusion, although they have had great difficulty finding ways of
measuring love's multifarious influence.
Another participant in the banquet, Agathon, focuses not so much on how
love promotes individual well-being, but on how it affects interpersonal relation-
ships: "Love is in the first place supreme in beauty and goodness himself, and in
the second the cause of like qualities in others." It promotes "peace among men
. . . it is love who empties us of the spirit of estrangement and fills us with the
spirit of kinship; who makes possible such mutual intercourse as this." When it
is Socrates' turn, he introduces the theme of Sublimation, or the changes in the
objects to which love is attached. He Claims he was taught the mysteries of love
by a woman who explained to him that we "first fall in love with one particular
beautiful person and beget noble sentiments and partnership with him." Later
we "become a lover of all physical beauty" and then discover that beauty of
soul is more valuable than beauty of body. Contemplation of examples of moral
beauty leads love eventually to attach itself to love of wisdom, or "many beauti-
ful and magnificent sentiments and ideas," until it arrives at its final goal, the
"contemplation of absolute beauty," which is unchanging and eternal.
To sum up, the participants in the Symposium conclude that love promotes
health in body and mind, that it does so by creating harmony among conflicting
elements inside the seif, that therefore it can be promoted by music and rhythm,
that it evokes good qualities in others and promotes peace and happy relation-
ships among people, that in excess or attached to the wrong objects it can cause
great distress, and that it Starts out in life by being attached to physical objects
but then extends its aims to more abstract goal objects. Many of these same
themes have been echoed by personality theorists whose contributions were re-
viewed in Chapter 2. In the rest of this chapter we will review the extent to
which these themes have been found to characterize love as modern psycholo-
gists have attempted to measure it. Unfortunately, progress in this area has so
far been somewhat disappointing. We know much less about the afriliative mo-
tives in the scientific sense than we do about the achievement motive, the power
motive, or the avoidance motives. The value of reviewing the themes covered in
Plato's Symposium on love is that it keeps in front of us what psychologists
ought to be investigating as we review what they have been investigating.
• THE SEXUAL MOTIVE

The Biological Basis of Sexuality in Males
As both Freud and Socrates suggest, a useful place to Start the study of love is
with its physical component—with the contact gratifications that make up the
sexual response. Sexual arousal or pleasure in this type of contact depends to a

considerable extent on the presence of hormones—androgen in the case of men
and estrogen and, to a lesser extent, androgen in the case of women. These hor-
mones play an important part in determining sex differences in anatomy and in
responsiveness to various Stimuli. Embryos containing the Y, or male sex, chro-
mosome in each cell produce a testis that secretes the male sex hormones, or an-
drogens, which promote the development of the male genitalia and secondary
sex characteristics. If androgen is not secreted in the prenatal period, the embry-
onic gonad turns into an ovary, which releases the female estrogen. There have
been cases reported of female embryos that have been accidentally exposed to
large amounts of androgen (see Money & Ehrhardt, 1972). Whereas children
with this prenatal experience grow up as women and are treated as such, they
appear more masculinized in several ways. They act like tomboys, prefer tradi-
tional male toys, are more aggressive and competitive in sports, and have a
lesser interest in children.
The importance of the chief androgen, testosterone, for sexual activity in the
male has long been known. Castration in animals, which removes the testis that
produces testosterone, greatly diminishes sexual activity in animals, and injecting
testosterone subsequently restores it to its former level (see Grünt & Young,
1952). Furthermore, male primates showing more testosterone in blood plasma
are more dominant and have more sexual access to females (see Rose, Holaday,
& Bernstein, 1971). Yet there is obviously no simple, automatic connection be-
tween hormones and sexual behavior, for the brain is also involved in sexual re-
sponsivity, and what an animal is experiencing also influences its actions. For
example, Rose et al. (1971) reported a case in which a defeat in the struggle for
dominance in a primate hierarchy led to lower levels of testosterone excretion.
In this instance an experience led to diminished testosterone production, which
led to less competitive behavior.
The brain is clearly involved in the complex mechanism by which the body
regulates hormone production and sexual responsivity. The hypothalamus, a
portion of the midbrain, or limbic System, releases gonadotrophic hormones,
which stimulate the gonads to produce testosterone, which feeds back and influ-
ences another part of the hypothalamus to increase sexual responsivity. So what
the organism experiences influences the hypothalamus at the outset of the cycle,
as well as its responsivity at the end of the cycle, when it is affected by the tes-
tosterone released (see Klein, 1982).
The influence of higher mental processes is even greater in humans. For ex-
ample, Bremer (1959) showed that only one half of a sample of males castrated
in adulthood lost sexual interest altogether right away. Others remained sexually
active for several years. The amount of sexual experience the males had previ-
ously had proportionately delayed loss of interest, which suggests that the sexual
motive involves learning based on a natural incentive, as we have argued, so
that it continues to influence behavior even after the natural incentive is re-
moved. Androgen production increases at puberty in males, as does sexual inter-
est. Because of its association with sexual interest in such biological phenomena
as castration and puberty, androgen has been called the "libido hormone"
(Money & Ehrhardt, 1972). But is there any evidence of stable individual differ-
ences in sexual drive, as indicated, perhaps, by the fact that some men regularly
produce more androgen?
A number of facts raise this possibility, although they do not conclusively
establish it. Klaiber, Broverman, and Kobayashi (1967) have reported that the
amount of androgen excreted in urine in young males is significantly correlated
with various measures of physical size and strength, such as ehest, waist, hip,
and right bieeps circumferences, as well as with body weight. In other words,
males who exerete more androgen are bigger and stronger. They are also better
at simple perceptual motor tasks, such as naming objeets as fast as possible, dis-
playing what Broverman, Klaiber, Kobayashi, and Vogel (1968) call an automa-
tization cognitive style. They are less good at tasks that require delay and inhibi-
tion, such as finding figures embedded in a complex design. Broverman et al.
(1968) do not report on whether the highly androgynized males are more sexu-
ally active, but Eysenck (1973), in another connection, has found evidence that
suggests they might be. He found that men and women who score high on his
measure of extraversion are more sexually active. Table 9.1 summarizes some
typical findings showing that among unmarried German students, extraverts are
more sexually active than introverts.
Extraverted males appear likely to be more highly androgynized for two
reasons: Many of the items on the extraversion scale are similar to the items on
Sheldon's Somatatonia Scale (Sheldon & Stevens, 1942), which Sheldon found
to be closely associated with the bigger, stronger, mesomorphic body type char-
acteristic of more androgynized males. In other words, both extraverts and so-
matatonics love physical adventure, need and enjoy exercise, are energetic and
assertive, and so on. Furthermore, Eysenck (1973) has shown that extraverts
are characterized by adrenergic dominance, as indicated by the "lemon test."
He measured the amount of saliva produced by four drops of lemon placed on
the tip of the tongue. The glands that produce saliva are innervated both by ad-
Table 9.1.
SEXUAL PRACTICES OF INTROVERTS AND EXTRAVERTS AMONG
UNMARRIED GERMAN STUDENTS (after Eysenck, 1971, after Giese & Schmidt,
1968)
Males Females
Sexual Practice Introverts Extraverts Introverts Extraverts
Masturbation at present 86% 72% 47% 39%
Petting 57% 78% 62% 76%
Coitus 47% 77% 42% 71%
Mediän frequency of coitus per month 3.0 5.5 3.1 7.5
(sexually active students only)
renergic and cholinergic fibers, which act in Opposition to each other. Adrener-
gic Stimulation produces "dry mouth" and cholinergic Stimulation, a more copi-
ous, watery solution. Drops of lemon produce much less saliva from extraverts
than from introverts, indicating that in the former the adrenergic control is
stronger.
Broverman, Klaiber, Vogel, and Kobayashi (1974) have summarized evi-
dence indicating that adrenergic dominance goes with the automatization style
characteristic of highly androgynized males. There is also a strong connection
between androgens and adrenergic dominance. For example, injection of testos-
terone restores adrenergic dominance. Male sex offenders who have been cas-
trated so that their supply of testosterone is shut off show reduced destructive-
ness and sexual aggressiveness (Hawke, 1950). But injections of testosterone
reinstated their destructive aggressive behavior, indicating increased activity of
the sympathetic nervous System and adrenergic dominance. So muscular an-
drogynized males seem likely to have stronger sexual motives, because they
share temperamental and hormonal characteristics with extraverts, who are
more sexually active.
What is not in question here is whether some males are more sexually active
than others: Kinsey, Pomeroy, and Martin (1948) demonstrated that long ago.
What needs clarification is the role of the androgens in producing these differ-
ences in activity, as well as some method of measuring differences in strength of
the sexual motive that is better than the count of sexual activities. For here, as
in every other area of activity we have studied, action is determined not only by
motives, but also by habits or skills and expectations or values. Thus, how often
a man has sexual intercourse is an imperfect indicator of the strength of his sex-
ual motive. To take an extreme case, it is worthless as an indicator of the
strength of sexual drive among monks who have taken a vow of celibacy. All we
can conclude so far is that the amount of androgen excreted might be an ap-
proximate indicator of the strength of the natural incentive on which the male
sexual motive is built.
The Biological Basis of Female Sexuality

Table 9.1 shows that female extraverts, like male extraverts, tend to be more
sexually active, yet we argued in the case of males that this might be linked to
adrenergic dominance associated with greater production of androgens. Does
this imply that androgens might be a factor in the female sexual response? Yes,
it does, for there is even better evidence of such a connection in females. After a
hysterectomy and removal of the ovaries, or during menopause, women are
often given testosterone, which increases their sexual interest (Foss, 1951). The
reason seems to be that testosterone is very much like another androgen women
secrete naturally—androstenedione—which is produced by the adrenal gland
and which is also closely related to women's sexual responsivity (Klein, 1982).
It has been speculated that it is the production of this androgen in low levels
more or less continuously that makes the human female sexually responsive
throughout the menstrual cycle, whereas among lower animals, females are only
responsive when estrogen levels are high.
Women's sexual interest and activity are also controlled somewhat by estro-
gen and progesterone levels. The frequency of sexual intercourse (Udry & Mor-
ris, 1968) and of heterosexual dreams (Baron, 1977) is greater during the period
when estrogen is building up, reaching a peak approximately fifteen days after
the onset of menstruation, at a time when conception is most likely to occur.
The levels of both decrease later in the cycle, when progesterone secretion in-
creases. Furthermore, it has been reported that taking a synthetic form of pro-
gesterone for birth control decreases sexual desire (Grant & Meyers, 1967).
Is the Sexual Response a Motive?

Not initially. Male infants Start having erections the moment they are born, yet
it is not appropriate to speak of them as having a sexual motive until consider-
able learning has taken place that ties certain cues to the pleasurable responses
evoked by natural incentives. The pleasure, in this case, appears to be mediated
by the sex hormones. Initially, rhythmic contact sensations, as from rubbing the
genitals or other parts of the body, release the hormones; in time, however, ei-
ther naturally or through learning, other types of Stimulation produce the same
effect. Many cues get attached to these sources of positive effect, so among
adults it is easy to demonstrate that the sexual response has all the functional
characteristics of a motive: that is, it energizes behavior, sensitizes the person to
certain Stimuli, and causes learning. Many studies have demonstrated that view-
ing pornographic movies is arousing to most men and women (Mann, Ber-
kowitz, Sidman, Starr, & West, 1974), as indicated by such direct measures as
increased size of the penis in males or increased vaginal pulse pressure in fe-
males (Wilson & Lawson, 1976b, 1978). Listening to sexually explicit tapes has
the same effect (Heiman, 1975). Furthermore, arousal energizes sexual activity
when it can occur. Men who were experimentally aroused by erotic Stimuli were
more likely to engage in sexual intercourse afterward if they were married (Cat-
tell, Kawash, & DeYoung, 1972) or in masturbation if they were not married
(Amoroso, Brown, Pruesse, Ware, & Pilkey, 1971).
Increased sensitivity to various tactile and visual Stimuli oecurs during
arousal of the sexual motive. Masters and Johnson (1966), in their detailed
study of human sexual behavior, give many illustrations of how this happens
through the foreplay leading up to sexual intercourse, but it has obviously been
difficult to document with experimental measures just how and to what extent
this happens. Other evidence for the sensitizing effect of the sexual motive can
be found in the greatly increased interest both boys and girls show in erotic
Stimuli at puberty.
The sexual motive also selects out appropriate sexual responses, which is
what happens when Masters and Johnson (1966) make use of the sexual urge to
help partners become sexually more adequate. Furthermore, in adolescence the
sexual motive organizes more effective courting behavior and causes married
couples to act in ways that are sexually more satisfying. The popularity of books
like The Joy of Sex (Comfort, 1974) testifies to the fact that the sexual motive
stimulates many people to learn better ways of satisfying it.
Homosexuality
Psychologists and other students of human behavior have long wondered why
the sexual motive in some people is attached to someone of the same, rather
than the opposite, sex. Socrates thought it was perfectly natural and edifying to
both parties for men to Start out by making love to boys. Freud, the Christian
church, and many others thought such attachments were abnormal or a sign of
depravity. At the present time no one knows for sure how or why homosexual
attachments occur. Biology might play a role. For example, it has sometimes
been argued that some kind of hormonal imbalance, particularly at the prenatal
level, is responsible for homosexuality (see Bell, Weinburg, & Hammersmith,
1981), and that less androgynized, more feminine males might be less likely to
initiate the behaviors that would result in their attracting women. However, at
the present time there is no evidence for hormonal differences between adult ho-
mosexual and heterosexual males (Meyer-Bahlburg, 1977).
Learning could also play a role. For example, case studies summarized in
Bieber et al. (1962) suggest that homosexual males were more likely to have a
strong, dominant mother and a cold or absent father. Thus, they grew up fear-
ing women and seeking the love from males that they failed to get from their fa-
ther. The most obvious flaw in this argument is that cultures in which mothers
are dominant and fathers absent—for example, urban black families—do not
produce a greater number of male homosexuals.
Learning clearly may play a role, because it occurs quickly in connection
with sexual arousal. For example, Rachman (1966) demonstrated that a for-
merly neutral Stimulus could readily acquire sexually arousing properties if it
were associated with an erotic Stimulus. He had male subjects view pictures of
nude females at the same time they were looking at women's shoes. After a few
conditioning trials, the shoes acquired the property of producing sexual arousal.
Thus, people of the same or opposite sex could become associated in this way
with sexual arousal through learning. No one knows whether in fact this leads
to homosexual preference, for most boys engage in homosexual activities first,
but grow up to be heterosexual.
Storms (1981) has pulled together a number of findings that support the in-
ference that early puberty, with its accompanying changes in sex hormone pro-
duction, may be responsible for homosexual attachments in males. From the
ages of eight to thirteen, boys play predominantly with boys or form homosocial
groups. If the sexual drive happens to mature during this period, Storms argues
that according to learning theory there would be more opportunities for it to be-
come attached to boys than after age thirteen, when heterosexual groupings are
more common. In support of the hypothesis, he notes that 60 to 80 percent of
homosexual males report sexual arousal and activity before the age of thirteen,
as compared with 20 to 30 percent of heterosexual males. The fact that homo-
sexuality is less common in women (D. H. Rosen, 1974) supports the thesis,
since sexual maturity (as represented by orgasm) occurs more often later in
women, and at a time when they are no longer playing primarily with girls.
Whatever the explanation for homosexual preference ultimately turns out to
be, the learning must be of a very Special kind, which is difficult to reverse. En-
ergetic attempts to decondition or decrease the arousal value of homosexual
Stimuli have, for the most part, been unsuccessful in shifting homosexuals to-
ward heterosexual intercourse (see Birk, Huddieston, Millers, & Cohler, 1971),
although pairing homosexual Stimuli with shock decreases preference for them,
• MEASURING THE SEXUAL MOTIVE IN FANTASY
Several studies of the effects on fantasy of arousing the sexual motive have been
carried out. They show that sexual arousal has an effect on, but does not influ-
ence, fantasy in a simple and direct way. Exposing young men to slides of nude
females or to an attractive female folk singer decreased overt references to sex in
imaginative stories they wrote afterward (Clark, 1955; Kaiin, 1972). In contrast,
when young men were exposed to the same erotic Stimulation in a relaxed
drinking party, sexual references in fantasy increased markedly, suggesting that
sexual thoughts had been aroused but inhibited under normal testing conditions.
Müssen and Scodel (1955) demonstrated that exposure to the nude slides
evoked more sexual references when the tester was a "young looking, informal,
permissive graduate Student" rather than a somewhat stern professorial man in
his sixties. However, sexual arousal was shown in several studies to increase in-
direct or symbolic references to sexual activities under less relaxed testing condi-
tions (see Figure 1.4 and Table 5.2). The complexity of these findings has dis-
couraged investigators from developing out of sexual arousal studies a scoring
System for the need for Sex (n Sex) that would measure individual differences in
the strength of the sexual motive.
May's Scoring System for Sex Differences in Fantasy Patterns

Robert May (1980), however, has developed a scoring System that may fill this
need, although he makes no such claim for it. He started with the observation
that men and women characteristically teil very different types of stories to a
picture of a male and female couple performing as trapeze artists (McClelland,
1963). Here is a typical male story:
The picture suggests a dynamic, intimate relationship between the man and
woman—hence the light is around their bodies and the rest is dark. This picture is a
climax to a period in which they have come to understand each other. Both are
completely lost in the thought of their union. They are totally occupied. From such
heights they can only go down. (McClelland, 1963)
Note that a period of climactic union (which May called Enhancement) is fol-
lowed by t4going down" (which May called Deprivation). In contrast, in the
women's stories the sequence was more often one of Deprivation followed by
Enhancement, as this example shows:
Mary is learning to do a change between trapezes. She is about to Swing to the next
and her teacher, old Mr. Picken, is going to become instantly ill, and she is going to
catch him from falling by a lucky chance. He will then train her to circus stardom.
(McClelland, 1963)
The men's stories rise and fall, and the women's stories fall and rise. To es-
tablish a scoring System, May first locates in the story the pivotal incident, "the
dramatic turning point in the story" (May, 1980). Words or phrases (units) are
scored for Deprivation (D) if they refer to physical tension, physical discomfort,
härm or injury, exertion or striving, falling, failure, and other unpleasant feel-
ings or outcomes. Units are scored for Enhancement (E) if they refer to physical
satisfaction, accomplishment, rising, flying, success, and other positive outcomes.
'The numerical scoring of units is done on the basis of their position before and
after the pivotal incident, and is arranged such that a story moving from Depri-
vation to Enhancement receives a positive score (+1) and a story moving from
Enhancement to Deprivation receives a negative score (—1)" (May, 1980).
Thus, individuals can get very high positive scores, indicating their stories follow
the D-E, or female, sequence more often or very high negative scores indicating
their stories usually follow the E-D, or male, sequence. Several studies have con-
firmed the fact that even in very different samples, men's stories yield an average
negative score (indicating a predominant E-D sequence) and women's, an aver-
age positive score (indicating a predominant D-E sequence), the differences being
highly significant (May, 1980).
Furthermore, this thematic difference in sex role style is not limited to our
time or our culture, since it occurs in folklore and mythology in other times and
places. The E-D sequence has been called the Icarus complex by Henry Murray
(1955) after the Greek story of the young man who attempted to fly up to the
sun on wings made of feathers stuck in wax, only to fall into the sea as the wax
melted from the heat of the sun. The D-E sequence has been called the Perseph-
one complex by McClelland (1963) after the ancient Greek myth of the young
girl who, while playing with her friends, was seized by Hades, the King of the
Underworld, who took her down under and kept her there until her mother,
Demeter, forced Zeus to have his brother Hades release her. While Persephone
was kept a prisoner below, the earth was barren, but after her return there was
great rejoicing. It was springtime, and plants and animals began to thrive
again—in a particularly dramatic representation of the theme (see McClelland,
1975).
Ogilvie (1967) found that several elements making up the Icarus theme tend
to occur together in the myths of preliterate cultures all around the world.
Goleman (1976) showed that fiction in the most populär men's and women's
magazines in the United States displayed the themes characteristic of each sex.
In the stories for men, the hero is an active, adventuring type who is out search-
ing for some treasure like the Golden Fleece, as in the legend of Jason and
Medea. The hero almost succeeds with the help of a woman, but in the end she
often betrays him—just as Medea betrayed Jason in the Greek legend—and the
hero fails. In the stories for women, a woman finds her hometown and her boy-
friend dull and uninteresting. She is attracted to and carried off by a mysterious
stranger, who in fact turns out to be a crook and gets her into all kinds of diffi-
culties. In the end she betrays him, returns home, marries her boyfriend, and
lives happily ever after. The stories for men follow the E-D sequence, and those
for women, the D-E sequence.
But what reason is there to connect these themes with sexuality? In the first
place, May points out that they reflect differences between the sexes in their sex-
ual experience. The fact that a man has testicles and a penis
. . . gives sexual feelings an external and clearly defined locus . . . sensations experi-
enced in the penis, and its tumescence and detumescence, become the barometer
of sexual interest. Because it is visible, accessible, and sensitive the penis rapidly
becomes the primary bodily focus of attempts to heighten and then resolve
tension. . . . Its outward and mobile quality also makes the penis (and testicles) a ve-
hicle for exhibitionistic pride . . . but this same outwardness makes it vulnerable.
Concern about the intactness and functioning of the genitals is a convenient meta-
phor for a variety of fears about physical intactness. Likewise there is a difficult
business about will power and control that gets involved. Pride over the activity of
the penis is counterbalanced by the eerie recognition that these movements are not
under one's conscious control. . . . Put simply, the male genitals function to embody
and delineate the themes of external focus, motion, aggressive extension outward,
and prideful control versus shameful failure. (May, 1980)
In contrast, May (1980) argues that the female genitals are not so easily
known:
They do not stand out to be seen. They are more likely to be experienced as part of
a vital but ambiguous "inside." . . . The difference in sexual anatomy gives the
woman more opportunity and incentive to pay attention to and puzzle about what
goes on inside her. . . . Possessing a vagina and womb requires imagining, and expe-
riencing, various transactions across that threshold. Menstruation, intercourse, and
child birth are all events in which the boundary is crossed—something goes in or
comes out.
Often pains or difficulties can lead to pleasure, as in the experience of bear-

ing a child, so "for women the major themes are of internal focus, taking in,
and closing and holding, and the need for a faith that things will come out all
right" (May, 1980). Thus, the E-D and D-E themes appear to be based on dif-
ferent experiences connected with sex organs. Erikson (1951) had reached a sim-
ilar conclusion earlier in his study of sex differences in children's play. He found
that boys typically built tall towers and worried about their collapse, or made
roadways or tunnels along which little cars could move. Girls built more en-
closed Spaces with typical family scenes that might be threatened by disruption
from the outside. He summed up his findings explicitly: "the spatial tendencies
governing these constructions closely parallel the morphology of the sex organs:
in the male, exte mal organs, erectable and intrusive in character, serving highly
mobile sperm cells; internal organs in the female, with vestibular access, leading
to statically expectant ova" (Erikson, 1951).
Erikson's findings have been criticized, but they have been confirmed re-
cently in a careful replication of the original study by Cramer and Hogan
(1975). Some of these differences may be due to social learning. Fathers may en-
courage their sons to be more interested in cars, and mothers may encourage
their daughters to enjoy household activities. Yet it is difficult to attribute all
the dhTerences to social role expectations. As May (1980) puts it, "Are girls en-
couraged to be interested in entrances and boys in towers?" It is even harder to
imagine that their parents or others could have socialized them into thinking in
E-D or D-E terms.
Evidence That Deprivation-Enhancement Scores Measure the Sexual Motive

Is there any evidence that D-E scores reflect differences in the strength of the
sexual motive in men and women? There is some, but it is not conclusive. The
sex differences in fantasy patterns become more distinct as children grow older;
they approach a peak at puberty, when the Output of the sex hormones in-
creases. Cramer and Hogan (1975) tested children a year apart just as they were
entering puberty and found that of the twenty-eight whose D-E patterns shifted
during the year, twenty of them shifted more in the direction characteristic of
their sex. Furthermore, May (1980) reports that homosexual males scored more
in the feminine direction than any other group of males tested. However, he did
not find that D-E scores varied in women as a function of the sex hormones re-
leased during the menstrual cycle. And McClelland and Watt (1968) report that
women who work outside the home show less of the D-E pattern than house-
wives, although it seems unlikely that the former would have a weaker sex drive
than the latter.
Probably the most convincing evidence of the link between the D-E pattern
and sexuality is provided by a study that demonstrated that the pattern shifted
under the influence of a sexually arousing film (Bramante, 1970). As Table 9.2
shows, the film shifted the males toward the male pattern and the females, to-
ward the female pattern. Thus, if we follow the logic employed in other studies
of motivational arousal, we might infer that males who think more strongly in
E-D terms under normal testing conditions are thinking the way they do when
they are sexually aroused, and therefore they have a stronger sexual motive. The
same could be said of women who show a stronger D-E pattern. The case is
strengthened by findings from two other studies. Kaiin (1972) reports that an at-
tractive female folk singer increased sexual and aggressive themes in fantasy for
males under disinhibited conditions, and that themes of negative personal after-
math of sex also increased significantly. In other words, sexual arousal in males
produced a theme of enhancement followed by deprivation. And Sara Winter
(1969) reported that nursing, which produces a form of sexual arousal in
women, was associated with an increase in themes of the D-E female type.
Table 9.2.
EFFECTS OF A SEXUALLY AROUSING FILM ON MALE (E-D, - ) AND
FEMALE (DE, + ) RELATIONSHIP PATTERNS TO TRAPEZE PICTURE
(after May, 1980, after Bramante, 1970)
Mean D-E Score

Males Females
Kind of Film N Mean N Mean Difference
Control comedy film (C; Laurel and Hardy) 25 -0.59 25 +0.77 1.36
Romantic, sexually arousing film (A) 29 -1.89 31 +1.90 3.79
Difference (A - C) -1.30 +1.13 2.43;/? < .05
If the D-E pattern measures n Sex in men and women, it might be expected
to relate somehow to sexual activities. Unfortunately, clear evidence for such a
connection does not exist, partly because research has not focused on this point
and partly because sexual activities are so much determined by nonmotivational
factors such as inhibitions and opportunities. May's (1980) case studies of males
or females showing scores that are extreme for their sex do fill out the picture of
what such people are like. Men who score high in the male direction tend to be
more "macho." They value self-conscious toughness and bravado and admire
their fathers as tough and competent. However, they are often uneasy about
their manliness and worry about whether they can really perform physically and
sexually as required: "They envision men as inherently tougher than women and
thus properly the leaders and initiators in relations with women" (May, 1980).
Women high in the female pattern, on the one hand, see their mothers as
demanding that they be nice and "ladylike." and they resent the passivity that
seems to be required of women. On the other hand, they are less concerned
about achievement and more with nurturance, succorance, and endurance. In
other words, the two extreme scores are associated with extremes in sex role
Stereotyping—the active, aggressive male and the receptive, nurturing fe-
male—together with some doubts and dislike for the pattern they show.
Furthermore, May (1980) reports that "in both sexes the people with the
more extreme and traditional fantasy patterns report that they feit isolated, lone-
ly, and cut off from peers in early adolescence." They feit they had failed to es-
tablish good affiliative relationships with others. It is not certain just what this
means, but one possibility worth exploring is that their stronger sexual drive ei-
ther caused them to overaspire for satisfactory love relationships or made form-
ing affiliative relationships more difficult for them. May (1980) summarizes Sulli-
van's view that ideally in prepuberty, children develop a close relationship with
a chum or an intimate friend of the same sex with whom they can feel sympa-
thy and uncomplicated love: "The early adolescent transition is difficult. Lust
impels one towards someone different from oneself. Thus it disrupts the chum-
ship, and also impels one into situations that are intensely anxiety-provoking
and fraught with potential humiliation."
If we speculate that extreme E-D or D-E patterns reflect differences in sex-
ual drive, then the alienation and difficulties that men and women with extreme
patterns report are understandable in Sullivan's terms. For if, in their case, lust
comes too early or in too great strength, it could well cause awkwardness and
anxiety in developing normal affiliative relationships. However, such speculation
needs to be carefully checked out by further research. Do androgynized males,
for example, show the E-D pattern more strongly on May's measure? And are
males with a strong E-D pattern more sexually active when opportunities and
values permit? May did not regard his score as a measure of sexual drive, but of
differences in sex role style, and the Suggestion that it might reflect sexual moti-
vation needs much better confirmation than presently exists.
• THE NEED FOR AFFILIATION
People clearly like interacting with other people, and some like it more than
others. The conceptual problem is how to get from the physical rhythmic con-
tact sensations characteristic of sexual pleasure to pleasure in more distant, non-
physical types of interactions. A link has been provided by the work of Condon
(1979), who has conducted microanalyses of interpersonal interaction in sound
motion pictures, frame by frame. He has demonstrated that motion of various
body parts is carefully articulated with the Speech patterns of the Speaker and
with the body movements and Speech of the listener: "In essence the listener
moves in synchrony with the speaker's speech almost as well as the Speaker does
. . . as rapidly as within fifty milliseconds the whole organization of change of
the body motion of the listener reflects the organization of change in the incom-
ing speech" (Condon, 1979).
These interactional patterns have been observed in film sequences from sev-
eral different cultures and in newborns: "All of the infants exhibited a marked
synchronization of body movement with both tape recorded and live human
voices. Interactional synchrony thus begins as early as the first day of life"
(Condon, 1979). Clearly, rhythm and harmony are involved, just as Plato ar-
gued thousands of years earlier. Condon likens interactional synchrony to a kind
of dance in which two persons unconsciously engage. If the dance goes well
there is pleasure; if interaction goes badly—if the action and reaction are not in
synchrony—there is dissatisfaction and, in extreme cases, pathology such as is
found in autistic children. Thus, we may conceive of contact gratifications as a
specific subtype of a larger set of rhythmic interpersonal interactions that, if
they are present and in harmony, produce pleasure, and if they are absent or
not in harmony, produce unhappiness or anxiety.
Methods of Measuring the Need for Affiliation

Whether or not it is exactly what people seek in interacting with others, cer-
tainly depriving individuals of the opportunity for interaction arouses in them
what has been called the need for ajfiliation, or the need to be with people. Ship-
ley and Veroff (1952) studied the effects on fantasy of two types of affiliative
arousal. In one case students in a fraternity, seated in a group, rated the attrac-
tiveness to them of a number of characteristics, such as argumentative, entertain-
ing, modest, and sincere. Then each member of the group stood up, and the
others recorded which two adjectives from the list best described him. This pro-
cedure was designed to make more salient for each subject the extent to which
he was liked or disliked by other members in the group. In the second instance,
thirty-five freshmen were located who had not been chosen for admission to any
fraternity in a College in which 95 percent of the freshmen were admitted to fra-
ternities. Thus, these students had been specifically denied the opportunity for
interaction of a type that was strongly valued in this particular College. The sto-
ries men in these two groups wrote to pictures were contrasted with stories writ-
ten under control conditions by subjects who had not undergone the sociometric
rating procedure just described and who had been accepted by fraternities. The
stories written after a concern for affiliation had been aroused contained many
more references to rejection, loneliness, Separation, seeking forgiveness, or peo-
ple's changing their ways to preserve an interpersonal relationship.
Similar shifts in story content were found when the concern for affiliation
was aroused by the sociometric rating procedure among another group of stu-
dents at the University of Michigan (Atkinson, Heyns, & Veroff, 1954). These
investigators also found shifts in various subcategories related to the problem-
solving sequence taken from the scoring System for n Achievement (see Figure
6.6). They defined affiliation imagery as the concern over establishing, maintain-
ing, or restoring a positive, affective relationship with another person or persons.
They also found increases in the aroused over the control group for such sub-
categories as instrumental activity, anticipatory goal State, and obstacles in the
environment to establishing a positive relationship. The frequency with which
people include these categories in stories written under neutral conditions consti-
tutes their n Affiliation score. Rosenfeld and Franklin (1966) demonstrated that
the same type of arousal increases the frequency of the same categories for
women. The French Test of Insight described in Chapter 7 also yields an
n Affiliation score that is very similar in its correlates to the n Affiliation score
obtained from stories written to pictures (see French & Chadwick, 1956).
Evidence That the Need for Affiliation Score Measures a Motive

Do people who score high in n Affiliation behave in the three distinctive ways
that indicate they are more motivated for affiliation? While no studies have dem-
onstrated that n Affiliation is associated with physiological arousal, those who
score high in it are energized to act more often in an affiliative way. The best
evidence confirming the point is presented in a study by Constantian (1981) and
reviewed in Chapter 13. When students were equipped with electronic pagers
and were beeped randomly throughout the day for a week, those high in n Affil-
iation, as contrasted with those low in n Affiliation, were found to be more often
talking with or writing a letter to someone.
The need for Affiliation also sensitizes people to affiliative cues. Atkinson
and Walker (1956) presented to subjects some slides of objects so blurred that
they could not be recognized. Images were projected in each of four quadrants
on the slide, and the task was to State in which quadrant the image was clearest.
The fuzzy images were either of faces, neutral objects like lamps, or plates about
the size of a face. Subjects high in n Affiliation saw the quadrant in which the
faces appeared as most clear more often, even though they could not recognize
the Stimuli. This was particularly true if the affiliation motive had also been
aroused by the sociometric judging procedure used in previous studies. For sub-
jects high in n Affiliation, visual cues related to affiliation or those present in the
human face are more salient.
The n Affiliation score also passes the third criterion for determining
whether it is measuring a motive. As contrasted with individuals low in n Affili-
ation, those who score high tend to learn affiliative associations faster. The task
that showed this effect involves learning relationships among people. For exam-
ple, subjects are told that there are four staff members in a library—A, B, C,
and D—and they are to learn the relationships among them. After reading
about an event involving the staff, they are asked to mark on an answer sheet
which of several simple Statements such as "A likes B" and "C dislikes B" is
correct. If they mark the right answer with a Special crayon, a T for "true" ap-
pears on the sheet telling them that their response is correct. Otherwise, an F
for "false" appears. They are given opportunities to show their knowledge by re-
peatedly choosing among such Statements about the people involved on succes-
sive trials. Subjects high in n Affiliation learn the network of social relationships
in such tasks faster than subjects low in n Affiliation, as Figure 9.1 shows. For a
sample of seventy-four adults, the total number of correct responses made in five
trials correlated .20, p < .05, with n Affiliation; —.08 with n Achievement; and
.08 with n Power, the latter two being insignificant correlations. The tendency of
people high in n Affiliation to learn social networks faster is not strong, but it
has been confirmed in a second sample of young adults. This is a further exam-
ple of the fact that a motive facilitates learning only that material specific to
achieving its goal. Neither a high n Power nor a high n Achievement led to sub-
jects' learning social networks faster.
CHARACTERISTICS OF PEOPLE
WITH A STRONG NEED FOR AFFILIATION
Performing Better When Affiliative Incentives Are Present
Individuais with a strong affiliative motive will also perform better on tasks that
do not involve affiliative content if the incentive in the Situation is shifted from
achievement to affiliation. Normally in a laboratory Situation, subjects high in
n Achievement recall more interrupted tasks than subjects low in n Achieve-
ment (Atkinson, cited in McClelland, Atkinson, Clark, & Lowell, 1953). Howev-
er, Atkinson and Raphelson (1956) demonstrated that this is true only if the in-
8. /
High A? Affiliation f/
shi
c /
2 7 /
Ü
a£ M
00
c ^*^L ^
'a
1 '
6
8
;es i n L
-'Low n Affiliation
/
»uodi
5
rect Res
o 4
U /
<<—
o /
1 3
3
zccd
o>
^>
1 1 1 1 1
1 2 3 4 5
Trial
Figure 9.1.
Mean Number of Relationships Correctly Identified on Successive Trials for Subjects High and Low
in n Affiliation.
structions and approach to the subject suggest the importance of doing well.
They deliberately sought to minimize such achievement cues by approaching the
subjects in a very relaxed way and simply asking them to help out the experi-
menters by doing the tasks. In this instance the recall of interrupted tasks was
no longer associated with high n Achievement, but with high n Affiliation. In
other words, the subjects for whom the incentive of pleasing the experimenter
was more important tended to recall the tasks in which they had failed to please
the experimenter by not completing them.
French (1955) reported a very similar result. In presenting a digit-symbol
Substitution task to a group of officer candidates at a U.S. Air Force base, she
varied the incentives for working at the task by giving different instructions to
different groups. In general, subjects with high n Achievement performed
better when the instructions stressed the importance of achievement (see Table
7.1). She also assayed n Achievement and n Affiliation after the men had fin-
ished working on the task and again found that those who scored highest in
n Achievement had shown the greatest gains in Performance previously in two
out of the three conditions. However, in the relaxed condition there was no cor-
relation between the postperformance n Achievement score and past perfor-
mance. Instead, there was a strong relationship between the postperformance
n Affiliation score and having done better at the task previously. She reasoned
that this was because, in the relaxed condition, the female experimenter was
friendly and asked for cooperation, so those who would have done better in this
condition would be those for whom the incentive of cooperating in getting the
experimenter's approval would be most likely to be salient. Thus, those in
whom affiliation motivation was aroused by the cooperation incentive were the
ones who had done better at the task.
The effect on Performance of shifting from achievement to affiliation incen-
tives is dramatically shown in Table 9.3. Atkinson and O'Connor (1966) were
attempting to confirm once again that subjects high in n Achievement and low
in Test Anxiety preferred intermediate risks, performed better, and showed less
persistence when the odds of success were very small than subjects low in
n Achievement and high in Test Anxiety. However, the results did not conform
to expectations, as Table 9.3 demonstrates. Instead, the subjects high in n Affili-
ation showed the characteristics supposedly associated with high n Achieve-
ment. They preferred intermediate risks, showed less persistence at a very diffi-
cult task, and tended to perform better than subjects low in n Affiliation. The
researchers reasoned that what had happened was that the way the task to be
performed was administered differed from the usual practice. The subjects were
all male, and they were tested individually by a female experimenter while she
stood by and watched their Performance. Thus, an affiliation incentive was
introduced in the sense that the subjects would be more likely to want to
please her or get her approval than when performing in groups for a male test
administrator.
In any case, the results make sense if it is assumed that in this Situation, the
more difficult the task, the more approval the subjects might expect for perform-
ing it. That is, in terms of the Atkinson model presented in Chapter 7, the in-
centive value of approval equals 1 — Ps. If this were true for subjects high in
n Affiliation, multiplying the incentive for approval (1 — Ps) times the probabil-
ity of success (Ps) would yield the preference for intermediate risk for subjects
high in n Affiliation—which is exactly what is shown in Table 9.3.
The important theoretical point about such an analysis is that it shows that
motives other than the achievement motive not only can lead to faster perfor-
mance if the incentive is changed, but can also produce other behavioral charac-
teristics like preference for moderate risk, which we have identified as associated
with n Achievement. We must know not only the motivation of the performers
in a Situation, but also the incentives for which they are working. Female exper-
imenters seem more likely than male experimenters to introduce affiliation incen-
tives for male subjects. Jopt (cited in Heckhausen, 1980) has also shown that in
Germany, a female test administrator produced better Performance from male
sixteen-year-olds who were high in an aspect of n Affiliation than a male test
administrator did.
Table 9.3.
RELATIONSHIP OF n ACHIEVEMENT AND n AFFILIATION TO PERFORMANCE WITH A
CLOSELY SUPERVISING FEMALE EXPERIMENTER (after Atkinson & O'Connor, 1966)
Prefer Fast
Intermediate Substitution Low Persistence
Motive N Risk Performance When Ps = .05
n Achievement — Test Anxiety
High 9 44% 67% 67%
Moderate 19 53% 59% 47%
Low 9 56% 22% 44%
p value n.s. n.s. n.s.
n Affiliation
High 11 82% 73% 73%
Moderate 12 50% 42% 58%
Low 12 25% 42% 25%
p value <.05 n.s. .10 - .05
Gallimore (1981) has reviewed evidence that affiliation incentives are more
important to students of Hawaiian ancestry than individualistic achievement in-
centives are. He has obtained results for this cultural group that are very like
those in Table 9.3. Students of Hawaiian ancestry with high n Affiliation Orient
more to teachers, are able to relate to others more effectively, do better at cer-
tain types of tasks such as reading achievement, and prefer tasks of intermediate
difficulty. In none of these cases is their n Achievement score related to these
same Performance characteristics. All that is needed to explain these results is to
assume, as in the case of Table 9.3, that for Hawaiian students the incentive as-
sociated with doing more difficult tasks is social approval or affection rather
than the achievement incentive.
The way in which affiliation incentives affect the Performance of subjects
high in n Affiliation is dramatically illustrated in Figure 9.2. In this study
(McKeachie, 1961), the grades College students received in a course were related
to the students' motive scores and characteristics of the teacher's classroom. As
the graph shows, students with high n Affiliation worked harder and got better
grades in a classroom where the instructor was judged to be warm and friendly
(taking a personal interest in students, calling them by name, and so forth). Stu-
dents with high n Affiliation, in contrast, did not do as well in classrooms where
such affiliation incentives were not present. The fact that students with low
n Affiliation in less friendly classrooms also did better means only that subjects
high in motives other than n Affiliation, like n Achievement or n Power, did
better in other types of classroom environments.
w 50 -
20 -
f
c
b 10 -
Teacher Warm Teacher Less Warm

Figure 9.2.
Relationship of Teacher Warmth and n Affiliation to Grades in Psychology Class (data from
McKeachie, 1961).
Maintaining Interpersonal Networks

Since individuals high in n Affiliation learn social relationships more quickly, are
more sensitive to faces than to other objects, and engage in more dialogue with
others, they might be expected to also show signs of maintaining their connec-
tions with other people. This has turned out to be the case. Lansing and Heyns
(1959) reported that individuals high in n Affiliation make more telephone calls,
write more letters, and pay more visits to friends than individuals low in n Affil-
iation. Boyatzis (1972) confirmed, using a slightly different measure of affiliation
motivation, that they tend to make more telephone calls. Lundy (1981a) found
that high-school students with high n Affiliation tended to join more social
clubs, although the same relationship does not occur for adults (McClelland,
1975), probably because many voluntary organizations for adults represent Spe-
cial interest groups where the primary incentive is not merely social.
People are very important to individuals high in n Affiliation. They prefer
friends to experts as working partners (French, 1956), and, when given feedback
by the experimenter about how a group is working out, they prefer feedback on
how well the group is getting along together rather than on how well they are
performing at the task (French, 1958b). In judging welfare clients seen on slides
and heard on a tape, they tend to pick more positive adjectives than do subjects
low in n Affiliation. The correlation between n Affiliation and such a positive
bias score in adjective descriptions of people in trouble is .22, N = 134,

p < .01 (McClelland & Klemp, 1974).
However, a positive bias toward people is strongly influenced by other fac-
tors such as the expectation of success from reaching out to others, as it should
be according to the Atkinson model for predicting approach tendencies. Table
9.4 shows the mean percentage of positive acts toward others displayed by
women in small groups as a function of their n Affiliation score and their expec-
tation that they were in a friendly group (Fishman, 1966). High n Affiliation did
not lead to more positive acts if a woman thought she was in an unfriendly
group that was not likely to reciprocate.
Furthermore, being in a friendly group by itself does not lead to signifi-
cantly more positive acts, regardless of n Affiliation score. If we think of positive
sociometric expectancy as equivalent to a high probability of success of an af-
filiative act, then the Atkinson model applies here, too: the affiliative motive
times the probability of success of an affiliative act yields the strongest approach
tendency, or the largest proportion of affiliative acts actually emitted.
The importance of other factors can also explain variations in relationships
found with the n Affiliation score in different samples. For example, Rokeach
(1973) reports in a College sample that high n Affiliation is associated with
a high value placed on a world at peace and true friendship—both of which
suggest the importance to such individuals of positive interpersonal networks.
However, in another sample of adults aged thirty-one of a more varied socioeco-
nomic background, these correlations were insignificant (McClelland, Constan-
tian, Pilon, & Stone, 1982). Instead, n Affiliation correlated with the value
placed on happiness. We can only speculate as to the reason for these differ-
ences, but it seems possible that the adults, as contrasted to the College students,
thought the possibility of a world at peace was so low that even those with high
n Affiliation did not rank it very high in their value scheme. They were reacting
Table 9.4.
MEAN PERCENTAGE OF POSITIVE ACTS IN A SMALL GROUP AS A
FUNCTION OF n AFFILIATION AND POSITIVE SOCIOMETRIC
EXPECTANCY, OR PROBABILITY OF AFFILIATION (after Fishman, 1966)
Female Subjects'
n Affiliation
Low High P difference
Low positive sociometric expectancya 25.8% 23.4%

High positive sociometric expectancy 26.1% 36.8% <.001
a
The extent to which a subject saw others as friendly and thought they liked her; the extent to which they saw her
as friendly.
like the subjects in Table 9.4 with high n Affiliation, who did not react posi-
tively if they thought they were in an unfriendly environment.
Cooperation, Conformity, and Conflict

Those concerned about other people might be expected to be more cooperative
and to conform more to the wishes of others, but the data on this point do not
permit clear generalizations. Individuais high in n Affiliation, particularly males,
believe that goodwill is more important than reason in solving human problems
(McClelland, 1975); women high in n Affiliation responded more to a partner's
request to "please slow down" in working on a task together than women low
in n Affiliation (Walker & Heyns, 1962), but the same result was not obtained
for men.
Attempts to relate n Affiliation to conformity with and without social sup-
port or in conflict with the majority have yielded complex results (Hardy, 1957;
Samelson, 1958). People high in n Affiliation tend to go along with the opinions
of a stranger who disagrees with them so long as the stranger is attractive, but
not if the person who disagrees is unattractive (Burdick & Burnes, 1958). Fur-
thermore, both deCharms (1957) and Harris (1969) found that increasing affilia-
tive incentives in the environment increased cooperative Performance in a group
more for those low in n Affiliation than for those high in n Affiliation. This is
reminiscent of Wendt's (1955) finding that pacing work increases Performance
more for those low in n Achievement than for those high in n Achievement.
But what is missing is any clear evidence that individuals high in n Affiliation
are generally more cooperative or conforming.
However, many studies demonstrate that they act whenever possible to
avoid conflict. When subjects were asked to work together to solve a problem,
communicating with each other by means of written slips of paper, the partici-
pants high in n Affiliation wrote fewer references to potentially disruptive deci-
sions about what the group ought to be doing (Exline, 1962). At a political
level, Hermann (1980) found that the n Affiliation score in Russian political
leaders' speeches was correlated .47, p < .05, with favoring detente with the
West: they wanted to avoid confrontation with the West. People high in n Affili-
ation also want to change people more who disagree with them. Subjects high in
n Affiliation made more suggestions for changes to strangers the greater the
views of the strangers differed from their own, but the same was not true for
those low in n Affiliation (Byrne, 1962).
People high in n Affiliation avoid competitive games such as roulette (see
Figure 8.5). When participating in a Simulation of international diplomacy, indi-
viduals high in n Affiliation tend to be quite passive: they carry out fewer con-
flicting or cooperative acts, teil fewer lies in newspaper reports, and so on
(Terhune, 1968b). The need for Affiliation in males is negatively associated with
playing checkers or chess (r = - . 2 9 ; N = 78; p < .05; McClelland, 1975).
Young adults with high n Affiliation avoid joining Special interest groups
(r = -.32;/? < .01; McClelland, Constantian, Pilon, & Stone, 1982). They
also avoid talking about people in negative terms (r = — .30; p < .01). As
would be expected, n Affiliation is regularly negatively correlated with n Power,

which promotes competitiveness. For example, among a sample of 685 mostly
older males in business, the correlation is — .13, p < .001; for a sample of 125
older women, the correlation is — .32, p < .001.
Managerial Behavior
People who try to avoid conflict and criticism should not make very good man-
agers, and, as findings presented in Chapter 8 have already indicated, men with
high n Affiliation tend not to succeed in management. They do spend more time
with subordinates (Noujaim, 1968), but this may only mean they want to be on
good relations with them, which is not always possible for a manager who must
make difficult decisions at times. At any rate, young men with high n Affiliation
tended not to get promoted as often to higher levels of management at AT&T
(McClelland & Boyatzis, 1982). Small manufacturing firms or research and de-
velopment firms headed by men high in n Affiliation tended to be less successful
(Kock, cited in McClelland & Winter, 1969/1971; Wainer & Rubin, 1969).
The one exception to this general picture is the motivational profile of man-
agerial integrators, among whom Litwin and Siebrecht (1967) found n Affiliation
scores to be somewhat higher for those judged more rather than less effective.
Lawrence and Lorsch (1967) also found that effective integrators had higher
n Affiliation scores than ineffective integrators. A typical integrative manage-
ment job is that of the employee relations manager, whose objective is to get
management and labor to work together. Obviously, it is functional for a man-
ager in this job to like being with people and to spend time trying to get people
to resolve their differences and go along together, both of which are characteris-
tics that would be theoretically expected of a person high in n Affiliation. In
contrast, more typical management Jobs require being competitive, trying to in-
fluence others, and making hard decisions that may hurt people's feelings, all of
which involve actions difficult for people high in n Affiliation to take, since they
are primarily interested in avoiding conflict.
In groups that meet to study their own interactions (called self-analytic
groups), effective helpers had higher n Affiliation scores than other participants,
which suggests they might make good managers (Kolb & Boyatzis, 1970). Effec-
tive help was defined as an instance of someone reporting that he or she had
tried to help another, who said at the end of the Session that the help had been
significant. But remember that the goal of such groups is to understand group
process, and the behavior of the participants is not aimed at getting work done,
as it is in most management Jobs. Thus, the skill that those high in n Affiliation
showed would be more useful in an integrator type of management position.
Furthermore, individuals very high in n Affiliation were most often classified as
nonhelpers—that is, they made no attempt to help other members of the group.
It was those with moderate n Affiliation who were the most effective helpers.
McClelland (1975) has described the characteristics of a Special motivational
syndrome in which n Affiliation is high but n Power is also high and inhibition
is low. He calls it the "personal enclave" motivational pattern, because it is
characteristic of people in countries like Mexico and Italy, where n Power and
n Affiliation are both high and where paternalistic family firms are also com-
mon. Close affiliative ties are maintained within firms, which are often domi-
nated by a Single family, and power motivation and competitiveness pit the co-
hesive unit against Outsiders. Men and women with the personal enclave
motivational pattern seek strength from inspirational figures, find security in
close personal ties, and are unusually aware of threats from the outside. Perhaps
managers with this motivational pattern are more successful in countries where
high values are placed on affiliation, although no evidence has yet been collected
on this point. However, it may be as true in these countries as it is in the
United States that high n Affiliation leads to making particularistic judgments
about people in the firm that are viewed as unfair by others, thus leading to
poor morale and less productivity.
Fear of Rejection
Throughout the history of the n Affiliation score, investigators have suspected
that it represents primarily a fear of rejection (f Rejection) for a variety of rea-
sons. As was just noted, individuals high in n Affiliation act in various ways to
avoid conflict and competition as if they feared negative feedback from others.
Furthermore, the methods of arousing the motive in order to develop the scor-
ing System all involved the threat of rejection by peers. The original scoring sys-
tem focused on rejection and loneliness (Shipley & Veroff, 1952) and was only
later expanded to include more positive types of affiliative activities (Atkinson,
Heyns, & Veroff, 1954). Subjects who scored high according to the 1954 System
were more often judged by their peers to be approval seeking and egotistical, as
well as self-assertive and confident in contrast to others.
Furthermore, a number of investigators have reported that individuals with
high n Affiliation are less populär (Atkinson, Heyns, & Veroff, 1954; Crowne &
Marlowe, 1964; Shipley & Veroff, 1952). The explanation given for this has been
that individuals high in n Affiliation are anxious about their relations with oth-
ers, fear disapproval, and spend time seeking reassurance from others, which
makes them unpopulär. For example, Müssen and Jones (1957) reported that
late-maturing boys, who are anxious about their masculinity, score higher in
n Affiliation.
Crowne and Marlowe (1964) found that people high in n Affiliation tend to
score high on a measure of the need for Social Approval, although others have
not duplicated this finding (Fishman, 1966; McClelland, Constantian, Pilon, &
Stone, 1982). But there is evidence that they are apprehensive about social eval-
uation. If they are told they are being observed during a Performance through a
one-way mirror, they are more anxious than individuals with low n Affiliation
scores (Byrne, 1961a). They also dislike a person more whose vievs differ from
their own (Byrne, 1961b). Presumably they wish to avoid the possibility of com-
ing in contact with someone who disagrees with them. However, if they also
have high self-esteem, they are less worried about social contact. Black males
high in n Affiliation and in self-esteem report less social distance in their rela-
tions with whites (Littig & Williams, 1978).
Some theorists, particularly Schachter (1959), have argued that there should
be a connection between the motive to affiliate and fear of rejection, because
people affiliate in response to threat. He found that threatening female College
students with a future shock led them to spend more time with others than
alone while they were waiting for the experiment to Start. Of course, the ten-
dency to affiliate does not indicate the presence of an affiliative motive: spending
time talking with others may be simply a means of reducing anxiety. This is dis-
cussed further in Chapter 10.
Because of the association of the n Affiliation score with fear of rejection,
several attempts have been made to get separate scores for approach and avoid-
ance aspects of the affiliative motive (Byrne, McDonald, & Mikawa, 1963;
Mehrabian & Ksionzky, 1974). deCharms (1957) simply divided the score into
two parts, one of which was the sum of all story categories dealing with positive
affiliation themes and the other, the sum of all subcategories for themes dealing
with negative, fear-of-rejection themes. The partitioning was logical but not
strongly justified by previous research, since arousal studies had shown that both
positive and negative themes increased even when the arousal involved primarily
fear of rejection (Rosenfeld & Franklin, 1966).
Furthermore, attempts to use these different scores have not produced a
particularly meaningful set of findings (Conners, 1961; Fishman, 1966). Laufen
(1967) attempted to derive separate scoring Systems for the two aspects of the
affiliative motive by presenting subjects in advance of the TAT with a question-
naire that focused either on positive or negative affiliative themes (see Heck-
hausen, 1980). But he found that only the questionnaire dealing with negative
affiliative themes (/'Rejection) succeeded in producing differences that could be
scored in the stories written afterward, as compared with what was obtained
under control conditions. Boyatzis (1972, 1973) employed a semiprojective tech-
nique and an a priori scoring System for distinguishing between / Rejection and
n Affiliation. Subjects were first asked to write a story of the usual type about a
person like John mentioned in a brief introductory sentence such as "John has
started a new job. He and his family have just moved into a new neighbor-
hood." After writing a story about John, the subjects are presented a number of
incidents that might arise in John's life and are asked how they think John
would feel about them. For example, "After John received an invitation from
Harry and Joanne (some of his new neighbors) for dinner, (a) he was pleased at
the opportunity to meet some of his new neighbors (Affiliative Interest), or (b)
he wanted his neighbors to like him and hoped that they could become close
friends (Rejection Anxiety)."
It was hoped that employing this technique would avoid some of the biases
inherent in having the subjects make Statements about themselves indicating ei-
ther affiliative interest or rejection anxiety. Results obtained with this instrument
support the notion that these two aspects of affiliation motivation have different
correlates. As Table 9.5 shows, in two different studies, those scoring high in
Affiliative Interest reported having a larger number of close friends. This was
true of both men and women, although in the first study the greater interest in
close friends for women was shown only by the measure of the length of time
they had spent with their dosest friends living away. In contrast, those scoring
Table 9.5.
CORRELATIONS OF AFFILIATIVE INTEREST (AI) AND REJECTION ANXIETY (RA) SCALES
WITH REPORTED AFFILIATIVE BEHAVIORS (after Boyatzis, 1972)
Males Females
AI RA AI RA
Reported
r
a
Affiliative Study l Study 2 Study 1 Study 2 Study 1 Study 2 Study 1 Study 2
Behaviors (N = 23) (N = 75) (N = 23) (N = 25) (N = 52) (N = 76) (N = 52) (N = 76)
Number .36* .25** .17 -.02 .03 .22* -.05 .01
of close
friends
Length of .45* — .08 — .37* — .13 .17

time seen
dosest
friends
living
away
Attitüde -.12 .09 .53*** .25** -.20 -.03 .25* .17
similarity
to dosest
friends
Time .33* .13 .08 -.19 .13 -.15 .04 -.21"
spent
alone
Amount .16 — -.04 — — .11 — -.30*
enjoy
being
alone
a
Study 1 included twenty-three males and fifty-two females; Study 2 included seventy-five males and seventy-six females.
*p < .05.
**p < .01.
***p < .001.
high on Rejection Anxiety seemed most concerned about having close friends
who had attitudes similar to their own. The inference to be drawn from this fact
is that they wanted to be sure their friends agreed with them. Furthermore, they
appeared to avoid spending time alone, as if they constantly needed the reassur-
ance of the presence of others. In Constantian's (1981) study discussed in Chap-
ter 13, she found that subjects high in n Affiliation, when beeped randomly
throughout a week, were significantly less often found to be completely alone.
We will consider further aspects of the fear of rejection in Chapter 10 on the

avoidance motives.
THE INTIMACY MOTIVE

Methods of Measuring the Intimacy Motive
McAdams (1980) decided that a whole new approach was needed to measuring
the positive aspects of the affiliative motive, in view of the fact that the n Affilia-
tion score seemed to reflect primarily, although not exclusively, the fear of rejec-
tion. Returning to the characteristics of love as described by many authors ever
since Plato's Symposium, he decided that arousal conditions should stress open-
ness, contact, reciprocal dialogue, joy and conviviality, and caring and concern.
He obtained written imaginative stories in four situations that approximated
these conditions. In one, new members of a sorority or fraternity in College
were being welcomed in a celebration that addressed the goals of togetherness
and good feeling for one another. In another, students were obviously having a
good time at a dancing party, and those who had not had more than two
drinks (to avoid picking up the influence of alcohol) were invited to write sto-
ries. In still another condition, undergraduates had just participated in a psy-
chodrama in which they had become acquainted with and interacted with a
number of new people under very enjoyable circumstances. Stories written
under all these conditions were compared with stories written when the same
kinds of individuals were not under the influence of affiliative feelings, as in a
classroom setting. Finally, McAdams compared stories written by both members
of undergraduate couples who were very much in love (Peplau, Rubin, & Hill,
1976) with stories written by other undergraduates, most of whom were not in-
volved in a relationship.
In searching for scoring categories to characterize the stories written under
affiliative arousal, McAdams was much influenced by the writings of Martin
Buber (1965, 1970) on the I-Thou relationship, by David Bakan (1966) on the
difference between agency and communion, by Maslow (1954) on the distinction
between being-love versus deprivation-love, and by Sullivan's (1953) general
treatment of interpersonal relationships. McAdams found that the scoring cate-
gories listed in Table 9.6 distinguished the stories written under all four arousal
conditions from stories written under control conditions. The differences were
significant in three out of the four comparisons and in the same direction for the
fourth comparison. As can readily be seen, these categories come much closer to
the characteristics described by experts in the field from Plato to the present.
Note in particular that the categories for scoring for the intimacy motive do not
describe aspects of a problem-solving sequence, as did the original n Affiliation
scoring scheme based on earlier work with n Achievement. Instead, the catego-
ries reflect more often states of being, that is, the quality of interactions rather
than aspects of getting something done, which is involved in an achievement ori-
entation. This conforms to theoretical expectations, since, as McAdams (1982a)
Table 9.6.
INTIMACY MOTIVATION SCORING CATEGORIES (after McAdams, 1980)
A relationship or interpersonal encounter produces positive ajfect (love, friendship,

happiness, peace, or tender behaviors).
Dialogue occurs, as in "swapping stories" or discussing a relationship.
Psychological growth and coping results from an interpersonal encounter.
A character feels a sense of commitment and concern for another not rooted in guilt or a
sense of duty.
Two or more characters are engaged in a relationship that transcends time and space.
There is a reference to the reunion of people who have been apart.
Characters find they are in harmony with each other, "on the same wavelength."
A character surrenders to a power beyond the seif Controlling an interpersonal
relationship, as in "they feil helplessly in love."
Characters escape into intimacy, to a Situation in which they can experience peace,
liberation, and so on, together.
There are references to togetherness with the natural world, as in "they love the way the
air feels against their skins."
points out, the active pursuit of intimacy "undermines the striven-for goal State,
rendering it less preferable, less intimate. As Buber so eloquently put it, The
Thou encounters one by grace—it cannot be found by seekingV
Stories written after intimacy arousal describe people enjoying being togeth-
er, being committed to each other in a relationship that transcends time and
space and in which they show a commitment and concern for each other. Fur-
thermore, they refer to interpersonal harmony just as Eryximachus or Condon
(1979) do, as well as to union with each other, with the natural world, or with a
power beyond the seif. All in all, this scoring System captures the essence of
what is meant by love or the positive affiliative motive much better than the
n Affiliation scoring system does.
Contrast Between the Affiliative and Intimacy Motives

The intimacy motive score obtained by summing the presence of such character-
istics in stories written under neutral conditions measures a disposition that dif-
fers from n Affiliation. The correlation between the n Affiliation score and the
intimacy motive score is only .32, p < .05, as reported in McAdams and Pow-
ers (1981).
Table 9.7 compares how people who score high in each motive are rated by
their peers (McAdams, 1980). As would be expected, those with a strong inti-
macy motive are judged to be more warm, sincere, appreciative, and loving and
Table 9.7.
CORRELATIONS BETWEEN MOTIVE SCORES AND ADJECTIVE RATINGS
BY PEERS (after McAdams, 1980)
Adjective Intimacy Motive n Affiliation

Natural .59*** .14
Warm .54*** .27
Sincere .50*** .12
Appreciative .37* .26
Loving .34* .21
Sensitive .27 .13
Honest .49*** .21
Calm .44*** .04
Dominant -.43** -.05
Outspoken -.41** -.02
Self-centered -.40** -.13
Imagination —.41** —.25
*p < .05.
**p < .01.
***p < .001.
less dominant and self-centered, whereas none of these characteristics is as

strongly associated with the n Affiliation score.
The picture is clarified still further by self-ratings of various characteristics,
as shown in Table 9.8. Young adults high both in the intimacy motive and
n Affiliation regard themselves as more gentle, natural, loyal, contented, and re-
alistic than subjects low in the motives. In contrast, only subjects high in
n Affiliation regard themselves more often as unselfish, cooperative, sociable, and
thoughtful. Notice that many of these characteristics involve doing things for
others. They imply that such people think they go out of their way to be nice to
others. Those who score high in the intimacy motive do not have this opinion of
themselves. In fact, as the right section in Table 9.8 shows, they have a rather
negative view of themselves as frivolous, tactless, touchy, and cowardly. The im-
plication is that they have higher Standards about their relationships with others
and therefore are more likely to regard themselves as having fallen down in
their empathy for others' feelings. In contrast, individuals high in n Affiliation
seem not so sensitive to the feelings of others.
Adjective peer ratings in a later study (McAdams & Powers, 1981) confirm
this impression. On the one hand, subjects scoring high in the intimacy motive
are considered to be significantly more nervous, whereas those high in n Affilia-
tion are nearly significantly less nervous. On the other hand, strong intimacy
motivation goes with being judged more sincere, loving, and cooperative, all of
Table 9.8.
CORRELATIONS BETWEEN n AFFILIATION AND INTIMACY MOTIVE SCORES AND
SELF-DESCRIPTIVE ADJECTIVES CHECKED BY YOUNG ADULTS (TV = 78)
Correlations Significanta
for
Both .Motives n Affiliation The Intimacy Motive
c c
rb r rb r rb
Realistic .39 .40 Unselfish .31 .05 Self-confident .11 .27

Contented .31 .24 Cooperative .24 .07 Foresighted .01 .29
Loyal .30 .22 Patient .27 .10 Frivolous -.01 .26
Gentle .25 .23 Sociable .37 .12 Tactless .12 .38
Natural .23 .27 Stahle .34 .13 Touchy .11 .28
Steady .30 .05 Cowardly -.03 .27
Thoughtful .30 .05
Warm .24 .11
Positive feelings about seif Reaching ouit to others Sensitive to the feelings of others
and relationships
a
r = .28 at p < .05, and r = .30 at p < .01.
b
Numbers in this column show correlations with whether adjective is checked and n Affiliation score.
c
Numbers in this column show correlations with whether adjective is checked and intimacy motive score.
which are characteristics less often attributed to those high in n Affiliation. In

this study the subjects high in n Affiliation were considered to be more enthusi-
astic and expressive in contrast to those high in the intimacy motive, to whom
these characteristics were not attributed. In a study by McAdams and Constan-
tian (1982) it was found that students high in n Affiliation who when beeped by
a pager were found to be alone more often expressed the wish to be with some-
body than subjects low in n Affiliation. The same contrast did not appear for
those low and high in the intimacy motive. Once again it appears that high
n Affiliation implies a more active approach to interpersonal relationships, which
goes along with the need for contact and reassurance.
Evidence That the Intimacy Score Measures a Motive

Do people who score high in intimacy engage in more energetic intimacy-related
actions? Two research assistants observed and coded various behaviors during
psychodrama sessions, and most sessions were also videotaped so that other
measures could be obtained later from these recordings (McAdams & Powers,
1981). As the findings summarized in Table 9.9 show, students who were higher
in either n Affiliation or the intimacy motive engaged in more affiliative behav-
iors. They stood closer to other people, made more "we" references in talking
about their scenario, produced more laughter from the group, and issued fewer
commands for others to do things. The differences in the effects of the two mo-
tives were most apparent in the themes chosen for the scenarios. Participants
high as compared with those low in the intimacy motive much more often chose
themes of positive interpersonal affect and surrender of control. These and sev-
eral other indicators in action of the characteristics of the intimacy motive scor-
ing system were summed to get a total intimacy behavior index score, which
correlated very highly with a person's intimacy score in written stories, but not
with the n Affiliation score in stories. Perhaps the best indication of the effect of
intimacy motivation on behavior involves touching. Even though some of the
data were lost due to failure in the videotaping machine, it is apparent that par-
ticipants with a strong concern for intimacy touched other people more often in
an affectionate, nonthreatening way.
To turn to the second type of behavior indicating the presence of a motive,
other studies have also confirmed the fact that individuals scoring high in inti-
macy get involved in deeper relationships with others. Fourth- and sixth-graders
with strong friendship motivation (a simplified version of the intimacy motive
Table 9.9.
CORRELATIONS BETWEEN MOTIVE SCORES AND BEHAVIOR IN PSYCHODRAMA
SCENARIOS: N = 43 (after McAdams & Powers, 1981)
Behavior in Psychodrama Scenarios Intimacy Motive n Affiliation

Discrete behaviors
Physical proximity of protagonist to nearest other .42** .35*
Number of "we" references per minute of introductory phase .39* .30
Number of outbursts of laughter per minute of scenario .32* .31*
Number of commands per minute of introductory phase -.31* -.26
Themes of scenarios
Positive affect .68*** .33*
Reciprocal dialogue .55*** .33*
Surrender of control .45** .12
Nonthreatening touchinga .31| .08
Total intimacy index .70*** .27
a
This theme was judged from videotapes available for only twenty-eight subjects and is not included in the total intimacy index.
t/7 < .10.
*p < .05.
**p < .01.
***p < .001.
score) had more detailed knowledge about their best friend and described the re-
lationship in terms of greater depth than boys and girls low in friendship moti-
vation. Their friendships were also more stable over time (McAdams & Losoff,
1982). When College students were asked to recall in some detail what went on
in a number of friendship episodes, those high in intimacy motivation, as con-
trasted with those low, more often described a relationship as involving listening
(rather than agentic striving), self-disclosure, and mutual trust (McAdams,
Healey, & Krause, 1982).
McAdams (1979) found that the intimacy motive sensitizes individuals to
people's faces. He exposed subjects to schematic drawings of faces originally em-
ployed by Brunswik and Reiter (1938) to show that judgments of personal char-
acteristics changed markedly with variations in the placement of the eyes, nose,
and mouth. McAdams found that subjects high in the intimacy motive showed a
much greater sensitivity to variations in placement of facial features than sub-
jects low in the intimacy motive. Sixty-six percent of the former, as contrasted
with 31 percent of the latter (p < .01), varied their adjectival characterizations
of the faces more than the average amount as the configurations changed. This
was true for intimacy-related adjectives like friendiy/'hostile and sincere/insin-
cere, but not for adjectives unrelated to intimacy like intelligent/stupid or
hon est /dishon est.
The result fits well with theories that face-to-face contact is an essential part
of intimate relationships. The quality of face-to-face contact between mothers
and infants is related to the quality of the mother-child attachment a year later
(Blehar, Lieberman, & Ainsworth, 1977). Among adults, the intimacy of a rela-
tionship is signaled by face-to-face or eye-to-eye contacts (Argyle & Cook,
1976). Some believe that "eye contact" serves as an innate releasing mechanism
(or natural incentive) for maternal care-giving responses (Brazelton, Koslowski,
& Main, 1974) or for interpersonal intimacy in general (Patterson, 1976). If
this is true, it is not surprising that those high in intimacy motivation are
particularly sensitive to variations in facial cues, even those of schematic
figures.
The intimacy motive also has a selective effect on memory. In one study, in-
dividuals with a strong intimacy motive recalled more autobiographical incidents
involving intimacy experiences (McAdams, 1982b). In another study, subjects
were asked to listen carefully to stories read to them about each of two pictures,
"because we will return to them later on" (McAdams & McClelland, 1983). The
story to the first picture centered on a group of old College friends enjoying a
twenty-five-year reunion in the woods, and it contained references to friendship,
dialogue, warmth, and interpersonal harmony. The story to the second picture
contained no intimacy themes and dealt with a man who travels from Utah to
Mexico City and eventually comes to own a pizza parlor. The two stories were
approximately equal in length, and each contained thirty-three facts that could
be recalled. After listening to the stories, the subjects engaged in other activities
lasting about one hour and then were asked to recall each of the stories. Inti-
macy motivation was highly correlated with the number of facts recalled from
the intimacy story (r = .56; p < .001) but not at all with number of facts re-
called from the neutral story (r = .07; not significant). The difference between
the correlations is highly significant.
In another, similar study, subjects were shown a Single picture on a screen
and read a lengthy story containing thirty power facts, thirty intimacy facts, and
eighteen neutral facts (McAdams & McClelland, 1983). Subjects whose intimacy
motive scores were higher than their n Power scores recalled significantly more
intimacy facts relative to the power facts they recalled. The reverse was true of
subjects higher in n Power. The intimacy motive, like other motives, exerts a se-
lective effect on memory.
Work on the intimacy motive has only recently begun, but two recent
studies suggest it may be a very important human characteristic. If the theoriz-
ing about love that has been going on for centuries is correct, those who are
more capable of love or who love and are loved more should be more mature,
healthier, and happier. McAdams (1982a) reports that those scoring higher in
intimacy motivation as a measure of the capacity to love and be loved also score
as more mature in the Loevinger Sentence Completion Measure of Ego Develop-
ment (1976) and as having reached a higher stage of commitment in their reli-
gious lives.
Furthermore, McAdams and Vaillant (1982) found that the intimacy motive
score was significantly related to the lifelong personal adjustment of a group of
male College graduates. The psychosocial adjustment of fifty-seven male gradu-
ates of a prestigious College was rated when they were about forty-seven years of
age. They had originally been tested and interviewed in College, approximately
thirty years earlier, and had been recontacted for information every five years
since. Thus, it was possible to base the overall rating of adjustment on continu-
ous information over their lifetimes on success in their work (as reflected in in-
come, promotions, and Job enjoyment), in nonvocational adjustment (as reflected
in regulär pastimes and enjoyment of vacations), in marriage (if sustained at
least fifteen years with both partners happy), and in avoiding trouble (as re-
flected in low frequency of psychiatric visits, low number of days of sick leave
used, and misuse of alcohol or drugs).
On one of the visits when the men were approximately ten years out of Col-
lege, they were administered a TAT, which was retrieved and scored for various
social motives. Those who scored high on the intimacy motive at age thirty were
rated as significantly better adjusted at age forty-seven than those at age forty-
seven who had scored low on the motive at age thirty. There was a trend in the
same direction for those scoring high versus low on n Achievement at age thir-
ty, but there were no differences approaching significance for those scoring high
versus low on n Affiliation or n Power. These findings strongly suggest that the
capacity to love, as reflected in the intimacy motive score, enabled these men to
live happier, better-adjusted lives in their work, play, and interpersonal relation-
ships. The case for such an inference would have been better if the motive
scores had been obtained in College, but as it is they represent a period in the
men's lives some seventeen years before the final ratings of adjustment were
made, suggesting that the intimacy motive may have helped promote the better
adjustment that occurred afterward.
THE AFFILIATIVE MOTIVES AND HEALTH
Chapter 8 reviewed a number of studies showing that the inhibited power mo-
tive syndrome (at least moderate n Power that was greater than n Affiliation,
along with high inhibition) is associated with sympathetic nervous System activa-
tion and poor health. For the most part, the opposite result—better health—is
associated with the reverse motivational syndrome (n Affiliation greater than
n Power and low inhibition, or the relaxed affiliative syndrome). Some evidence
already suggests that the affiliative motives promote better health. In a longitudi-
nal study of blood pressure, the n Affiliation score obtained when male College
graduates were in their early thirties was significantly negatively correlated with
diastolic blood pressure twenty years later (r = —.26; p < .05; McClelland,
1979a). Inspection of Table 8.10 shows that those with the relaxed affiliative
syndrome report the lowest average severity of illness score (47.4), which is sig-
nificantly different from the average severity of illness score for all other subjects
(94.2).
Similarly, male prisoners whose n Affiliation score was greater than their
n Power score and who also reported little stress had experienced less severe ill-
nesses than all other subjects (McClelland, Alexander, & Marks, 1982). Figure
8.11 reflects this result in the lower average severity of illness score for all other
subjects, for most of whom the n Affiliation score was higher than the n Power
score. Their better health was also reflected in the fact that their immune de-
fense System was in better order: they tended to have higher concentrations of
immunoglobulin A in saliva (S-IgA), the body's first line of defense against viral
infections, particularly those of the upper respiratory tract.
These same results appeared more clearly in another study (Jemmott, 1982),
the results of which are summarized in Figure 9.3. In this case, motive scores
were collected from first-year dental students in the summer before they began
their academic work. Concentrations of S-IgA and illness reports for the past six
weeks were also obtained at different points throughout the year at periods of
high and low academic stress. The students were under great pressure in No-
vember, April, and June, because they had to pass very exacting examinations in
order to continue their work. In general, concentrations of S-IgA declined for
all subjects significantly during periods of high academic stress and recovered at
the time of a period of low stress in July following the first year. The data for
the two most extreme contrasting motivational groups are also included in Fig-
ure 9.3. Notice that those with the relaxed affiliative syndrome (higher n Affilia-
tion than n Power, low inhibition) showed significantly higher concentrations of
S-IgA throughout the year, indicating that their immune System was functioning
better than those with the inhibited power motive syndrome, who had a lower
level of immune defense, which weakened throughout the academic year and
failed to recover afterward. As might be expected from this difference, subjects
0.25 -
\
0.24 - Relaxed Affiliative
0.23 - \ Syndrome
0.22 - (/V=15)
0.21 -
0.20-
0.19-
0.18-
0.17-
S E
0.16 —
c ^ 0.15 -
.2 5 0.14 —
0.13- Inhibited Power
0.12- Motive Syndrome
B
B
_L _L _L
Low: High: High: High: Low:
September November April June July
Degree of Stress
Figure 9.3.
Mean Salivary Immunoglobulin A Secretion Rate During High and Low Stress Periods for All
Subjects and for Two Motive Groups (after Jemmott, 1982).
with the relaxed affiliative syndrome reported a lesser increase in the number of
colds they experienced in the April-to-June period as compared with those with
the inhibited power motive syndrome.
There is other evidence that aroused affiliative states are associated with
signs of good health. In a study conducted by McClelland and Kirshnit (1983),
subjects' motive scores were obtained on one day (assessment day), along with
saliva samples to check on concentrations of S-IgA. A few days later (treatment
day), the subjects came to the laboratory in small groups to view films designed
to arouse different motivational states. One film, Triumph of the Axis, was a
documentary showing Hitler's early successes in World War n and his cruel
treatment of Jews. It was designed to arouse power motivation and had that ef-
fect in imaginative stories written afterward. As Figure 9.4 shows, it did not de-
press S-IgA concentrations overall. However, it did have that effect for those
who were high in the inhibited power motive syndrome to Start with.
The other film was a documentary on the life and work of Mother Teresa of
Calcutta, a nun who has devoted her life to caring for the poor, the sick, and
the dying in the slums of Calcutta. It was designed to arouse the affiliative mo-
tives and had that effect in stories written afterward. As Figure 9.4 also shows,
- 30 - Film on Mother Teresa
Triumph \
oft he
^1
Axis
20-
<
00
-
o
c
o
o
I
U
10 -
90 Minutes of
Intervening Tasks
1 1
Assessment Treatment
Day Day
Figur'e 9.4.
Effects of Two Films on Salivary Immunoglobulin A (after McClelland & Kirshnit, 1982).
it produced a significant increase in S-IgA concentrations immediately, suggest-

ing that Mother Teresa's capacity for loving had evoked a similar response in
the viewers, which had a beneficial effect on their body's defense against disease.
The finding needs further confirmation and clarification, but it is consistent with
the overall picture derived from a number of similar experiments: the affiliative
motives, or the capacity to love and be loved, are somehow associated with bet-
ter health, just as Eryximachus contended over 2000 years ago.
ORIGINS OF THE AFFILIATIVE MOTIVES

The longitudinal study that discovered child-rearing practices consistently associ-
ated with n Achievement and n Power scores did not succeed in finding any
such practices associated with higher adult n Affiliation or intimacy motive
scores that meet rigorous criteria for cross-validation (McClelland & Pilon,
1983). Only two clues to the origins of these motives turned up. The mother's
report when her child was five years old that she used praise more often as a
technique of socialization was significantly associated with the adult intimacy
motive score. Although the correlation was low (r = .26; p < .05), it was
fairly consistent across subdivisions of the sample (male or female, and middle
or lower class), and it is certainly consistent with theoretical expectations that
positive encouragement should develop the capacity for loving. In contrast, if
the mother reported at the same time that she was unresponsive to a child's cry-
ing, this practice was significantly associated with a higher n Affiliation score in
adulthood (/* = .27; p < .05). This was especially true for individuals from a
middle-class background and for girls, although the results were not significant
for boys or individuals from a lower-class background. Once again the finding,
as far as it goes, is consistent with the belief that high n Affiliation is associated
with high / Rejection, since it might be expected that babies whose mothers re-
fused to go to them when they were sufFering would be likely to grow up with a
strong unconscious need for reassurance or / Rejection. Such an interpretation is
also consistent with Youngleson's (1973) report that institutionalized children
who receive less mothering than noninstitutionalized children score higher in
n Affiliation, particularly its negative / Rejection aspect.
The fact that findings on the origins of the affiliative motives are so much
less definitive than those for n Achievement and n Power suggests either that
later experiences are more important for shaping these motives, or that the
Sears' group, which defined the dimensions of child rearing to be investigated,
was more "doing" than "being" oriented (Sears, Maccoby, & Levin, 1957). That
is, most of the over one hundred dimensions of parent behavior that they rated
dealt with something the parent did or failed to do to or for the child. On theo-
retical grounds we might suppose that it was more the relationship of the par-
ents to the children that would shape the affiliative motives than any particular
techniques they employed. Thus, further studies of good parent-child relation-
ships (see Sroufe & Waters, 1977) are needed to get a clearer picture of how the
capacity to love develops more in some people than in others.
• RELATIONSHIP OF SEXUALITY
TO AFFILIATION AND INTIMACY
There are many reasons for thinking on theoretical grounds that the develop-
ment of the sexual motive should relate to the development of the affiliative and
intimacy motives. The reasons are based on information as diverse as that ob-
tained from case studies of individuals with sexual dysfunctions and those ob-
tained from Harlow's studies of the inabilities to function sexually of monkeys
that have been deprived of the early rhythmic, caressing body contact character-
istic of normal mother-infant relationships (see Chapter 5). However, empirical
studies that confirm these relationships using the methods described in this
chapter have not yet been made.
Researchers working with the affiliation and intimacy motive scores have
not been interested in examining their relationship to sexual experience or sexual

behavior. The dosest they have come is the finding in a national sample survey
that the reported frequency of a husband's affectionate interaction with his wife
(which suggests sexual activity) is not related to his n Affiliation score (Veroff,
Depner, Kulka, & Douvan, 1980). Nor is a mother's treatment of childhood
sexuality related to adult n Affiliation or intimacy scores (McClelland & Pilon,
1983), although punishment for masturbation and sex play might be expected to
inhibit the development of the intimacy motive. Perhaps the sexual motive is
quite independent of the other two despite their supposed theoretical relation-
ship. However, these findings are by no means definitive. They only call atten-
tion more sharply to the fact that more work needs to be done on the origins,
development, and interrelationship of the sexual and affiliative motives.
NOTES AND QUERIES
1. Freud feit that the libido (or sex drive) underwent various stages in its de-
velopment, which could be characterized by attachments at the oral, anal,
phallic, and genital levels. In Chapter 8 we reviewed the Stewart System for
coding the extent to which adults are oriented around one or another of
these stages. This chapter suggests that testosterone might indicate the ex-
tent of sexual motive arousal in males. Would it be a reasonable test of
Freud's theory, then, to examine the amount of testosterone produced in
males in response to cues characteristic of the stage toward which they are
most strongly oriented, according to the Stewart coding System? For exam-
ple, would you expect that Stage I-oriented males would be more "turned
on," as measured in this way, by power-oriented reading than, say, Stage II
males? Or would such a difference appear only for more sexually explicit
Stimuli, showing, for example, that Stage I males are more turned on by
oral sex than Stage II males? What would it mean if there were no differ-
ences in testosterone produced by various cues in adults oriented toward
one stage or another? Would Freud's theory have been disproved?
2. What does it mean that boys who are late in reaching sexual maturity have
a higher n Affiliation? Does it establish a connection between sexuality and
love, or not? It could mean only that late-maturing boys feel left out and
concerned about establishing interpersonal relationships. May also reports
that young men with a strong E-D pattern also feel left out. How might
these facts be connected?
3. Most people can report when they feel they have gotten along well with
somebody. They may say there were "good vibes" in the relationship, mean-
ing much the same as Condon does when he observes harmonious reciprocal
voice and movement interactions. Harlow's work with real and substitute
mothers discussed in Chapter 5 shows that the ability to establish such har-
monious relationships in adulthood depends on good mother-infant interac-
tions. If the affiliative or intimacy motives are built on the pleasure from
harmonious interactions, would it not mean that there should be a correla-
tion between good mother-child interaction—as measured by Sroufe and
Waters (1977), for example—and adult n Affiliation or intimacy motive
scores? Speculate on what aspect of the mother-child interaction would be
more likely to lead to strengthening the affiliative versus the intimacy
motive.
4. Why do you think it has been difficult to get consistent results showing that
n Affiliation is related to cooperation or conformity behavior? What other
factors should influence whether such behavior occurs? Try to design an ex-
periment taking these other factors into account that would show that
n Affiliation is related to cooperation and conformity.
5. If moderate risk taking characterizes people high in n Affiliation under cer-
tain conditions, as reported in this chapter, how can it be concluded, as sug-
gested in Chapter 7, that moderate risk taking characterizes people high in
n Achievement? Does this suggest a reason why types of behavior are sel-
dom good indicators of the motives behind them? Or can you think of be-
haviors that might pretty uniquely indicate the presence of a motive like the
achievement or affiliative motives?
6. If appeal to different types of incentives can produce more approach behav-
ior from people strong in different motives, does it follow that teachers, for
instance, can produce better Performance from all their students simply by
employing a wide variety of incentives? Or does an appeal to one type of in-
centive facilitate Performance for people with one type of motive and inhibit
it for others? Review McKeachie's (1961) work mentioned in this and the
previous chapter.
7. Chapter 8 stressed the fact that motives show alternative manifestations.
Apply this concept to the affiliative motive. For example, a person strong in
n Affiliation might have many friends, spend a good deal of time with a few
friends, or telephone or write letters often. If these were genuine alternative
expressions of n Affiliation, how would we get a composite measure that
would demonstrate this fact?
8. This chapter and the previous one have asserted that managers high in
n Affiliation would respond more to individual needs and have a harder
time sticking to universalistic rules, but this hypothesis has never been em-
pirically tested. Design an experiment that would test it.
9. Explain the fact reported by McAdams and Constantian (1982) that subjects
high in n Affiliation but not in intimacy motivation more often expressed a
wish to be with somebody when they were alone.
10. In examining the reports students gave of what they were thinking and
doing when they were paged at random intervals during a week, it was
found that women reported having interpersonal thoughts nearly twice as
often as men, yet the two sexes did not differ in average levels of intimacy
motivation and n Affiliation (McAdams & Constantian, 1982). What could

this mean? Which would be the better measure of motive strength—the
TAT or the pager reports? What other variables might influence either
measure?
11. Meditation has been associated with relaxation (see Benson, 1975) and with
an increase in n Affiliation (see Alexander, 1982). Would you expect medita-
tion to be connected with an improvement in health? Look for evidence to
back up your conclusion.
12. If love and loving are associated with health, would you expect loving cou-
ples to be sick less often? How would you test such an hypothesis?
13. An undergraduate once carried out a study in which he found that brown-
eyed College males were more sexually aroused (that is, told stories with
more sexual content) by a picture of a beautiful blue-eyed woman than by a
picture of an equally beautiful brown-eyed woman, and vice versa: blue-eyed
males were more aroused by the picture of a brown-eyed woman. If this
finding were confirmed, how would you explain it? Does it have any impli-
cations for the view that eyes have a Special significance in arousing the af-
filiative motives?
14. Chapters 4 and 5 argued that motives are based on positive and negative
emotions of various types. Mc Adams and Constantian (1982) found that
subjects with a strong intimacy motive reported more positive afFect when
interacting with someone than subjects with a weak intimacy motive. No
difference in affect was reported between the two motive groups in nonin-
teracting situations. Does this confirm the hypothesized relationship between
motive and affect, or could the relationship found be explained in some
other way?
15. In commenting on Wilsnack's study of the effects of drinking on women's
thought content, Wilson and Lawson (1976b) State: "However, it is unclear
what the sexual content of these dubious projective tests means without vali-
dating the tests against appropriate objective measures." In what sense is
thought content "dubious"? Their proposed appropriate objective measure of
sexual arousal is a vaginal photoplethysmograph, which, when inserted,
measures vaginal pulse pressure and blood volume. What if the two mea-
sures of sexual arousal do not agree? Can it be assumed that the physiologi-
cal measure is always a better measure against which the thought measure
must be validated? Think of what can happen during rape, for example.
10
The Avoidance Motives
• Generalized Anxiety as a Motive

Self-report Measures and Their Correlates in Behavior
The Incentive Value of Being with Others as a Means of Reducing Anxiety
Measuring the Need for Security in Fantasy
The Effect on Behavior of Maslow's Needs for Safety and Self-esteem
• Fear of Failure
Critique of the Test Anxiety Measure of Fear of Failure
• Measuring Fear of Failure in Fantasy
The Heckhausen, German Measure
The U.S. Measure
Characteristics of People with a Strong Fear of Failure
• Comparison of Measures of Fear of Failure

• Origins of the Fear of Failure
• Fear of Rejection
Relationship Between Fear of Failure, Fear of Rejection, and the Approval Motive
• Fear of Success
Measuring Women's Fear of Success in Fantasy
How Fear of Success Affects Competitive Behavior in Women
Fear of Success Among Black Men and Women
• Fear of Power
Measuring Fear of Power in Fantasy
The Association of Fear of Power with Atypical Role Behavior in Men
Fear of Power Among Women
• Other Fears
Correlates of VerofTs Need for Power Score as a Measure of Fear of Weakness
Fear of Intimacy
• Conclusion
373
E . or historical reasons, as explained in Chapter 3, academic psychologists first
conceived of motives as eiforts to avoid discomfort and to reduce strong Stimula-
tion, whether caused by hunger, thirst, pain, electric shock, conflict, or frustra-
tion. They found it easy to observe the effects of such strong Stimulation on
their favorite subjects—namely, animals in the laboratory—and they believed
that psychoanalysts working with patients had confirmed their view that anxiety
reduction was a kind of master motive. From this point of view it did not make
much sense to think in terms of different kinds of avoidance motives: Individu-
ais simply learned different ways of reducing their anxiety, and these might be
as varied as the number of people studied. Thus, one person might reduce his or
her anxiety by chewing gum, another by jogging, and a third by going to the
movies.
It did not make sense to try to define and measure a gum-chewing motive, a
jogging motive, or a movie-going motive, so not much attention has been given
to sophisticated measurement in this area such as has characterized the work on
measuring n Achievement or n Power. In the beginning, scholars were satisfied
with simple reports by individuals as to how anxious they feit. Evidence of
the existence of other types of more specific avoidance motives began to crop
up in the study of approach motives, as noted in the past three chapters. All
of this work will be reviewed, but it does not as yet add up to a very clear
or coherent picture of what the avoidance motives are and how they should be
measured.
GENERALIZED ANXIETY AS A MOTIVE

Many questionnaires have been developed that measure the extent to which peo-
ple say they feel nervous, anxious, or worried. They include items like "Do you
consider yourself rather a nervous person?" or "Do you worry over possible
misfortunes?" or "Do you get nervous in places such as lifts, trains, or tunnels?"
These items are from Eysenck's Measure of Neuroticism (1957a), an instrument
widely used in Great Britain. In the United States the Taylor Scale of Manifest
Anxiety (1953) has been the instrument of choice, as noted in Chapter 3. It asks
the subject to respond to items like "Every few nights I have anxiety dreams"
or "I am easily upset."
Spielberger, Gorsuch, and Lushene (1970) simplified the measure even fur-
ther by asking subjects to indicate the extent to which they feit nervous, irrita-
ble, or anxious versus calm and relaxed. By rating adjective descriptions of such
characteristics they indicated how they generally feit—which Spielberger called a
measure of trait anxiety—or how they feit at the particular moment—which he
called a measure of State anxiety. He reasons that "anxiety as a personality trait
(A-trait) would seem to imply a motive or acquired behavioral disposition that
predisposes an individual to perceive a wide ränge of objectively nondangerous
circumstances as threatening, and to respond to these with A-state reactions dis-
proportionate in intensity to the magnitude of the objective danger" (Spielberger,
1966). The distinction is the same as the one made in Chapter 6 between a mo-
374
The Avoidance Motives 375
tive disposition, considered to be a relatively stable aspect of personality, as

compared with aroused motivation, a temporary aspect of personality (Zucker-
man, 1976). All these instruments, as well as several others, measure the same
disposition—the willingness of the person to admit to anxiety in its various
forms. They have the same behavioral correlates and correlate highly with each
other (see Feij, 1975; Shiomi, 1978). Those who score high on one tend also to
score high on the others.
There is also a substantial body of knowledge showing that individuals who
score high in Manifest Anxiety display the characteristics that are an essential
feature of being more motivated. They are more activated or energized. Particu-
larly when stressed, they show various signs of greater autonomic activity such
as pupillary dilation, increased heart rate, increased sweating as picked up in
the galvanic skin response, faster breathing, more oxygen uptake, and so on
(Eysenck, 1947). However, self-reports of anxiety and physiological signs of
anxiety are by no means perfectly correlated. Some people who show all the
physiological signs of anxiety do not admit to feeling anxious or nervous (see,
for example, Salter, Meunier, & Triplett, 1976). So once again we run into the
problem of response bias characteristic of all self-reports of motivation.
More anxious than nonanxious people also show the selective sensitivity
characteristic of greater motivation. They report Stimulation increasing in inten-
sity to be painful sooner than subjects low in anxiety do (Shiomi, 1978), and
they do not see as well in the dark (Eysenck, 1947).
Finally, they display the third essential feature of being more motivated:
They learn some types of material faster under certain conditions than people
low in anxiety. (See, for example, Figure 3.7.) In general, anxious people are
better at learning easy material and poorer at learning more difficult material
(see Weitzner, 1965). The explanation of this fact in terms of Hullian theory
was that drive (in this case represented by a score on the Manifest Anxiety
Scale) was kind of a blind, nondirective force that served to activate whatever
responses were already strongest in the Situation. If the strongest response hap-
pened to favor learning, as in simple tasks, higher drive would facilitate learn-
ing, but if the naturally stronger responses were incompatible with doing the
task, higher drive would interfere with learning.
However, there were several crucial objections to such an interpretation. To
begin with, Eysenck (1957a) reported that increasing another irrelevant drive
(namely, hunger) during conditioning did not facilitate acquiring the eyelid re-
sponse the way higher anxiety did. How could this be true if any increase in
drive blindly activated the strongest natural response in the Situation, which had
been demonstrated by the anxiety research to be the eye blink? Also, as Chapter
3 pointed out, Weiner (1966) demonstrated that the reason more anxious sub-
jects did more poorly on difficult tasks was because they failed more often on
them.
This led to a new and much simpler interpretation of the learning results.
Anxiety is not a blind force, but a directive one. In comparison to nonanxious
individuals, more anxious individuals will learn more quickly to do whatever
will reduce their anxiety. In the conditioned eye blink experiment, an air puff to
the eyeball is unpleasant, and more anxious subjects learn more quickly to antic-
ipate the puff and to blink their eyes in order to avoid the unpleasantness (Tay-
lor, 1951). In simple learning tasks, more anxious subjects experience much suc-
cess and continue to learn more quickly in order to reduce the anxiety that
would come from failure. In more difficult tasks, they experience more failure and
respond by trying to avoid doing the task that results in poorer Performance.
If the goal of the anxiety motive is some kind of escape, measurement in
this area is deficient, because it concentrates on the anxiety responses rather
than on what relieves them. A motive is better defined in terms of its objective,
or the incentive that it seeks or that would satisfy it. There is some evidence
that what more anxious people seek is some kind of security, or the reassurance
that comes from being told what to do or knowing what to do. For example, if
they are talking to someone who nods or says "Mhm-hm" every time they say a
plural noun, they begin quite unconsciously to say more plural nouns (Doherty
& Walker, 1966). Or if they are Standing still with their eyes closed and the ex-
perimenter suggests strongly that they are swaying forward, they are more likely
to sway further in response to the Suggestion than subjects low in anxiety
(Eysenck, 1947).
The Incentive Value of Being with Others as a Means of Reducing Anxiety

Schachter (1959) carried out a series of experiments that illustrates dramatically
the incentive value of being with others for anxious people. He aroused State
anxiety by having a number of female College students participate in an experi-
ment concerned with the effects of electric shock. In one condition the women
were told by a very formal, authoritative experimenter in a white coat that the
shocks would really hurt but would do no permanent damage. In the other con-
dition the women were told by an informal experimenter that the shocks would
be very mild and would only tickle or tingle. In both conditions the subjects
were then told that they had to wait awhile before the experiment began and
they could either wait in separate rooms alone or with others.
Schachter found that those threatened with severe shock were much more
likely to choose to spend the waiting time with others than those who expected
only a mild shock. In other words, anxiety increased the incentive value of being
with others. He also tried in a variety of ways to find out exactly what was reas-
suring about being with others and concluded that the subjects wanted to be
with others to talk over the Situation and reduce their anxiety and uncertainty
as to how they should react. He did not include a trait measure of anxiety, but
presumably if he had, those higher in trait anxiety would have wanted even
more to be with others. He did, however, locate an indicator of individual dif-
ferences in anxiety levels. Firstborn or only children chose more often to be
with someone than others did. He reasoned that firstborn or only children are
under more pressure from their parents, tend to be more anxious, and tend to
have been more often and more quickly rewarded by the presence of another
when they cried. For when there are more children in the family, the mother
cannot as readily respond to a child's trouble by going to him or her immediate-
ly. All the evidence points to some kind of need for security, reassurance, or re-
duction of uncertainty as being characteristic of people with high anxiety levels.
Measuring the Need for Security in Fantasy

Since self-report measures of motives are generally weak, and since the measure
of the anxiety motive is particularly weak because it does not mention its incen-
tive, it would seem desirable to develop better measures using the more sophisti-
cated methods of measurement outlined in Chapter 6 and used in other areas.
Two such attempts have been made, although neither has been developed very
far. As has been described, Walker and Atkinson (1958) had soldiers write sto-
ries to pictures under different levels of aroused anxiety created by their near-
ness to the explosion of an atomic bomb. In the control condition men wrote
stories at a remote desert camp before they knew they were to participate in
atomic maneuvers. Another group wrote stories at the camp after they had been
briefed fully on the fact that an atomic explosion was to take place in ten hours'
time. The third group wrote stories just after the bomb had gone off, while the
atomic cloud was boiling up directly overhead and while they were in a State of
"acute alertness for the wail of a siren" that would announce the declaration of
an emergency, meaning that everyone had to leave as soon as possible. Other
men wrote stories at times farther removed from the explosion.
A scoring System for the stories was developed based on previous work with
the n Achievement score. The researchers first identified whether a story was
fear or anxiety related in any way, as indicated by the presence of threat imag-
ery, which was scored when there was mention of threat of bombing, physical
härm, or attack, as well as the emotional response of fear. Table 10.1 illustrates
some of the scoring categories they developed to cover various aspects of the
anxiety-reduction action sequence. Note that the measure in this case is oriented
around relief of fear—wanting to remove threat, doing something to remove it,
overcoming obstacles to avoid threat, or experiencing relief from threat. In this
sense it is a more adequate measure than the Manifest Anxiety Scale, which fo-
cuses exclusively on the threat or anxiety itself.
Furthermore, an overall need for Security (n Security) score obtained by
summing these characteristics showed validity, as Figure 10.1 illustrates, for it
increased in strength as the bomb threat became more salient and decreased as
it receded in time.
In addition, this measure should not be subject to the response biases in-
volved in self-reports of anxiety. For example, Epstein (1962) has shown that re-
ports of anxiety vary considerably with the experience of a potentially threaten-
ing event like sport parachute jumping. The experienced jumpers report less and
less anxiety but do not show a correspondingly regulär decrease in physiological
signs of anxiety. In other words, their bodies continue to show signs of anxiety,
although subjectively they report very little anxiety. Despite the promise of the
Walker and Atkinson measure, it has been used only in the one study.
Table 10.1.
SAMPLE SUBCATEGORIES SCORED ONLY IF THE STORY CONTAINS THREAT IMAGERY
(after Walker & Atkinson, 1958)
Category Brief Definition Example

Need (N) Someone wants something that, if "They want shelter from the bomb.'
attained, would remove the threat of
physical härm.
Instrumental Activity (I) Someone performs or thinks of "They are running away from the
performing an act, the function of fire."
which is to reduce the threat of
physical härm. If the outcome of the
act is successful, score I-f; if
unsuccessful, score I — ; and if
doubtful, score I?
Nurturant Press (NuP) Forces from the environment provide "First aid will be given."
relief.
The Effect on Behavior of M aslow's Needs for Safety and Self-esteem

A somewhat more widely used measure of n Security was developed by Aronoff
(1967). He employed a sentence completion test to measure Maslow's hierarchy
of needs (see Chapter 2), for he was convinced that two groups of people he was
studying on the island of Saint Kitts in the British West Indies differed in the
extent to which their basic needs had been satisfied. Maslow's theory argues that
basic needs for food, drink, safety, security must be satisfied before higher order
needs for love and self-esteem can become salient.
Aronoff observed that most of the men on the island cut sugar cane for a
living, and very few become fishermen, despite the fact that there is an excellent
market for fish and those who fish make a better living. He thought an explana-
tion for this fact might be that n Security on the island was great and was best
satisfied by cutting cane rather than by fishing. Certainly working in a cane-
cutting gang provides more security than fishing. The men work down the rows
in a field like a wedge driven in deeper at one end. They work at a pace set by
the head cutter and cannot get ahead or behind the person on either side of
them. They need not assume responsibilities for their efforts. They are paid for
what the whole group accomplishes, so if they are tired or ill, they will benefit
nevertheless from what the other men have cut that day. Furthermore, they get
a lot of individual support from each other, as they continually talk and make
jokes as they work. There is "little room for individual achievement. It is impos-
sible for a person to decide, by himself, that on a given day or week he will cut
more cane, because the pattern of cutting arranges and regulates the speed at
which each man works. . . . Even if an individual cutter could cut more on a
Remote 10 Hours Within 10 Hours 2 Weeks

Army Camp Before 30 Minutes After After
(Control) Atomic After Atomic Atomic
Explosion Atomic Explosion Explosion
Explosion
Source of Stories
Figure 10.1.
Percentage of Soldiers Scoring High in n Security Before, During, and After Exposure to an Atomic
Explosion (Form B Results) (after Walker & Atkinson, 1958).
given day, the System of payment is such that he could not benefit exclusively,
but would only be augmenting the income of his slower followers" (Aronoff,
1967).
In contrast, the fisherman's income is totally dependent on how much effort
he puts out. He must decide how often he goes out, where he fishes, how much
he wants to invest in traps, and so on. Aronoff reasoned that men would engage
in such activity only if their basic security needs had been met so that, in Mas-
low's terms, they were now free to pursue needs higher up on the hierarchy,
such as the need for greater personal self-esteem. The Sentence Completion Test
Aronoff used to examine whether this line of reasoning was correct included
cues like "Money is . . ." or "A wife . . ." or "I am sad because I . . . " Safety
or security needs were scored whenever the subject mentioned reliance on exter-
nal authority, the need for care, reciprocity, or safety gratification. Examples of
completions in this category are "A wife . . . take care of me" or "A friend . . .
if I need anything he does help me." Completions scored for self-esteem include
such Statements as the following: "I am good at . . . one and everything" or "I
want . . . to live for some purpose before time has expired" or "A wife . . . is
the second in the home." Such an interest in who or what is better or worse is
very reminiscent of what is scored for n Achievement.
Aronoffs results, as summarized in Figure 10.2, confirm his expectations.
A much higher proportion of the cane cutters showed a high need for Safety
(n Safety) or n Security than the fishermen, who scored higher in Self-esteem.
Furthermore, he found some evidence that might explain why the cane cutters
had a greater n Security. He recorded the number of men in each occupational
group who had experienced the death of a parent by the age of twelve, the
death of a large number of siblings in the family while they were young, or
both. He reasoned that those who had experienced a higher number of deaths in
their family while they were young would be likely to have a stronger n Safety.
As expected, more cane cutters than fishermen had experienced a severe loss of
close family members in childhood.
Aronoff s measure has been used in other studies as well. For example,
[ H Cane Cutters
J Fishermen
100
80
I
o
c
60
40
20
Deaths in Safety Positive

Family Needs Self-Esteem
(Early Loss)
Figure 10.2.
Percentages of Cane Cutters and Fishermen Scoring High in Early Loss, Safety Needs, and
Self-esteem (after Aronoff, 1967).
Ward and Wilson (1980) presented to women undergraduates a moral dilemma

as if it were a legal case for discussion. The subjects were instructed to read a
case and to try to reach a verdict about what to do as if they were members of
a Jury. They then wrote a two-page essay summarizing the reasons for their par-
ticular verdict. Their reasoning could then be coded according to the level of its
moral maturity using criteria established by Kohlberg (1969).
Some of the subjects carried through this procedure alone. Others were sub-
jected to group pressure in the following way: After reading the case and stating
their opinion, they discussed it with two other women who were confederates of
the experimenter and who had been instructed to use the arguments and ratio-
nale characteristic of a lower stage of moral reasoning. Thus, each subject was
exposed for twenty minutes to strong expressions of opinion and arguments de-
signed to lower the level of their moral reasoning. Then they wrote their final
essay giving the reasons for their verdict, just as the subjects did who had gone
through the whole process alone.
The subjects had taken the sentence completion test, so their safety and self-
esteem needs could be assessed, as in the previous study. As Figure 10.3 makes
clear, the women's safety and self-esteem needs interact with group pressure in
different ways. Those who score high in n Safety lower their level of moral rea-
soning in response to group pressure, whereas those high in Self-esteem do not.
It is clear that those motivated by high n Security rely more on others for sup-
port and guidance in what they do, just as those who score high on general anx-
iety scales do.
• FEAR OF FAILURE

Much more attention has been given in the motivation literature to a more
specific type of anxiety—namely, people's anxiety over whether they are per-
forming well. In his general analysis of motivational determinants of risk taking,
Atkinson (1957) conceived of an incentive to avoid failure as the mirror image
of the incentive to approach and achieve. So his formula for the motivation
to avoid failure paralleled the formula for the motivation to achieve. That is,
he conceived of the tendency to avoid failure as a product of the motive to
avoid failure times the probability of failure times the negative incentive value of
failure, which was conceived to grow larger in direct proportion to how easy the
task was at which the person failed. That is, people should be much more un-
happy if they failed at easy tasks than at difficult tasks. So the negative incentive
value of failing at an easy task should be much greater than for failing at a hard
task.
The way the formula works is illustrated in Table 10.2, which is constructed
to parallel exactly Table 7.2, which shows how the approach aspect of the
achievement motive operates. Notice that for a hard task, where the probability
of failure is high, the negative incentive value is relatively slight, so the product
Safety Oriented Esteem Oriented
Figure 10.3.
Average Moral Maturity of Reasoning Among Safety- and Esteem-oriented Women When Alone
and When Subjected to Group Pressure (data from Ward & Wilson, 1980).
of the variables indicates only a slight avoidance tendency. The two variables
have opposite values for an easy task, again yielding a slight avoidance tenden-
cy. However, maximum avoidance should occur for tasks of moderate difficulty,
just as they should yield the maximum approach tendency for those attracted by
the incentive value of success in the parallel case.
To check the applicability of his model, Atkinson needed some way of mea-
suring Fear of Failure (f Failure). For this purpose he employed the Test Anxi-
ety Questionnaire developed by Sarason and Mandler (1952). It was made up of
forty-two items like the following: "Before taking an individual intelligence test,
to what extent are you (or would you be) aware of an 'uneasy feeling'?" or
"When you are taking a course examination, to what extent do you feel that
your emotional reactions interfere with or lower your Performance?" Scores on
this test correlate moderately highly with scores on the Taylor Manifest Anxiety
Scale (r = .53; Raphelson, 1957) and indicate more specifically how fearful the
person is about performing in test situations. Atkinson decided that the purest
index of/Failure would be a high Test Anxiety score and low n Achievement,
indicating no approach tendencies toward Performance. In contrast, the purest
Table 10.2.
AROUSED MOTIVATION TO AVOID FAILURE AS A JOINT FUNCTION OF
MOTIVATION (Maf\ EXPECTANCY (/}), AND INCENTIVE (If OR -Ps)
(after Atkinson, 1957)
Maf X Pf X If(or -PJ = A voidance

Task A (hard) 1 X .90 X -.10 -.09
Task B 1 X .50 X -.50 = -.25
Task C (easy) 1 X .10 X -.90 = -.09
index of an approach- or success-oriented achievement motive would be a high

n Achievement score and a low Test Anxiety score. So he and many others
working in this tradition have regularly classified subjects into four groups, de-
pending on whether they scored high or low on each of these measures.
Our attention in this chapter will be focused primarily on those considered
high in / Failure because they score high in Test Anxiety and low in n Achieve-
ment. As one would expect if the measure were adequate, such subjects showed
signs of greater physiological arousal when they were performing a difficult eye-
hand coordination task than subjects performing the sam'e tasks who were high
in n Achievement and low in Test Anxiety. That is, the more anxious subjects
perspired more during the task, as shown by decreased resistance of the skin to
the passage of an electrical current (Raphelson, 1957).
A key question is whether subjects considered high in/Failure as measured
this way will avoid moderately difficult tasks most, as they ought to according
to the Atkinson model. To put it the other way around, would they be most at-
tracted to very easy or very difficult tasks? Obviously, according to Table 10.2,
they should avoid all tasks somewhat, but it was further assumed that they
would often have to choose among possibilities for extrinsic reasons, and in that
case their lesser avoidance of extremes in probability of failure should show up
as a positive preference for those extremes.
Several studies supported this expectation. Consider, for example, the evi-
dence summarized in Table 10.3 that subjects high in/Failure were much more
unrealistic than those low in / Failure in choosing the occupations they wished
to enter. College students were asked to estimate their own ability, as well as
the ability they thought was needed for various occupations (Mahone, 1960).
When their own occupational choices were examined, it was discovered that
those high in / Failure much more often had estimated the ability needed for
the occupation to be much higher than their own ability. In terms of the model
in Table 10.2, they were expressing a preference for an occupation in which, in
their own terms, their probability of failure would be quite high, since they did
not have the ability for it. In contrast, as pointed out in Chapter 7, those high
in n Achievement and low in Test Anxiety more often picked occupations for
which there was little discrepancy between the ability they perceived they had
and the ability that was required for the occupation. Furthermore, those high in
/Failure were more often inaccurate in estimating their own ability (as com-
pared with actual academic Performance), and they even chose occupations that
did not realistically fit in with interests they said they had. All in all, it looked
as if they were avoiding realistic choices because they had more negative valence
for them, just as the Atkinson model would predict.
Feather (1961) obtained results on persistence after failure that fit the
model. The results have been included in Figure 7.6. The subjects classified as
low in n Achievement in this figure are also high in Test Anxiety, meaning in
Atkinson's terms that they are high in / Failure. Notice that if the task is con-
sidered very difücult and they fail, they are much more likely to persist in work-
ing on it than if the task is considered easy and they fail. The reason in terms of
the model in Table 10.2 is that if they fail at Task C (an easy task), the proba-
bility of failure subjectively increases for them, and it has an avoidance value
more like that for Task B, which yields the maximum avoidance. So they would
be less likely to keep working at the task. In contrast, if they fail at a very diffi-
Table 10.3.
FEAR OF FAILURE RELATED TO LACK OF REALISM IN VOCATIONAL ASPIRATION
(data from Mahone, 1960)
Percentage of Subjects in Each Category

High f Failure Low f Failure
Occupational Choice High Test Anxiety, Low Test Anxiety,
Characteristics Low n Achievement High n Achievement
Positive goal discrepancya
High 61 31
Low 39 69
X2 = 5.69; p < .01.
Inaccuracy in estimate of own ability

High 67 44
Low 33 56
X2 = 3.52; p < .04.
Realism of choice in terms of interest

Realistic 32 56
Unrealistic 68 44
X2 = 3.50; p < .04.
a
The diiference between the subject's own estimated ability and perception of the ability needed for the occupation chosen.
cult task like Task A in Table 10.2, it increases the probability of failure but
moves the avoidance tendency even farther from Task B, which has the maxi-
mum avoidance value for them. So they would be more likely to persist after
failure on a difficult than an easy task, according to the model.
However, in the most direct test of the model, subjects high in / Failure
generally continue to show some preference for moderately difficult tasks, even
though considerably less than those high in n Achievement and low in Test
Anxiety. For example, in the classic study by Atkinson and Litwin (1960) on
the distances from the peg at which subjects choose to stand in a ringtoss game
(see Table 7.3), about 43 percent of the choices made by those high in /Failure
(high in Test Anxiety, low in n Achievement) are for moderate distances, as
contrasted with 21 percent for very close or easy distances and 36 percent for
very far or difficult distances. This would seem to be a direct contradiction of
the model, which predicts that such people should above all avoid the middle
distances. The explanation given for such a discrepancy is either that the sub-
jects were not extreme enough in their/Failure, or that extrinsic motivational
factors have influenced their choices. Another possibility is that the measure of
/Failure employed is defective.
So far as actual Performance is concerned, subjects high in Test Anxiety act
in ways that suggest they are afraid of failure and are seeking to do whatever
will reduce the experience of failure. As contrasted with those low in Test Anxi-
ety, they typically do more poorly on laboratory tasks such as digit-symbol Sub-
stitution (Mandler & Sarason, 1952), but occasionally they perform better if the
threat of immediate negative evaluation is remote. Thus, they do less well on ap-
titude tests but tend to get better grades in College (Mandler & Sarason, 1952).
The explanation apparently is that in doing course work they work harder to
avoid failure but are not disrupted by the threat of immediate failure, as they
are in a timed test Situation.
Whether they think they are likely to be negatively judged is very important
to subjects high in/Failure, as the data summarized in Table 10.4 show. In this
experiment the subjects were shown a word every two seconds and were asked
to learn to anticipate and say what word was Coming up before it showed in the
window in front of them. In the control, or neutral, condition, they were simply
told how to perform the task. With no particular understanding as to what the
task was all about, the subjects high and low in Test Anxiety did not perform
differently. However, if the subjects were told that Performance on the test
would yield a measure of their intelligence, the subjects high in Test Anxiety
performed much more poorly than those low in Test Anxiety. In contrast, if
they were reassured in advance that "these kinds of lists are hard and so it's no
surprise or matter of concern if you progress slowly at first and make mistakes,"
they performed actually somewhat better than those low in Test Anxiety.
Furthermore, Weiner has shown (as reported in Table 3.2) that subjects
high in / Failure perform better if they believe they are succeeding. Thus, if they
are "ahead of the game" they will perform better, because that is the best way
to avoid failure, but if they are behind they do worse, because that seems the
best way to get out of the failure Situation. The conclusion is similar to the one
Table 10.4.
NUMBER OF CORRECT RESPONSES FOR FINAL TRIAL BLOCK ON
SERIAL LEARNING TASK (after Sarason, 1971)
Difference
Conditions High Test Anxiety (H) Low Test Anxiety (L) (H — L)
Control (neutral) 47.8 46.7 1.1

Achievement orientation 34.1 65.1 -31.0**
Reassurance 58.8 42.3 16.5*
*p < .05.
**P < .01.
reached about the efFect on Performance of a more generalized type of anxiety,

as measured in the Taylor Manifest Anxiety Scale. Individuais will do whatever
seems to them to be the best way of avoiding anxiety in a Situation.
Critique of the Test Anxiety Measure of Fear of Failure

A number of objections have been raised to using Test Anxiety as a measure of
fear of failure. It has the usual weakness of all self-report measures: people who
are afraid of failure are not always willing to admit they are. This shows up in
the fact that some subjects low in Test Anxiety nevertheless show signs of
strong physiological arousal under stress (Salter et al., 1976). Furthermore, as
already noted, subjects classified as high in fear of failure by this measure do
not actually avoid taking intermediate risks the way they ought to according to
theory. Perhaps the most troubling aspect of the way the Atkinson group mea-
sures the fear of failure is that it does not allow a person to be high both in
n Achievement and in fear of failure. By the Atkinson method, in order for
people to be very high in fear of failure (H), they must be low in n Achieve-
ment (L). Such people are classified as LH. To be sure, a person can be classi-
fied as high in n Achievement (H) and high in Test Anxiety (H), or HH, but,
according to the Atkinson scheme, such a person can be neither as high in
n Achievement as one who is classified as HL or as high in fear of failure as
one who is classified as LH.
MEASURING FEAR OF FAILURE IN FANTASY

The Heckhausen, German Measure
In view of the general superiority of fantasy measures of motive strength,
several attempts have been made to break down the scoring categories for
n Achievement into those that seem more oriented around failure in contrast to
those oriented toward success (see deCharms & Dave, 1965; Moulton, 1958).
However, none of these scoring schemes seem to work very well.
In Germany, Heckhausen (1963) derived a scoring System based on charac-
teristics of stories told by individuals who either did or did not adopt defensive
goal-setting strategies. In the level-of-aspiration experiment, in which subjects set
goals for their Performance on the next trial after learning their score on the
previous one, they are considered to be afraid of failure if they set a goal that is
the same as or lower than their present Performance. If their goal is somewhat
higher than their present Performance, they are considered to be activated by a
positive hope for success. Heckhausen also included more pictures of authority
figures, such as teachers or bosses, who might be considered to be demanding
achievement from others.
He found that the stories written by two groups of individuals with con-
trasting goal-setting strategies differed in characteristic ways. Those who set pos-
sible goals for themselves told stories dealing more often with positive urges to
achieve, with expectations of success, and with concentrated efforts to achieve
success. He summed these categories into what he called a Hope of Success
(h Success) score, which for all practical purposes correlates so highly with the
n Achievement score constructed as explained in Chapter 7 that it can be con-
sidered a substitute for it. Subjects who adopted a defensive goal-setting strategy
wrote stories that dealt more often with a need to avoid failure, as in the State-
ment "He hopes that the teacher won't notice the mistake." Other scoring cate-
gories included instrumental acts designed to avoid failure, negative feelings
about work, and criticism or scolding from those in charge of the work. These
categories were summed to give an / Failure score, which correlated neither
with the U.S. n Achievement score nor with the German h Success score at a
very high level.
The German measure of/Failure has many advantages over the Atkinson
System of measurement through the Test Anxiety Questionnaire. It permits a
person to be high both in h Success and in / Failure. It is not subject to re-
sponse bias, and, as Figure 10.4 shows, it is associated with marked avoidance
of tasks of moderate difficulty, as the Atkinson model predicts. In this experi-
ment, Schneider (1978) took the trouble of varying such extraneous factors as
the level of the subject's n Affiliation and whether the experimenter was present
or absent. Thus, the curve showr. is for the pure case in which the person is
high in/Failure, low in n Affiliation, and making his or her choices alone with-
out an experimenter's being present. The result is exactly what the Atkinson
model would predict. That is, subjects high in / Failure prefer very easy or very
difficult tasks and positively avoid choosing tasks of intermediate difficulty.
The behaviors associated with the German / Failure measure make a coher-
ent theoretical picture (Heckhausen, 1980). As compared with subjects low in
/ Failure, those who score high in this measure better recall tasks they have
completed as compared with those they have failed to complete. They take lon-
ger to do their homework, solve fewer problems under time pressure, and have a
shorter future time perspective than those high in h Success. They also perform
better after success feedback, which confirms the expectation developed in the
last section that they will work harder whenever that is the most obvious way to
1 2-3 4-5 6-8 9

Task Difficulty Level
Figure 10.4.
Percentage of Choices of Tasks of Different Difficulty Levels by Subjects Higher in / Failure Than
h Success and Low in n Affiliation with Experimenter Absent (data from Schneider, 1978).
avoid the criticism of failure. Furthermore, if they fail at a task they have cho-
sen to do it is less attractive for them, whereas under similar conditions the task
is more attractive for those high in h Success. Even failure at a task that has
been assigned to those high in / Failure makes it less attractive for them, al-
though failure at such a task for those high in h Success has no significant effect
on its attractiveness for them.
The U.S. Measure

The chief drawback of the Heckhausen System for measuring / Failure was that
it was published in German and therefore was not used by U.S. researchers.
Quite independently of Heckhausen, Birney, Burdick, and Teevan (1969) derived
a new scoring System for/Failure by contrasting stories written when the fear
of failure was aroused by performing very difficult tasks. For College students
they used a speed-reading machine, in which sentences to be read aloud could
be exposed for ever briefer periods of time until the subjects had to fail. For a
group of eighth-graders they aroused the fear of failure by giving them a very
difficult math test.
Stories written when the fear of failure was aroused in these ways showed a
marked increase in what they called Hostile Press, defined as reprimands for
personal actions; legal or jurisdictional retaliation for actions taken; deprivation
of affiliative relationships, "as when a person feels lonely or is afraid of being re-
jected" (Birney et al., 1969); and hostile environmental forces such as fires or
earthquakes. Figure 10.5 shows how the percentage of stories showing Hostile
Press increased significantly when the fear of failure was aroused either among
eighth-graders or College students. The result is similar to one obtained by the
Heckhausen group: When their / Failure measure was subjected to a factor anal-
c
I J Lower/ Failure Arousal Group
"CSS
« 40 — | j Higher/Failure Arousal Group
OStl
X
% 30 _
o
—
§ 20 —
• •
1 10 —
Pei
Eighth-graders College Students

(Math Test) (Speed-reading Test)
Age Group and Method of Arousal of Fear of Failure
Figure 10.5.
Percentage of Stories Showing Hostile Press Among Groups in Which Fear of Failure Has Been
More or Less Aroused (after Birney, Burdick, & Teevan, 1969).
ysis, they discovered two factors, one of which dealt with story content related
to avoiding failure, and the other with story characteristics related to criticism
for failure. They found that the criticism categories were relatively more fre-
quent among those who scored higher in the overall measure of/Failure (Heck-
hausen, 1980). Thus, both independent investigations established the fact that
high / Failure is strongly associated with being criticized or punished.
The U.S. scoring System also included other categories, such as the desire
for relief from Hostile Press, instrumental actions to avoid criticism, and so on.
The behaviors associated with the overall Hostile Press score obtained by sum-
ming these categories have been examined and found to be very similar to those
associated with other measures of/Failure. For example, subjects high in Hos-
tile Press conform more to group pressure, just as subjects high in Manifest
Anxiety do. When they were asked to judge which of two lines was longer, after
hearing five stooges give the incorrect judgment, they more often also gave the
incorrect response, even though it was obviously wrong (Birney et al., 1969).
Characteristics of People with a Strong Fear of Failure

Table 10.5 summarizes some of the relationships found in this study for various
types of Performance situations. In comparison with those low in Hostile Press,
subjects high in Hostile Press took longer to learn to trace mazes correctly when
the mazes were presented to them for very short periods. They volunteered less
often to play darts "to show how good you are at it." They played an interper-
sonal game competitively less often than those low in Hostile Press. In this
game there are two players, each of whom is trying to collect the maximum
number of points. How many points a person can make, however, depends on
what the other player does, as well as on what the person's own choices are.
The following matrix is typical of the points awarded for different combinations
of moves (Birney et al., 1969):
Points Awarded for Player (Player 1, Player 2)

Player 1
Choice A Choice B
Player 2 Choice A 3, 3 4, 0
Choice B 0, 4 1,1
If Player 1 picks Option A, he or she gets either 3 points or 0 points, depending

on whether Player 2 picks Option A or B. If Player 1 has picked A, Player 2
gets 3 points for picking A and 4 points for picking B. If Player 1 picks B, he
or she gets 4 or 1 points, depending on the Option Player 2 picks. Player 2 is
faced with a similar set of choices, and neither player knows in advance how the
other will choose. The best strategy is for both players to pick the same options,
but that means trusting each other, for if one player picks A and the other picks
B, one player may lose out altogether. It is possible over a sequence of moves to
determine whether a player is primarily defensive—trying to avoid total loss, as
when Player 1 picks Option B—or whether the player is actively and competi-
Table 10.5.
PERFORMANCE OF SUBJECTS SCORING HIGH AND LOW IN HOSTILE PRESS IN VARIOUS
ACHIEVEMENT-RELATED ACTIVITIES (after Birney, Burdick, & Teevan, 1969)
Hostile Press
Score
Activity High
High Low
LOW />difference
Average number of trials needed to trace mazes correctly (speeded task) 122.5 102.2
Percentage who volunteer to play darts "to see how good you are" 25% 68%
Percentage who play interpersonal game competitively 221%
1% 43%
43% < .05
Percentage who get above-average grades
In elementary school 76% 24%
In high school 71% 29%
In College 70% 30%
tively trying to defeat the other person by choosing options that would cause the
other player to get 0 points.
Subjects high in the Hostile Press score, indicating a high fear of failure, do
not play this game competitively, but more often try to avoid losing. Finally, as
Table 10.5 shows, subjects scoring high in Hostile Press more often get better
grades at the elementary, high-school, and College levels than those low in Hos-
tile Press. This too is consistent with results obtained with other measures of
/Failure.
• COMPARISON OF MEASURES OF FEAR OF FAILURE

Most significantly in terms of Atkinson's theory, individuals who score higher
on the U.S./Failure measure also avoid moderate risks and show preference for
extreme probabilities of success. Ceranski, Teevan, and Kalle (1979) reported an
experiment in which they used three different measures of fear of failure to see
which was most closely associated with this theoretically important prediction.
The results are summarized in Table 10.6. The Hostile Press measure of fear of
failure is the only measure in these situations that successfully predicts at a sig-
nificant level which subjects will adopt a defensive goal-setting strategy and
show a preference for extreme probabilities of success. A much higher propor-
tion of subjects high in Hostile Press, as compared with those low in Hostile
Press, set Performance goals that were either the same as, less than, or very
much higher than the level they had achieved on a given trial. That is, they set
either very easy or very difficult goals for themselves and avoided setting moder-
ately difficult goals. In a different part of the experiment they also showed a dis-
tinct preference for working at tasks described either as very easy or very hard.
The Test Anxiety measure by itself did not succeed in predicting these prefer-
ences, and the Atkinson measure also failed, although the trends were in the ex-
pected direction.
The result of all this work is that there are three different measures of fear
of failure, all of which show similar relationships to behaviors, such as defensive
goal setting, poor Performance under some conditions and better under others,
avoidance of competition, and susceptibility to positive suggestions from others.
The measures also significantly correlate with each other, although at a fairly
low level (Birney et al., 1969). Hostile Press scores correlate with the German
/ Failure measure, as well as with the Atkinson method of determining / Failure
scores. The Atkinson method has been used most often because it involves less
work, but it is probably the least satisfactory of the three for reasons already
given. The German / Failure measure probably is the most carefully worked out
and thoroughly tested in terms of its relationships to various behaviors, but it
has not been used by U.S. researchers. The U.S. Hostile Press measure has not
been used except by those who devised it, and whereas it has clear advantages
over the Atkinson measure, it raises some theoretical questions that have never
been resolved. One problem is reflected in the fact that one of the scoring cate-
gories for Hostile Press involves fear of rejection, or imagery that is also scored
Table 10.6.
RELATIONSHIP OF THREE MEASURES O F / FAILURE TO DEFENSIVENESS AND AVOIDING
MODERATE RISKS (data from Ceranski, Teevan, & Kaue, 1979)
Percentage of Subjects Percentage of Subjects

Using a Defensive Goal- Preferring Extreme
setting Strategy in a Probabilities of Success
Measure off Failure N Scrambled-Words Test" in Eight Dijferent Tasksb
Birney, Burdick, and Teevan measure
High in Hostile Press 55 62% 53%
Low in Hostile Press 68 21% 9%
Z7 difference <.001 <.001
Test Anxiety measure
High in Test Anxiety on Test Anxiety
Questionnaire (TAQ) 69 35% 31%
Low in TAQ 63 38% 21%
P difference n.s. n.s.
Atkinson measure
Low in n Achievement, high TAQ 33 36% 33%
High in n Achievement, low TAQ 37 27% 22%
P difference n.s. n.s.
a
The discrepancy between the goal set for next trial and Performance on the previous trial was either negative, 0, or -1-3.5 or more.
b
These subjects preferred tasks in which probabilities of success were described as either very easy (.85 to .92) or very difficult (.08 to .15).
for n Affiliation. This introduces a significant relationship to the n Affiliation

score but leaves unanswered the question as to whether this is an artifact of the
scoring System or a real relationship between fear of failure and fear of rejection.
The German scoring System does not introduce this complication. Another
problem lies in the fact that the scoring System for Hostile Press is named for
the source of the fear rather than for the incentive that would resolve it. As we
have seen, this is a general problem for all anxiety measures. It may simply re-
flect the fact that we are still unable to specify very precisely what the incentives
are for avoidance motives.
• ORIGINS OF THE FEAR OF FAILURE

A question that naturally arises in reviewing the research on / Failure measure-
ment is why the self-report approach adopted by the Atkinson group works as
well as it does, given the failure of self-reports to measure approach motives ad-
equately (see Chapter 6). The ans wer may lie in a fundamental difference be-
tween approach and avoidance motives. Usually it is some force in the external
environment—a parent, a teacher, or a peer—that by punishing a child forms

the basis of an avoidance motive. If children do not get food into their mouths
properly, their mothers scold. If they do not put their shoes on properly, their
nursery-school teachers scold. If they cannot throw a ball, friends may laugh.
All of these experiences build up a fear of failure, but they are all based on eas-
ily identifiable external events followed by internal responses of feeling upset. So
a self-report measure may work better for avoidance than for approach motives
because the events underlying the former are more easily identified. The Stimu-
lus value of a "moderately diiücult task," as well as the satisfactions arising
from mastering it that form the basis of the achievement motive, are not so eas-
ily identified as a sharp rebuke for performing poorly and the anxiety that im-
mediately accompanies it.
Studies of the origins of the fear of failure support such a hypothesis. Smith
(1969) found that mothers of sons who feared failure set higher Standards of
achievement for their sons and, at the same time, took a less favorable view of
their son's ability to achieve those Standards than other mothers. They were set-
ting their sons up for criticism and punishment. It is as if they were saying,
"You must do better, but I know you can't."
Birney et al. (1969) approached the same issue in a different way. They
asked both parents and their children how the parents reacted when the chil-
dren succeeded or failed. They categorized the responses they got into two con-
trasting patterns. In one the parents reacted neutrally for success but punished
the child for failure, and in the other the parents rewarded success and were rel-
atively neutral about failure. Of the students high in Hostile Press or / Failure,
83 percent had parents who feil into the first pattern—predominantly punish-
ment for failure—as contrasted with only 54 percent of the students scoring low
in Hostile Press, a highly significant difference. Other studies (for example, see
Hassan, Enayatullah, & Khalique, 1977) support the general conclusion that
parents who are rigid, authoritarian, and punitive tend to produce children with
a strong / Failure. What such findings also suggest is that the sources of such
avoidance motives are external and easily identified by the child, so they can be
reported on a Test Anxiety Questionnaire with reasonable accuracy.
FEAR OF REJECTION
Studies of n Affiliation reviewed in Chapter 9 strongly suggest that it is primar-
ily a measure of Fear of Rejection (f Rejection), whereas the intimacy motive is
a better index of positive affiliative tendencies. Nevertheless, some attempts have
also been made to get a purer measure of the fear of rejection. Sorrentino and
Sheppard (1978) followed the procedure adopted by Atkinson to get a purified
measure of/Failure. As an equivalent to the Test Anxiety Questionnaire used
in that procedure, they turned to a self-report measure of/Rejection developed
by Mehrabian (1970). It asks subjects to respond to Statements like "I prefer not
to go to a place if I know that some of the people who will be there don't like
me" or "I enjoy discussing controversial topics like politics and religion." A per-
son responding true to the first Statement and false to the second would be
scored as high in / Rejection. Subjects who score both high in n Affiliation and
low on this scale were considered by Sorrentino and Sheppard to be low in
/ Rejection (or high in Hope of Affiliation, or h Affiliation), whereas those who
score low in n Affiliation and high on this scale are considered high in / Rejec-
tion. Subjects with other combinations of scores on these two measures are con-
sidered intermediate in / Rejection.
Sorrentino and Sheppard were able to carry out a natural experiment that
demonstrated that individual differences in / Rejection levels made a real differ-
ence in the speed with which individuals swam under different competitive con-
ditions. In one instance the swimmers were told to swim as fast as they could,
each one having been given a handicap by the coach based on past Performance.
The goal was to try to beat all other individuals competing. In the other in-
stance the swimmers with the same handicap were told they were part of a team
and that the objective was for their team's score to be better than the team
scores of other competing groups. As Figure 10.6 makes clear, subjects high in
/Rejection swam slower when they were working for a team goal than when
they were competing as individuals. It is as if their fears that they would be crit-
icized by other members of the group actually interfered with their ability to
swim. In contrast, those high in n Affiliation (or low in / Rejection) swam faster
when they were part of a group than when they were competing individually.
The differences in swimming times, while small in absolute terms, are highly sig-
nificant statistically and would have important effects on winning different kinds
of races in competitive athletics.
130
I Jj Individual Competition
2
135 Group Competition
ii 140
I 145
High Affiliative Interest High /Rejection

Figure 10.6.
Average Swimming Speeds of Subjects High in Affiliative Interest and in / Rejection in Individual
and Group Competition (data from Sorrentino & Sheppard, 1978).
A number of other studies have been carried out using a self-report measure
of/Rejection developed by Mehrabian and Ksionzky (1974), which may be re-
garded as somewhat valid on the grounds that it reflects an avoidance motive
about which individuals may be more aware. Such an assumption seems justified
by results obtained with it, such as those illustrated in Figure 10.7. In the exper-
iment shown, Student participants were told they were about to take part in a
study of how strangers interact with each other. While they were waiting to
meet the strangers they filled out a social Status survey, which was scored; they
were told that in terms primarily of the income level and education of their par-
ents, they were at Positions 2, 5, or 8 on a 10-point scale representing social Sta-
tus. They were all told that the stranger they were about to meet was at Point 5
on the same scale. Thus, they knew whether the stranger was of a higher or a
lower Status than themselves. Then they chatted for a few minutes with the
stranger, who was a confederate of the experimenter instructed to act in certain
Standard ways. Afterward the subjects reported how much they had liked the
stranger on a scale of 1 to 9.
As Figure 10.7 shows, students who were high in/Rejection liked the
stranger of lower Status more than the stranger of higher Status. It might be in-
ferred that they were more nervous in the presence of a higher Status person be-
cause of their fear of being rejected, so they liked the individual less. In con-
trast, those low in / Rejection liked the higher Status stranger better. Whereas
the Mehrabian and Ksionzky measure of/Rejection thus has some promise, its
o
_JSlranger of Higher Status
Z 5.00 — J Stranger of Lower Status

o
Met (Seal
1 4.00 -
o
Average Liiking
—
g
High/Rejection Low/Rejection
Figure 10.7.
Liking for a Stranger of Higher or Lower Status by Individuais High or Low in / Rejection (after
Mehrabian, 1971).
exact Status is unclear, partly because it is obviously open to response bias and
partly because it does not correlate with the n Affiliation score (Karabenick,
1977), which has been shown to reflect a fear of rejection.
Relationship Between Fear of Failure, Fear of Rejection,

and the Approval Motive
Why should anyone fear failure? If a boy is scolded by his mother for doing
poorly in school, why should he care? On theoretical grounds it might be pre-
sumed that he is afraid of her disapproval and rejection. In this case, fear of
failure simply translates into fear of rejection. There are hints throughout the
empirical literature that the two avoidance motives may be connected. Recall
the discussion in Chapter 3 of findings that the mere presence of another person
can introduce evaluation apprehension that affects Performance in ways that
look as if the person is simultaneously fearing failure and disapproval or rejec-
tion. (See Table 3.3.) Furthermore, the previous sections reviewed several studies
showing that reassuring subjects, or even telling subjects what to do, reduced
their fear of failure and enabled them to perform better. And Birney et al.
(1969) found that arousing the fear of failure experimentally increased thoughts
of rejection specifically, so they included these thoughts as a category in their
Hostile Press scoring System. Perhaps fear of failure and fear of rejection are
closely related, and both derive from a need for social approval.
As Chapter 5 pointed out, Crowne and Marlowe (1964) have developed a
measure of the need for Social Approval, so such a hypothesis can be tested. It
is conceivable that people who answer items on a questionnaire always in the
socially desirable direction may be avoiding making unfavorable comments
about themselves because they fear disapproval or rejection. Supportive evidence
for such an inference derives from the fact that the approval motive score ob-
tained from such a questionnaire correlates highly significantly with the n Affili-
ation score (Crowne & Marlowe, 1964), which we have determined is primarily
a measure of/Rejection. Furthermore, the behaviors associated with a high
need for Social Approval are very similar to those associated with a strong
/ Failure and a strong / Rejection, as measured by the n Affiliation score. Sub-
jects who score high on any of these measures tend to engage in defensive goal
setting, to yield to group pressure, and to more readily shape what they say in
response to the nod and the "Mhm" from the experimenter in verbal condition-
ing experiments.
But then we run into an apparent snag. If these measures are all tapping the
same or similar characteristics, why does the need for Social Approval score
correlate negatively with the Manifest Anxiety score (Crowne & Marlowe,
1964)? That is, why do those who mention only desirable qualities in themselves
score so very low in admitting to any anxieties? They deny they are anxious.
That makes sense, but how can we argue then that they are really anxious about
being rejected?
At a deeper level the results are not really contradictory. A person who
fears rejection may deny being anxious in the search for approval. What is at
fault is the self-report measure employed, which reflects more the strategy of
dealing with the fear (denial) than a measurement of the extent of the fear itself.
So we can only conclude that the fear of failure, the fear of rejection, and the
need for approval could probably all be shown to be related to each other, if we
only had better measures of each of them that had tested their relationships in a
definitive study. One other bit of evidence Supports this conclusion. In Germany,
Laufen attempted to develop a method of scoring stories for h Affiliation and
/Rejection by having subjects first fill out a questionnaire containing Statements
relating to either n Affiliation or/Rejection. Only the exposure to/Rejection
Statements produced significant changes in the thought content of stories in the
direction of increasing thoughts about rejection (Heckhausen, 1980). The scoring
keys for the stories he developed were subsequently used in another study by
Jopt (cited in Heckhausen, 1980), who found that/Failure and / Rejection
scores summated to produce the best Performance from sixteen-year-old male
students working for a female teacher. That is, the boys worked harder for such
a teacher if they scored high on the Heckhausen / Failure measure and the
Laufen / Rejection measure than if they were high on either one of these scales
alone. The same result was not obtained when the teacher was male. This sug-
gests that under certain conditions, fear of failure and fear of rejection combine
readily to produce an effect, although it does not prove they are aspects of the
same characteristic.
• FEAR OF SUCCESS
Measuring Women's Fear of Success in Fantasy
Trained in the Atkinson tradition for studying achievement motivation, Horner
(1968) became curious as to why women did not always behave the same way as
men in achievement situations and, in particular, why they scored higher in Test
Anxiety. She noted that many psychologists believed women had much more
anxiety over appearing aggressive and competitive than men did (Horner, 1973).
So she designed a test to tap this anxiety by asking subjects to write a story
about a verbal lead such as the following: "After first-term finals, Anne finds
herseif at the top of her medical school class." For men she substituted the
name John for Anne in the verbal lead. She found that women wrote many
more negative stories to the Anne cue than men did to the John cue. The stories
written by women dealt with the negative consequences of success or even with
denial of the Situation described by the cue. Here are some examples:
Anne is an acne faced bookworm. She runs to the bulletin board and finds she is at
the top. As usual "she smarts off." A chorus of groans is the rest of the class's
reply.
Well it certainly paid off. All the Friday and Saturday nights with my books, who
needs dates, fun—Fll be the best woman doctor alive.
Anne, the girl who swallowed the canary in her younger days, is pretty darned
proud of herseif. But everyone hates and envies her.
Anne is a code name for a non-existent person created by a group of med students.
They take turns taking exams and writing papers for Anne. (Horner, 1973)
In general, these stories indicated that women feared social rejection based
on Performance success much more than men did. Horner called the score ob-
tained by summing these negative characteristics a measure of fear of Success
( / Success). She validated the measure by checking whether women who scored
high in / Success would in fact avoid winning in direct competition with men.
She arranged for a large number of students of both sexes to work at a number
of verbal and arithmetic tasks in a group competitive Situation to see who could
get the best scores. Some of the subjects then worked under noncompetitive con-
ditions. Whereas most of the men did better under competitive than noncompet-
itive conditions, only 23 percent of the women high in/Success performed bet-
ter under competitive than noncompetitive conditions, as contrasted with 93
percent of the women low in / Success. Fear of Success did not have the same
effect on the Performance of men. In other words, the women who had written
stories about the negative consequences of success were holding back in competi-
tion with men as compared with how they would perform normally when they
were not in competition.
Horner's original data showed that 65 percent of the women as compared
with only 9 percent of the men, wrote stories that were scored for / Success.
However, this sex difference tended to disappear in later replications of her find-
ings (Hoffman, 1974; Zuckerman & Wheeler, 1975). Furthermore, several inves-
tigators reported a failure to find the expected relationship between / Success
scores and Performance of women in a mixed-sex classroom Situation (Feather &
Simon, 1973). One problem may have been that the original findings on fear of
success in women had been widely written up in national magazines, so many
College women knew about them. The women's movement was gathering mo-
mentum at the same time, and it seemed reasonable to infer that many women
were consciously avoiding writing negative stories to the Anne cue to prove they
were not afraid of getting ahead in the world.
Horner, Tresemer, Berens, and Watson (1973) then developed a new scoring
System for/Success based on the stories written after succeeding or not suc-
ceeding in competition with a particular male. That is, feedback on Performance
was manipulated so that each woman in the arousal condition appeared to have
beaten her male competitor. In the control condition, similar women wrote sto-
ries after working on the same task but not in competition with anyone. The
scoring categories that differentiated the stories written under these two condi-
tions had to do with the problematic expression of instrumental Performance
activity, with problems' being solved without effort (called "relief")> and with
approach to other people introduced into the story. These new indicators of
/ Success are less obvious than the old ones and therefore are harder for individ-
uals to change consciously in line with their values. At any rate, perhaps for
this reason the results with this measure have been more consistent in picking
out those people who will do less well in competition with others.
How Fear of Success Affects Competitive Behavior in Women

Shinn (1973) used the new measure of/Success to study the effects of the
merger of two formerly single-sex secondary schools. She found that before the
merger, when the boys and girls were tested separately in their segregated
schools, they did not differ in average levels of/ Success, nor did the/Success
score predict who would do poorly in working competitively on a verbal task
like anagrams. The girls did do better than the boys on the verbal task on the
average, in accordance with a general tendency of women to do better on verbal
tasks. However, the results were quite different a year after the merger had
taken place, when the boys and girls were in the same classrooms competing
with each other. The girls had lost their superiority in performing a similar ver-
bal task—unscrambling words—and their / Success score had risen dramatically
over what it had been when they were in the single-sex school. The boys had
not shown a similar rise in / Success scores. Furthermore, after the merger, only
30 percent of the girls high in / Success scored above average on the scrambled-
words task, as compared with 72 percent of those low in / Success, which is a
significant difference. In other words, fear of success was inhibiting the perfor-
mance of the girls now that they were in a Situation where they were competing
with boys, whereas before, when they were competing with girls, high / Success
had not inhibited their Performance. A similar shift did not occur for the boys.
Their / Success score was not associated with poor Performance. In other words,
it looked as if it was specifically the fear of the consequences of competing with
boys that inhibited the Performance of girls.
Karabenick (1977) showed that high/Success diminished Performance for
women competing with men only when they thought the task was masculine.
He described anagrams as related either to masculine or feminine abilities and
interest and stated further that either men or women were particularly adept at
solving them. Different groups of subjects worked at the task, believing that it
was masculine or feminine. They either worked alone or in competition with a
member of the same or opposite sex, who was described as performing a little
below average on the same task. Thus, the subjects would tend to believe that
they had a good chance of beating the other person. The crucial question was
whether they performed better under competitive or noncompetitive conditions
when they were working against a member of the same or opposite sex on a
task they thought was either masculine or feminine. The results showed that the
/Success score was related to a difference in Performance under competitive
conditions under only one condition—when women high in/Success were
competing with men on a masculine task, they tended to perform less well
(r — —.25; p < .01). If they were working on exactly the same task in compe-
tition with men but thought it was a feminine task, their / Success had no affect
on their Performance. Nor did male / Success scores relate in any way to their
Performance in competition with women or other men.
The inhibiting effects on Performance of fear of success in women shows up
not only in laboratory tasks but also in attitudes toward family and career
throughout life. Stewart (1975) obtained / Success scores from stories written by
women in two different Colleges in their freshman year and then observed how
these scores related to what the women did over the next fourteen years. Fear of
Success scores were significantly higher in College B than in College A, and
Stewart found considerable evidence that the women in College B experienced
much more conflict about whether pursuing a career would interfere with hav-
ing a happy marriage and family. In College A the women were much more
single-mindedly oriented toward a career, and they were less concerned about
marriage. Thus, it was not surprising to discover that / Success scores among
College B women were associated significantly with getting married early, with
having children soon, with not working, and with "job dabbling," that is, with
playing around with working rather than pursuing a consistent career path.
Stewart interpreted this as meaning that the women in College B, who were
afraid of the consequences of success, were so worried about working and ap-
pearing unfeminine that they quickly got married and had children to prove
their femininity.
She also found signs that among women who single-mindedly pursued a ca-
reer without thinking about family, high / Success could be associated with a
lack of career persistence. For example, among the substantial number of
women from both Colleges who had received a Doctor of Philosophy degree,
high / Success was associated with not working full-time ten years after College,
whereas there was no such relationship for women who had attained only a
Master's degree. The explanation is that women who are afraid of the negative
consequences of success are much more likely to worry about working after hav-
ing concentrated on a career enough to get a Doctor of Philosophy degree than
are women who have not spent so much time in professional development. As
another sign of the same trend, Stewart discovered that among women who had
never married or not had a child early, high / Success was significantly related
to a lack of career persistence; this was not true, however, among women who
had married or had a child. She explains the finding as meaning that the women
with families had proved their femininity, so their fear of success would no lon-
ger hamper them in pursuing a career. On the other hand, Single women who
feared success would still be worried about whether they were working so much
as to exclude the possibility of having a happy married life, so they would keep
interrupting their careers in order to give more balance to their lives.
Jackaway and Teevan (1976) reasoned that if fear of success results from a
fear of failure in interpersonal relationships, then there should be a positive cor-
relation between the Hostile Press / Failure measure and the/Success measure.
They found this to be the case, the correlation being .44 for male and .53 for fe-
male high-school students. The relationship was higher for women, as would be
expected from the larger number of their stories that mentioned affiliative failure
consequent upon success, although the difference between the correlations for
men and women did not reach significance. As Jackaway and Teevan (1976)
point out, the correlation is due in part to an overlap in the scoring Systems,
since some of the same events in the stories are scored by the two different Sys-
tems. In particular, as mentioned earlier, the Hostile Press System scores for af-
filiative failure, which also is scored in the / Success System. It would be worth
checking the correlation with the German fear of failure measure before suggest-
ing, as they do, that "for those people (especially women) whose affiliative and
achievement needs are interrelated, fear of failure and fear of success may be
nearly equivalent since fear of rejection thus becomes tantamount to fear of fail-
ure" (Jackaway & Teevan, 1976).
Fear of Success Among Black Men and Women

Fleming (1974) found that fear of success functions quite differently among
black men and women. At the outset, using verbal cues like Horner's, she dis-
covered that only 29 percent of black College women gave / Success story com-
pletions, as contrasted with 67 percent of black College males. Thus, the results
were the opposite of what Horner had been finding for white College students.
The black women were much lower in/Success than white women, and the
black men were much higher than white men. She explained this reversal as due
to racial discrimination and oppression. Historically black men have been in
much greater danger for being assertive and "uppity" than black women. Often,
in fact, it was easier for black women to find housework and support the family
than it was for black men, who were systematically discriminated against. Thus,
black women had no reason to fear the negative consequences of achievement;
in fact, they were taught to expect rewards from working hard, whereas among
black men, trying to get ahead could often lead to trouble, disappointment, and
rage.
However, in later studies using the Horner-type cues, the differences be-
tween black men and women tended to disappear, just as the opposite difference
had disappeared among whites. Nevertheless, a closer analysis of the stories
written by black men and women, as well as application of the new scoring sys-
tem for/Success, still led Fleming (1975b) to conclude that black men feared
success more than black women, although the evidence was not as conclusive as
might be wished for. The following are typical stories written by black College
men:
Sam is finding himself. He is out of sight. . . . Sam is thinking that the white boys
will obviously without a doubt be uptight.
Sam will be wanted by all of the Wall Street Firms. A lot of white boys will jump
out of Holyoke Center. The professors will of course make sure that Sam is not
number one the following term. (Fleming, 1975a)
Fleming (1974) found, furthermore, that/Success scores were negatively re-

lated to Performance when black men were competing directly with white men.
First, she obtained / Success scores under neutral conditions and three weeks
later had a white male compete directly with a black male on an arithmetic test
to see who would do better. She then reported that the black had won, and she
administered a second / Success measure followed by further Performance alone
on the anagrams task. She found that those black men whose / Success scores
had increased most after beating a white male Opponent did perform signifi-
cantly less well on the final anagrams test (r — —.46; p < .001). Their high
/ Success had interfered with their Performance. She also reports that achieve-
ment motivation training of the type described in Chapter 14 increased fear of
success motivation in black males rather than achievement motivation.
In Fleming's (1974) study the black women competed with black men
rather than with white women. She found that high/Success was significantly
related to poorer Performance in competition with black men only among black
women of lower-class background. She feit that this was true because the con-
flict between career and marriage is more severe among lower-class than among
middle-class blacks. For example, among the women of lower-class background,
75 percent of those high in/Success had mothers who were better educated
than their fathers, as contrasted with only 38 percent of those low in/Success.
This difference suggests that in the families of women with high / Success there
was more conflict over what women should be doing—that is, working in line
with their better education or taking care of the family. These differences of
opinion also showed up in the fact that whereas none of the parents of the
women low in/Success were divorced, 38 percent of those high in/Success
had divorced parents.
The black students were also asked what values their parents had stressed in
bringing them up. Table 10.7 reports the correlations of/Success scores among
black men and women with values stressed either by their mothers or fathers.
The most striking finding so far as the women are concerned is that those high
in / Success had mothers who emphasized one characteristic and fathers who
emphasized another. For those of working-class background, the mothers
stressed achievement and social assertiveness while the fathers stressed self-
reliance. For those of middle-class background, high / Success was associated
with mothers who played down self-reliance and whose fathers stressed achieve-
ment. In this case it looks as if ambivalence over being assertive (high / Success)
derives directly from the fact that they were getting opposite messages about the
importance of such behavior from their fathers and mothers.
In the case of black men, the only significant result is that fathers both
from the middle class and the working class who insisted on social assertiveness
tended to have sons who were low in/Success. To put it the other way around,
if the fathers cautioned their sons against being assertive or "uppity," the sons
tended to grow up with high / Success. This finding Supports Fleming's general
argument that high / Success in black males is associated with fears of the nega-
tive consequences in a white society of being an assertive black man.
What if a black male high in / Success is placed in a competitive white en-
vironment? Table 10.8 contrasts how black men high in/Success behave in a
predominantly white College as compared with how they behave in a predomi-
nantly black College; both of the Colleges were located in the deep South (Flem-
Table 10.7.
CORRELATIONS O F / SUCCESS SCORES AND SUBJECTIVE REPORTS OF PARENTAL
TEACHING EMPHASES AMONG BLACK MEN AND WOMEN IN COLLEGE (after Fleming, 1974)
Wotnen Men
Working-class Middle-class Working-class Middle-class
Background Background Background Background
Parental Teaching Emphases Mother Father Mother Father Mother Father Mother Father
Achievement orientation .46* .08 .06 .30** -.07 .04 -.23 -.14
Self-reliance -.22 .51** -.28* -.04 .22 .16 .04 -.09
Social assertiveness .36* —a .20 .15 -.38* a -.43**
a
Not enough cases to compute.
*p < .05.
**p < .001.
ing, 1983). In the white College black men tend to be more successful than
usual. They hold more Offices in extracurricular activities, they are described as
more outspoken, and they think of themselves as more extraverted. In a white
environment, their high / Success has led to social success. The study suggests
that they have been sensitive to the criticism of being too assertive in a white
environment and, as a consequence, have been accepted by their white peers.
Some blacks more oriented toward the Black Power movement might regard
them as "Uncle Toms" who "go along in order to get along." But notice that
this behavior also has a cost. The men who are high in / Success in a white en-
vironment also have needed more medical assistance for psychosomatic disor-
ders. In contrast, black men high in n Achievement do not do particularly well
in a white College in the South; they find it harder to do well and feel fatigued
and disoriented (Fleming, 1983).
Black men who fear success do not behave in markedly different ways in a
black College than other men. They are oriented somewhat more toward male
domination, perhaps to compensate for the fact that they fear they will be re-
jected by women if they outperform them. That is, in a black College they are
competing with black women—not with whites, as in a white College environ-
ment. Thus, their fears over the negative consequences of success have more to
do with sex role relationships than with getting along in a white world, where
competition among males dominates.
Fleming's studies underline the fact that fear of success has to do with the
negative consequences of a person's Standing out in some way or being assertive
in relation to another group that is perceived as, or that is, more powerful and
in a position to reject or punish the person for being assertive. What people are
Table 10.8.
CONTRASTING BEHAVIORS ASSOCIATED WITH H I G H / SUCCESS AMONG
BLACK MALES IN PREDOMINANTLY WHITE AND BLACK COLLEGES
(after Fleming, 1983)
Correlations with f Success

Behavior White College Black College
Offices held in extracurricular activities .41**
Described as outspoken .49** —
Extroverted self-concept .37**
Times medical assistance sought for
psychosomatic complaints .34* —
Male-dominated career choice — .31*
Belief that marriage will help career — .31*
*p < .05.
• V < oi.
afraid of and how it will affect their Performance depends very much on
whether they are male or female and black or white. But in several different
types of situations it has been shown that the more recent experimentally de-
rived / Success score is significantly related to poor Performance in competition
with those in a position of greater power.
FEAR OF POWER
Measuring Fear of Power in Fantasy
For reasons just given, it seems not unreasonable to suppose that fear of success
might be related to fear of power of others, another avoidance motive for which
a scoring System has been developed by Winter (1973). He was interested in get-
ting separate measures for approach and avoidance aspects of power motivation.
He decided to classify any story as representing fear of Power (f Power) if it
contained any doubts about the direct expression of assertiveness. These doubts
could be reflected in any one of three ways: "(1) explicit Statement that the
power is for the benefit of some other person or cause; (2) guilt, anxiety, self-
doubt, or uncertainty on the part of the person concerned with power; or (3)
irony and skepticism about power as shown by the story writer's style" (Winter,
1973). The usual scoring System for n Power was applied, and if the major
theme of any story expressed doubts of this kind, all the categories scored for
this story were put in the "Fear of Power" column. If a story did not show any
of these doubts, the scoring categories for n Power were placed in a "Hope of
Power" (h Power) column. In general, the h Power a n d / P o w e r scores do not
correlate significantly with each other, and the h Power score correlates so
highly with the overall n Power score that it might be considered a substitute
measure for it. The / Power score correlates only moderately (in the ränge of
.30 to .50) with the overall n Power score. Thus, it deserves to be studied in its
own right. We might wonder also if/Power is the same as the socialized power
motive (s Power) discussed in Chapter 8, since both include power imagery di-
rected at helping others. There is a moderate relationship: the socialized power
score and the/Power score correlate .49, and the personal power score (p Pow-
er) and h Power score correlate .46 (McClelland & Watson, 1973), but the rela-
tionships are low enough to indicate that the measures are not identical and de-
serve separate treatment.
The Association of Fear of Power with Atypical Role Behavior in Men

Winter (1973) associates high/Power with male paranoia. He reviewed clinical
evidence that male paranoids feel weak in a conflict with powerful parents and
then associate these feelings of weakness and Submission with homosexual
wishes and fears. Such men want power, which they hope to get from homosex-
ual attachment to a powerful male, but they fear the effects of the desire for
power (since it may lead to feared homosexuality) and also fear the power of
other people over them. Winter demonstrated that among males, paranoid
schizophrenics average much higher on his/Power score than nonparanoid
schizophrenics or control subjects, whereas there was no difference among these
groups in their average h Power scores. He further found that high / Power was
not related to the usual assertive expressions of power motivation, such as hav-
ing more prestige possessions, holding office more often, or playing with compet-
itive efficiency in a game of cards.
In contrast, high / Power did seem to be related to a desire for autonomy as
reflected in answers to questions such as how much freedom students should be
given in choosing their subjects of study. Also, students who handed their pa-
pers in late at the end of a course or requested an incomplete grade scored sig-
nificantly higher in / Power, suggesting to Winter that they resisted being
pushed around by course requirements. When he asked students to give the fa-
vorite words or phrases they used when swearing, those who used "asshole,"
some Variation thereof, or any homosexual expression scored significantly higher
in / Power than those who did not use such expressions. Those who used these
phrases were if anything, lower in h Power. Thus, he concludes that those "high
in fear of power have homosexual desires or fears" and that, following Freud,
"fear of power seems to result from a reversal in the Oedipal stage. Here the
boy identifies with his mother and becomes thereby a sexual object for the fa-
ther" (Winter, 1973). His conclusions are based on very few subjects and seem
rather far-reaching in terms of the data on which they are based, but they are
suggestive and deserve to be checked.
Behaviors associated with high h Power and / Power were examined in a
larger sample of adult males of mixed socioeconomic background (McClelland,

1975). The results are summarized in Table 10.9. So far as the supposed lack of
competitiveness of men high in / Power is concerned, the evidence is mixed. On
the one hand, they are not so interested in traveling to new places or playing
the field sexually as those high in h Power. On the other hand, they do not re-
port playing competitive games like checkers or chess significantly less often and
do report getting involved in arguments more frequently, which is not true of
those high in h Power. Such a correlation does not teil us how effective they are
in arguments, so it does not directly disconfirm Winter's finding that they do
not play a competitive card game as well.
The evidence for the greater interest in autonomy of those high in / Power
is somewhat stronger. They report that they have more difficulty getting up in
the morning to face the demands of the day, and they also report that they en-
gage in more rituals before going to bed at night. Recall that both Chapter 2
Table 10.9.
CORRELATIONS OF h POWER AND / POWER SCORES WITH VARIOUS BEHAVIORAL
CHARACTERISTICS AMONG ADULT MEN: N = 85 (after McClelland, 1975)
Behavioral Characteristics h Power f Power

Assertiveness
Frequency of arguments .04 .30**
Frequency of playing competitive games .04 -.09
Travel to new places .20t .05
Playing the field sexually .18t .03
Autonomy
Frequency of difficulty in getting up in the morning .01 .24*
Number of rituals before going to bed .09 .19t
n Power score for those predominantly
At Stage I (intake); N = \\ .95*** .78**
At Stage II (autonomy); N = 12 .37 .66*
At Stage III (assertion); N = 13 .82** .04
At Stage IV (mutuality); N = 33 .80*** .52**
Mystical "at oneness"
Transpersonal contacts (with God, nature, and so on) .04 .32**
Psychological mindedness .03 .25*
Preferred metaphor for death as an "infinite ocean" .06 .28**
Deviant stature .17 .22*
ip < .10.
*p < .05.
**p < .01.
***p < .001.
and Chapter 8 pointed out that ritualistic behavior of this type is theoretically
associated with the second stage of psychosexual development, which is oriented
around breaking away from dependence on the parents and becoming indepen-
dent or autonomous.
Further confirmation of the orientation toward autonomy is found in the
correlations of the total n Power score with the h Power and / Power scores for
men oriented predominantly toward each of the four stages of psychosocial ma-
turity. In this analysis, men were assigned to one stage or another, depending on
the stage for which they received the highest Standard score in Stewart's coding
System (see Table 8.5). Thus, twelve men were classified as being oriented pri-
marily toward Stage II (autonomy); the n Power scores of these men were much
more likely to b e / P o w e r (r = .66; p < .05) than h Power (r — .37; not sig-
nificant). The reverse is true for the men at Stage III (assertion). In their case,
the n Power scores were nearly all h Power (r = .82; /> < .01) rather than
/ Power (r = .04; not significant). What this indicates is that / Power is
strongly associated with the autonomy stage in psychosocial development and
h Power, with the assertive stage in psychosocial development.
What is new in the findings presented in Table 10.9 is the association of
/ P o w e r scores with various measures of mystical "at oneness." Men high in
/ P o w e r report having had more transpersonal experiences such as a religious
experience of the presence of God; a feeling of being "at one with nature"; or an
extrasensory experience, such as communication from someone not present or
dead or the feeling of being healed by someone. They also think that one of the
better metaphors for death is an "infinite ocean." Finally, men high i n / P o w e r
score higher on the psychological mindedness scale, which includes such items
as "Sometimes I think of natural objects as possessing human qualities" or "The
rieh internal world of ideals, or sensitive feelings, reverie, of self-knowledge is
man's true home." In Chapter 8 the argument was presented that such feelings
of mystical union or of psychic reality, being ultimate, are associated with femi-
ninity, the mother, or the female life-giving principle. Thus, these results con-
firm Winter's hypothesis that men high in / Power may have identified more
with their mothers. The results also suggest a reason why these men may engage
in less agentic striving than is typical of the traditional male role. They may fear
that acting assertively in this way would break up their pereeived union with
the environment (nature, the "divine ground," mother).
The final correlation in Table 10.9 suggests that atypical stature in men may
have something to do with the development of high / Power. This measure is
the absolute deviation of a man's stature from the average stature for men (sev-
enty inches). The correlation means that men who deviated markedly from this
average—being either very tall or very short—are more likely to have higher
/ P o w e r scores. The relationship is not strong, but once again it suggests some
kind of inadequaey lying behind the fear of power in men.
Fear of Power Among Women

Among a similar group of adult women of mixed socioeconomic background,
/ P o w e r scores had no more significant correlates in behavior than would be ex-
pected by chance. Fear of Power did correlate significantly with psychological

mindedness (r = .29; p < .01), as it did for the men in the sample, a fact that
takes on significance primarily because Stewart (1975) found in a quite indepen-
dent sample of women that those high in / Power were significantly more likely
to mention that problems they had met in life were of a psychological nature.
But why should the fear of power turn women inward—or men, for that mat-
ter? Perhaps they are afraid of the consequences of assertiveness in the real
world, in which it is all too often met with retaliation. So to avoid the push and
shove of competitive interaction with people, they retreat to the world of psy-
chic reality, where they have more control over what is happening.
This is reminiscent of the evidence reviewed in the "Fear of Success" sec-
tion that white women draw back from assertiveness in competing with white
men. In one group of subjects, half of whom were men and half of whom were
women, there was a significant correlation (r = .33; p < .05) between the
/ P o w e r score and the/Success score (McClelland & Watson, 1973). In two
samples of adult women College graduates, Stewart (1975) found that/Success
was significantly associated with h Power and not/Power. While more evidence
is clearly needed to understand the relationship of these two motives, such a re-
sult is not inconsistent with the general argument being presented. For it is the
women who are most assertive (high in h Power) who might be most likely to
fear the consequences of their assertiveness for interpersonal relationships (high
in / Success). But the Situation is quite different for men, since they have no rea-
son to fear their assertiveness. So among men it may be those who are ambiva-
lent about expressing power (high / Power score)—perhaps for some of the psy-
chodynamic reasons mentioned by Winter—who will also be ambivalent about
successful achievement (high / Success score). Both men and women high in
/ Power and / Success are in conflict over feelings of weakness, compensate for
those feelings through assertiveness, and fear the retaliation or rejection that
may come from such assertiveness.
OTHER FEARS
Correlates of Veroffs Need for Power Score as a Measure of Fear of Weakness
As mentioned in Chapter 8, Veroff has interpreted the measure of n Power he
developed as indicating anxiety about the ability to exercise power or fear of
weakness. His conclusion is justifiable on two grounds. First, to derive his cod-
ing System, he obtained stories written while Student candidates for office were
awaiting election returns, that is, when they were anxious about whether or not
they would be able to exercise power. So we might expect his scores to reflect
that anxiety. In contrast, the methods of arousing power motivation used by
Winter (1973) and Uleman (1972) encouraged subjects to feel confident of their
own strength. Second, the characteristics of people scoring high in Veroff
n Power suggest they are concerned about being weak.
Table 10.10 summarizes some key results from a national sample survey un-
dertaken in 1976 (Veroff, Depner, Kulka, & Douvan, 1980). So far as the men
are concerned, the Interpretation seems straightforward. Those scoring high in
Veroff n Power are more often the sons of blue-collar workers who earn good
money themselves, but from a Job of low prestige that they do not regard as
very interesting. However, they feel good about their competence at work, re-
port that they very often interact affectionately with their wives (which, as
Veroff (1982) remarks, "comes close to asking about their sexual activity"), and
feel that being a father is very important. In other words, it looks as if they are
compensating in the family area for feeling that they lack prestige and impor-
tance in the Job area. The fact that they stress their competence at work despite
its unimportance, as well as in the sexual area, suggests protest masculinity. In
effect, they are saying, "Even though my job does not amount to much in my
own eyes or the eyes of the world, I am a good worker, make good money, and
am a good (sexually competent) husband and father." The fact that they admit
to drinking too much and taking drugs also suggests that there is some anxiety
about being powerful, since, as we reported in Chapter 8, drinking alcohol is
one way men increase feelings of power. None of these behaviors, except drink-
ing, characterizes men who have a score high in Winter n Power, who have a
more confident approach to exercising power.
The results for women are less clear-cut but can be interpreted in a similar
way. Those high in Veroff n Power are particularly likely to mention satisfac-
tions obtained from work outside the home (as contrasted with housework),
which seems to contribute to their high sense of self-esteem. They do not ex-
Table 10.10.
SELF-REPORTED CHARACTERISTICS OF MEN AND WOMEN WITH A STRONG FEAR OF
WEAKNESS: HIGH VEROFF n POWER SCORE (after Veroff, 1982)
Area of Assessment Characteristics of Men (N = 508) Characteristics of Women (N = 700)

General well-being Rating high in self-esteem
Background Being a son of a blue-collar worker
Work Having a high income Mentioning ego satisfactions from work
Being in an unprestigious occupation Preferring work to leisure
Perceiving work as not interesting Feeling high job satisfaction
Feeling good about work competence
Family Having high marital interaction Feeling low satisfaction with housework
Seeing fatherhood as fulfilling a major
value
Behavior Having a drinking problem causing family Reporting low reliance on informal
trouble support in crises
Taking drugs to relieve tension
press satisfaction with various aspects of family life the way other women do.
And since they do not rely on others for support in crises, they appear more in-
dependent than other women, who find more satisfaction in interpersonal rela-
tionships. In short, they are strongly work oriented, which could be interpreted
as compensation for feelings of weakness or unimportance as a wife and mother
or as a worker in comparison with men. That is, since prestige in our society
goes with success primarily in a person's occupation, these women may feel infe-
rior in strength in this respect in comparison with men and compensate by con-
centrating on job success. However, there are no signs of anxiety associated with
this focus on the world of work, so the interpretation that it is the result of a
fear of weakness is on shakier ground than in the case of men. The most that
can be concluded is that the results for women can be interpreted as meaning
that high Veroff n Power indicates a fear of weakness as a woman, which leads
to a compensatory striving for job success, which is not characteristic of women
in general.
Fear of Intimacy
Recently Gilligan and Pollak (1982), in a study of themes in fantasy, reported
that men's stories more often than women's stories include images of violence in
connection with affiliative relationships. They suggest that this may be inter-
preted to mean that men fear the consequences of intimacy, just as Horner had
earlier argued that women feared the consequences of success. The finding de-
serves further exploration, since it is consistent with the expectation that all so-
cial motives have avoidance aspects.
Obviously there are many more specific fears—for example, fear of the dark,
fear of high places, fear of being closed in, or fear of school—that have been
studied and that might be considered for discussion in a chapter on avoidance
motives. They have not been included here on the grounds that the learning as-
sociated with them does not qualify as a general motive disposition affecting a
variety of behaviors.
CONCLUSION
The State of knowledge about the avoidance motives is not very satisfactory.
Several different avoidance motives have been measured, and individuals who
score high in these measures behave in understandably different ways. Yet we do
not know which are the most important avoidance motives, how to measure
them best, or how they relate to each other. We are not even entirely sure
whether avoidance motives differ theoretically in significant ways from approach
motives. Much more systematic work needs to be done in this area comparable
to the work that has been carried out on approach motives like the need for
power and the need for achievement, for which adequate measures have been
developed.
NOTES AND QUERIES
1. Make a list of the situations that make you anxious. What do you seek in
such situations? Are your goals similar in each Situation?
2. According to AronoflPs analysis, what would be likely to happen on the is-
land of Saint Kitts if living and public health conditions improved? After
you have made your predictions, consult Aronoff (1971).
3. Construct a model like Atkinson's that accounts more directly for the fact
that subjects high in Test Anxiety do better when they think they are doing
well and worse when they think they are doing badly.
4. If individuals high in / Failure do less well on exams or laboratory tasks,
why should they generally get better grades in school?
5. List the grades you want to get for each of the courses you are currently
taking. Then list the grades you have received in similar courses in the past,
and decide whether your goals for your current courses are defensive, as de-
fined in the chapter, showing a high / Failure.
6. Why would subjects high in / Failure be more responsive to what others
think?
7. If a coach urges a team to do better for "dear old Siwash," what effect is
this incentive likely to have on a person high in n Achievement, high in
n Power, high in n Affiliation, and high in / Rejection?
8. Try to design a study that would show definitively whether or not fear of
rejection, fear of failure, and the need for approval are different aspects of
the same avoidance motive.
9. Suppose a female executive is asked to play golf with her male boss. Explain
some of the motivational factors possibly present in the Situation that would
affect how well she played.
10. Describe at least two conditions in which a man high in / Success could be
expected to perform less well.
11. In Chapter 2, homosexuality was associated with oral aggressiveness (late
Stage I); in Chapter 9, with early sexual maturity during homosocial group-
ings; and in this chapter, with Stage II (autonomy strivings). How could all
these relationships possibly be true? Is it helpful to think in terms of what
might lead to homosexual attachments and what motives might arise after
they have been formed?
12. If psychic mindedness in men is associated with identification with the femi-
nine principle and related to high / Power, why should psychic mindedness
in women also be related to high / Power?
13. If high Veroff n Power represents a fear of weakness, at least in men,
should it not be related to Winter's / Power, which also is associated with
feelings of inadequacy?
14. Would you expect the Hostile Press score to be more related to the/Power
score or the / Success score?
15. Perhaps all the avoidance motives that have been identified are aspects of
the same anxiety, or "tension reduction," motive of such great theoretical
importance to the behaviorists (see Chapter 3). One indication that this
might be true would be that they all influence behavior in the same way.
Review the behavioral correlates of the difFerent avoidance motives to exam-
ine the extent to which the behaviors they lead to are the same or difFerent.
Is there any Single type of behavior characteristic of one avoidance motive
that is not characteristic of other avoidance motives, so a case for keeping
avoidance motives separate can be made?
11
Motivational Trends
in Society
• Analyzing the Reasons for the Growth and Decline of Civilizations

Freud on the Motivational Origins of Civilization
Basic Motivational Structure and Culture
Objections to Motivational Analysis of Historical Events
Measuring Collective Motives
• The Collective Concern for Achievement, Entrepreneurship, and Economic

Growth
A Motivational Explanation for the Growth and Decline of Ancient Civilizations
Reasons for the Loss of Achievement Motivation and Economic Decline
Waves of Achievement Concern and Economic Expansion in the Middle Ages
How Collective Achievement Concerns Predict Rates of Economic Growth Among
Modern Nations
Other Factors Affecting the Relationship of n Achievement Levels to Rates of
Economic Growth
Achievement Motive Levels and Political Protest
• The Collective Concern for Affiliation and Civil Rights
• The Collective Concern for Power

The Imperial Power Motive Pattern and Empire Building
Consequences of the Imperial Power Motive Pattern for Public Health
• Historical Shifts in Collective Motive Levels

Motivational Explanations for Events in English History from 1400 to 1830
Motivational Factors in the History of the United States from the Founding of the
Republic to the Present
Motive Levels and the Success of Automobile Companies
• Origins of Collective Motivations

Ideological Factors Affecting Collective Achievement Concerns
Environmental Factors Influencing n Achievement Levels
Sources of the Collective Concern for Power
Sources of the Collective Concern for Affiliation
• Difficulties in Interpreting Motive Trends in Society

The Meaning of Collective Motive Scores
415
• ANALYZING THE REASONS FOR THE
GROWTH AND DECLINE OF CIVILIZATIONS
Cultures, like individuals, differ greatly not only from each other, but from
themselves at different moments in time. Some are peaceful, others aggressive;
some rieh, some poor; some expansive and mobile, others stay-at-home. Anthro-
pologists, historians, economists, political scientists, and philosophers have often
tried to figure out why. Why were the Romans such geniuses at military and
civic organization, and the Greeks not? Why were the Greeks so successful eco-
nomically for some hundreds of years before Christ only to disappear for a time
as a nation of importance in world history? Why did the Roman Empire rise
and fall? When a second flowering of civilization oecurred on the Italian penin-
sula during the Renaissance, why was it in the arts rather than in military sci-
ence, as at a much earlier period? What caused the British Empire to expand
over the entire face of the globe in the nineteenth Century and to decline almost
equally rapidly in the twentieth? Why were the British more successful than any
other European nation?
Answers by historians to such questions tend to be given in terms of partic-
ular events in history, such as a battle that was won or lost, suddenly favorable
terms of trade, or the discovery of a new economic resource to exploit. Such
events are of great importance, but they appear to become influential because of
national aspirations or collective motivations. Yes, the British were favored in
the nineteenth Century by being relatively free from land invasion, but they were
also more enterprising than other peoples. And in the twentieth Century they
seem to have lost their "drive," although they oecupy the same territory and are
still free from land invasion. Or consider the ancient Romans. Once they got or-
ganized, they operated from a base that had many natural advantages, but first
they had to come back from defeat after defeat on land and at sea at the hands
of Hannibal and other Carthaginian generals. Why did they bounce back again
and again? Scholars inevitably turn to explanations in terms of character or col-
lective motivation. Some nations or cultures appear to be more "motivated"—to
be more active in military, economic, or aesthetic spheres—at some times than
others. What has psychology had to say about the role of collective motivations
in history and society?
Freud on the Motivational Origins of Civilization

The formal discussion was started in a serious way by Sigmund Freud
(1930/1958) who asked the most fundamental question of all: Why should peo-
ple create any kind of civilization? Why should they not live like other animals,
in simple social Orders in which primitive urges for food, sex, or aggression are
directly and immediately gratified? Why do they construet elaborate governmen-
tal struetures, norms governing interpersonal relations, artistic and intellectual
monuments to people's pride?
He found at least part of the answer in a biologically determined sexual
urge that a son has to sleep with his mother and to destroy his arch-rival, his
416
Motivational Trends in Society 417
father. Freud called this basic motivational conflict the Oedipus complex after
the Greek mythical hero who did unwittingly kill his father and sleep with his
mother. Because he loves his father and for other reasons, every son eventually
observes the incest taboo, represses acting out his sexual interest in his mother,
and seeks substitute satisfaction for his thwarted sexual drives in work and in
creating a social structure through which his urges can be satisfied later and less
directly. The son learns that to get what he wants—let us say sex—without de-
stroying himself and the family, he must "grow up," accept the rules on which
his parents insist, go to school, learn a trade, and observe the rules of courtship
and marriage.
According to Freud, millions of young men acting out this primitive moti-
vational drama succeed in creating the elaborate edifice we call society. By this
line of reasoning the most complex social institutions are interpreted as the cre-
ation of simple motivational urges and the conflicts among them. Capitalism, for
instance, which is often seen as the motivational force behind modern economic
development, has been interpreted as an acting out of the urge to "accumulate
filthy lucre," or more literally, feces (Fromm, 1947). Sexual or libidinal urges in
boys get satisfaction first from various parts of the body; then the boys transfer
their attachment to the mother (see Chapter 2). Some peoples get fixated at the
"holding-on" stage (see Table 2.4), when pleasure is gained especially from accu-
mulating and playing with feces. Such peoples are motivated especially to accu-
mulate and to become capitalists, according to some followers of Freud.
Other theorists have drawn attention to other types of motivations for civi-
lized activities that derive from the unfolding of sexual and aggressive instincts
in the nucleus of the family. N. O. Brown (1959), for instance, has argued that
the fundamental motive for creating civilization is the fear of death, of one's
own mortality. Social institutions become ways of protecting people from think-
ing about death or guaranteeing immortality. D. Bakan (1966) believes that the
fundamental motives for the two sexes differ: Men are agentic, seeking dominion
and control. Women, on the other hand, seek communion or the kind of interre-
latedness that permits them to rear their children successfully.
Basic Motivational Structure and Culture

But how could such general motivations account for variations in culture
growth? Ralph Linton (1945), an anthropologist, and Abram Kardiner (1945), ä
psychoanalyst, reasoned in the 1940s that a certain social or economic condition
might favor the development of people of a certain personality type who would
in turn create a Special type of culture. Child-rearing practices differ greatly
from one culture to another and even in the same culture from one period in its
history to another. It seemed logical to suppose that some society might stress
toilet training, for instance, in a way that would create more people who, when
they grew up, would continue to be interested in hoarding and presumably
therefore in capitalistic enterprise as a way of satisfying this urge.
Many case studies were written of particular societies (see, for example,
Erikson, 1950/1963) that showed that the type of adult motivations that charac-
terized a society did indeed seem to reflect the way children were reared. For in-
stance, Kardiner (1945), in studying the people of Alor, an island in the Portu-
guese East Indies, found that the adult males seemed to be suspicious and dis-
trustful, primarily because they had been fed irregularly and abandoned by their
mothers in their first month of life because the mothers had to go out to work
the family garden daily to survive. An economic condition (daily gardening by
the mother away from home) created a child-rearing practice (frequent desertion
of children) that produced a personality type (suspicious and distrustful) that
made it difficult for the Alorese to collaborate and produce a thriving economy.
Wolfenstein and Leites (1950) analyzed family relationships in a number of
U.S. films of the 1940s and discovered what appeared to be an unusual Variation
in the way the classic son-mother-father triangle was pictured. In these films the
father figure or older man typically appeared as bumbling and inefFectual,
whether as an ineffectual or corrupt sheriff who could not keep order in a West-
ern town or as an ignorant, perhaps immigrant father who did not understand
what was going on. The hero or son figure therefore typically had to take over
and set things right. The women to whom the hero was attracted often appeared
to be "bad women," but on further acquaintance they turned out to be whole-
some types who just happened to be substituting for a friend as a nightclub
singer for the evening. Thus, the intensity of the oedipal conflict is eased for the
U.S. son, because his father is pictured as somebody too inefFectual to be worth
hating and because the son is sexually attracted to bad women (not at all like
mother), although they really are like mother underneath.
It does not take any great leap of the imagination to explain the participa-
tion of male students in the revolts of the 1960s in just these terms. The young
men were showing the characteristic U.S. disrespect for authority figures (fa-
thers, policemen, and College administrators), whom they regarded as bumbling
and ineffectual and from whom they must take over the reins of power if society
were to improve. Similarly, the young men appeared to be consorting with bad
women in sex or drug orgies and "be-ins," yet on further examination these bad
women turned out to be wholesome types from the next block. Of course, some
further explanation would be needed as to why the U.S. version of the oedipal
conflict led to Student apathy in the 1950s and revolt in the 1960s, but these an-
alysts would argue that to observe that Americans in the family rebel success-
fully against their fathers is a Start at explaining why there is so much distrust
of authority in the United States.
A group of scholars used this approach extensively shortly after World War
II in an attempt to find the underlying motivation for the rise of Naziism in
Germany and particularly the German persecution of the Jews (Adorno, Frenkel-
Brunswik, Levinson, & Sanford, 1950). They discovered it in an authoritarian
personality syndrome characterized by "dominance-subordination, deference to-
wards superiors; sensitivity to power relationships; need to perceive the world in
a highly structured fashion; excessive use of stereotypes; and adherence to what-
ever values are conventional in one's setting" (Greenstein, 1969). After carefully
examining case histories of highly authoritarian and nonauthoritarian people,
they stressed the importance of early experiences in the family in developing this
particular cluster of traits:
When we consider the childhood Situation . . . we find reports of a tendency toward

rigid discipline on the part of the parents with affection which is conditional rather
than unconditional, i.e., dependent upon approved behavior on the part of the child.
Related to this is the tendency . . . to base [family] interrelationships on rather
clearly defined roles of dominance and Submission. . . . Forced into a surface Sub-
mission to parental authority, the child develops hostility and aggression which are
purely channelized. The displacement of a repressed antagonism toward authority
may be one of the sources and perhaps the principal source, of his antagonism to-
ward out groups. (Adorno et al., 1950)
In other words, children who are reared strictly and rigidly, often in a way that
suggests rejection by the parents, turn their hostility away from their respected
parents toward outgroups, such as Jews and other minorities. German families
were considered to rear their children more often in these ways and therefore to
produce the motivational characteristics of the authoritarian personality syn-
drome, which in turn was responsible for the rise of Naziism in Germany.
Objections to Motivational Analysis of Historical Events

Critics of this type of motivational analysis of collective phenomena have been
numerous. They have contended—sometimes quite persuasively—that to attrib-
ute psychological motivations to historical or social events is at worst absurd or
at best unnecessary and circular. At times motivational analysis of social char-
acteristics does appear circular. The reasoning runs as follows. A society re-
quires a certain personality type for its survival, so parents bring up their chil-
dren so that they have motives to act in ways required by society. German
society required order and discipline; therefore, German parents must have
trained children to want order and discipline. In the United States, people are
disrespectful of male authority figures. Therefore, the culture must structure the
mother-son-father triangle in ways that make sons want to belittle their
fathers.
Is anything added to knowledge by such Statements? Is not the desire to at-
tribute "motives" to events a primitive, animistic way of thinking, as we pointed
out in Chapter 2? In much the same way, a young child might State that an
apple feil from a tree because "it wanted to." We might well ask, Why does it
"want to" now and not ten minutes ago? Why did the Germans produce Nazi-
ism in the 1930s and not before? The trouble with defining motivations as re-
sponses to requirements of the social System is that it does not explain change.
Motives become "explanations" for whatever happened that cannot be accounted
for by other causes.
The only way out of this circularity is to find some independent way of
measuring the alleged collective motivation. Then observers can check to see
whether the motives in fact existed that they are using to account for social
events. They might be able to show, for instance, that some motivational charac-
teristics of the authoritarian syndrome were in fact higher in Germany at the
time of Naziism than they had been previously or than they were in other coun-
tries, which did not produce Naziism. A study of motives in society avoids the
circularity accusation only when motives can be measured and their contribution
to what happened in society assessed independently of the events themselves.
Some scholars have argued that independent assessment of motivation is un-
necessary even if it can be done. Most economists, political scientists, and histo-
rians have managed to discuss quite sensibly what has happened in history with-
out reference to human motivations; at most they make a simple assumption
about the desire to pursue rational self-interest, which appears sufficient to ac-
count for most of the social events with which they are dealing. For example,
Robert Waelder (1960), a psychoanalyst, was once asked by a social scientist
studying the rise of Naziism to give his explanation of what motivated this de-
velopment. He replied that he thought the success of Prussian militarism in
bringing about the unification of Germany had much to do with creating faith
in strong-arm methods. Furthermore, the Germans had tried the democratic
process and found it unable to cope with problems of inflation and disunity in
the 1920s. In other words, Waelder was arguing that the immediate past experi-
ences of the German people, combined with a simple motivational theory about
people promoting their own self-interest, were sufficient to account for the rise
of Naziism. He goes on to say, "I was then interrupted by my host, a noted an-
thropologist. This was not what I had been expected to contribute. As a psycho-
analyst I should point out how Naziism had developed from the German form
of child rearing. I replied that I did not think that there was any such relation-
ship; in fact, political opinion did not seem to me to be determined in early
childhood at all. This view was not accepted and I was told that the way the
German mother holds her baby must be different from that of mothers in de-
mocracies" (Waelder, 1960).
Waelder's view that childhood motivations are largely irrelevant for explain-
ing complex political events is widely shared. But they may be irrelevant for
several reasons that are worth examining a little further. They may be irrelevant
because motives formed in childhood are simply less important in governing
adult behavior than motives arising out of the concrete historical Situation in
which the adults find themselves. Thus, many social scientists who reject child-
hood analyses of social character are quite willing to accept similar analyses
based on contemporary motivational pressures. An economist might argue that
certain groups of people were forced to leave the farm and go to the city be-
cause of population pressure in the rural areas. They left because there was too
little to eat for so many mouths to feed. They did leave. Therefore they must
have been forced to want to. Although this form of motivational analysis is
much more populär than the childhood form, it is just as circular. It is assumed
that population pressure—or scarcity of food—will make people want to mi-
grate. In fact, that is not always the case. Some peoples migrate under such con-
ditions and others do not. What is needed here, as in the study of childhood
motives, is some direct measure of what people want. Only then will the ob-
server be able to predict who will migrate and who will not when there is little
to eat (see McClelland & Winter, 1971).
Granted that some motivational analysis is ideally desirable and even neces-
sary, is not the easiest way to find out about people's motives simply to ask
them what they want? Thus, one implication of Waelder's argument about the
German people would appear to be that if a public opinion poll had been con-
ducted in Germany, say in 1930, it would have been quite possible to find out
that the German people wanted Naziism in one form or other. Childhood moti-
vations may be considered irrelevant, because just asking people what they want
is a better guide to their actions. It is a fact that public opinion surveys have
been found useful in identifying wants that have important social effects. For ex-
ample, such surveys are now routinely used to make economic forecasts, because
the number of people who say they want to buy a car will in fact predict how
many cars will be bought in the following year—a fact that is of tremendous im-
portance not only to the auto industry, but to economic planners as well. Politi-
cians have also learned to watch attitude surveys to see what their constituents
want and to try to shape their campaign promises and public Performance in the
light of that information.
Here it is useful to distinguish, as explained in Chapter 8, between demands
and motive dispositions. A demand is an immediate situational pressure that
evokes about the same motivational intent in all people exposed to it. Thus, a
hurricane produces a desire to seek safety; an insult, a desire to retaliate; a bank
failure, a desire to withdraw one's money; or cutting prices in half, a desire to
buy. Such demands create intents that can be accurately reported by people in
public opinion polls, and they will predict actions in the short run but not in
the long run. So far as individuals are concerned, demands will predict immedi-
ate responses, as when an experimenter asks a subject to take off his or her coat,
but any given demand will not predict whether a person will usually wear a
coat. Measures of motive dispositions are necessary to explain general long-term
"drifts" of behavior over varieties of demand situations—who is more likely to
wear a coat over the long term or who is more likely to resist a request to take
off his or her coat.
Reasoning by analogy, we would expect then that measures of collective
motivation would be better at predicting long-term social trends than at predict-
ing what Will happen at any particular time and place. Motivational concerns of
nations or cultures, like motivational concerns of individuals, should predict ag-
gregates of actions—say, the number of homicides over a five-year period—
rather than particular actions, such as a riot in Detroit. To get at the reasons
for a riot in Detroit, it might be better to conduct a public opinion poll to find
out what people in the area wanted in that particular place at that particular
time.
Measuring Collective Motives

As we have seen, individual psychology escaped the charge of circularity in
making motivational analyses by not simply inferring motives from actions, but
by insisting on checking these inferences independently by studying fantasy or
what was going on in a person's mind. Thus, at the individual level, it is not
enough to infer that because a person is studying very hard in school, he or she
therefore must have a strong need to achieve. Instead, we must observe thoughts
or fantasies and see how prominent the theme of achievement is in them. Only
then can we State, based on evidence other than the act of studying itself, that a
person has a strong need to achieve, for studying hard may be motivated by a
variety of motives—the need to achieve; the desire for social approval; or devo-
tion to parents, who are expecting a strong scholastic record. It is impossible to
teil from a Performance itself what motivated it. And asking people what their
motives are may also not always be helpful, because they will often give the an-
swer that appears socially desirable at the moment. Instead, it is necessary to
ask individuals to teil a number of stories, which may be coded for a variety of
motivational themes, to find out what is generally of concern to them. This was
explained in Chapter 6.
The same line of reasoning can be applied to assessing collective motives. It
is circular reasoning to assume that because of the Nazi dictatorship, the Ger-
mans had a strong need for power. It would have been unsafe and possibly mis-
leading to ask the German leaders or the general public in a public opinion poll
what their motives were. Instead, it is necessary to find out what the German
people had on their minds by studying their collective fantasies—as they occur
in populär plays or novels, stories for children, or often-retold folk tales. Such
material has been coded for the three most widely studied individual mo-
tives—namely, n Achievement, n Affiliation, and n Power.
The behavior of individuals scoring high in these motives may give some
clues as to what nations will be like when their collective fantasies are focused
on achievement, affiliation, or power. But we obviously cannot jump from indi-
vidual to collective psychology in any simple, straightforward way without care-
ful empirical checks. What, in fact, are nations with a high concern for affilia-
tion like? Will they protect civil rights more? Will they be less aggressive toward
their neighbors? Will they develop better local transportation Systems or faster
direct-dialing telephone Systems? Will they be more likely to Sponsor child-
centered forms of education?
A number of empirical studies have begun to provide answers to questions
like these. But before we review them, it should be noted that the "need" termi-
nology used to describe motives at the individual level may not be appropriate
at the collective level. When a large number of achievement themes are found in
children's stories of a given country, it is not altogether clear what the measure
reflects in the country. It may mean that because there are a lot of individuals
in the culture whose fantasies turn to achievement, some individuals with high
n Achievement are more likely to be involved in writing children's stories. In
this sense, the achievement content of the stories may reflect the mobile or suc-
cessful personality types in the culture. Or it may be considered not so much a
reflection of the motivation of individuals as an index of a prevailing ideology
that shapes everybody's thoughts to a greater or lesser degree in that nation. In
this sense we might conceive of collective "needs" that bear only an indirect re-
lationship to the distribution of individual needs in the country. Most authors
continue to use the "need" terminology to signal the relationship of collective to
individual motives, as we shall see, but they do not intend to suggest that the
measures have identical meanings or identical effects in action at individual and
collective levels.
THE COLLECTIVE CONCERN FOR ACHIEVEMENT,

ENTREPRENEURSHIP, AND ECONOMIC GROWTH
The case best worked out for the influence of motivational levels on society is
the one that links the achievement motive to rate of economic growth. It grew
out of studies made of the activities of individuals scoring high in n Achieve-
ment. As noted in Chapter 7, they tend to be attracted to moderately challeng-
ing tasks, to do better at such tasks, to take personal responsibility for a Job at
work, and to seek feedback on the quality of work—all with the goal of con-
stantly trying to do better. It seemed likely that such men would make good
business entrepreneurs, because the business role demands taking carefully calcu-
lated risks, using feedback to improve the profit picture, innovating to promote
efficiency, and so forth. As data summarized in Chapter 7 make clear, individu-
als high in n Achievement in fact are more likely to turn out to be successful
entrepreneurs.
It seemed reasonable to infer that if a culture's collective fantasies were ori-
ented around achievement, that culture should contain a larger number of entre-
preneurs. The hypothesis was checked by examining the folk tales from thirty-
nine preliterate societies (Kulakow, cited in McClelland, 1961). Folk tales in
many ways are like the imaginative stories written by individuals and scored for
n Achievement. They are short, with a beginning, a middle, and an end; they in-
volve fantasy; and they have the added advantage of representing an oral tradi-
tion, meaning that they have been told and retold by different individuals from
the culture over time. This should mean that if there is a dominant focus on a
theme like achievement in the culture, it should get into the stories. The oniy
disadvantage of using folk tales is that they are not as structured as the stories
written by individuals, so the subcategories of the scoring System cannot readily
be applied. Therefore, only the number of achievement images was counted in
the twelve folk tales collected for each culture. Then, from ethnographic rec-
ords, it was determined whether there were any full-time entrepreneurs in the
cultures. Among twenty-two preliterate societies whose folk tales contained an
above-median amount of achievement imagery, 74 percent contained at least
some full-time entrepreneurs, as contrasted with only 35 percent of the cultures
with below-median amounts of achievement imagery, a difference that is highly
significant. This was a clear instance of a hypothesis derived at the individual
level that was confirmed at the collective level, that is, by comparing a measure
of collective motivation with an index of a type of activity in the culture.
A Motivational Explanation for the Growth and Decline

of Ancient Civilizations
Another study pushed the reasoning one Step further: if collective achievement
motivation was associated with the presence of more entrepreneurs, their ener-
getic activity should be associated with signs of greater economic growth. The
hypothesis was first tested by examining achievement motive trends and stages
of economic development in the history of ancient Greece. Three periods in
Greek history were distinguished—a period of growth, roughly from 900 B.c. to
475 B.c.; a period of climax, roughly from 475 B.c. to 362 B.c.; and a period of
decline, from 362 B.C. to 100 B.C. Comparable samples of Greek literature were
selected from each of these periods and scored for n Achievement. The selec-
tions centered around six different themes: the relationship of people to their
gods, farm and estate management, public funeral celebrations, poetry, epigrams,
and war Speeches of encouragement. For example, in connection with the last
theme, all instances of a man's encouraging others to fight in Homer's Iliad
were scored for the early period, a similar speech by Pericles as recorded by
Thucydides was scored for the period of climax, and a speech by Demosthenes
was scored for the later period. All of these selections, consisting of 8,440 lines
or over 80,000 words, were mixed together and scored blindly—in the sense that
the scorer did not know from which period the selections came—for n Achieve-
ment (Berlew, 1956; McClelland, 1958b).
Historians agree that the period from 475 to 362 B.C. represented the peak
of Athenian Greek commercial success, leaving out a very brief period of expan-
sion under Alexander later; quantitative support for this consensus was obtained
by constructing maps of the area within which Athenian Greece traded in the
sixth, fifth, and fourth centuries B.C. (Berlew, 1956). Such maps could be made
because Greece traded largely by exporting surplus wine and olive oil in earth-
enware jars to various places overseas, and these were discarded when empty.
Archaeologists have identified places where the remains of such vases have been
found and determined which Century they represent from markings on them.
Locating the distances of these pottery remains from Athens gives a rough index
of Athenian trade area, which shows a steep rise in the fifth Century B.C. as
noted by historians. See Figure 11.1, which plots the average n Achievement
score per hundred lines of text in the early, middle, and late periods. Note that
the n Achievement level was highest before the expansion of Athenian commer-
cial civihzation and had declined significantly during the period of economic cli-
max, thus foreshadowing the subsequent economic decline in the area. The gen-
eral hypothesis is confirmed that collective levels of high n Achievement precede
economic growth, and collective levels of low n Achievement precede economic
decline.
Independent confirmation of this relationship was obtained in a different
way. As noted in Chapter 7, Aronson (1958) discovered that individuals high in
n Achievement used more S shapes and diagonals and fewer multiple waves in
"doodles" or free-line drawings than individuals low in n Achievement. Again,
by analogy from the individual to the cultural level, it was reasoned that vases
in the early period of Greek history should contain more S shapes and diagonals
and fewer multiple waves than vases from later periods in Greek history. Exami-
nation of 242 vases from the periods of growth, climax, and decline showed the
same decline in this index of collective n Achievement level as that obtained
from scoring literary productions. Approximately 65 percent of the vases from
the early period showed an above-median number of diagonals or S shapes, as
contrasted with about 45 percent of the vases from the period of climax and 15
percent of the vases from the period of decline (McClelland, 1958b). Figure 11.2
4.0
3.5
3.0 <S
/ \Trade Area 25
8 n Achievement
Level
\ - ÜI
2.0 H z:
..5 g |
~ 2
<§
1.0 c
•S* 1
0.5
1 1 1 1 1 1
800 700 600 500 400 300 200 100
B.C. B.C.
Time (in Centuries Before Christ)
Figure 11.1.
Average n Achievement Level Plotted at Midpoints of Periods of Growth, Climax, and Decline of
Athenian Commercial Civilization as Reflected in the Extent of Trade Area (after McClelland,
1961).
shows a vase from an earlier period containing many diagonals and S shapes
and few multiple waves as contrasted with a vase from a later period showing
the reverse pattern.
Davies (1969) subsequently extended this type of analysis to explain the col-
lapse of ancient Minoan civilization on the island of Crete, from which no deci-
pherable written records exist. He obtained an n Achievement index score for
each pot in three samples of pottery representing three different periods in the
history of Minoan civilization. He found once again that by this index, the aver-
age level of n Achievement was significantly higher in the early period, before
the building of great cities and palaces in the middle period (see Table 11.1).
There was some rebuilding after great earthquakes in 1640 B.c., but following
the eruption of a volcano at Thera in 1450 B.C., which caused tidal waves to
flood Minoan seaports, the civilization went into a steady decline. Previous his-
torians have supposed that these external disasters were responsible for the de-
cline of Minoan civilization, but Davies argues that the real cause was loss of
the achievement motive, or the psychic energy needed to build and maintain a
thriving commercial empire, for the natural disasters such as the tidal waves did
not come far enough inland to affect many of the cities directly.
What is especially exciting about such findings is that changes in the collec-
tive concern for achievement preceded the rise and fall of a civilization. It is
common sense to infer that levels of achievement concern were high during the
periods of climax of these civilizations, for would that not explain why they
Figure 11.2.
Greek Vases from Earlier and Later Periods Illustrating S Shapes and Diagonals Characteristic of
High n Achievement (Left) and Multiple Waves Characteristic of Low n Achievement (Right). The
left, black-figured Greek vase is from the early fifth Century B.c. and the right, red-figured vase,
from the late fifth Century B.c.
Table 11.1.
MEAN POTTERY INDEX OF n ACHIEVEMENT LEVELS FOR THREE
PERIODS IN MINOAN CIVILIZATION (after Davies, 1969)
Mean Pottery
Period Number of Vases Index of n Achievement0
Early (2600-1950 B.C.) 37 5.2^
Climax (1950-1650 B.c.)

Earthquake (1650 B.C.) 67 3.1
Late (1600-1400 B.C.)
Eruption of Thera (1450 B.c.)
Collapse of Minoan civilization 24 2.7
a
S shapes plus diagonals minus multiple waves or zigzags.
b
Significantly higher than for later periods.
were doing so well at that time—expanding trade and building great palaces and
cities that could be discovered centuries later by archaeologists? But such rea-
soning falls into the trap of simply naming a motive to account for a certain
type of activity. Actual measurement of achievement motive levels shows that
they were higher before the periods of climax and presumably accounted for the
increased entrepreneurial activity that accumulated the resources to build the

palaces and cities in the period of climax.
Reasons for the Loss of Achievement Motivation and Economic Decline

But why did achievement motive levels decline? Two explanations have been ad-
vanced. McClelland (1961) drew on the knowledge gained from studying the be-
havior of mothers of sons high in n Achievement (see Chapter 7) to argue that
as Greek merchants accumulated more and more resources as a result of their
increased entrepreneurial activity, they used the resources to support household
slaves to look after their children. These slaves, particularly the pedagogue who
walked the child to school, were so available for child care that it seemed likely
they would spoil the children and not make the kind of demands for achieve-
ment that mothers make who have sons high in n Achievement.
McClelland (1961/1976) advanced a similar argument to explain why the
South lost the U.S. Civil War. There too, successful merchants or planters, pre-
sumably high in n Achievement, acquired household slaves, who might well
have spoiled the "young master" to promote their own self-interest, thereby un-
dermining the achievement demands of the parents. This would mean that suc-
ceeding generations in the South would have lower n Achievement levels than in
the North, where there were no household slaves, with the consequence that the
South would in time become less developed economically than the North. Most
historians agree that this was the case at the time of the Civil War and that one
of the main reasons the South lost the war was that it had lesser economic re-
sources to sustain it in the long run.
Unfortunately, such theorizing, while it appears reasonable, has never been
directly tested. Another explanation for the decline in n Achievement levels fol-
lowing increases in prosperity is simpler and derives directly from Atkinson's
model of the aroused tendency to achieve as a function of the probability of
success and the incentive value of success (see Table 7.2). According to that
model, as the probability of success increases, the incentive value of success de-
creases, so when tasks are easy or commercial success is assured, the level of
aroused achievement interest declines. The model is designed to explain the
temporary arousal value of tasks, but it might be supposed that if tasks regu-
larly had little arousal value, the level of dispositional n Achievement would
also decline. There is some empirical evidence for such a supposition in a study
by Andrews (1967) in which he found that levels of n Achievement declined
over three years for businessmen working in a Company that did not present
moderate achievement challenges to them. It is difficult to know just how find-
ings at the individual level apply to the collective level, but one simplifying as-
sumption is to suppose that collective concerns also follow the Atkinson model.
Then one could argue that the perception of national prosperity from commer-
cial success lowers the collective interest in achievement, just as it does at the
individual level.
What is needed to validate such a claim is a systematic study showing that
periods of prosperity are followed by a decline in the collective concern for
achievement. Several instances of such a relationship are found throughout this
chapter, and one careful study has been made of the effect of profitability in a
Company on lowering subsequent levels of n Achievement. (This will be dis-
cussed later.) But more attention needs to be given to testing directly the appli-
cability of the Atkinson model at the collective level.
Waves of Achievement Concern and Economic Expansion in the Middle Ages

The explanation for the rise and fall of ancient civilizations was so simple and
straightforward, and its empirical confirmation so striking, that other investiga-
tors immediately began testing the motivational explanation for economic
growth in other times and places. Perhaps the level of achievement concern is
important only at very early stages of national development. Could it also ac-
count for such complex historical events as the commercial opening up of the
New World by the Spanish Conquest in the sixteenth Century or the Industrial
Revolution in England in the nineteenth Century? Did the voyage of Christopher
Columbus signify a high level of achievement concern in Spain at the time?
Why were voyages not being sponsored by other countries, such as England and
France, at the same time? Why did the English voyages come later? Was the
English level of achievement concern low earlier, and did it increase? These are
the kinds of questions that can be empirically answered using the new methodol-
ogy of coding populär literature for motivational concerns.
Cortes (1960, in McClelland, 1961) coded populär literature in Spain before,
during, and after the period of maximum commercial expansion into the New
World. Once again the n Achievement levels were estimated by coding a large
amount of literary material: about 100,000 words from fifty-six different authors
in each of three time periods. As Figure 11.3 illustrates, he found exactly the
same relationship as Berlew had obtained earlier for ancient Greece: n Achieve-
ment levels were high before the period of rapid commercial expansion of Spain
into the New World and declined prior to the collapse of Spain's economic em-
pire. The voyages of discovery were sponsored by a country whose n Achieve-
ment level had been high, giving it the energy and resources to pursue further
economic advantage in the New World. And again, prosperity—perhaps particu-
larly the accumulation of gold—lowered subsequent levels of n Achievement.
To look for motivational factors underlying waves of economic expansion in
England, Bradburn and Berlew (1961) examined a large amount of textual mate-
rial over a longer period of time, roughly from 1400 to 1830 A.D., representing
seven different epochs. They coded 150 pages of comparable text material from
forty to fifty authors for each time period, representing usually a half Century.
Their objective was to select passages from material that had roughly the same
form and function throughout the entire historical period studied. So they
scored selections from plays (about 2,240 lines per period), accounts of sea voy-
ages (about 2,500 lines per period), and street ballads (about 1,308 lines per pe-
riod). They chose street ballads to represent the thoughts and feelings of ordi-
nary people who listened to them recited in the streets, and plays and accounts
of sea voyages to represent what was on the minds of a better-educated class of
people. As in the case of ancient Greece, the amount of achievement imagery
Climax, Decline,
Economic Growth, 1200-1492 1492-1610 1610-1730
11 - 22
10 20 o
8 18 <u
ear)
n Achievement Level
the
8. 8 16
00
8.
Pons
es 14
12 CD
g T3
o
10 c
Q o
o (_
u c
6 ÖO
c
'S.
D.
4
Shipping
r;
1200 1300
1
1400 1500 1600 1700
Year
Figure 11.3.
Average n Achievement Level in Literature at Midpoints of Periods of Economic Growth, Climax,
and Decline in Spain from 1200 to 1700 (after McClelland, 1961).
per one hundred lines for each of these three categories of material was aver-
aged to get an overall index of the n Achievement level for that time period.
Then a rough index of economic activity in England was obtained by exam-
ining gains in coal imports at London. To get a time series reflecting quantita-
tively approximate changes in rates of expansion of the English economy was
difficult, but Bradburn and Berlew finally found that reports of coal imports at
London were the best continuously available figures that would approximately
reflect how much fuel the expanding economy was using at different time peri-
ods. When the coal figures first become available, coal was important not only
for domestic fuel but also for the dressing of meat, washing, brewing, and dye-
ing—all key activities in the developing English economy. The amount of coal
imported increased regularly from the earliest to the latest period considered,
and the higher the level of imports, the greater in absolute terms the increase to
the next time period.
What the investigators wanted to know was whether the gain was greater or
less than could be expected given the level of economic activity in the previous
period. To find out, they used a linear regression formula to predict the gain
from the previous level and then observed whether the actual gain was more or
less than expected in terms of the overall trend. Since this statistical method of
estimating rates of growth is used throughout this chapter, it is important that
it be understood. An analogy may make the point clear. It is common to calcu-
late what a person's expected grade in College should be based on an intelligence
or aptitude test score. People with more scholastic aptitude generally get better
grades. Given this relationship, it is possible to estimate that a given person,
Susan, with her aptitude should be able to get an A. If she gets a B, we say she
is an underachiever. She is doing well, but not as well as could be expected
given her aptitude. Similarly, Bradburn and Berlew estimated whether the level
of economic gain in a given period represented overachievement or under-
achievement in relation to its starting level, given the general relationship
between levels of economic activity at the outset of the periods (analogous
to aptitude) and subsequent gains in economic activity (analogous to grades
achieved).
As shown in Figure 11.4, according to this measure, gains in coal imports
were above expectation in the early period (representing overachievement), then
dropped considerably (representing underachievement), and finally rose sharply
toward the end of the eighteenth Century. The n Achievement levels also shown
in Figure 11.4 behave as if they are having an influence on subsequent rates of
economic growth. They are high in the sixteenth Century, preceding the first
rapid rate of economic growth; then decline preceding a drop in the rate of eco-
nomic growth; and finally rise again sharply, predicting the great increase in
economic activity in England preceding the Industrial Revolution toward the
end of the eighteenth Century.
Furthermore, the English n Achievement level in the earliest period (1400
to 1550) is lower than later (1550 to 1600), whereas in Spain the n Achievement
levels dropped sharply over the same time period (Figure 11.3). Thus, a compar-
ison of the n Achievement levels in the two countries from 1400 to 1600 pre-
dicts that entrepreneurial voyages of exploration should be more frequent in the
earlier part of this period for Spain than for England, and more frequent for En-
gland than for Spain in the later period. This was actually the case, as shown by
the voyages from Spain to South America in the early part of the seventeenth
Century and the voyages from England to North America in the early part of
the sixteenth Century.
Findings for England in Figure 11.4 differ from the earlier ones for Greece
and Spain in that there are two waves of high n Achievement followed by in-
creases and decreases in rates of economic growth. Economic historians have
pointed out that the great improvement in the British economy in the early
nineteenth Century required many concrete changes, such as agricultural im-
provement through crop rotation; enclosures of land, permitting a marked in-
crease in sheep population; the development of an extensive System of transport
through turnpike roads; improved credit; an increase in available capital, and an
educational System to turn out the kind of workers needed by industry. Figure
11.4 suggests that a heightened level of achievement motivation in the country
as a whole led individuals to make all these changes in various sectors of the so-
6.0
Above-
Normal
Increase
o =>
5.0 + 1.0 2 fe
Rate of
Economic Growth
A-
>
I
2c s
4.0 A \
ZI °
§ ö
3.0 -1.0
Ü £
o
Below-
Normal
Increase
I _L
1550 1600 1650 1700 1750 1800
Year
Figure 11.4.
Average n Achievement Levels in Samples of English Literature (1500-1800) Compared with Rates
of Acceleration and Deceleration in English Economy, Measured by Coal Imports at London (data
from Bradburn & Berlew, 1961).
ciety. In other words, the motive measure comes closer to answering the ulti-
mate question as to why all these changes were introduced. They were the
means of improving how the society functioned and were designed to satisfy an
increased collective concern for improvement.
How Collective Achievement Concerns Predict Rates

of Economic Growth Among Modern Nations
So much for historical changes in rates of economic development. What about
modern nations? Can variations in their rates of growth be predicted from levels
of collective achievement concern? To test such a possibility, some form of pop-
ulär literature had to be found that would yield roughly comparable data from
countries widely different in level of modernization, industrialization, religious
orientation, and so forth. Stories read by children in the public schools were
chosen, because they seemed roughly comparable from country to country in

purpose, in length, in imaginativeness, and in the extent to which they were rel-
atively uninfluenced by historical events of temporary significance. They seemed
comparable to the folk tales used in the study of preliterate cultures in the sense
that they often represented an oral tradition in which recurrent cultural themes
were shaped by educational authorities to produce stories that everyone would
agree should be read by children. Even though the same plot may be present in
the textbooks of several countries, such as Aesop's fable of the Fox and the
Raven, it is retold in ways that can stress, for example, either achievement or
power. Consider, for example, the two versions of the well-known story about
the hare and the tortoise in Table 11.2. The plot is the same, but in Portugal
the tortoise Starts the action by wanting to do better than the hare in a race
(n Achievement). In England the hare Starts the action by boasting of his prow-
ess (n Power) and making fun of the tortoise.
In most countries educational authorities take great care to make sure that
children read in school what is considered right and proper. What educational
authorities think is right for children to read may well represent the motiva-
tional ideas uppermost in the minds of people in a country. At least, this was
the assumption made in coding children's stories for n Achievement, n Power,
and n Affiliation from twenty-three countries in 1925 and thirty-nine countries
in 1950 (McClelland, 1961). In each case, twenty-one stories were selected at
random from at least three different readers from the second, third, and fourth
grades.
As a measure of rate of economic growth, McClelland (1961) employed fig-
ures on kilowatt-hours of electricity used, since national income figures were not
available in the early period, nor were they comparable from country to coun-
try. The advantage of the electricity figures is that they are available in compa-
Table 11.2.
ALTERNATIVE VERSIONS OF THE STORY OF THE TORTOISE AND THE
HARE
One day the tortoise said to the hare, "Say, Ma'am, / bet that I can get to that bramble
patch first."* The hare replied, "Say, mister, surely you must be off your rocker . . ."
(Portugal)
A hare one day made fun of a tortoiseb because of his slow pace and boasted of his own
great speedb in running. The tortoise only laughed and said, "Let us try a race. I will run
with you five miles. The fox will be the judge."
(England)
a
Scored for achievement imagery.
Scored for power imagery.
rable form from country to country, they obviously are a key ingredient in all
types of economic activity, and they correlate highly with national income fig-
ures where both sets of figures are available. To get an estimate of whether the
economy was expanding more rapidly or slowly than expected, the level of elec-
tricity consumed at one time period was related to the gain to the next time
period across all countries, and the resulting regression equation was used to
predict whether a given country was gaining more or less than expected as
compared with the average of all countries.
Table 11.3 summarizes the correlations between collective n Achievement
levels at two different time periods and collective rates of economic growth also
at two different time periods. Note first of all that n Achievement levels in chil-
dren's readers around 1925 correlate significantly with the rate of economic
growth measured in relative gains in electricity consumed per capita between
1929 and 1950, but not with gains in electricity consumed at a still later time
period (1952 to 1958). Furthermore, when stories from a larger sample of coun-
tries were coded for n Achievement around 1950, the resulting index of collec-
tive achievement motive levels predicted gains in electricity consumed between
1952 and 1958 or 1950 and 1967, but not earlier (1929 to 1950). In both cases,
high n Achievement levels preceded more rapid rates of economic growth. And
the n Achievement level in 1950 was not the product of more rapid rates of
growth earlier (1929 to 1950), meaning that people's n Achievement levels did
not rise as a response to greater economic opportunities, as economists often as-
sume is the case (Papanek, 1962).
A more concrete illustration of just what these correlations represent is
given in Table 11.4. Note that many more countries high in n Achievement in
1950, like Turkey, Greece, and the Soviet Union, show an above-expected
Table 11.3.
CORRELATIONS OF NATIONAL n ACHIEVEMENT LEVELS WITH GAINS IN ELECTRICAL
POWER CONSUMPTION IN KILOWATT-HOURS (after McClelland, 1961/1976)
Gain in Log gain Log gain

n Achievement n Achievement kwh per Capita in kwh in kwh
(1925) (1950) (1929-1950) (1952-1958) (1950-1967)
n Achievement (1925): N = 22 .26 .53* .10
n Achievement (1950): N = 39 .03 .43** .39* (N = 40)
Log gain in kwh (1929-1950) .13
Log gain in kwh (1952-1958)
*p < .05.
**p < .01.
Table 11.4.
RATE OF GROWTH IN ELECTRICAL OUTPUT (1952-1958) AND NATIONAL n ACHIEVEMENT
LEVELS IN 1950 (after McClelland, 1961)
Deviations from Expected Growth Rate« in Standard Score Units

National National
n Achievement Above n Achievement Below
Lewis (1950)b Country Expectation Levels (1950)* Country Expectation
High 3.62 Turkey + 1.38
n Achieve- 2.71 Indiac -hl.12
ment 2.38 Australia + 0.42
2.33 Israel + 1.18
2.33 Spain + 0.01
2.29 Pakistan0 + 2.75
2.29 Greece + 1.18 3.38 Argentina -0.56
2.29 Canada + 0.06 2.71 Lebanon -0.67
2.24 Bulgaria + 1.37 2.38 France -0.24
2.24 U.S.A. + 0.47 2.33 Union of S. Africa -0.06
2.14 W. Germany + 0.53 2.29 Ireland -0.41
2.10 U.SS.R. + 1.62 2.14 Tunisia -1.87
2.10 Portugal + 0.76 2.10 Syria -0.25
Low 1.95 Iraq + 0.29 2.05 New Zealand -0.29

n Achieve- 1.86 Austria + 0.38 .86 Uruguay -0.75
ment 1.67 U.K. + 0.17 1.81 Hungary -0.62
1.57 Mexico + 0.12 1.71 Norway -0.77
0.86 Poland + 1.26 1.62 Sweden -0.64
1.52 Finland -0.08
1.48 Netherlands -0.15
1.33 Italy -0.57
1.29 Japan -0.04
1.20 Switzerlande -1.92
Correlation of n Achievement level (1950) X 1.19 Chile -1.81
Deviations from expected growth rate = .43, 1.05 Denmark -0.89
p < .01. 0.57 Algeria -0.83
0.43 Belgium -1.65
a
The estimates are computed from the monthly average electrical production figures, in millions of kilowatt-hours, for 1952 and 1958,
from United Nations, Monthly Bulletin of Statistics, January 1960, and Statistical Papers, Series J, World Energy Supplies, 1951-1954 and
1955-1958.
The correlation between log level 1952 and log gain 1952-1958 is .976.
The regression equation based on these thirty-nine countries plus four others from the same climatic zone on which data are available
(China-Taiwan, Czechoslovakia, Romania, and Yugoslavia) is log gain (1952-1958) = .9229 log level (1952) + .0480.
Standard scores are deviations from mean gain predicted by the regression formula (M = —.01831) divided by the Standard deviation
of the deviations from mean predicted gain (SD = .159).
bßased on twenty-one children's stories from second-, third-, and fourth-grade readers in each country.
c
Based on six Hindi, seven Telugu, eight Tamil stories.
d
Based on twelve Urdu and eleven Bengali stories.
e
Based on twenty-one German Swiss stories, mean = .91; twenty-one French Swiss stories, mean = 1.71; overall mean obtained by
weighting German mean double to give approximately proportionate representation to the two main ethnic population groups.
growth rate in electricity consumed, whereas many more of those low in

n Achievement, like Uruguay, Belgium, and Switzerland, show a below-expected
economic growth rate. The case of Switzerland highlights the fact that we are
predicting rates of gain in economic activity rather than levels of economic ac-
tivity. Switzerland is a wealthy country with a high level of economic activity,
but it shows a slow rate of gain relative to other countries at the same high
level, like the United States. The lead time between 1950 achievement motiva-
tion levels and subsequent growth is so short—certainly as compared with the
lead time in ancient Greece, which was over two hundred years—that we must
assume the effect is not taking place through teaching children to be more
achievement oriented who later grow up to become active entrepreneurs. (See
Chapter 7.)
Other Factors Affecting the Relationship of n Achievement Levels

to Rates of Economic Growth
Is it not possible that collective n Achievement levels merely reflect some other
societal characteristic that is really responsible for more rapid economic growth?
For example, might they not reflect educational levels of the work force, which
in turn produce more rapid economic growth? Perhaps better-educated people
understand better the great importance of achievement to improve the Standard
of living of all the people by economic means. Southwood (1969) has carried out
an extensive investigation that checks out such possibilities. He obtained infor-
mation for each country on such structural variables as occupational stratifica-
tion, amount of secondary school education, population size, union membership,
degree of open political competition allowed among parties, amount of political
unrest, and literacy. He correlated each of these variables with n Achievement
levels, rates of economic growth, and also political unrest, the variable of chief
interest to him.
None of the other national characteristics assessed explains away the rela-
tionship between collective n Achievement levels and rates of economic growth.
For example, consider the influence of educational level. The percentage of
high-school graduates in a country in 1955 is not related to rate of economic
growth from 1952 to 1958 (r = - . 0 9 ) nor to the n Achievement level in 1950
(r = —.22). If the nations are split in terms of whether they are above or
below average in the percentage of high-school graduates, the collective
n Achievement level relates significantly to economic growth both among better-
educated and poorer-educated countries.
However, investment in education does seem to accelerate economic growth
if collective n Achievement levels are high. McClelland (1966) found that when
investment in higher education was high relative to the economic level of a
country in 1950, more countries developed rapidly if they were high in
n Achievement level, but not if they were low in n Achievement level. This is
important, because policy makers may reason that investment in education is
important for promoting economic growth, because more knowledge obviously is
required to produce a technologically advanced industrial society. But the data
show that simply increasing the knowledge base in a population does not auto-
matically lead to more rapid economic growth: the level of achievement motiva-
tion must also be high for people to make more energetic use of such knowl-
edge. Chapter 13 will show how, at the individual level, the contribution of
greater skill to Performance depends on the level of motivation. Just increasing
know-how does not lead to better Performance unless motivation for that perfor-
mance is high. The same principle applies to the economic Performance of
countries.
Or consider group discrimination. In Southwood's (1969) data, the degree of
discrimination against minority groups as of 1965 is significantly correlated with
the rate of economic growth between 1952 and 1958 (r — .38; p < .05) and
with a low level with n Achievement in 1950 (r = .20). Could it be that dis-
crimination against minorities forces them into business and increases
n Achievement, which increases the rate of economic growth? At first the hy-
pothesis appears reasonable, because many persecuted minorities like the Jews or
Quakers in the West, Jains or Marwaris in India, or the Gurage in Ethiopia
have been successful in business. Hagen (1962) has argued that it is their loss of
Status that leads such minorities to try to regain respect through the only chan-
nel open to them, namely, increased business activity.
There are two problems with this hypothesis. One is that there is no system-
atic evidence that discriminated minorities in general counterstrive. As Weber
(1904/1930) pointed out, the Catholics in England were a persecuted minority,
but they did not become successful in business, and there seem to be as many
cases of this type as of the counterstriving variety. Secondly, it is difficult to de-
termine whether the achievement motive is the cause or result of group discrimi-
nation. Certainly there is ample evidence that minorities successful in business
are persecuted because they are perceived as getting rieh at the expense of oth-
ers. McClelland (1961) points out that Hermes, representing the image in myth
of the rising Athenian businessman, is pictured as tricky and dishonest, which is
exactly the way such businessmen were perceived by the Athenian aristoeraey.
Such an image has characterized minorities successful in business ever since (see
LeVine, 1966), whether they be Jews in the West or the Ibo in Nigeria.
The only systematic way to make sure the n Achievement-economic growth
relationship is not the accidental produet of its association with other variables
is to enter all the variables that predict economic growth into a multiple regres-
sion equation and see whether n Achievement contributes to explaining rates of
economic growth when the effect of the other variables has been removed. That
is, all the factors like group discrimination that are significantly associated with
rates of growth are entered into an equation along with collective n Achieve-
ment levels to see which factor best predicts rates of growth, as well as whether
that factor continues to be significantly associated with rates of growth when the
influence of other factors has been partialled out or controlled. When South-
wood followed this procedure, he found that the n Achievement variable comes
out as being important compared with other factors in predicting rates of eco-
nomic growth and continues to contribute significantly to the variance in rates
of growth after the contribution of other variables has been removed (McClel-
land, 1961, 1976).
Achievement Motive Levels and Political Protest
High n Achievement does not always lead to rapid economic advance smoothly.
It can result in violence. Southwood (1969) reasoned that high achievement con-
cerns combined with low opportunities to achieve would produce frustration,
anger, and political protest. This, in fact, is what he found, as shown in Figure
11.5, which plots the relationship between collective n Achievement levels and a
measure of internal war only for countries below average in the proportion of
students going to high school. For these countries, the higher the achievement
aspiration of the country, the greater the amount of internal violence, presum-
ably because young people with high aspirations do not have the opportunity to
advance for lack of education. In contrast, for countries above average in the
proportion of students in high school, there is no relationship between the col-
lective n Achievement level and domestic violence. In fact, for a somewhat more
sensitive index of political disorder—that is, turmoil (the number of strikes, gov-
ernment crises, riots, and demonstrations)—the relationship is even negative
with collective n Achievement levels. Among the more educated countries, the
higher the n Achievement level in 1950, the less the turmoil (r = —.23).
10
Argentina
India
8 -
Iraq
- 6 -- Hungary
/
South Africa
Brazil /*• Lebanon
^ 4 Pakistan • S
Iran •
Turkey •
^S^ U.S.S.R.
~
Chile • Spain
Italy • • Mexico •
Svvit/crland Uruguay Portugal
• 1 1 • • 1 1 1
0.5 1.0 1.5 2.0 2.5 3.0 3.5
n Achievement (1950)
Figure 11.5.
Relationship of Internal War to Collective n Achievement Levels (1950) Among Countries Low in
High-School Education (data from Southwood, 1969).
Among the poorly educated countries with low educational opportunities, the
greater the n Achievement level in 1950, the greater the turmoil (r = .46;
p < .05).
McClelland (1975) also found a significant association between n Achieve-
ment levels and violent protests in the history of the United States. Most of the
protest measures involved strikes, riots, or demonstrations over lack of economic
opportunity for labor or minority groups being discriminated against. In decades
in U.S. history when n Achievement levels were high, the index of violent eco-
nomic protest also was high, and vice versa. A strong urge to get ahead, if it is
blocked, tends to lead to violent protest.
The fact that some minority groups with high n Achievement levels tend to
get ahead faster may lead to discrimination against them, as has been noted.
This is vividly illustrated by what happened to the Ibo in Nigeria. LeVine
(1966) has shown that Ibo boys report a much higher percentage of achievement
dreams than Hausa-Fulani boys, who come from a dominant tribe in northern
Nigeria. The Yoruba, another major tribal group in Nigeria, falls in between
these two extremes in achievement motivation. The Yoruba were more modern-
ized because of earlier contact with the West. For instance, in the 1920s, of the
twelve doctors in the country, eight were Yoruba; four European; and none, Ibo
or Hausa-Fulani. By the 1930s forty-nine, or 33 percent, were Ibo; seventy-six,
or 48 percent, were Yoruba; and one was Hausa-Fulani. Many have testified to
the extraordinary speed with which the Ibo had moved upward in all walks of
life in Nigeria, from university professors to poets and novelists. LeVine (1966)
comments further: "Apart from this success in becoming a major part of the
modernized, professional and governing elite of Nigeria, Ibo entrepreneurs, fan-
ning out all over the country, have also enjoyed a substantial measure of suc-
cess. The Ibo have a reputation for being willing to take any sort of job, no
matter how menial, when they first enter a town and then working their way
up, living frugally and accumulating resources until they become wealthy." Ibo
people were persecuted and killed, particularly in the Hausa-Fulani part of the
country; they protested against the discrimination and ultimately revolted be-
cause they feit blocked. They finally seceded from the country, trying to set up
the separate State of Biafra. Eventually they were overpowered by the central
government of Nigeria. Their history illustrates dramatically the dynamic rela-
tionships between n Achievement levels, economic advance, discrimination, and
violent protest.
• THE COLLECTIVE CONCERN FOR AFFILIATION

AND CIVIL RIGHTS
Societal characteristics associated with collective levels of n Affiliation have not
been extensively studied, but what evidence there is suggests that they are very
much what we would expect from knowing how individuals high in n Affiliation
behave. Table 11.5 lists the main cultural characteristics associated with a high
concern for affiliation expressed in folk tales. In such societies, wife beating is
uncommon, infanticide is not practiced, daughters tend to live closer to their

parents, children are rated to be more often obedient, and adults regularly help
each other more. In short, cultures concerned about affiliation show that they
care more for people or respect what are generally called civil rights.
The analysis of the correlates of national n Affiliation levels in children's
stories leads to similar conclusions. Countries high in the concern for affiliation
tend to have lower psychogenic death rates from homicide, suicide, ulcers, cir-
rhosis of the liver, and hypertension. Barrett and Franke (1970) correlated chil-
dren's reader n Affiliation scores with national death rates from these causes
both in 1925 and in 1950. In nine out of ten time-lagged comparisons (from
n Affiliation in one period to death rates in the next), the correlations, although
small, were negative, suggesting that high n Affiliation predicts less violence and
the kinds of interpersonal aggravation that accompany heightened psychogenic
death rates of this type.
McClelland (1961) also found that national n Affiliation levels were associ-
ated with a greater interest in having children. As direct evidence for this con-
clusion he notes that in the countries studied in 1950, collective n Affiliation
levels were significantly associated with the birth rate (r = .41; p < .05). Peo-
ples more concerned about affiliation tended to have more children. However,
the relationship was reversed for the 1925 sample of countries: at that time, peo-
ples more concerned about affiliation had fewer children. He explains the shift
by demonstrating that infant mortality was a key variable strongly related to the
birth rate in 1925 but not in 1950. At the earlier time period, countries with a
high birth rate also had a high infant mortality rate, but this was not true by
1950, when public health measures had reduced infant mortality rates every-
where. Countries high in n Affiliation in 1925 had a lower infant mortality rate
and also a lower birth rate, meaning they had fewer children and took better
Table 11.5.
CROSS-CULTURAL CORRELATES OF n AFFILIATION IN FOLK TALES (after
data in McClelland, Davis, Kaiin, & Wanner, 1972)
Cultural Characteristic r with Collective n Affiliation N

Wife beating -.50* 17
Infanticide -.50* 21
Residence distance of daughter -.61** 14
Frequency with which children are obedient .33* 31
Dependency on others for help in adulthood .34* 36
*p < .05.
**p < .01.
care of them. By 1950, when infant mortality was no longer an issue, peoples
concerned with affiliation could express their concern for children more directly
by having more of them.
A high collective concern for affiliation appears to have served as a protec-
tion of individual rights against violent acts by the government. In The Achiev-
ing Society, McClelland (1961) observed that ruthlessly totalitarian regimes
nearly always have been marked by high collective levels of n Power and low
collective levels of n Affiliation. Children's readers showed this pattern in Ger-
many in 1925, before the rise of Hitler; in Japan before the rise of the military
dictatorship in the 1930s; in Russia during the rule of Joseph Stalin; and in
Spain both in 1925 and in 1950, when the dictator Francisco Franco was on his
way to power or was in power. Of twelve clearly totalitarian regimes, 92 percent
showed this motivational pattern, as contrasted with only 18 percent of coun-
tries judged not to be controlled by dictators.
Since that time, figures for China have also become available, and they fit
the same pattern. In 1925 Chinese children's readers were high in n Power (1.50
Standard deviations above the mean for all countries) and low in n Affiliation
(2.01 Standard deviations below the mean for all countries), facts that are cer-
tainly not at variance with the common judgment that the country was under
the control of a military dictatorship at that time. By 1955, after the Communist
takeover, China's readers were showing one of the highest n Power levels re-
corded in any country at any time (3.27 Standard deviations above the mean),
and n Affiliation was still below the world average (0.55 Standard deviations
below the mean). These figures again indicate what most analysts would agree
was true—namely, that the Chinese Communist totalitarianism of the 1950s
was, if anything, even stronger and more severe than the earlier military
dictatorship.
Such findings make good theoretical sense in terms of known correlates of
n Power and n Affiliation at the individual level. As Chapter 8 made clear, in-
dividuals with high n Power want to be important and to have their way. A
strong power motive, however, can be held in check by a high affiliative motive
that leads to concern for others' welfare, which might be expected to prevent a
government from ruthlessly riding over the rights of individuals. In fact, if high
n Affiliation is not present, governments can be expected to be violent in trying
to get their way. Countries showing a higher n Power than n Affiliation in their
children's readers in the 1950s, such as Iran, are more likely to be politically
unstable and have a higher proportion of deaths from domestic violence (Feier-
abend & Feierabend, 1966). In contrast, countries that have been relatively
democratic world powers, such as Great Britain in 1925 or 1950 or the United
States in 1950, are characterized by high levels of n Power and n Affiliation.
The desire to impose the national will in these countries is held in check by a
concern for the welfare of others. While few social scientists have been willing
to place much faith in such indicators of collective motivation levels, high
n Affiliation in populär literature has considerable Utility in predicting that the
civil rights of individual Citizens of a country will be protected from govern-
mental oppression.
THE COLLECTIVE CONCERN FOR POWER

The Imperial Power Motive Pattern and Empire Building
Individuais with the imperial power motive profile (high n Power greater than
n Affiliation, and high Activity Inhibition) tend to be more effective managers
and empire builders (see Chapter 8). So it seemed reasonable to investigate
whether nations characterized by this motive profile are better organized empires
or have what political and sociological theorists speak of as greater "system ca-
pability." That is, governments vary in the extent to which they extract re-
sources from their members and use them effectively to Service individuals or
groups in the society or defend it against attack from outside.
McClelland (1975) selected two measures of governmental System capability:
the percentage of gross national product (GNP) taken by government away
from private expenditure and the percentage of GNP devoted to defense. Table
11.6 shows the proportionate amounts spent in these two ways by various coun-
tries, classified according to whether they displayed the imperial power motive
syndrome or some other motive combination. Every country with the imperial
power motive pattern collected a greater proportion of GNP for government use
than the median for all countries in the sample, as contrasted with half of the
countries with other motivational patterns, a difference that is statistically signif-
icant. Countries with the imperial power motive pattern also spent nearly 5 per-
cent of their GNP on the average for defense, as contrasted with about 3.5 per-
cent spent by other countries. In short, as expected, nations with the imperial
power motive pattern showed signs of having developed more System capability
than other nations in the sense that they collected more of their GNP in taxes
and spent more of it on defense. McClelland (1975) reasons that such countries
are typical of expansive, well-organized nations in the past like the great Roman
Empire organized in the period after the birth of Christ.
Consequences of the Imperial Power Motive Pattern for Public Health

Since individuals with the imperial power motive pattern also show more sus-
ceptibility to cardiovascular disease (Chapter 8), it seemed worth investigating
whether nations with such a motive profile showed higher death rates from this
cause. The results are summarized in Table 11.7. Once again, a prediction from
the individual level is confirmed at the collective level. Countries high in n Pow-
er and high in inhibition show the highest average death rates from heart dis-
ease and hypertension—significantly higher than those low in both factors for
hypertension or than those low in n Power and high in Activity Inhibition for
heart disease. In the latter instance, it appears that higher self-control leads to a
more regulär life, which protects against heart attacks if n Power is also low.
The averages for each quadrant should be compared with those for individuals
similarly classified in Figure 8.10. There, too, individuals low in n Power have
lower average diastolic blood pressure, suggesting they should be less susceptible
to death due to cardiovascular disease.
Table 11.6.
EXPENDITURES FOR GOVERNMENT (1959-1960) AND FOR DEFENSE (1950s) AMONG COUNTRIES WITH THE IMPERIAL
POWER VERSUS OTHER MOTIVE COMBINATIONS (1950) (after McClelland, 1975)
Low Activity Inhibition0 High Activity Inhibition0
Country
Percentage ofGNP Percentage o/GNP |
(Imperial
0
Group Country For Government For Defensef Syndrome) For Government For Defense §
High Belgium 28.9 3.0 Argentina 31.4 2.6
n Power > Canada 36.5 4.4 China — 3.6
n Affiliation Finland 39.5 2.0 Denmark 31.4 2.6
Iran — 5.6 W. Germany 41.3 3.8
Italy 31.8 3.0 India — 1.9
Mexico 20.6 0.7 Iraq 38.0 8.2
New Zealand 34.2 2.2 Pakistan — 3.9
Spain 25.9 2.2 Switzerland 36.8 2.9
Taiwan 29.2 12.3 U.S.S.R. 44.2 10.4
Union of S. Africa 33.4 0.8 U.S.A. 36.3 9.6
Mean 31.1 3.62 Mean 37.1 4.95
Low Australia 36.1 2.7 Austria 34.8 1.5
n Power or Chile 17.6 2.8 Bulgaria 30.3 —
n Affiliation France 33.0 5.9 England 33.3 6.7
> n Power Hungary 27.1 — Greece 22.2 5.1
Netherlands 40.0 4.0 Ireland 24.1 1.4
Norway 40.0 3.5 Israel 28.0 6.3
Portugal 22.2 3.2 Lebanon — 2.4
Sweden 38.1 4.7 Japan 40.5 1.6
Syria — 4.9 Poland 30.2 —
Tunisia — 1.8 Turkey 23.1 2.4
Mean 31.8 3.72 Uruguay — 1.0
Mean 29.6 3.16
a
Below or above the median number (46) of the word not in twenty-one children's stories.
b
Reciprocal of the figure for private consumption, after Russett and others (1964), Table 48; median for whole sample = 30.3.
c
After Russett and others (1964), Table 23; median for whole sample = 2.70.
Table 11.7.
DEATH RATES PER 100,000 INHABITANTS IN VARIOUS COUNTRIES FROM HEART DISEASE
(A) AND HYPERTENSION (B) AROUND 1968 AS A FUNCTION OF NEED FOR POWER AND
CONTROL (1950) (after McClelland, 1976)
A ctivity Inhibition
Low High
Group Country A B A B
Italy 246.0 33.0 West Germany 252.0 25.9
Union of 248.0 24.8 Switzerland 247.0 23.6
South Africa England 367.0 22.8
Belgium 320.0 23.9 U.S.A. 339.0 22.8
Finland 330.0 19.6 Argentina 156.0 22.8
New Zealand 306.0 13.6 Denmark 363.0 12.5
High Canada 255.0 13.6
n Power Spain 132.0 12.8
Taiwan 38.2 8.7
Average 277.4 18.8 Average 287.3 21.7
Hungary 302.0 33.0 Austria 326.0 28.4
Low Sweden 350.0 21.2 Uruguay 189.0 27.2
n Power Portugal 135.0 18.8 Ireland 374.0 25.2
Norway 290.0 16.9 Poland 140.0 20.0
Australia 328.0 15.5 Japan 76.0 18.2
France 200.0 11.9 Bulgaria 176.0 16.0
Netherlands 228.0 11.5 Greece 131.0 12.5
Chile 92.0 8.9 Israel 194.0 11.2
Average 240.6 17.2 Average 200.7 19.8
Note: Mexico is omitted because a very high percentage of deaths was not classified. Mann-Whitney U-tests, heart disease: high high
low high, p = .07; hypertension: high high > low low, p = .04.
High n Power and low inhibition in individuals is associated with heavy

drinking (see Figure 8.7). McClelland, Davis, Kaiin, and Wanner (1972) also
found a similar relationship at the cultural level by examining drinking ratings
for cultures classified according to whether they were high or low in n Power
and Activity Inhibition scores in folk tales. As Figure 11.6 shows, the propor-
tion of heavy-drinking societies is highest in those cultures for which collective
power motive scores are high and Activity Inhibition scores are low. Cultures
high in n Power and high in Activity Inhibition are least often heavy-drinking
societies, a result that is also paralleled at the individual level.
In reviewing such studies, we are reminded of early social psychologists who
spoke about a "group mind" as if a nation or a group had a mind of its own
that led to characteristic group activities. Certainly motive levels as assayed in
100
U Low Inhibition
80 I j H i g h Inhibition
II 60
•2 u
40
20
N= 10 12 12 10
High n Power Low n Power
Figure 11.6.
Percentage of Heavy-drinking Societies as a Function of n Power and Inhibition Scores in Folk
Tales (after McClelland, Davis, Kaiin, & Wanner, 1972).
group products—whether of nations or preliterate cultures—are associated with

group outcomes very similar to the actions associated with similar motive dispo-
sitions assayed at the individual level. The parallels are so close that it seems
reasonable to infer that the measures of collective motivation are somehow tap-
ping central motivational or ideological tendencies among individuals in the
group.
• HISTORICAL SHIFTS IN COLLECTIVE MOTIVE LEVELS

Motivational Explanations for Events in English History from 1400 to 1830
Several studies have investigated the relationship between historical events
and shifts in collective motive levels. The demonstration that high collective
n Achievement levels preceded periods of economic expansion initiated this
mode of analysis. Giliberto (1972, cited in McClelland, 1975) extended it by
coding the samples of populär literature collected from various periods in En-
glish history for n Power (Winter System) and n Affiliation.
One of his most striking findings is illustrated in Figure 11.7, which plots
average n Power scores from 1550 to 1800 obtained from samples of plays and
street ballads. Historically this Covers a time when thece was a power struggle
8.00
Ballads
7.00 -
(Lower Orders)
6.00
5.00
•2 4.00
3.00
2.00
_L _L _L J_ J_
1550 1600 1650 1700 1750 1800
1501-1575 1576-1625 1626-1675 1676-1725 1726-1775 1776-1825
Tudor Elizabethan Puritan Recovery Oligarchie Peace
Dominance Glory Revolt Rule at Home
Time Periods in English History
Figure 11.7.
Levels of n Power in Lower and Upper Classes: England, 1500 to 1800 (after McClelland, 1975).
going on between the king and his party and Parliament, which was more
broadly representative of the people. In the earliest period, the Tudor kings, rep-
resented particularly by Henry vm, had Consolidated the power of the aristocra-
cy, which ruled the people in an authoritarian manner. If we assume that the
excerpts from plays reflect more the motive levels of the ruling class, labeled the
"upper Orders" in Figure 11.7, then the power motive level of this class was
much higher than that of the "lower Orders," as reflected in the street ballads
they produced and listened to, during the period of Tudor dominance. Subse-
quently, the people struggled against the ruling class and succeeded in gaining
power during roughly the period of the Puritan revolt (1626 to 1675). The fact
that they gained ascendance over the aristocracy is reflected in the much higher
level of n Power in street ballads around 1650 than in the plays, which presum-
ably reflected the thoughts and feelings of the ruling class. It is not possible to
know from these data whether shifts in n Power levels led to the successful re-
volt of the people's party or vice versa, but it is interesting to note that the mo-
tivational changes fit well with what was happening in the country, if we assume
that the two types of populär literature represent the states of mind of different
social classes.
Giliberto pooled the n Power scores from all sources and plotted them for
various periods in English history along with the n Affiliation and n Achieve-
ment scores, as shown in Figure 11.8. Note first that the collective level of
n Power is much higher than the level of n Affiliation during most of this peri-
od, suggesting that there should have been a good deal of internal violence, op-
pression, and violation of civil rights if cross-national findings are applicable
here. This was certainly the case. The famous British historian Trevelyan (1942)
describes the period of Tudor dominance in the 1500s as characterized by "se-
nile, ferocious feudalism." Around 1650, during the Civil War and its aftermath,
the English were busy killing each other in the name of religion; even as late as
6.00
Expressing Power ./
~ 5.00 - y / ^ ^ v ^ Power
/
f
•=5 4.00 -
A
/i Achievement A,
D.
S 3.00 \
o
c \
o
I 2.00
o ^ \
Affiliation \
1.00-
1 I 1 1 1 1
1550 1600 1650 1700 1750 1800
1501-1575 1576-1625 1626-1675 1676-1725 1726-1775 1776-1825
Tudor Elizabethan Civil War, Recovery Oligarchie Economic
Dominance Glory Puritanism Rule, Growth
Wesleyanism
Time Periods in English History
Figure 11.8.
Levels of n Power, n Affiliation, and n Achievement in Populär Literature: England, 1500 to 1800
(after McClelland, 1975).
1750, working-class people were regularly executed for petty crimes such as rob-
bing someone of a Shilling or breaking into a house with the intent to steal
(McClelland, 1975). According to the motivational data, only the periods
around 1600 during the reign of Queen Elizabeth and around 1800 could be de-
scribed as eras of good feeling in which internal violence should have been less
prevalent. In general, historians could argue that these were indeed periods of
lesser violence against human life.
The second point Giliberto noted about the curves in Figure 11.8 is that the
two sharp rises in n Affiliation in, 1600 and 1700 were both followed by periods
of religious reform on behalf of the oppressed, which were followed by a drop in
n Affiliation. Such a sequence of events makes sense in motivational terms, be-
cause n Affiliation at the individual level is strongly associated with a concern
for people: if the people were suffering, as they were during the period of Tudor
dominance or during the English Civil War, it would seem only natural to those
high in n Affiliation that Steps should be taken to alleviate their condition. In
the early period during the Civil War, the Steps taken involved putting the
Champions of the common people in charge in Parliament; at the later period
after the Wesleyan Methodist revival, around 1750, the Steps involved pushing
for laws to protect civil rights and to help the poor.
But notice that while the impulse to help the oppressed may derive from a
heightened n Affiliation, if it is combined with a high level of n Power, the
n Affiliation level immediately drops. This switch makes sense in terms of the
motivational dynamics involved, because acting powerfully on behalf of other
people means thinking in terms of what was defined as socialized power in Chap-
ter 8, and a high score for socialized power is regularly associated with a iow
level of n Affiliation (McClelland, 1975). In commonsense terms, if there is a
powerful concern to help other people, it may become so strong and determined
that in the end it overrides sensitivity to whether people are getting hurt. The
devotion to a larger good, or social justice, can become so strong that it once
again violates individual rights. It was just such a passion for social justice that
led the Puritans to kill their English brothers and sisters, or Northerners in the
United States to kill their Southern brothers and sisters in the Civil War in the
name of protecting the rights of blacks. Such a dynamic sequence is suggested
by the English data, although two instances are not enough to demonstrate its
generality.
Motivational Factors in the History of the United States

from the Founding of the Republic to the Present
McClelland (1975) made a similar analysis of motivational trends in U.S. history
on a more fine-grained basis. For every decade from 1780 to 1970, he obtained
scores for three social motives from stories in school textbooks, hymns, and pop-
ulär fiction written during that decade. Since the types of material varied in the
frequency with which the motives appeared in them over time—for example,
power motive images appeared seven or eight times as often as achievement or
affiliation images in fiction—the mean motive scores per decade were converted
to Standard scores for each type of material (mean Standard score = 50;
o
o
B
H
O
E
>^
X
c
o
ure
puB ' U O I J O I J 'SUUIÄH u o "8
P9SBQ S9J0DS p9ZIpJBpUBJS UB9JAJ s; 32
SD = 10), and the three Standard scores for a decade were then averaged to
give the best estimate of the motive level for that time period.
The results are plotted in Figure 11.9 at the midpoints for each decade. The
n Achievement curve relates to periods of economic growth, as would be expect-
ed. It was consistently high after the Civil War before and during the period of
the most rapid economic growth in U.S. history. deCharms and Moeller (1962)
had earlier shown that the same rise in n Achievement, scored from school text-
books alone, was associated with innovativeness in the general population as
measured by the number of patents issued. Figure 11.10 shows how the rise in
n Achievement in the period from 1870 to 1900 was followed by a sharp rise in
the patent index from 1890 to 1930, as well as that the subsequent fall in
n Achievement from 1900 to 1930 was also followed by a fall in the patent
index from 1930 to 1950. The patent index can be considered an index of the
entrepreneurial drive to find new and better ways of doing things throughout
80
70
Patent Index
60
3
O
cu 50
- 40 < 2,
o £
C <U
0> >
30
20
10
_L _L _L _L _L _L
1810 1830 1850 1870 1890 1910 1930 1950
Midpoint of 20-Year Period
Figure 11.10.
Mean Frequency of Achievement Imagery in Children's Readers and the Patent Index in the United
States, 1800 to 1950 (after deCharms and Moeller, 1962).
this time period. The finer-tuned and more comprehensive n Achievement levels
in Figure 11.9 confirm the deCharms and Moeller findings and also suggest that
the low level of n Achievement from 1900 to 1930 might have had something to
do with the Great Depression in the 1930s. The data also suggest that the re-
cent rise in n Achievement may foreshadow a period of relative economic pros-
perity in the last quarter of the twentieth Century.
The shaded portions in Figure 11.9 indicate the periods in which n Power
was greater than n Affiliation. What is notable about them is that they occur
regularly before wars in U.S. history. McClelland observed that if one allows a
lead time of fifteen years, shaded areas predict war and nonshaded areas, peace,
in eleven out of thirteen comparisons before World War i (p < .05); one of the
"exceptions" was the Spanish-American War, which lasted three months and
might not be considered a war at all. After World War n, if the lead time is
shortened to around five to ten years, four out of five of the predictions of war
and peace are correct. In short, in the history of the United States, there is a
statistically significant association between whether or not n Power is greater
than n Affiliation and whether the country subsequently goes to war.
In commonsense terms, this relationship can be interpreted as meaning that
when the country is feeling assertive or ruthless (high n Power, low n Affilia-
tion), it is likely to respond to a challenge or provocation by going to war,
whereas the same or a similar challenge when the spirit of the country is more
affiliative does not lead to war. Once again, the relationship between the impe-
rial power motive pattern and violence is confirmed at the collective level.
The relationship found in English history between the n Power/« Affiliation
ratio and social reform can also be observed in U.S. history. Table 11.8 lists the
decades in U.S. history when n Power was at least moderately high (Standard
score = 46 or more) and n Affiliation balanced it (not over ten points higher or
two points lower). In the decade following each of these instances, n Affiliation
dropped markedly, creating the imperial power motive pattern in which n Pow-
er was greater than n Affiliation. This led in the next decade to a wave of social
reform, starting with the period of Jeffersonian democracy around 1805 and
ending with the Crusade for civil rights around 1965.
This is clearer evidence of the dynamics suggested by events in English his-
tory. A relatively high n Affiliation, if combined with a moderate-to-high
n Power, leads to a concern for the rights of the oppressed—in the U.S. case in-
cluding the poor, women, and blacks—which converts into a powerful move-
ment for social reform accompanied by a drop in sensitivity to whether people
are getting hurt (lower n Affiliation). This then creates the imperial power mo-
tive pattern, which has regularly been associated with going to war against an
outside enemy. It is suggestive that Crusades for social reform have six times fol-
lowed this motivational dynamic in U.S. history and that Crusades for social re-
form have so often been followed by the country's going to war. McClelland
notes that the balanced n Power/« Affiliation ratio is followed by the imperial
power motive syndrome six out of seven times, whereas an imperial power mo-
tive pattern develops only once after nine other motive combinations, a differ-
ence that would rarely have arisen by chance.
Table 11.8.
MOTIVE PATTERNS, SOCIAL REFORM, AND WAR IN U.S. HISTORY: STANDARDIZED MOTIVE SCORES FROM POPULÄR
LITERATURE (after McClelland, 1975)
Balanced Imperial
Midpoint Midpoint of Midpoint of Social Next Time
of Decade n Power n Affiliation Next Decade n Power > n Affiliation Next Decade Reform Period of War
1785 62 71 1795 64 > 59 1805 Jeffersonian 1812, War of 1812
democracy
1815 50 57 1825 57 > 40 1835 Jacksonian 1848, Mexican War
populism
1835 46 44 1845 47 > 37 1855 Abolitionism 1860, Civil War
1885 46 49 1895 48 > 43 1905 Muckraking, 1917, World War I
Crusade for
social justice
Shortened
lead time
1925 47 53 1935 55 > 48 1935 New Deal 1940, World War II
1955 49 53 1965 63 > 47 1965 Crusade for 1967, Vietnam War
civil rights,
War on
Poverty
A final inference to be drawn from Figure 11.9 is that n Affiliation appears

to rise most often in decades after a war, starting with the Revolutionary War.
It makes sense to infer that after people have spent some time killing and get-
ting killed, they should have a heightened concern for the welfare of others.
Note in Figure 11.9 that the highest peaks of n Affiliation were associated with
the Revolutionary War and the Civil War, both of which involved killing broth-
ers, sisters, and friends. In fact, in six out of the eight decades after a war in the
history of the United States, n Affiliation was high, as contrasted with only two
decades out of ten following a period of peace, a difference that is statistically
significant. This sets the stage for a new cycle of concern for others, which, if it
is accompanied by high n Power, will lead in time to the imperial power motive
pattern, social reform, and another instance of going to war. While all these re-
lationships need confirmation in the history of other nations, it is noteworthy
how much a knowledge of collective motive levels contributes to an understand-
ing of significant events in the history of the United States.
Motive Levels and the Success of Automobile Companies

Nations are not the only kind of collective enterprise. Perhaps indexes of collec-
tive motive levels in other types of organizations could predict organizational
Performance over time. Diaz (1982) reasoned that it might be possible to assess
the motivational spirit of a firm by coding the annual letters of the chief execu-
tive officer to stockholders in the Company. So he coded the annual letters to
stockholders for n Achievement and n Power for two U.S. automobile manufac-
turers (1952 to 1980) and three Japanese automobile manufacturers (1962 to
1976 and 1980). One of his objectives was to discover whether there was a moti-
vational reason why the Japanese automobile makers seem to be doing so much
better than their U.S. counterparts in recent years.
Table 11.9 highlights the differences in efficiency. The Japanese automobile
manufacturers come out better in every comparison. Figure 11.11 may explain
why. All three Japanese manufacturers (Nissan, Toyota, and Honda) scored
higher in collective achievement motive levels at nearly every point in time than
their U.S. counterparts, General Motors and Chrysler. All the studies of the
n Achievement level in individuals and nations suggest that it should translate
into greater efficiency and a stronger entrepreneurial spirit, which eventually
should lead to a more rapid improvement in Company Performance. Many ex-
planations have been advanced for why Japanese automobile manufacturers have
shown such a large competitive advantage over U.S. firms in recent years, but
one of the simplest explanations is that provided in Figure 11.11. The Japanese
firms have been activated by a higher level of achievement motivation than their
U.S. counterparts.
Diaz (1982) was also interested to see whether n Achievement and n Power
levels within a Company could be used to predict rates of growth for the Com-
pany at different time periods, so he paired Chrysler with General Motors in the
United States and Nissan with Toyota in Japan. Since the companies in each
pair were directly competing with each other for the same market, he computed
for each year the ratio of sales, assets, and return on investment for each pair.
Table 11.9.
COMPARISON OF THE EFFICIENCY OF JAPANESE AND U.S. AUTOMOBILE
MANUFACTURERS (after Lohr, 1982, p. 29)
Japan United States

a
Manufacturing (machine stamping operations)
Parts stamped per hour 550 325
Labor power per press line 1 7 to 13
Time needed to change dies 5 minutes 4 to 6 hours
Average production run (days) 2 10
Time needed to build a small car (hours) 30.8 59.9
Personnel (average automobile plant)
Total work force 2360 4250
Average number absent (vacations, illness, and so on) 185 500
Average absentee rate (percent) 8.3 11.8
a
Source: Harbour & Associates.
ö 0.50
o
U
0.40
S 0.30
2 £
un o
o U
0.20
2? 0.10-
L
0.00 I
1952 1960 1968 1976 1980
Year
Figure 11.11.
Level of Achievement Concern in Chief Executive Officer's Letters to Stockholders in Japanese and
U.S. Automobile Companies over Time (after Diaz, 1982).
Thus, Chryslers sales fluctuate between 16 and 37 percent of General Motors'

sales between 1952 and 1980. What he was estimating was how well Chrysler
was doing relative to General Motors. This method of Computing comparative
Performance has the advantage of Controlling for changes in demand and market
conditions that occur from year to year, since these should affect both compa-
nies in the same way. Then he also computed the comparative advantage of
Chrysler over General Motors in n Achievement and n Power levels for each
year, smoothed these ratios for adjacent years, and correlated each year's motive
ratio score with comparative economic Performance scores from all the other
years. Thus, for example, comparative n Achievement in 1960 would be corre-
lated with the comparative economic Performance of the two companies before
1960 and from 1961 to 1980.
The correlations obtained showed a very regulär cyclical pattern, as Figure
11.12 shows. If we consider first predicting forward from the motive level in a
given year to subsequent Performance (here measured in terms of relative as-
sets), it is clear that both comparative n Achievement and comparative n Power
levels predict gains in relative assets maximally about seven years later. The ef-
r
.60
Relative n Achievement/
Versus Relative
Assets
.30 -
Relative n Power Versus Relative Assets
I I I I I 1 I I I I
8.0 12.0
Motive .
Assets-»- _ . -»-Assets
Level
Time Lag (Years)
Figure 11.12.
Chrysler and General Motors: Relative n Achievement and n Power Correlated over Time with
Relative Assets (after Diaz, 1982).
fect of comparative motivational advantage is not feit in the near future but has
an impact several years later. In fact, the correlation between comparative
n Achievement levels and relative assets seven years later is .57, p < .05. And
remember that this correlation is based on every seven-year discrepancy in dates
starting with 1952.
Next consider the backward associations: correlations between motive levels
and Company Performance in previous years. It is easier to think of these corre-
lations as the relationship between Performance and subsequent motive levels.
As Figure 11.12 shows, these relationships are predominantly negative. This
means that good Performance is associated with low levels of n Achievement
and n Power in subsequent years and vice versa. The maximum negative corre-
lations occur over a two-year period. The immediate effect of comparative advan-
tage in the economic Performance of a Company is to lower n Achievement and
n Power in the succeeding two years. This fits earlier explanations given for the
decline in n Achievement. In terms of the Atkinson model (see Chapter 7), as the
probability of success increases to a high level, the challenge to achieve decreases.
Figure 11.13 shows a similar curve for Nissan and Toyota, the two Japanese
.80
.40
/^Relative n Achievement Versus Relative Assets
| .00
o
U
Relative n Power Versus Relative Assets

-.40
-.80 I I 1 I I I 1 I 1 1
-8.0 -4.0 0.0 4.0 8.0 12.0
Motive
Assets-• , , -»•Assets
Level
Time Lag (Years)
Figure 11.13.
Nissan and Toyota: Relative n Achievement and n Power Correlated over Time with Relative
Assets (after Diaz, 1982).
firms compared. Once again, the relationship of comparative n Achievement ad-

vantage to comparative economic Performance is exactly what we would expect,
and it is similar to what was obtained in the United States—with one exception.
Comparatively high Performance is associated with an immediate drop in
n Achievement levels over the next two years, and comparatively high n Achieve-
ment levels are associated with a gain in relative assets in subsequent years, with
a peak occurring earlier than for the U.S. companies. Note that the correlation
between n Achievement levels and relative assets reaches a generally high level
after four years in Japan, whereas in the United States at this time lag the rela-
tionship is still insignificant. The Japanese high level of n Achievement trans-
lates maximally into improved Company Performance after four rather than
seven years, as in the United States. Apparently there is less organizational iner-
tia in Japanese companies, so entrepreneurial drive can translate itself into re-
sults more quickly.
What are we to make of the fact that comparative n Power levels in Japan
are negatively associated with comparative economic Performance in subsequent
years rather than positively, as in the United States? Perhaps there is some dif-
ference in the way Japanese and U.S. firms are structured or managed such that
power motivation promotes organizational Performance in the United States and
inhibits it in Japan. It is impossible to be certain how this result should be inter-
preted, but a clue may lie in the difference in the nature of the power motiva-
tion expressed in the two countries. As noted in Chapter 8, socialized power
motivation (high n Power, high inhibition) is associated with management suc-
cess in the United States. High n Power and low inhibition is related to impul-
sive assertiveness in trying to establish personal control over others and does not
go with managerial success.
Diaz also scored the letters to stockholders for inhibition and found that in
the Japanese letters to stockholders the word not appeared only one half the
number of times per hundred sentences as in the U.S. letters. This finding is
contrary to the populär belief that the Japanese are a highly controlled people,
and it may reflect some translation or linguistic complications; if it is taken at
face value, however, it may explain why power motivation has such different ef-
fects in the two countries. For if the power motivation in the United States is
more socialized, it should lead to better management practices and better perfor-
mance, whereas if it is less socialized and more personal in orientation in Japan,
it should lead to less effective management. And these are the relationships actu-
ally found in the two countries. Whatever the meaning of the n Power results,
those for n Achievement are clear and consistent for both countries: high
n Achievement levels in companies, as in nations and in individuals, lead to
more entrepreneurial behavior and better economic Performance later.
• ORIGINS OF COLLECTIVE MOTIVATIONS

Ideological Factors Affecting Collective Achievement Concerns
If collective motive levels are important influences on what happens in history,
questions naturally arise as to where they come from. Why was Greek literature
so much higher in the number of achievement images in early, as contrasted

with later, periods? Was it the influence of some person or group of persons in
leadership positions who happened to be high in n Achievement? It is impossible
to know for certain.
Diaz (1982) has shown that motive levels in letters to stockholders of com-
panies shifted markedly depending on who was chief executive officer at the
time. So we might infer in this case that it was the achievement or power moti-
vation of individuals that somehow created a climate that changed business per-
formance. If this were generally the case, one might adopt a "great leader" view
of history in which happenings in the life of individual leaders would create mo-
tive levels that shape the course of history. But unfortunately we do not know
the motive levels of the chief executive officers of the automobile companies as
tested in the usual way, and it is entirely possible that what they wrote in their
letters to stockholders was as much a product of the general climate in the com-
pany as it was of their own individual motive levels. Furthermore, the motive
levels in letters to stockholders supposedly written by the same man (but per-
haps written by an assistant) varied from year to year, which would suggest the
chief executive officer was reflecting the climate in the Company rather than cre-
ating it from his own stable motive dispositions. In the case of even larger units
such as nation-states, it is even less clear that the motive levels of individual
leaders shape the collective motive levels for the society. For example, there is
not a close relationship between the motive levels for various decades in recent
U.S. history as shown in Figure 11.9 and the motive levels of U.S. presidents
during those decades, as summarized in Table 8.8.
If the ideas of an individual leader are not the main determinant of collec-
tive motive levels, perhaps major belief Systems current in the society have an
influence. Attention focused originally on religion as a source of achievement
motive levels. The impetus to this idea was provided by Max Weber's (1904/
1930) classic analysis of the Protestant origins of the modern capitalistic spirit.
He had observed that in countries like England and Germany, Protestant groups
seemed to work harder, to save more money, and to enter successfully into busi-
ness more often than similar Catholic groups. A number of studies have con-
firmed the fact that in the seventeenth, eighteenth, and even the nineteenth Cen-
tury, business leaders were more often drawn from Protestant than Catholic
minorities, particularly from small dissenting groups such as the Quakers
or later the Methodists in England (McClelland, 1961). Why should this
be so?
Weber found an answer in the doctrines of Protestantism—particularly in its
stress on the rationalization of all of life and the devotion to one's calling or vo-
cation. Reformation leaders such as Martin Luther and John Calvin had turned
against the authority of the Catholic church in Rome, particularly its authority
to guarantee people's salvation by selling them indulgences to make up for past
sins by good works. Thus, to a certain extent, Protestant believers were more
"on their own" than Catholic believers. They got the impression that "God
helps those who help themselves" and that therefore they could in a certain
sense, create their own salvation or, more correctly, the conviction that they
were saved. They could not gradually accumulate good works to their credit, as
in Catholicism, but only by systematic self-control, which at every moment

"Stands before the inexorable alternative, chosen or damned," could they create
the conviction that they were one of the elect of God (Weber, 1904/1930).
Weber argued that such a rigid rationalization of a person's own conduct
explains some of the extra determination and vigor with which Protestants pur-
sued capitalistic enterprise. McClelland (1961/1976) observed that the key ideo-
logical element may have been the stress on people's continually improving their
own conduct, even in small ways, from day to day "till we shall have arrived at
a perfection of goodness, which indeed, we seek and pursue as long as we live"
(Calvin's Institutes of the Christian Religion). Thus, the ideology of the Protes-
tant Reformation may have succeeded in raising achievement motivation among
Protestant groups, which in turn led them to develop capitalistic enterprise more
successfully than Catholics.
No direct test of this hypothesis has been made by comparing the achieve-
ment content of populär literature from Protestant or Catholic sources at critical
periods in the sixteenth through the nineteenth centuries. By the twentieth Cen-
tury, however, McClelland (1961) found no consistent differences in n Achieve-
ment levels among Protestant and Catholic groups of individuals in the United
States and Germany. Nor did children's readers from Protestant and Catholic
countries differ in their average n Achievement levels. He concluded, therefore,
that the higher n Achievement levels of Protestants could not be attributed so
much to the doctrines of the Reformation as to the way in which the Reforma-
tion aroused individuals and put them on their toes, so to speak, in their search
for salvation in nontraditional ways.
The key element seems to be some kind of an "inner worldly activism,"
which is supported institutionally by a church that also opposes traditional reli-
gious authority. To check this idea, McClelland (1961) studied the religious be-
liefs and practices of forty-five tribes whose folk tales had already been coded
for n Achievement levels. Of the cultures with high n Achievement in their folk
tales, 65 percent stressed individual over ritual contact with the divine, as com-
pared with only 23 percent of the cultures low in n Achievement. In other
words, in high n Achievement cultures, individuals less often gain favor with the
divine by participating in traditional, ritualistically prescribed ceremonies. In-
stead, they have more direct access to the divine through mystical communion
or individualistic interpretations of written traditions. McClelland also found
that in high n Achievement cultures, religious experts (priests, shamans, and so
on) are much less often necessary for the Performance of an individual's reli-
gious duties. Across cultures, those that stress inner over external religious au-
thority tend to have higher collective levels of n Achievement.
McClelland (1961) also reports an interesting contemporary confirmation of
the importance of religious reform in raising n Achievement levels. In the high-
lands of Mexico in the State of Chiapas, he located a village that had recently
been converted to a form of radical Christianity. The new believers acted in
many ways like the original Protestants in Europe at the time of the Reforma-
tion. They read the Bible in their native language; they burned the Catholic folk
images of saints and painted their churches white; they sang hymns; they be-
came more interested in health and schooling; and above all, they feit them-
selves superior to the traditional folk Catholics in other villages nearby, who
had to depend on the priests to make contact with God. Children in this village
were tested for n Achievement levels some eight to ten years after most of the
adults had been converted. Their scores were contrasted with those of school-
children in a nearby village that had remained traditionally Catholic, although
it, too, might have received an achievement impetus from the fact that the gov-
ernment had just given the peasants their own land to farm. The children from
the village recently converted to Protestantism showed a number of signs of
higher n Achievement. At least in this case, religious reform—in its early stages,
anyway—apparently had raised n Achievement.
It is not necessary to assume that Protestantism per se ; s the critical factor.
Minority religious groups throughout the world having no connection with
Christianity—such as the Jains in India (McClelland & Winter, 1969/1971) or
the Gurage in Ethiopia (McClelland, 1977a)—have produced disproportionately
large numbers of businessmen and almost certainly have shown higher levels of
n Achievement. What seems to be critical is some religious or other reason for
believing—perhaps only at the moment of reform—that one's group is superior
to more tradition-oriented groups in which authority rests in the institution
rather than in the individual, who has become a convert to the new way of life.
We can also use such a general interpretation to explain the fact that countries
like the Soviet Union and mainland China, which conceive of themselves as
originators of the great Communist ideological reform, score significantly higher
than the world average in n Achievement levels in school textbooks. On the
other hand, Communist countries such as Poland and Bulgaria, which did not
originate the reform but accepted it from without, are significantly below aver-
age in n Achievement levels. (See Table 11.4.)
Comparable studies have not been made of the possible religious or ideologi-
cal sources of collective levels of n Affiliation or n Power.
Environmental Factors Influencing n Achievement Levels

In the early twentieth Century it was commonly asserted that certain peoples
were more energetic or entrepreneurial because of their genetic endowment or
because they happened to live in a stimulating climate (McClelland, 1961).
Thus, the British, who were very successful economically at the time and who
had an empire that stretched around the world, were thought to have the ad-
vantage of belonging to the white race and also of living in a temperate climate,
which stimulated them to work harder. In contrast, blacks living in Africa were
thought to be less successful because they were racially endowed with less en-
ergy and because they lived in the enervating tropics. History has already dem-
onstrated the absurdity of such conclusions for the obvious reason that nations
of the same racial background and inhabiting the same piece of real estate are
much less energetic and successful today than they were two or three genera-
tions ago. If we take collective n Achievement levels as indicative of one very
important type of psychic energy, it is clear that race and climate can have little
to do with national economic success, since n Achievement levels have varied

widely over time for the same country. For example, collective n Achievement
in England was well above the world average in 1925, signaling that country's
strong economic Performance up to World War 11, and well below the world av-
erage in 1950, foreshadowing its more recent relative economic decline.
A more sophisticated view of the influence of environmental factors on na-
tional success has been advanced by Arnold Toynbee (1947), the British histori-
an, after an exhaustive comparative study of the rise and fall of a great many
civilizations. He observed that nations tended to rise when the challenge of the
physical or social environment was "just right"—neither too great (as in an ex-
cessively harsh environment) nor too small (as on the fabled tropical island
where everything needed is easy to get). This view comes very close to stating
Atkinson's position, as outlined in Chapter 7, that moderately challenging tasks
are most likely to arouse achievement motivation (see Table 7.2). Thus, Toyn-
bee's hypothesis seems reasonable on theoretical grounds, and he is able to give
many illustrations of how it explains what happened in history. For example,
the climatic and social challenge to the Scots in Great Britain was just right, so
they did well. But when they migrated to the hüls of North Carolina the chal-
lenge was too great, and they did less well. The difficulty with the hypothesis is
that it is not really testable, because Toynbee proposes no way of measuring the
ease or difficulty of the challenge. Sometimes it is social, sometimes economic,
sometimes climatic. He had a natural tendency to explain whatever the outcome
was in terms of whatever degree of challenge was necessary to explain it.
McClelland (1961) did find some evidence for the hypothesis so far as col-
lective n Achievement levels are concerned, if climatic conditions are taken as
an index of the degree of challenge. Using data from preliterate tribes, he
showed that cultures whose folk tales were high in n Achievement were much
more likely than those low in folk tale n Achievement to live in areas where the
mean annual temperature varied between forty and sixty degrees Fahrenheit,
where the mean daily or monthly temperature Variation was greater than fifteen
degrees Fahrenheit, where mean annual rainfall was less than sixty inches, and
where the quality of soil was not very good. These findings can be interpreted as
meaning that moderate challenge from the physical environment promotes the
development of n Achievement, but they can also be interpreted to mean that
groups high in n Achievement tend to seek out such environments.
Several studies point to the importance of challenge in the social structure
as a source of achievement motivation, or at least as a necessary complement to
it if it is to affect entrepreneurial behavior. Swanson (1967) has demonstrated
fairly conclusively that Protestantism won control in the Reformation in those
European communities in which the central governing authority was limited or
in some way more democratic. Regimes in which power was strongly central-
ized, sometimes by divine right, tended to remain Catholic. When authority was
divided up between representatives of various groups or guilds, the society was
more likely to be won over to Protestant reform. If achievement motivation de-
veloped in these Protestant centers, as McClelland (1961/1976) has argued,
Swanson sees it more as a response to the better economic opportunities pro-
vided in a governing System in which more people had a chance to get ahead.
The critical studies have not as yet been made that would show whether
changes in n Achievement levels occurred before, during, or after the changes in
governmental structure.
But certainly an open, competitive structure seems to be regularly associated
with higher n Achievement levels. For instance, the Ibo of Nigeria, as already
noted, have high n Achievement. LeVine, who studied them, feit they developed
high n Achievement because of the most characteristic feature of their Status
System—the title society: 'This consisted in the most developed form of a series
of ranked titles, the entry to which was contingent upon acceptance by existing
title holders, payment of a set entrance fee and providing a feast for members of
the society. Membership was open to anyone of free birth, but the fees and
feasts effectively limited title holding to those of some wealth. This was increas-
ingly true as a man progressed to higher titles" (LeVine, 1966). Thus, an Ibo
male could attain high Status by his own efforts and particularly by accumulat-
ing wealth. Parents, knowing this, would be likely to give their sons the kind of
achievement and self-reliance training that would develop the high levels of
achievement motivation actually found among the Ibo.
In contrast, LeVine noted that the individual Hausa males achieved Status
primarily by attaching themselves to an important leader who could provide re-
wards out of the resources he controlled in return for obedience, loyalty, and
Service. Since Hausa society rewards obedience and loyalty rather than achieve-
ment and self-reliance, it is not surprising that the Hausas score low n Achieve-
ment. Once again, an open, competitive structure in which self-reliance is re-
warded, as in the democratically ruled cities at the time of the Reformation,
seems to be an important source of encouragement for developing achievement
motivation.
One other piece of evidence corroborates this correlation. In preliterate so-
cieties, cultures with bilateral kinship Systems are more likely than those with
other kinship Systems to have high n Achievement levels in folk tales (r = .36;
N = 37; p < .05). The key aspect of a bilateral kinship System from the point
of view of the child is that he or she sees both the mother's and father's side of
the family as more or less equally important. Thus, children have more options.
They are not locked into a patrilineage or matrilineage. They do not have to be
what their mother's or father's people expect them to be. At the individual level
it is authoritarian upbringing, particularly on the part of the father, that pro-
duces low n Achievement in sons (Rosen & D'Andrade, 1959). (See Figure
7.13.) For a boy to develop high n Achievement, there must be some stress on
self-reliance, some opportunity for him to set his own goals. At the societal level
the equivalent would appear to be some kind of institutionalization of an open
structure that permits people to rise in a variety of ways, balances institutional
powers against each other, or at the very least does not restrict the individual to
one accepted, traditional way of behaving.
Sources of the Collective Concern for Power

Table 11.10 summarizes what little is known about societal characteristics asso-
ciated with high n Power levels either in the cross-cultural sample employing
Table 11.10.
CORRELATIONS OF SOCIETAL CHARACTERISTICS WITH HIGH n POWER
(after McClelland, Davis, Kaiin, & Wanner, 1972; Southwood, 1969)
Societal Characteristics Correlation with High n Power

Cross-cultural sample (n Power from folk tales):
N = J<5a
Colder climate 47***
Percent dependence on hunting .27*
Percent dependence on agriculture -.27*
Socioeconomic simplicity scale .41***
Lineal descent _>44***
Cross-national sample (n Power from children 's textbooks

around 1950): N = 38-40b
Gains in electric power use (1937-1954) .34**
Percent union membership (1966) -.38**
a
After McClelland, Davis, Kaiin, & Wanner (1972).
b
After Southwood (1969).
*p < .10.
**p < .05.
***p < .01.
folk tales or in the cross-national sample employing stories in children's text-

books. The results make good sense and fit in well with the limited amount
known about the origins of the power motive in individuals. What seems to en-
courage the development of high n Power is a demand for individual assertive-
ness, as in struggling to keep warm or in hunting as contrasted with agriculture,
which often relies on collaborative efforts. The socioeconomic simplicity scale in
the table is a factor score loading high on such variables as high degree of de-
pendence on hunting, a low degree of jurisdictional hierarchy, impermanence of
settlement pattern, and small size of local Community. In other words, in such
societies, individuals are very much on their own and not part of some large
supportive network. A lineal descent pattern is negatively associated with collec-
tive n Power, presumably because societies oriented around lineal descent are
precisely those that maintain a large supportive network dependent on the lin-
eage, so individuals are not expected to be assertive on their own. This pattern
of correlations is consistent with the findings reported in Chapter 8 showing that
parental permissiveness for assertiveness in the areas of sex and aggression, as
well as the lack of tight parental control, are associated with high n Power.
The only two societal variables significantly associated with collective
n Power levels among nations were prior gains in electric power use (positively
associated) and the percentage of union membership (negatively associated).
Both can be interpreted as consistent with the picture that has emerged so far,
for rapid gains in electric power usage means more power is placed in the hands
of the individual to make things or to turn on gadgets like radios, light bulbs,
irons, or mixers. What technology does is give people more power and encour-
age their assertiveness in using it, which might promote higher levels of collec-
tive n Power. On the other hand, union membership provides the kind of soli-
darity for working people that means they will get support whether they are
individually assertive or not. It might be considered the equivalent in the ad-
vanced technological world of the lineage System in preliterate society. Both
may tend to diminish the need for individual assertiveness and in fact so con-
strict individuals that they tend toward lower n Power. At best, such findings
are only suggestive. They need much further investigation.
One other factor associated with high collective levels of n Power has been
identified (McClelland, 1976). A major challenge to individual assertiveness is
unemployment, particularly for men in the period before women entered the
labor force in large proportions. In the history of the United States, relatively
high levels of unemployment have regularly been followed by higher n Power
relative to n Affiliation scores, as Table 11.11 shows. This table rates employ-
ment opportunities from 1820 to 1884 as better or poorer based on whether im-
migration to the United States was lower or higher than the average for approx-
imately one hundred years. Thereafter the ratings for employment opportunities
are based on actual unemployment figures. The n Power and n Affiliation scores
are those shown in Figure 11.9.
In nearly every instance, poorer employment opportunities are associated
with a higher n Power score in that decade, and better employment opportuni-
ties are associated with a lower n Power score in that decade. The difference is
highly significant. Some figures from national sample surveys of motive levels in
the United States confirm and modify this conclusion (Veroff, Atkinson, Feld, &
Gurin, 1960; Veroff, Depner, Kulka, & Douvan, 1980). In 1976, 59 percent of
black males were high in n Power as contrasted with only 48 percent of white
males; the difference might be attributed to the fact that there was certainly
much more unemployment among black than white males. However, there was
not a comparable difference in the 1957 sample. This suggests that an ideologi-
cal factor combined with a structural one to produce a rise in n Power among
black males. In 1957, before black consciousness had been raised, unemployment
may not have meant the same thing to blacks as it did in 1976, when they feit it
to be even more unfair. This serves as an important reminder that environmen-
tal factors like unemployment take on meaning in terms of the ideological fac-
tors also present at the time.
Sources of the Collective Concern for Affiliation

The only cross-cultural clue to what produces a high collective n Affiliation is a
positive correlation between signs of male solidarity in a culture and n Affilia-
tion levels in folk tales (r = .36; TV = 21; p < .10; McClelland et al., 1972). In
other words, in societies where men tend to live and work together, affiliative
concerns tend to be higher. The only other bit of evidence on this point, as al-
Table 11.11.
RELATIONSHIP OF EMPLOYMENT OPPORTUNITIES TO RELATIVE STRENGTH OF
n POWER IN U.S. HISTORY (after McClelland, 1976)
n Power in That Decade Is

Epoch Employment Opportunities Less Than n Affiliation Greater Than n Affiliation
1820-1824 Poorera
1830-1834 Poorer
1840-1844 Poorer
1850-1854 Better
1860-1864 Poorerb
1870-1874 Better
1880-1884 Better
1890-1894 Poorer
1900-1904 Better
1910-1914 ?d
1920-1924 Better
1930-1934 Poorer
1940-1944 Better
1950-1954 Poorer
Note: + = Confirmation of hypothesis; — = Nonconfirmation of hypothesis. The association of competition for Jobs with predominance
of power motivation is unlikely to be due to chance (p < .01).
a
Estimates are based on whether alien immigration for the five-year period is Iower or higher than averages for five-year periods from 1820
to 1913 (Easterlin, 1968).
b
This estimate may be incorrect, because fewer people migrated to the United States at this time because of the Civil War.
c
From here on, estimates are based on whether the unemployment rate was 5.5 percent or more in four out of five years, or less.
(Easterlin, 1968).
d
Motivational data are unavailable for this decade.
ready noted, is that killing in organized wars tends to raise collective n Affilia-
tion levels, at least in the history of the United States. But overall, the lack of
information about sources of the affiliative motive at the collective level is
marked and paralleled by a similar lack of information at the individual level.
• DIFFICULTIES IN INTERPRETING
MOTIVE TRENDS IN SOCIETY
For ease of exposition, the review of studies of historical trends in collective mo-
tive levels has emphasized the findings that can be readily interpreted, but they
are only part of the story. In every study, difficulties have arisen that have been
resolved in what seems to be a reasonable way that gives meaningful results, but
this does not mean the difficulties can be overlooked (McClelland, 1981b). It is
worth reviewing some of the problems encountered in this type of research, if

only to make it clear that the conclusions reached depend on decisions that are
open to dispute.
A major problem arises over the lead times between motivational shifts and
subsequent historical events. In the connection between n Achievement levels
and subsequent economic growth, the lead time varies from centuries in the
study of Ancient Greece to a few years among contemporary nations (Table
11.4). How can the same variable—collective n Achievement level—be supposed
to have the same effect after a Century as opposed to five years? Part of the dif-
ficulty may lie in determining in the studies of remote periods of time just when
the n Achievement levels were high. The scores are plotted at midpoints in
century-long periods, but that does not really indicate how high the level re-
mained at the end of the period or into the next one. In more recent periods it
has been possible to get estimates of motive levels more pinpointed in time;
when that is done, the lags to subsequent social changes get shorter, perhaps be-
cause the assessment of when the motive level existed is more accurate.
But that explanation does not account for the fact that in U.S. history, the
prediction of wars from collective n Power/« Affiliation ratios assumes a lead
time of about fifteen years up to World War I and five to ten years thereafter. Is
readjusting the lead time just a method of fitting the data to a hypothesis estab-
lished in relation to an earlier time period? Perhaps, but with increased commu-
nication and transportation, motive changes should translate into social changes
more rapidly. The comparison of the Performance of Japanese and U.S. automo-
bile companies clearly demonstrates that motive levels affect collective perfor-
mance after different lead times, depending on how well organized a collectivity
is. At least whenever adjustments in lead time have been made, they have al-
ways been toward shorter lead times in more recent periods.
Another problem arises over how to interpret different estimates of collec-
tive motive levels obtained from different sources of populär literature. A major
difference in n Power levels obtained from different sources in England between
1600 and 1700 was interpreted in terms of a class difference between the aristoc-
racy and the common people, but no systematic attempt was made to see if dif-
ferences in other motive levels from these two sources also correlated with dif-
ferences in other types of behavior between the two classes. We might even
question whether it is correct to assume that plays represent the interests of the
upper classes more than street ballads.
Furthermore, often differences in motive levels from different sources have
not been interpreted. For example, n Achievement level, as estimated from chil-
dren's textbooks and hymns, appeared to be high in the United States between
1880 and 1900, but according to the fiction sampled it was low (McClelland,
1975). While it may seem reasonable to estimate the level of n Achievement to
be high based on all sources, that still leaves the question of interpreting the
low level for fiction. Could it be that fiction was read more by women and that
their n Achievement level was low? Or that it was read more by the educated
classes, who had a lower level of n Achievement than less-educated people, who
were participating more in the religious life of the Community, showing their
higher level of n Achievement in the hymns that they wrote and sang and turn-
ing more often into successful entrepreneurs? Such questions can only be an-
swered by more extensive research into class and group differences in motive
levels reflected in different types of populär literature.
Finally, many questions can be raised about the measures of collective per-
formance that have been used in these studies. Economists interested in measur-
ing the rate of economic growth have never been satisfied with using indexes
like coal imports at the port of London or the consumption of electric power.
From their point of view, economic growth must be estimated in terms of all its
constituents, including contributions from agricultural, industrial, and Service
sectors. When some of these more typical measures of gross national income are
used, different results are obtained, and it is not so clear that n Achievement
levels precede periods of economic growth.
McClelland (1961) argues that aggregate measures of national income levels
are not as accurate as index measures, particularly for periods one or two hun-
dred years ago or for underdeveloped countries in the modern period, for which
many figures are unavailable. However, there is room for disagreement on this
point. Even in the case of the automobile companies, Diaz (1982) employed
three different measures of economic Performance: sales, assets, and return on
investment. The interrelationship between motive levels and assets, as shown in
Figures 11.12 and 11.13, was smoothest. The same types of relationship were
obtained for sales and return on investment, but more irregularities showed up
in year-to-year correlations. It can be argued that assets represent the best over-
all estimate of the growth of a firm, especially since companies may adjust
yearly profits to avoid taxes or to invest more heavily in future products. It can
also be argued that sales and return on investment are equally important indica-
tors of a firm's success and that collective motive levels do not associate as
closely with such indicators. These matters can only be settled by more studies
of this type on other companies and by a closer theoretical analysis of the best
way to measure economic success. The studies reported in this chapter have
only begun to open a field of investigation that needs further examination to es-
tablish more carefully the rules for obtaining data and interpreting them.
The Meaning of Collective Motive Scores

The most important problem in interpreting collective motive scores arises out
of trying to understand just what characteristic or characteristics they reflect.
The practice of scoring folk tales, Speeches, and other types of populär literature
for different types of motive imagery grew up as a natural extension of scoring
stories produced by individuals. But what do these collective motive scores
mean? In what sense can we think in terms of the motive level of a nation or a
culture? Do collective and individual motives function in the same way? Do the
two types of scores get at the same variable?
Recall that the füll motive scoring Systems derived for individuals had to be
abandoned in scoring Speeches or populär literature. Instead, only the number of
motive images per Standard unit of text are counted. Doesn't that suggest the
scores obtained in these two ways are not equivalent? Winter and Healy
(1981/1982) refer to imagery scoring as the running text System and have sys-
tematically examined what kind of results it gives in comparison with more
complete scoring Systems whenever the two methods have been applied to the
same material. They found that the simplified running text System reflects the
effects of experimental arousal of motives, although not as well as the füll scor-
ing Systems. Correlations of the simplified with the füll scores for individual pro-
tocols run between .40 and .70. Split-half reliabilities for the running text scores
vary between .43 and .70 for the various motive Systems. Reliability is, if any-
thing, higher for the running text System, probably because such scores are
based on larger samples of text than are available in the protocols obtained from
individuals. For example, McClelland (1961) found that in the cross-national
study of children's textbooks, correlations of n Achievement scores in odd and
even sets of stories were .67 in 1925 and .59 in 1950. While the odd-even reli-
ability coefficients were considerably lower for n Affiliation and n Power, these
values taken together are higher than those usually obtained for individuals
under the set to "be creative." Split-half reliabilities for public Speeches and
press Conferences run in the ränge of .62 to .77 (Hermann, 1980; Winter &
Healy, 1981/1982).
Other methodological issues relate to just how to go about sampling mate-
rial to get a good estimate of collective motive levels. McClelland (1975) has
suggested ten different rules for getting representative and reliable estimates of
collective motive levels from populär literature, such as making sure the litera-
ture is populär, choosing selections from several different authors and texts,
making sure the coding is done blindly so that the coder does not know what
time period is represented or what hypotheses are being investigated, and mak-
ing sure each motive index is based on at least ten sets of one hundred lines or
ten thousand words. Figure 11.14 illustrates how two such data sets provide
very similar n Power data points for different periods in English history. The in-
terpretation of power motive trends would be the same no matter which sample
of data had been used.
Winter and Healy (1981/1982) also show that the running text scoring Sys-
tems predict individual behaviors associated with the motives about as well as
the füll motive scoring Systems do. But they make a further important point:
when the running text System is applied to material for which it is designed—in
this case, a long intensive interview—the n Power score obtained predicts indus-
trial leadership behavior better than the n Powe. score obtained from the Stan-
dard picture-story exercise. Thus, it is possible that the running text scoring Sys-
tems may be more reliable and more valid if applied to standardized interviews
than the füll scoring Systems are when applied to protocols obtained under
test administration conditions, which are known to influence the results (see
Chapter 6).
The main reason for believing that the two types of scores are getting at
something similar is that the behaviors associated with collective and individual
motive scores are so similar. High n Achievement at either level is associated
with more entrepreneurial activity and greater economic success; high n Power
5.00
J 4.00
Sample B: 11 Alternate Sets of
100 Lines for Each Data Point
S 3.00 \
\
\
Sample A: II Sels of 100

2.00 Lines for Each Data Point
o
0u
1.00
_L
1576-1625 1626-1675 1676-1725 1726-1775 1776-1825
Periods in English History
Figure 11.14.
Comparability of n Power Scores Obtained from Two Samples of Running Text in Drama at
Different Periods in English History (after McClelland, 1975).
plus low Activity Inhibition is associated with heavier consumption of alcohol at

either level; high n Power plus high Activity Inhibition is associated with empire
building at either level; and so on.
What are we to make of such similarities? One possibility is that the collec-
tive motive score is an index of the percentage of people in a society or group
who are high in the motive in question. An early comparison failed to show a
relationship between mean levels of n Achievement for groups of individuals in
a nation and n Achievement levels as coded in children's stories (McClelland,
1961). But the samples of individuals tested were in no way representative.
Veroff et al. (1960) and Veroff et al. (1981) did obtain motive scores from
representative samples of U.S. adults in 1957 and again in 1976. The percentage
of individuals scoring high in each of the three social motives in 1957 and 1976
can be compared with national motive levels obtained in the 1950s and 1960s
from populär literature, as shown in Table 11.12. Only a comparison of shifts in
motive levels is meaningful, since the Standard scores for populär literature are
based on eighteen decades in U.S. history and the samples of individual scores,
on only two decades. Both types of measures agree that for both sexes com-
Table 11.12.
COMPARISON OF CHANGES IN MOTIVE LEVELS IN THE UNITED STATES
ESTIMATED FROM REPRESENTATIVE SAMPLES OF INDIVIDUALS AND
FROM POPULÄR LITERATURE (after Veroff, Atkinson, Feld, & Gurin, 1960; Veroff
& Veroff, 1981; McClelland, 1975)
Motive Measured 1957° 1976° Dijference

n Achievement
Men (percentage high) 47 53 +6
Women (percentage high) 41 60 + 19
Populär literature (Standard score) 54 55^ +1
n Affiliation
Men (percentage high) 58 39 -19
Women (percentage high) 48 50 +2
Populär literature (Standard score) 53 47b -6
n Powerc
Men (percentage high) 46 53 +7
Women (percentage high) 48 49 +1
Populär literature (Standard score> 49 63b + 14
a
The number of respondents in 1957 was 595 men and 774 women; in 1976 it was 508 men and 700 women.
b
In 1965, the midpoint of the 1960 to 1970 decade.
c
Scored according to the Winter (1973) System.
bined, there was a gain after the 1950s in national n Achievement levels and in
national n Power levels and that there was a loss in the national n Affiliation
level. Furthermore, if the comparison is restricted to the results for men, the
two types of measures agree that the gain was largest for n Power and next
largest for n Achievement, and that the loss was very large for n Affiliation.
However, the populär literature index was poor at picking up a big increase in
individual n Achievement levels, perhaps because the increase in n Achievement
for women resulting from the women's movement had not yet affected the popu-
lär literature that was being sampled in the 1960s. And the drop in the populär
literature n Affiliation score did not reflect the fact that women, if anything, in-
creased in n Affiliation between 1957 and 1966 (Veroff, Douvan, & Kulka,
1982).
These findings are based on too few data points to be decisive. Furthermore,
there is good reason to believe that the collective motive indexes might not be
an index of the percentage of people in the population normally high in the mo-
tive at a given time period, for data on individual motive levels summarized in
Chapters 7 to 9 suggest that they are reasonably stable over periods as long as
fifteen to twenty-five years. Shifts in collective motive indexes are often quite
sharp and rapid, so when the Standard score for n Power in populär literature in
the United States goes from 49 in the 1950s to 63 in the 1960s, it seems unlikely
that the proportion of individuals stably high in n Power had gone up by that
much. In fact, the data based on n Power scores from representative samples of
individuals in Table 11.12 suggest that this was not the case.
However, it is certainly not impossible for the proportions of individuals in
a country high in a given motive to change markedly over time, as the data on
n Achievement for women in Table 11.12 indicate. This is due in part to new
cohorts entering the adult population and old ones leaving it through death, and
in part to the effect of Special forces like the women's movement on Special
groups within the population. Veroff (1982) reports that the increase in
n Achievement levels for women from 1957 to 1976 appears to be due to a co-
hort effect—to the maintenance of a high level of n Achievement among women,
as they grow older, who belonged to the generation most affected by the
women's movement—and to the dying out of an older generation among whom
n Achievement had been very low in 1957.
On the whole it seems preferable to think of collective motive scores as
more analogous to aroused motivational states, which can shift quite rapidly as a
result of influences like the women's movement or a period of major unemploy-
ment. These aroused motivational states may create effects on group behavior
much like the effects produced in individuals by the chronic states of arousal
that we have labeled motive dispositions. Such states of collective motive arousal
may be conceived of as part of the ideology of the collectivity—whether it is a
culture, a nation, or a business firm.
Motivational ideology is not explicit in the sense that people can report ac-
curately what it is or answer questions in a public opinion survey that enable in-
vestigators to infer what it is. However, it is nevertheless very important in de-
termining actions a group takes, as the evidence reviewed in this chapter
indicates.
NOTES AND QUERIES

1. Locate a sample of your doodles and try scoring them for signs of
n Achievement using Aronson's (1958) coding System. Compare your results
with someone eise's. Do your results agree with your comparative level of
n Achievement as determined from coding your imaginative stories?
2. Go to a störe, examine carefully a wide variety of vases, decide which ones
you like best, and then see if they contain the designs characteristic of high
or low n Achievement according to the Aronson System. Do the vases you
like best contain the same designs you put into your doodles? Explain why
the designs you like in a vase might be different from the designs you spon-
taneously produce.
3. Populär songs can be considered to represent the collective motive concerns
of at least those who listen to them. Score for n Achievement the lyrics of
the ten most populär songs in the United States at two different time peri-
ods, one of which precedes economic growth (say, in the early 1960s)
and the other, economic decline (say, in the mid- 1970s). Do the expected
differences in n Achievement levels occur? If not, why not? If the songs
reflect the concerns of young people, might the lag in the effect of these
concerns be later, after the young people are playing a more central role in
society?
4. In the Communist Manifesto (1848), Karl Marx and Friedrich Engels State
the following:
The discovery of America, the rounding of the Cape, opened up fresh ground for the
rising bourgeoisie. The East-Indian and Chinese markets, the colonization of Ameri-
ca, trade with the colonies, the increase in the means of exchange and in commodi-
ties generally, gave to commerce, to navigation, to industry, an impulse never before
known. . . . Meantime the markets kept ever growing, the demand ever rising. . . .
This development has, in its turn, reacted on the extension of industry; and in pro-
portion as industry, commerce, navigation, railways extended, in the same propor-
tion the bourgeoisie developed, increased its capital and pushed into the background
every class handed down from the Middle Ages.
Examine critically this motivational analysis in terms of the empirical

studies reviewed in this chapter on the motivation behind economic
expansion.
5. Make a careful comparison of the motive levels of U.S. presidents in a given
decade (see Table 8.8) with the motive levels of that decade as determined
from populär literature (Figure 11.9). How much agreement do you find?
Try to explain whatever results you obtain. Look also for systematic leads
and lags: Do presidents' motive levels regularly precede or follow motive
levels in populär literature? You can also make a more detailed analysis by
comparing presidential motive levels with motive levels in different types of
literature (see McClelland, 1975).
6. Try to get figures from Amnesty International for numbers of deaths in a
country for political reasons, and correlate these figures (perhaps corrected
for population size) with collective n Affiliation scores. Is there a relation-
ship? Try to explain any major discrepancies.
7. It is suggested in this chapter that the Roman Empire developed in response
to a collective imperial power motive pattern. Design a study that would
test this hypothesis. Be specific about what you would do, and construct a
table including imaginary numbers that would confirm the hypothesis.
8. Inkeles and Levinson (1954) argue that the term group mind should refer to
the average or modal psychological characteristics of a representative sample
of individuals from the group. Give some alternative explanations of the
term, and discuss its possible relationships to characteristics of individuals in
the group.
9. Do you think collective motive levels have predictive power for periods of
social reform and violence only in democratic countries like the United
States and Great Britain? Why might this be so? Design a study that would
test whether similar relationships would hold in another type of country.
10. Compute the relationship between a president's motive levels (see Table 8.8)
and the percentage of the national budget he has proposed spending on
defense. If there are major discrepancies in what the president asks for
and what is voted, can it be explained in terms of discrepancies between
presidential motives and collective motives, as determined from populär
literature?
11. Design a study that would determine whether changes in n Achievement
level preceded or followed changes in governmental structure in Swiss can-
tons, some of which became Protestant and some of which remained Catho-
lic (see Swanson, 1967). How would you interpret the results, however they
came out?
12. Collective data suggest that in the United States, major periods of unem-
ployment are followed by a rise in n Power in populär literature. Design a
study that would test this relationship at the individual level. Be sure to in-
clude a way to test for possible sex differences. Would you expect sex
differences?
13. In 1982 unemployment levels were higher in the United States than they
had been since the 1930s. In terms of findings reported in this chapter on
the effects of motive levels in the history of the United States, construct a
scenario of how 1982 unemployment might affect motive levels, which in
turn might affect collective actions taken in the United States in subsequent
years.
14. From a study of country differences (Table 11.6 or McClelland, 1975), from
shifts in motive levels in the history of the United States, or from theory,
try to construct some hypotheses as to what factors (other than wars) might
increase collective concerns for affiliation. Formulating and testing such hy-
potheses could be very important for the future health and well-being of
people. Explain why.
12
Cognitive Effects on
Motivation
Cognitive Influences on Motive Arousal

Cognitive Effects on Incentives
Cognitive Influences on the Arousal Value of Cues
Is Motive Arousal Solely a Function of Cognition?
Alterations in Motive Arousal by Cognitive Dissonance
Motive-related Cognitions
Causal Attributions
Causal Attributions Related to Achievement Motive Strength
Effects of Causal Attributions on Subsequent Behavior
Evaluation of the Cognitive Theory of Motivation
Future Orientation and the Achievement Motive
Cognitions Affecting the Translation of Motivation Into the Impulse to Act

Self-confidence or Self-efficacy
Values
COGNITIVE INFLUENCES ON MOTIVE AROUSAL
Several theorists, notably Weiner (1980a) and Heckhausen (1980), have stressed
the great importance of cognitive factors in the motivation-action sequence. As
Weiner (1979) puts it, "Comprehension Stands with hedonism as among the pri-
mary sources of motivation." He feels that in motivation theory too much em-
phasis has been placed on affective arousal and not enough on the understanding
the person has of what is happening during a motivation-action sequence, which
determines whether affective arousal occurs or not.
A great many empirical studies have been carried out to clarify the relation-
ship between cognition and motivation, but before reviewing them it is worth re-
examining Figure 6.1, which identifies the key factors in a motivation-action se-
quence. Arousal demands (cues) typically contact an incentive, which, if it
relates to an existing motive disposition, leads to an aroused motive or motivation
to act. When, how, and whether this motivation gets converted into action is in-
fluenced by skills, cognitions (values), and opportunities, which determine
whether a particular kind of behavior occurs or not.
As noted previously, Weiner, Atkinson, and others use the term motivation
to describe the final excitatory potential for an act (the impulse to a given act)
after it has been influenced by expectations and values, whereas we use the term
motivation in the more restricted sense to refer to an aroused motive before it is
influenced by expectations and values that shape preferences for particular acts.
In some instances the two uses of the term motivation have, for all practical
purposes, the same meaning, as when an aroused hunger motive (that is, hunger
motivation) is associated with a strong impulse to eat (that is, motivation to eat
in the Atkinson sense). But when an aroused motive is not so simply and di-
rectly connected with the impulse to act in a particular way, it is important to
realize that aroused motivation is different from the impulse to act. For exam-
ple, power motivation may be aroused by watching a film of John F. Kennedy's
inaugural address as president (Winter, 1973), but what kind of impulse to act it
is associated with varies from person to person and is shaped in various ways by
each person's skills, values, and opportunities.
The distinction is important in trying to understand just how cognition in-
fluences motivation, because by definition it influences excitatory potential or the
impulse to act, but how it influences or determines the course of motive arousal
needs clarification. How experimenters explain a Situation to subjects and what
the subjects think is going on have important effects on what the subjects want
and do, as studies to be reviewed subsequently will show. But to infer that cog-
nition therefore determines motivation is to forget that the experimenter's expla-
nations may have an effect on the subjects' impulses to act in certain ways be-
cause most subjects are already motivated to comply in some way with what the
experimenter is asking them to do. It is not altogether clear the extent to which
explanations or cognitive factors create motivation (that is, arouse motives) or
merely Channel motives that have already been aroused in the Situation in the
direction of certain acts. So we need to examine how cognitions influence motive
arousal as well as later events in the motivation-action sequence.
Cognitive Ejfects on Motivation 475
Cognitive Effects on Incentives
To Start with the most obvious effect of cognition, growth in understanding in-
fluences what constitutes an incentive (see Chapter 5). At the outset, pleasure
is derived from variations around a simple Standard to which one is adapted
in many sensory or conceptual modalities. As one becomes used to the new
variations, more complex variations are needed to give pleasure in music, art,
or poetry. Similarly, children get pleasure from simple impacts (for example,
from dropping things), but as they grow up the impacts that give pleasure
become more and more social and symbolic until some people may get impact
pleasure only from influencing the course of national events. Or simple contact
gratifications in time become elaborated into the symbolic pleasures of roman-
tic love. The effects of these shifts in what constitutes an incentive or motive
arousal are obvious. Giving a child an opportunity to play with a toy car may
arouse the achievement motive, but giving the same opportunity to an adult
does not. Or watching a film of Kennedy's inauguration arouses the power
motive in adults but does not in children, who do not understand what was
happening.
Despite such examples, not much empirical work has been done on the ca-
pacity of incentives at different levels of cognitive complexity to arouse motives.
What has been done has dealt with the effects of Performance feedback on the
incentive value of tasks for the achievement motive. As noted in Chapter 7, peo-
ple high in n Achievement seek to perform tasks for which there is a moderate
probability of success, for only in such tasks do they maximize satisfaction from
doing them well. But "moderate probability of success" is a movable Standard:
it varies as the person gets success or failure feedback in the course of perform-
ing a task. How the person evaluates the feedback and consequently adjusts the
incentive value of the task should affect the amount of achievement motivation
aroused in the Situation. That is, if success on a very difficult task that arouses
little achievement motivation moves the perceived probability of success at the
task into a fifty-fifty chance of success, the task should arouse more achievement
motivation to do well on it in the future.
Heckhausen (1975a) carried out an experiment that indirectly suggests that
this effect occurs. He arranged for subjects to succeed at a task most of the
time. Afterward he asked the subjects to rate the extent to which they thought
the outcome was attributable to their own ability on this task, to their ability to
perform this task relative to others, or to other factors such as chance or effort.
As one would expect, most subjects rated their ability for this task higher after
success than they had beforehand. Success changed the incentive value of the
task for them, since it altered their perceived probability of success at the task,
which in turn should have raised the level of their achievement motivation.
This is a direct demonstration of how an understanding of what has happened
can modify the incentive value for performing a task, which affects motive
arousal.
However, there was one group of subjects for whom this effect did not
occur, as Figure 12.1 demonstrates. For subjects low in n Achievement (or high
| 1 Low Social Norm Ability
| ~ J High Social Norm Ability
<
2£
1 1
Hope of Success Fear of Failure

(High n Achievement) (Low n Achievement)
Figure 12.1.
Increase of Ability Estimates (Individual Norm) After Success for Subjects Classified High and Low
in n Achievement and in Social Norm Ability (after Heckhausen, 1975b).
in / Failure) who also regarded themselves as not very good at such tasks rela-
tive to others, there was no significant increase in estimated ability for the task
after success. We can infer that for them, actual success did not increase their
achievement motivation. On the other hand, subjects low in n Achievement who
feit they were good at the task relative to others were very encouraged by suc-
cess to believe that they had high ability for the task, which should have raised
the level of their n Achievement. In other words, both their motive level and
their beliefs about their competence relative to others influenced how a success-
ful Performance was evaluated, which in turn influenced the incentive value of
the task as shown by the judgment of their ability to perform it.
Many studies of this type have been carried out in the achievement area, as
we shall see, but most of them have focused on how shifts in perceived probabil-
ity of success influence the impulse to work hard at a task rather than in how
they alter the incentive character of the work Situation so as to arouse or dimin-
ish achievement motivation. Nevertheless, there is no doubt that alterations in
perceived probability of success are aflfected by the level of n Achievement and
affect the amount of achievement motivation aroused. (See the discussion later
in the chapter in connection with Figure 12.10 of the experiment from which
the data in Figure 12.1 were obtained.)
Cognitive Effects on Motivation \11
5 10 15 20 25 30 35 40 45 50 55 60 65 70
Start
of Film Time (in 15-Second Intervals)
Figure 12.2.
Skin Conductance Curves During Orientation and Film Periods Under Three Experimental
Conditions (Weiner, 1980a, after Lazarus & Alfert, 1964).
Cognitive Influences on the Arousal Value of Cues

According to Figure 6.1, the motivational sequence Starts when certain Stimuli,
cues, or demands contact an incentive (or elicit a goal in the mind of the sub-
ject), but cognitive factors play an important role in whether or not such con-
tact occurs. Lazarus and Alfert (1964) showed College students a film in which
adolescent boys undergo a tribal ritual in which their penises are deeply cut.
The students' viewing the film is sufficient to contact the pain incentive, which
leads to the arousal of the anxiety motive, as Figure 12.2 shows. In this instance
the measure of anxiety is the galvanic skin response, or the increase in electrical
conductivity of the skin due to imperceptible sweating from anxiety. The top,
"silent" curve shows the great increase in skin conductance reflecting anxiety
that accompanies the film when it is presented silently. However, in one ("denial
orientation") condition, before viewing the film the subjects received an elabo-
rate explanation on a sound track of the significance of the ritual by an objective
anthropologist in which he also stated that the procedure was not harmful. This
cognitive orientation significantly reduced the negative incentive value of the

film and the accompanying arousal of anxiety. The same denial commentary
presented on a sound track along with the film ("denial commentary" condition)
also reduced anxiety somewhat, but not as much as reorienting the subjects be-
fore they saw the film.
As this experiment demonstrates, a conscious understanding of what is
going on can modify the arousal effects of cues. In Chapter 6 a study was re-
viewed in which it was found that the cues associated with going without food
for as long as sixteen hours increased incentive value, as reflected in imaginative
stories about getting food rather than actually seeing and eating it (Atkinson &
McClelland, 1948). However, Sanford (1937) reported that subjects who con-
sciously fasted for equivalent periods of time did think more about food and eat-
ing it as time went along. In other words, knowledge that one was fasting
changed the nature of the incentive associated with arousal of the hunger mo-
tive. Schachter's (1971b) research on obesity, also reviewed in Chapter 6, dem-
onstrates that other cognitive factors, such as the time of day and the number of
external cues available in the environment, significantly affect the degree to
which hunger is aroused in overweight people (see Table 6.2). The importance
of such contextual understandings for arousal of the sexual drive is illustrated in
Table 5.2. In that study the presence of an attractive female folk singer did not
increase sexual arousal in young men, even when alcohol was also served, so
long as the experiment was conducted in a classroom. However, if the same pro-
cedure was followed in an apartment, sexual fantasies were increased. It seems
reasonable to infer that the students' understandings of what normally went on
in classrooms interfered with the capacity of the sexual cues to contact sexual
incentives and motives.
Several studies have demonstrated the way in which cognitive factors can
influence the capacity of instructions to arouse the achievement motive. As
noted in Chapter 6, when several groups of U.S. women understood that their
intelligence and leadership capacities were being tested, their average n Achieve-
ment score in stories written subsequently did not increase, as it did for U.S.
men who had been similarly instructed. Instead, the women's n Achievement
scores tended to rise when their social competence was being evaluated (see
Field, cited in McClelland, Atkinson, Clark, & Lowell, 1953). In other words,
for these U.S. women, motivational demands stated in terms of leadership and
intelligence did not have the same incentive value as for similar U.S. men. Their
cognitive understanding of what was important to them determined whether
achievement motivation was aroused or not.
Raynor and Teitelbaum (1982) have shown that even the time orientation in
the cues that elicit stories affects the extent of achievement motive arousal. They
asked subjects to write stories to sentence leads referring to a person who is
thinking about an event that defines "who he or she is becoming," "who he or
she is," or "who he or she has been." As Figure 12.3 makes clear, a cognitive
orientation toward the future elicits much more achievement motive arousal
than an orientation toward the past.
A subject's view of his or her competence for a task displayed in a picture
Cognitive Effects on Motivation 479
+ 2.50
+ 2.00 -
+ 1.50 -
.5 +1.00 -
+ 0.5 -
-0.5
Past Present Future
Orientation of Cue That Elicited Stories
Figure 12.3.
Mean n Achievement Score as a Function of Time Orientation of Verbal Cues That Elicited Stories
(after Raynor & Teitelbaum, 1982).
cue also affects the amount of achievement imagery in stories written to the pic-
ture. Shrable and Moulton (1968) had subjects rate their competence at tasks
displayed in various pictures, such as chopping a tree or playing a musical in-
strument. They found that more intelligent subjects wrote stories with more
achievement imagery to pictures displaying a task at which they feit they were
competent as contrasted to pictures displaying a task at which they feit less
competent. The reverse was true of subjects lower in intelligence. Clearly, peo-
ple's views of their general competence and their competence at a particular task
influence the extent to which particular cues can arouse achievement motivation.
What needs to be worked out in this area in more detail is what general cogni-
tive factors have the most influence in modifying the arousal value of demand
cues in different motivational areas.
Is Motive Arousal Solely a Function of Cognition?

Several studies have examined the way in which the subject's understanding of a
Situation influences the degree to which physiological cues are associated with
motive arousal. The original impetus for this type of research came as a reaction
to the behaviorist tradition, described in Chapter 3, in which drive Stimuli were

thought to involve physiological sensations that defined the drive. Thus, if an in-
vestigator could show that physiological sensations were not automatically asso-
ciated with drive arousal without a certain type of cognitive appraisal of the Sit-
uation, the behaviorist position would have been shown to have an important
limitation.
Valins (1966) studied the relationship of physiological cues to arousal of the
sexual motive. He showed male undergraduates a number of slides of seminude
females from Playboy along with a tape recording that in one condition pur-
ported to be the sound of the subjecfs heartbeat in response to each picture.
The auditory feedback was actually a Standard tape recording in which heart
rate went significantly up or down to half of the slides and showed no change
for the other slides. The slides that supposedly produced a heart rate reaction
are referred to as reinforced, whereas those that produced no Special response
are referred to as not reinforced. After the experiment was over the College men
were asked to rate the attractiveness of all the slides and were allowed to take
home the pictures they most preferred.
As Figure 12.4 shows, they rated slides as more attractive that in their own
minds they had associated with their heart beat's speeding up or slowing down.
They also chose more such slides to take home. In a control condition, subjects
observed the same slides and heard the same tape recording, but this time the
noises on the tape were explained as being related to something eise the subjects
were doing. As Figure 12.4 shows, if the subjects did not interpret the increased
or decreased noises associated with various slides as their own heart beats, there
was no difference in their preference for slides reinforced in this way as com-
pared with those not reinforced. Furthermore, the effect of the sound tape on
actual heart rate was the same under both conditions. Thus, physiological cues
were not related to sexual arousal unless the subject thought they should be.
That is, the subjects reacted to the information that their heart beats were going
up or down in response to certain slides by taking it to mean that they were
more "turned on" by certain women than others. This cognitive appraisal led to
an increase in the desire for these women, as shown by the higher attractiveness
ratings the subjects gave them and by their preferences for these pictures to take
home. Remember, however, that all the pictures were quite attractive and un-
doubtedly produced some sexual arousal, so what the cognitive information did
was shift preferences from one set of pictures to another, not create a sexual
arousal that did not exist before. It is doubtful if any amount of heart beat in-
formation could create a preference for a picture of a witch.
The experiment of this type cited most often by far is one conducted by
Schachter and Singer (1962). To produce physiological arousal, they injected
subjects with "suproxin," a drug that subjects were told would affect their vision
but that was actually either a placebo or a small amount of epinephrine (adrena-
line). Then, while the subjects waited for the drug's effects to occur, they were
put in a room with another subject who had supposedly received the same drug.
Actually, the second subject was a confederate trained to act either in a happy
or irritated way. In the "happy" condition the confederate behaved very cheer-
I
ides of
f^Reinforced Slides
Q s i i d e s Not Reinforced
7 0
•o
3
C
<L>
—l
p
ractiveness ]Ratings )
<
Control Noises Heart Rate Increase Heart Rate Decrease
Figure 12.4.
Eifects on Attractiveness of Pictures of Women When Changes in Heart Rate Are Associated with
Them (after Valins, 1966).
fully, laughing a lot, throwing paper airplanes, playing with a Hula-Hoop, and
generally encouraging the real subject to join in the fun. In the "irritated" con-
dition the confederate was filling out a questionnaire for the experimenter that
asked a lot of personal questions about sexual or financial matters. The confed-
erate expressed great irritation at having to answer such questions and generally
attacked the experiment and the experimenter, again encouraging the real sub-
ject to join in expressions of anger.
After this part of the experiment, the subjects filled out rating scales indicat-
ing how happy they feit and how irritated they feit. An overall "happiness"
score was computed by subtracting the irritation rating from the happiness
rating.
As Figure 12.5 shows, the two types of social treatment by the confederate
had no differential effect on how subjects feit afterward if they had received a
placebo injection. This is a little surprising, since it suggests that the experiment-
er's attempt to manipulate mood was ineffective. Schachter and Singer explain it
on the grounds that perhaps the subjects, after exposure to the angry confeder-
ate, did not show an increase in irritation because they were afraid to express
their irritation at the experimenter in füll.
The effects of treatment on mood were somewhat different if the subjects re-
ceived an epinephrine injection but were not informed of the physiological ef-
Happy Treatment
| Irritation Treatment
£ 2.00
i
"S ..50
1.00
.5 0.50
Placebo Epinephrine Epinephrine

Ignorant Informed
Condition
Figure 12.5.
Effects of Social Treatment on Mood of Subjects Receiving Injections Under Different Conditions
(after Schachter & Singer, 1962).
fects of the drug. When there is an actual physiological arousal (for example,
the increased heart rate from the injection of epinephrine), it looks as if the
treatment—representing what the subjects thought the effects should be based on
the behavior of another person receiving the drug—had a differential effect on
mood. Those exposed to a happy confederate rated themselves as somewhat hap-
pier than those who had been exposed to the irritated confederate. The differ-
ence shown in Figure 12.5 is not significant, but it was bolstered by coding the
behavior shown by the subjects in the waiting room. While being exposed to the
angry confederate, they engaged in significantly more acts of irritation than
those in the placebo condition, and while being exposed to the happy confeder-
ate, they engaged in somewhat more happy acts than those in the placebo condi-
tion, although not at a statistically significant level.
Schachter and Singer and others interpreted these results to mean that cog-
nitive appraisal determines entirely the motivational or emotional efFects of phys-
iological arousal. A person feels a physiological disturbance, as from an injection
of epinephrine, but the emotional label given to the State depends entirely on the
subject's understanding of the Situation. If people believe the State is associated
with happiness, they will feel happy; if they believe it is supposed to be associ-
ated with irritation, they will feel irritated. Peoples' observations and inferences
about themselves determine their emotions and motivations. As William James
put it long ago in describing the James-Lange theory of the emotions, people are
afraid because they observe themselves running; they do not run because they
are afraid.
The Schachter and Singer experiment has almost universally been cited as
support for this inference, although the evidence it provides is by no means con-
clusive. Marshall and Zimbardo (1979) attempted to replicate the happy treat-
ment effect and found, as had Schachter and Singer, that happiness did not in-
crease significantly after epinephrine injection over what it was after a placebo
injection. This raises a real question as to whether the cognition that the drug
was supposed to produce happiness in combination with the physiological
arousal from epinephrine in fact produced a happier mood than a person would
get without the physiological arousal. The subjects, in fact, did not really dis-
cover from the euphoric behavior of the confederate the cognition ('T feel
happy") to explain their internal arousal.
Furthermore, the biggest change in the effects of social treatment occurred
when the subjects were correctly informed of the effects of epinephrine on inter-
nal Symptoms. At one level this strongly Supports the contention that cognitive
understandings significantly modify the effects of physiological arousal on emo-
tions, but at another level it raises a major difficulty for the Schachter and
Singer point of view. Why should the subjects exposed to the angry confederate
feel so much happier after the experiment than those exposed to the happy con-
federate if they know what the effects of epinephrine are? This is the largest and
most significant difference in Figure 12.5, yet Schachter and Singer make little
of it, for from their point of view, once the subjects know accurately what the
effects of the drug are, the social treatments should have no differential effect or,
at any rate, not a reverse differential effect, on mood.
What this finding suggests is that there is a bias in a particular direction in
the effects of epinephrine on mood. As noted in Chapter 8 on the power motive,
there is reason to believe that the catecholamines, of which epinephrine is one,
subserve the power motive System. Thus, the stirred-up physiological reactions
that go with epinephrine are often associated with anger, aggression, and trying
to have impact on others. Once the subjects are allowed to identify correctly
these internal reactions—in the informed condition—they can express irritation
consistent with those internal cues obtained when they were exposed to the
angry confederate. In the ignorant condition they are more confused about the
meaning of the physiological signs, and when they act in an irritated way after
exposure to the irritated confederate, they are not behaving as consistently with
the internal cues as they are in the informed condition. So they feel less satisfied
in the ignorant than the informed condition.
We then need only to assume further that expressing anger in a way that is
consistent with internal cues has a cathartic effect: the subjects feel better from
having expressed anger that is consistent with their State of physiological arous-
al, and thus they rate themselves happier afterward. In contrast, exposure to a
happy confederate suggests a reaction that is totally inconsistent with the physi-
ological signs that normally go with anger and irritation, so in the informed
condition subjects remain confused and rate themselves less happy at the end of
the experiment than in the ignorant condition. Admittedly, such an interpreta-
tion is speculative and requires further checking, but the results do strongly sug-
gest that the physiological arousal caused by epinephrine—if it is understood—
affects the outcomes of the treatments in a way that suggests that its effects are
biased more in one direction than another. This adds to the skepticism that cog-
nitive appraisal, all by itself, can turn the physiological arousal due to epineph-
rine into happiness. However, cognition certainly has one important effect:
knowledge of the effects of epinephrine completely reverses the effects on moods
of the two types of treatments in combination with epinephrine arousal.
Alterations in Motive Arousal by Cognitive Dissonance

Later research demonstrated that physiological states were not simply the pas-
sive product of cognitive understandings, as suggested by Schachter and Singer,
but could themselves influence cognition. To create cognitive dissonance, Zanna
and Cooper (1974) asked College students to write a brief essay supporting the
view that inflammatory Speakers should be banned from the campus. Most stu-
dents disagreed strongly with this viewpoint, so writing such an essay could
create dissonance because they would find themselves saying things they did not
believe. In one condition the students were told the experimenter realized that
performing the task would involve going against their beliefs, and they were
given a free choice as to whether to do it or not. In the other condition the stu-
dents were simply asked to perform the task and were not given the choice of
not doing it. At the outset of the experiment they were given a pill, which dif-
ferent groups of subjects were told would either make them more tense, more
relaxed, or have no discernible effects. Ratings of mood states by the subjects
showed that the suggested effects of the pill occurred: those who were told it
would relax them feit more relaxed, and those who were told it would increase
tension feit more tense, even though the pill they received was the same and
should not in reality have had any effects.
Figure 12.6 presents the mean agreement with the view that Speakers should
be banned from campus for subjects exposed to various treatments (Zanna &
Cooper, 1974). Consider first the results for subjects on whom the pill had no
effect. Students who freely chose to write a counterattitudinal Statement ended
up agreeing more with that viewpoint than subjects who wrote similar State-
ments, but without having freely chosen to do so. Cognitive dissonance can ex-
plain the result: Subjects who found themselves voluntarily supporting a position
with which they strongly disagreed should be in a State of conflict or confusion.
Why would they do such a thing? To reduce the conflict or dissonance, they
moved toward agreement with the disavowed opinion so that they could feel less
conflict for having endorsed it. Subjects who had not freely chosen to write this
Subjects Have Free Choice to Parlicipate
Q 15 1 Subjects Do Not Have Free Choice to Participate

c
o
I
.£ 10
c
cd
-o
Tension None Relaxation

Effects of Pill
Figure 12.6.
Effects of Meaning of Taking a Pill on Attitudes When Subjects Do or Do Not Have the Free
Choice to Participate (after Zanna & Cooper, 1974).
Statement feit less dissonance, because they could explain their behavior by say-
ing that they were more or less forced to do it. So they would be under less
pressure to change their opinion to explain their stränge behavior.
Next, notice what happened when the pill created a State of tension in the
subjects: The cognitive dissonance effect disappeared. Now the subjects in the
free choice condition had an adequate physiological explanation for their feeling
of tension from doing something stränge. They no longer needed to modify their
attitudes to reduce the feeling of tension, for they perceived that it derived from
a physiological source. On the other hand, if the pill produced relaxation, the
cognitive dissonance effect was magnified. Now the subjects in the free choice
condition had two reasons for feeling stränge: they were voluntarily doing some-
thing inconsistent with their beliefs, and furthermore they were feeling very re-
laxed about it, which seemed even stranger. So they modified their opinion even
more toward a disavowed position because of the much greater feeling of disso-
nance or inconsistency.
The important inference to be drawn from this experiment is that physiolog-
ical states, or at least perceived physiological states, modified what the subjects
understood was going on. One physiological State (induced tension) destroyed
the cognitive dissonance effect obtained when the pill had no psychological ef-
fect; another (induced relaxation) enhanced the cognitive dissonance effect. So
the conclusion is the same as the one we drew from the Schachter and Singer
experiment: physiological cues influence cognitive perceptions, just as cognitive
perceptions influence the interpretation of physiological states.
There is no reason to conclude that cognitive factors alone (the knowledge
that one is doing something stränge voluntarily) create and control the disso-
nance the subjects are motivated to reduce. Perceived physiological states also
contribute to whether the negative incentive of dissonance occurs. Of course, we
can argue, as Zanna and Cooper do, that cognitive expectations produced the
perceived physiological states so that it looks as if all the effects in the experi-
ment were based on cognitions. However, since the physiological effects were
real in the sense that the subjects actually feit relaxed or tense, we can view
them as Controlling the dissonance effect directly, regardless of how they were
produced.
Other cognitive dissonance research has dealt more directly with its effects
on the arousal of other motives. In a typical study (Grinker, 1969), subjects par-
ticipated in an eyelid conditioning experiment in which air puffs are delivered to
the eye, causing it to blink. If a tone precedes the air puff slightly, in time the
eyelid will blink to the tone before the air puff is delivered. That is, the tone
elicits a conditioned avoidance response to the mildly painful air puff. After the
subjects were conditioned, they were told that the intensity of the air puff would
be increased in the next series of trials. In the control condition this was all the
subjects were told, and in the subsequent trials they showed a large gain in the
percentage of eye blinks to the tone to avoid the more painful air puffs. In the
dissonance condition the subjects were asked whether they wished to continue in
this more unpleasant series of trials. All freely chose to do so, but they showed
only a very slight increase in the percentage of eye blink avoidance responses to
the stronger air puffs. This result indicates that the cognitive dissonance manipu-
lation had directly decreased the pain experienced from the air puffs or de-
creased the arousal of the fear motive based on the pain. To justify their volun-
tarily undergoing such an unpleasant experience, the subjects somehow reduced
its unpleasantness. A different understanding of what they were doing had de-
creased the capacity of the cues in the Situation to arouse the anxiety motive.
In a more obvious manipulation of drive arousal, Mansson (1969) flrst cre-
ated thirst in subjects by having them eat a lot of dry crackers with a hot sauce.
After obtaining ratings on how thirsty they feit and other matters, he asked
them if they were willing to participate in a further experiment that involved
going without water for a period of time. In the low dissonance condition he
gave a strong justification for the deprivation procedure, explaining that it was
necessary to get good results, and he further noted that the time of deprivation
would be short. In the high dissonance condition he gave no justification for the
procedure and explained that the time of deprivation would be long. Thus, the
subjects in this condition were voluntarily agreeing to remain thirsty for a long
period of time for no discernible reason. In the moderate dissonance condition
either the justification was strong and the deprivation was long, or the justifica-
tion was weak and the deprivation was short. Data were also obtained from sub-
jects in a control condition who were not exposed to the dissonance treatment
and from subjects who, unknown to the experimenter, had refused to take part
in the second experiment.
Mansson then examined many different kinds of responses of the subjects to
see whether the cognitive manipulations had affected the thirst drive. He found
that subjects in the high dissonance condition, as contrasted with those in the
low dissonance condition, gave fewer water-related responses in stories written
to pictures, learned thirst-related words more slowly, rated themselves as less
thirsty, and actually drank less water when given an opportunity, as shown in
Figure 12.7. Note that the thirst manipulation worked: those in the control con-
dition who had eaten salty, highly spiced crackers drank much more than those
who had eaten crackers and peanut butter. Yet those in the high dissonance
condition drank much less than this, although they too had eaten the highly
spiced crackers. Those in the low dissonance condition drank much more than
those in the high dissonance condition, as did those who had refused to partici-
pate in the experiment. In short, those in the high dissonance condition gave
many different signs of being less thirsty than subjects in other conditions. The
cognitive pressure to explain their stränge, inconsistent voluntary behavior was
strong enough to interfere with the capacity of the dry sensations in the mouth
(physiological cues, in Figure 6.1) to arouse the thirst motive.
The type of cognitive manipulation involved in dissonance studies is pecu-
liar, but in theoretical terms its effects are no different from examples of similar
phenomena described earlier. For instance, in the Lazarus and Alfert genital Op-
eration film, the cues arising from viewing what appeared to be a painful Opera-
tion were blocked or diverted from arousing the fear motive by cognitive expla-
nations that the Operation was not so painful and fully justified.
Cognitive dissonance can either increase or decrease motive arousal, depend-
ing on other factors. Glass and Wood (1969) had subjects rate another person
both before and after they had apparently given the other person a strong elec-
tric shock. If they were not given a free choice as to whether they would partic-
ipate in shocking someone (low dissonance condition), there was no change in
their liking for the other person. However, if they voluntarily chose to shock the
other person, they disliked the other person more afterward, in order to justify
their stränge behavior, but only if they were also high in self-esteem. Subjects
low in self-esteem in the high dissonance condition actually liked the other peo-
ple more after shocking them. It is as if they knew they were no good and that
shocking another person simply confirmed their negative self-image, so they in-
creased the likability of the other person to justify still further their negative
picture of themselves. In contrast, those high in self-esteem found themselves
voluntarily doing something very inconsistent with their positive self-image, so
they had to increase their dislike for the person to explain to themselves why
they were doing it. Here a third variable, self-esteem, interacts with a cognitive
dissonance treatment to raise or lower the arousal of the motive to feel aggres-
sive toward another (an aspect of the power motive). Obviously, the effects of
cognitive variables on motive arousal are varied and complex.
"3
Low High High Moderate Low Refusal

Control Dissonance
Experimental Treatment
Figure 12.7.
Mean Water Consumption in Cubic Centimeters for the Low and High Control, High Dissonance,
Combined Moderate Dissonance, Low Dissonance, and Combined Refusal Groups (after Mansson,
1969).
• MOTIVE-RELATED COGNITIONS
Causal Attributions
Cognitive variables not only influence motive arousal; they also accompany and
follow the behavior associated with it. A very extensive research literature has
grown up around the explanations people give for what they have done in
terms of how such explanations derive from existing motive states or feed back
and affect motive states. The interest in this field developed from two sources.
One was the research on cognitive dissonance, which showed that if subjects ex-
plained or justified their behavior for external reasons, it left their internal mo-
tive states relatively unaffected, but if they could not find external justifications,
their internal motive states were likely to be altered in significant ways, as in
Mansson's thirst experiment. So explanations for behaviors or causal attribu-

tions seem to be important modifiers of motive states. The other source of inter-
est was research on the reasons subjects give for success or failure in an
achievement-related context. Taking a clue from an earlier analysis by Heider
(1958), Weiner and others (Frieze & Weiner, 1971; J. P. Meyer, 1980; Tasser,
1977; Weiner, 1980a; Weiner & Kukla, 1970) have examined the conditions
under which subjects attribute success or failure to ability, effort, task difficulty,
or luck.
Table 12.1 lists some of the cues people utilize in making inferences con-
cerning the reasons for success and failure. For example, if people are often suc-
cessful at a certain type of task, they are likely to attribute the outcome to abili-
ty. If Performance shows a wide Variation, people are likely to attribute it to
differences in effort. If they are unexpectedly successful, they are likely to attrib-
ute it to luck.
These are not the only reasons given for success and failure, although they
are the ones on which the most research effort has focused. Falbo and Beck
(1979) suggest that less than half the reasons people give can be classified into
these four categories. Subjects typically mention such other factors as calmness,
carelessness (which could indicate lack of effort), or personality problems. Fac-
tor analyses of all sorts of different reasons people give support the conclusion
that there are three general types of reasons given for success and failure. One
type involves an internal-external dimension. People attribute the outcome either
to factors in themselves such as effort or ability, or to external factors such as
task difficulty or help from another person. The second type involves a stability
dimension. People attribute the outcome either to some stable, relatively un-
changeable cause such as ability, or to a variable cause such as effort or luck.
The third type involves a controllability dimension. People attribute the outcome
to a factor over which they have some control, like effort, or over which they
have no control, like luck.
Table 12.1.
SOME CUES UTILIZED FOR INFERENCES CONCERNING THE CAUSES OF
SUCCESS AND FAILURE (Weiner, 1980a)
Causes Cues
Ability Number of successes, percentage of successes, pattern of success,
maximal Performance, task difficulty
Effort Outcome, pattern of Performance, perceived muscular tension, sweating,
persistence at the task, covariation of Performance with incentive value
of the goal
Task difficulty Objective task characteristics, social norms
Luck Objective task characteristics, independence of outcomes, randomness of
outcomes, uniqueness of event
Research on causal attributions linked up with motivation theory when it

was discovered that people with a strong or weak achievement motive gave dif-
ferent explanations for success and failure. Weiner (1980a) concluded that those
high in n Achievement attributed success to ability and effort and failure to bad
luck or lack of effort (see also W. U. Meyer, 1973). Those low in n Achieve-
ment (or high in / Failure) perceived success as due to luck and failure as due
to lack of ability. Following the lead of others who were defining emotions in
cognitive terms, Weiner concluded from these observations that the achievement
motive was a cognitive disposition, a pattern of explanations for Performance
that gives rise to affect. Thus, a woman high in n Achievement feels good about
performing well, because she attributes her success to high ability, whereas a
woman low in n Achievement does not feel particularly good after success,
because she attributes it to luck. "The achievement motive may be defined as
a capacity for perceiving success as caused by internal factors and failure by
unstable factors. . . . Thus the achievement motive is a cognitive, rather than af-
fective, disposition. . . . Affect follows cognitive appraisal" (Weiner, 1980a).
The way attributions influence motivation to achieve is described by Weiner
(1979) in scenarios like the following. Suppose some students have just received
grades on an examination. The first one says to himself, "I just received a D on
the exam. That is a very low grade. There really is something lacking in me.
What I lack I probably always will lack." As Weiner points out, attributing fail-
ure to lack of ability typically leads to hopelessness and a lack of trying to do
better, which would seem to characterize people low in n Achievement. In con-
trast, a second Student also receiving a D and evaluating it as a poor grade may
say to herseif, "I didn't really study for the exam. Next time I am going to try
harder, and because I know I have the ability, I am sure I will do better." This
causal attribution pattern should characterize the person with high n Achieve-
ment and should also be associated with doing better next time, for Heckhausen
(1975a) has shown that effort intended in a subsequent task is correlated .63
with effort actually expended. W. U. Meyer (1973) has demonstrated that sub-
jects who attribute poor Performance on the first trial of a task to lack of effort
tend to work faster on the second trial (see Figure 12.8). Notice, in contrast,
that subjects who attribute failure to the stable factors of low ability or task dif-
ficulty do not work harder or do better on the next trial. Weiner reasoned that
the differential effect of these causal attributions on Performance is similar to the
effect expected for individuals high and low in n Achievement. Subjects who at-
tribute failure to lack of effort tend to do better, just like subjects high in
n Achievement; those who attribute failure to lack of ability tend not to do bet-
ter, just like subjects low in n Achievement.
Weiner (1981) has also extended his attributional analysis of motivation be-
yond the area of achievement. He constructed a number of episodes (Weiner,
1980b) for students to read and react to, such as the following:
At about 1:00 in the afternoon you are riding a subway car. There are a number of
other individuals in the car and one person is Standing holding on to the center pole.
Suddenly this person staggers forward and collapses. The person is carrying a black
Low Attribution
6
? High Attribution
r 4
5
'S
&
Ability Task Effort Luck Ability

Difficulty Plus
Task
Difficulty
Atiributional Elements
Figure 12.8.
Intensity of Performance, in Seconds (Trial 1 Minus Trial 2) As a Function of Attribution to the
Four Causal Elements and to the Combined Stable Factors. High numbers indicate greater improve-
ment in speed (after Weiner, 1980a, after W. U. Meyer, 1973).
cane and apparently is ill. (Alternate form: the person is apparently drunk. He is
carrying a liquor bottle wrapped in a brown paper bag and smells of liquor.)
Try to assume that you actually are on the subway and try to imagine this
scene. Describe your feelings in this Situation.
Then the subjects rated the causes of the person's falling on the dimensions of
internal-external, controllability, and stability. In general, it was clear that the
students saw the cause of the drunk's falling as internal and controllable,
whereas they saw the fall of the ill person as uncontrollable. Furthermore, the
ill person elicited emotions of sympathy and desire to help, whereas the drunk
elicited disgust and unwillingness to help. In Weiner's view, the subject's cogni-
tive appraisal of the Situation (ill or drunk) and the reasons given for it com-
pletely determine the emotional response (sympathy or disgust) and the motiva-
tion that is subsequently aroused (to help or not help).
In terms of the model of the motivation-action sequence in Figure 6.1, there
is no reason to doubt that understanding affects what a person intends to do

after a motive is aroused. It may even feed back after an action and affect the
level of motive arousal itself, as we have seen. But this is quite different from
saying, as Weiner seems to, that the pattern of causal ascriptions is the motive.
To examine this possibility, it is first necessary to review carefully the evidence
on causal attributions of people differing in motive strength and the evidence on
the subsequent effects on behavior of such causal attributions.
Causal Attributions Related to Achievement Motive Strength

Unfortunately, the empirical findings in this area are not as clear-cut as one
could wish for or as Weiner's theory requires (Weiner, Russell, & Lerman,
1979). The best study has been performed by W. U. Meyer (1973), who asked
subjects for their Performance goal for a task; gave them feedback of results that
was higher, lower, or the same as compared with expectation; and then asked
them to rate the extent to which their Performance had been due to ability, ef-
fort, or luck. The results he obtained are shown in Figure 12.9. Consider first
the causal attributions given after success. Subjects high in n Achievement
(h Success), as compared to those low in n Achievement (/Failure), explained
success as more due to ability and effort, and less to luck. So far as failure is
concerned, those high in n Achievement attributed it more to lack of effort and
luck, whereas those low in n Achievement attributed it to lack of ability.
Jopt and Ermshaus (1977, 1978) found similar results for intellectual, but
not manual, tasks and also reported that subjects high in/Failure were more
likely than subjects high in h Success to attribute failure to task difficulty. Oth-
ers have found no attributional patterns associated with the achievement motive
(Schneider, 1977). Weiner et al. (1971) found a pattern of results that was some-
what different from Meyer's. For instance, they found that subjects high in
n Achievement attributed both success and failure more to ability, and that
those high in n Achievement attributed failure less, rather than more, to lack of
effort, as in Figure 12.9. Some of the findings reported by the Weiner group are
complicated by the fact that they regularly use the Mehrabian measure of
v Achievement as if it were an adequate measure of n Achievement, despite the
lack of either theoretical or empirical support for such a procedure (see Weiner
& Potepan, 1970). Entin and Feather (1982) even found that subjects low in
n Achievement attributed success to effort and ability in a pattern almost identi-
cal to that followed by those high in n Achievement. However, the task used in-
volved working only for two minutes on anagrams before Performance feedback
was given and causal attributions requested. Other studies have typically in-
volved the subjects in much longer, more demanding tasks.
What we must conclude from these results is that a number of variables
other than achievement motivation affect causal attributions and that therefore it
would be quite unsafe to use the pattern of causal attributions to diagnose
n Achievement levels, as would seem to be required by a theory that states that
the achievement motive is defined by the pattern of causal attribution. The most
general conclusion that can be drawn is that subjects high in n Achievement
Success Expected Failure Success Expected Failure

(Higher Than (Lower Than (Higher Than (Lower Than
Expected) Expected) Expected) Expected)
High n Achievement (h Success) 4 -

Low n Achievement (/Failure)
Success Expected Failure
(Higher Than (Lower Than
Expected) Expected)
Figure 12.9.
Average Stren^th of Causal Attributions to Ability, Effort, and Luck for Success- and Failure-
motivated Subjects If the Number of Correct Responses Is Higher or Lower Than Expected or as
Expected (after Heckhausen, 1980, after W. U. Meyer, 1973).
tend to attribute success to ability and failure to lack of effort, whereas subjects
low in n Achievement tend to attribute failure to lack of ability (Heckhausen,
1980).
Effects of Causal Attributions on Subsequent Behavior

Most studies of attribution simply ask the subject to think back and try to ex-
plain previous Performance. Heckhausen (1975a) went a step farther and exam-
ined how causal attributions affected what people intended to do next. He first
arranged for subjects to be successful on about 50 percent of the trials in per-
forming a task. After a period of rest, subjects were divided into two groups, in
one of which success was experienced on about 80 percent of the trials (the suc-
cess condition) and in the other, on about 20 percent of the trials (the failure
condition). They were also led to believe that there would be some further tasks
to accomplish. After each phase of the experiment they were asked to give a
self-evaluation, their intended level of effort, and their Performance goals for the
next phase of the experiment.
Heckhausen found that whereas subjects high in n Achievement might
attribute failure to lack of effort, in this Situation neither those high nor low in
n Achievement showed a gain in intended effort after a period of consistent fail-
ure. So there is another weak link in Weiner's presumed causal chain of events:
subjects high in n Achievement do not necessarily try harder after failure, even
though they may attribute their failure to lack of effort. However, their evalua-
tion of themselves depends more on how much effort they think they have made
than is true for subjects low in n Achievement (see Figure 12.10). If those high
in n Achievement feel they put out little effort, failing does not lower their self-
esteem, but their self-esteem falls if they think they tried hard. In contrast, for
the subjects low in n Achievement, how much effort they think they put out
makes little difference in the general tendency for them to view themselves more
negatively after failure.
We could presume that self-evaluation after failure would affect the willing-
ness of the subjects to continue with the task if they had a free choice. On the
one hand, subjects high in n Achievement who attribute failure to low effort
Low Effort Attribution
High Effort Attribution
<
c -l
o
Ss -2
I
ed
High n Achievement Low n Achievement

(h Success) (/Failure)
Figure 12.10.
Changes in Self-evaluation of Subjects After Failure as a Function of n Achievement Level and Low
Versus High Effort Attribution (after Heckhausen, 1975a).
might be more willing to continue working at the task. On the other hand, sub-
jects low in n Achievement appear to be just depressed by failure regardless of
the reasons they give for it. At times they are even defensively unrealistic. If
after failure they say that it means they have little ability for the task, they
show a significant gain in the eflfort they intend to put out as compared with
other subjects low in n Achievement who do not attribute the failure to low
ability (Heckhausen, 1975a). This is irrational, because if they think they have
little ability for a task, they ought to try less hard rather than harder. Apparent-
ly, trying hard under such circumstances is reported simply to make them feel
better or to make others think better of them for working hard against such
enormous odds. Or they are showing the greater attraction to very difficult tasks
predicted by Atkinson's model of achieving tendencies (Chapter 7).
In general, causal attributions have more effect on the subsequent behaviors
of subjects high in n Achievement than on those low in n Achievement. Figure
12.11 illustrates the point well (Haiisch & Heckhausen, 1977). Six- to eight-year-
old children built towers as the experimenter induced different expectations of
success. They were scored as high or low in n Achievement using Aronson's
"doodle" test (see Chapter 7). Toward the end of the experiment, subjects differ-
ing in expectations of success were told that they had succeeded or failed. Then
the experimenter measured the time between the subjects' picking up a block
and putting it on the tower to get an estimate of the enthusiasm with which the
child was carrying out the task. Figure 12.11 shows changes in the speed of
placing blocks on the tower as a function of the n Achievement level, expecta-
tion (probability) of success, and success and failure feedback.
The subjects high in n Achievement behaved as one would expect: If the
probability of success was low, actual success led to greater effort, whereas fail-
ure led to less effort. If the probability of success was high, success led to less
effort and failure, to more effort. This is exactly what one would expect from
Atkinson's expectancy-value model of achievement motivation (see Table 7.2).
For subjects high in n Achievement, cognitive information on how well they
were doing fed back and modified their motive arousal for the next task.
For subjects low in n Achievement, however, the Situation was quite differ-
ent. They did not take probability of success into account in relation to success
and failure, as those high in n Achievement did. Instead, failure simply de-
pressed their Performance. So subjects high in n Achievement seem more ratio-
nal: they give explanations for Performance that are more realistic, and they ad-
just their aspirations and efforts according to the feedback they get as to how
well they are doing. Subjects low in n Achievement take such causal attributions
and calculations less into account. When they get failure feedback, they may
react quite unrealistically in terms of the amount of effort they put out and the
reasons they themselves give for failure.
Evaluation of the Cognitive Theory of Motivation

Cognitive understandings certainly modify behavior and influence motive arous-
al, but can it be concluded that a causal attribution pattern defines a motive, as
L Low Probability of Success
§ 9 I I High Probability of Success
10
I
2 n
G
NJ
13
U
Success Failure Success Failure
Feedback Feedback Feedback Feedback
High« Achievement Low n Achievement
Figure 12.11.
Average Change in Speed of Movement to Work on a Task in Reaction to Success and Failure
Feedback as a Function of Low or High Probability of Success and High or Low n Achievement
(after Haiisch & Heckhausen, 1977).
Weiner suggested? Or that the achievement motive is a System for evaluating the
seif, as suggested by Heckhausen (1980; see also Heckhausen & Krug, 1982)?
The redefinition of motives in cognitive terms was part of the general shift in
psychology away from emotional variables. At first sight it makes good sense as
far as the achievement motive is concerned, because it seems to involve a partic-
ular way of people's appraising their own Performance. Those with a high
n Achievement score generally set realistic goals for their Performance (with
moderate probability of success in terms of their own Performance) and tend to
attribute success to ability and failure to lack of efFort. They therefore maintain
a positive attitude toward Performance, which facilitates it. Those low in
n Achievement (high in / Failure), in contrast, tend to have a highly negative
attitude toward Performance; avoid setting Performance goals, if possible; set un-
realistic, very high or very low goals, if they must set goals; avoid evaluation;
and attribute success to luck and failure to lack of ability. Thus, they are caught
in a cycle of defensive behavior that prevents them from doing well and feeling
good about Performance.
So it seems reasonable to conceptualize the achievement motive as a cogni-
tive disposition to judge one's Performance in one of these two ways. If this line
of reasoning is pushed to its logical conclusion, we might be able to dispense
with coding thoughts altogether and measure the strength of the achievement
motive from the patterns of goal setting and causal attributions people follow. In
fact, Heckhausen (1982) adopts such a viewpoint in discussing the cognitions
different students reported while taking a demanding oral examination.
However, the case for the achievement motive's being a cognitive disposition
is not persuasive for a variety of reasons. To begin with, the causes people with
high and low n Achievement assign for success and failure vary considerably,
depending on the circumstances. So it would be hazardous to use causal attribu-
tion or goal-setting patterns as a way of measuring n Achievement strength. For
example, as already noted, Entin and Feather (1982) have found that under
some circumstances even individuals low in n Achievement attribute success and
failure to effort and ability. In the second place, while Weiner (1980a) argues
that the important effects of achievement motivation on behavior can be ex-
plained in attributional terms, this is not always the case. For example, one of
the best-supported findings in the attributional literature is that subjects low in
n Achievement (or high in / Failure) attribute failure to lack of ability. So, ar-
gues Weiner (1980a), this explains why they "quit in the face of failure" (medi-
ated by the belief that failure is caused by a lack of ability, which presumably is
uncontrollable and unchangeable). But subjects low in n Achievement do not al-
ways quit in the face of failure. If they fail at a task with low probability of suc-
cess, they are less likely to quit than subjects high in n Achievement (see Figure
7.6). Heckhausen (1975a) found in the study just reviewed that subjects high in
/ Failure who attributed their failure to a lack of ability were precisely those
who said they intended to try harder next time. Cognitive interpretations of per-
formance vary too much as a function of particular conditions to be considered
a reliable index of motive strength.
Figure 12.12 (from Heckhausen, 1975b) provides another illustration of the
inability of an attribution pattern to explain the effects of the achievement mo-
tive on behavior. As compared with individuals low in n Achievement, those
high in n Achievement prefer working at tasks of moderate probability of suc-
cess and getting feedback at how well they are doing at such tasks (see Chapter
7). This turns out to be the case for the experimental results reported in Figure
12.12, as shown by the right-hand graph. Subjects high in n Achievement prefer
to work on tasks of moderate probability of success. Furthermore, one of the
best-replicated findings of attribution research is that subjects high in n Achieve-
ment attribute success to ability. However, as the graph on the left-hand side of
Figure 12.12 shows, those who consider themselves to have high ability for a
task are less likely to prefer feedback on tasks of moderate difficulty than sub-
jects who consider themselves to have low ability. Thus, we cannot predict one
of the main behaviors explained by achievement motivation theory if we choose
to use people's belief in their own ability as an index of achievement motive
strength.
In any case, as pointed out in Chapter 6, if one wants to argue that a par-
ticular measure (for example, a goal-setting or causal attributional pattern or a
70 70
Low Ability Attribution
High n Achievement (h Success)
60 60 —
A
50 50 - / \
u / \
/ \
'S 40 40
/ \
30 30
' ^ \ \
/ Low n Achievement^*
20
High Ability Attribution (/Failure)
/
10 10 -
_L 1 1 1
Low Middle High Low Middle High
Subjective Probability of Success Subjective Probability of Success
Figure 12.12.
Percentage of Subjects Choosing Performance Feedback on Tasks of Low, Moderate, and High
Probability of Success as a Function of Perceived Degree of Ability Possessed for the Task (Left)
and High and Low n Achievement (Right) (after Heckhausen, 1980, after Heckhausen, 1975b).
self-evaluative System) is tapping a motive, one must demonstrate that persons

high in that measure behave as if they are more motivated—that is, act in a
more aroused way, focus more sensitively on some cues than others, and learn
relevant responses faster. Thus, for example, it would have to be shown that
subjects who generally attribute success to ability or evaluate their Performance
positively would learn certain types of material more rapidly, just like the sub-
jects high in n Achievement do (see Figure 6.13). Such subjects should also be
shown to work harder only when an achievement incentive (a moderate proba-
bility of success) is present. The fact is that cognitive evaluations of Performance
have been studied only in the context of an ongoing experiment or episode, so
little real attention has been given to whether or not they can be taken as evi-
dence of differences in motive strength. So far, the functional criteria for deter-
mining whether a motive disposition is being measured have not been systemati-
cally applied to cognitive indexes of motive strength.
Analytically, the difficulty with the attributional model of motivation is that
it Starts with behavior that is already ongoing and asks the subjects for explana-
tions of it afterward, rather than trying to account for what generated the be-
havior. A Student receives a D on an examination, or a person sees someone fall
in the subway. How the person interprets such events has consequences for what
the person chooses to do next, but it does not explain why the behavior oc-
curred in the first place. Why did the Student take the examination? Why did he
or she care about the grade received? What led to the particular causal explana-
tion given? If the Student were high in n Power, we could readily understand
why a D would be upsetting, because it signifies loss of recognition and respect,
which are important to the Student. Why should a person react with sympathy
when observing an ill person fall in the subway? Again, if the observer were
high in n Affiliation, we would be more likely to expect sympathy than if the
person were low in n Affiliation and cared not much at all about other people.
Still another difficulty with a wholly cognitive theory of motivation lies in
the fact that Nicholls (1978) has demonstrated that causal attributions cannot
really be made by children below the age of five. It may seem unreasonable to
expect them to verbalize such causal explanations, although they might feel
them. In fact, Weiner (1979) attributes causal reasoning to eight-week-old in-
fants, but it seems doubtful that infants, or even young children, have enough
sense of seif for them even to feel attributions like "I should try harder" or "I
did well because I am good at this sort of thing." In fact, Heckhausen (1980)
assigns the origins of the achievement motive to an age (around two and a half
to three and a half years) when a sense of whether one has done well first ap-
pears in a child, because he defines the motive as a self-evaluative disposition.
But it seems unlikely that the achievement motive depends on a sense of seif
that develops at this age in view of evidence on the very early origins of the
achievement motive as reviewed in Chapter 7.
We must also remember that cognitive understandings do not predict
long-term drifts in behavior the way motives do. For example, Chapter 8 re-
viewed a study by McClelland and Boyatzis (1982) that showed that the impe-
rial power motive pattern predicted the tendency to move upward in the man-
agement hierarchy over the long term at AT&T. Yet in the same study, a large
number of cognitive variables, obtained from personality inventories, completely
failed to predict managerial success over time (Bray, Campbell, & Grant, 1974).
For example, Self-confidence and Ascendancy from the Guilford-Martin Inven-
tory of Factors did not predict managerial success, nor did v Achievement,
v Autonomy, v Dominance, v Endurance, nor any of the other eleven scales
on the Edwards Personal Preference Schedule predict long-term success in
management.
It is important to realize that the items that make up these scales refer to
just such attributional patterns Wiener claims should motivate behavior. For ex-
ample, items from the Edwards scale such as "I like to do my best in whatever
I undertake" or "I like to work hard at any job I undertake" endorse the im-
portance of effort, and an item such as "I like to supervise and to direct the ac-
tions of other people whenever I can" refers to belief in one's ability. Yet beliefs
like these, even when aggregated into scales, do not in the long run predict anal-
ogous behaviors. They represent factors that can modify or direct power striv-
ings, but they are not the motives themselves. In the short run in experimental
situations in the laboratory, the effect of such value Statements may be notable,
because the experimenter relies on motives that the subject brings to the Situa-
tion. In life and over the long run, where we cannot assume such motives to
generate spontaneous behavior, the cognitive variables have little predictive
power. Their chief contribution is in steering behavior once it has been gener-
ated by a motive or incentive.
Cognitive understandings, however, play an important role in the
motivation-action sequence. Causal attributions, perceived instrumentalities, and
self-evaluative Systems do generally differ for people high and low in n Achieve-
ment, and these cognitive variables explain why they behave differently in vari-
ous situations—in particular, why people high in / Failure perform in a self-
defeating manner. Such cognitive variables modify what a person chooses to do
next. They enter in to affect excitatory potential or the final impulse to act in a
given Situation. They also lead to emotional changes (positive or negative affect
over Performance), which feed back to alter motive arousal and motive disposi-
tions over time. For example, if people high in n Power get good grades, the
fact that they understand they are good grades and are due to their ability and
effort leads to the emotion of pride in the recognition that they will receive,
which strengthens the incentive character of good grades and perhaps the power
motive itself in the long run. Or if a person high in n Affiliation understands
that the person who has fallen in the subway is ill, that leads to the emotion of
sympathy and the act of helping. The motive comes first, and the cognitive un-
derstandings direct or Channel it and raise or lower it. It is more correct to say
that certain cognitive dispositions characterize a motive than to say they are the
motive.
The study of cognitive variables intervening in the motivation-action se-
quence performed a very important function in clarifying just how motives influ-
ence Performance, as Dweck and Wortman (1982) point out. For example,
achievement motivation theory predicts, according to Atkinson's model (Table
7.2), that as people high in n Achievement succeed at a task, it will become less
interesting to them (it has a high probability of success), and they will set a
higher level of aspiration for that or some other task (with a lower probability
of success). But what goes on in their minds as they make this shift? Research
on cognitive variables suggests that they view their success as due to ability; this
increases their self-confidence, so they choose a more demanding task, perhaps
because they want to continue to get information on how capable they are
(Trope, 1975). Without information on such cognitive variables, there is a gap in
our understanding of how motives get translated into various actions.
Future Orientation and the Achievement Motive

Another cognitive orientation that interacts with the achievement motive has
been extensively studied by Raynor and Entin (1982b). They observed that most
Performances in life are perceived as part of some overall framework and as
Steps on the way to a goal. This is particularly true for students, for whom writ-
ing a successful term paper may be perceived as related to getting a good grade
in a course, which is related to graduation, which is related to getting into grad-
uate or Professional school and going on to a career as a teacher or a lawyer.
Raynor (1968) reasoned that individuals high in n Achievement ought to work
harder at a task that they perceived as important for future success than at a
task of lesser importance. He asked students to rate the importance/helpfulness
of a good grade in an introductory psychology course for future career success
(Raynor, 1968). When he sorted out the grades in the course at the end of the
Semester according to students who were high or low in n Achievement and
who perceived the course as of low or high importance to them, he obtained the
results summarized in Figure 12.13. As expected, the perceived instrumentality
of the course made more of a difFerence to the subjects high in n Achievement.
They got a considerably higher average grade in the course if they thought it
was important than if they thought it was not. The perceived importance of the
course made little difFerence to subjects low in n Achievement.
Raynor then moved on to construct a formal extension of the Atkinson
achievement motivation model to cover contingent paths in which success in an
immediate Step is necessary "to earn the opportunity to move on to the next
step." As Bandura (1982) summed it up in another connection, "self-motivation
is best summoned and sustained by adopting attainable sub-goals that lead to
large future ones. Whereas proximal sub-goals provide immediate incentives and
I I Low Perceived Instrumentality

I—I of the Course
4.00 — I—I High Perceived Instrumentality
I—I of the Course
3.00
2.00
Low High
n Achievement n Achievement
Figure 12.13.
Grades Attained in a Course as a Function of n Achievement and Perceived Instrumentality of the
Course (after Raynor & Entin, 1982b).
guides for action, distal goals are too far removed in time to effectively mobilize
effort or to direct what one does in the here and now." Raynor's model pre-
dicted that success in a contingent path should make more of a difference to
subjects high than low in n Achievement. When subjects were asked to perform
three-step arithmetic problems, results were obtained (Entin & Raynor, 1973)
that confirmed the prediction, as shown in Figure 12.14. If subjects could not
move on to the next Step in the task unless they had reached a criterion in per-
forming an earlier step, those high in n Achievement performed much better
than those low in n Achievement. But if the subjects could move on to the next
step regardless of how they performed on the first step, there was no difference
in the overall Performance of subjects high and low in n Achievement. The cog-
nitive understanding of what Performance means clearly interacts with the mo-
tive to determine Performance.
The contingent path variable also explains some of the confusion over
whether individuals high in n Achievement attribute success more to ability
than subjects low in n Achievement. It turns out that they do, but only in con-
tingent paths (Entin & Feather, 1982). Subjects high in n Achievement seem to
care less about noncontingent paths, and their ability attribution reflects this
fact.
c
o High « Achievement
_>,
5 | 35
-Step Arit hmetic 1
Problems Solved (
^ ^ ^ ^ Moderate n Achievement
O
^~~^^-^^^^Low n Achievement
If -
E | 25
Mean N
<
1 1
Noncontingent Contingent
Kind of Path
Figure 12.14.
Mean Number of Complex Arithmetic Tasks Solved as a Function of n Achievement Level and
Whether Moving On Is Contingent on Success at the Previous Step (after Entin & Raynor, 1973).
The greater importance of contingent paths to subjects high in n Achieve-

ment is also shown by the fact that if on such paths the probability of success
declines from step to Step, they work much harder than those low in n Achieve-
ment. The latter tend to slack off as compared with how hard they would work
in a contingent path with increasing probability of success (Raynor & Harris,
1982). This result can be explained as being due to the fact that since subjects
high in n Achievement believe their success is due more to ability, they will
continue to work harder even as the task gets more and more difficult; as the
task gets more difficult, subjects low in n Achievement tend to slacken their ef-
forts, because they believe they have less ability.
To examine how subjects high and low in n Achievement engaged in future
planning, Pearlson and Raynor (1982) asked students to complete a future plans
questionnaire in which the first step was to write in a future goal they sought.
Then, on successive pages, they were to write in the Steps that would lead to
that goal, the activity that constituted each step, the positive and negative out-
comes that might result from each activity, the chances of succeeding at each
activity, and the importance of the future goal to them. In general, subjects high
in n Achievement were able to define more Steps leading to the future goal.
They had a more differentiated cognitive map of how to get from where they
were to where they were going, and this was particularly true if the goal was of
high importance to them. The importance of the goal, like the contingent path
variable, interacts with the achievement motive to influence what the subjects
do—in this case, the extent to which they have planned ahead (see also Pearl-
son, 1982).
Raynor (1982) has extended his analysis to explain the effects of aging
viewed as progress toward a closed, contingent path, at least so far as those pur-
suing careers are concerned. They Start out life seeing many Steps as leading to
a career goal in the anticipated future. Then, as they move through the Steps
necessary to their achieving their career goal, the probability of success in-
creases—at least to moderate levels—and stimulates achievement motivation.
Then, as they approach retirement, they move from "becoming" to "having
been": there are no immediate next Steps required for attaining a future goal,
and their n Achievement level declines. There is some support for this model
from the two national sample surveys of achievement motivation that were con-
ducted in 1957 and 1976 (Veroff, Atkinson, Feld, & Gurin, 1960; Veroff, Dep-
ner, Kulka, & Douvan, 1980), in the sense that n Achievement levels are signifi-
cantly lower for men and women over the age of sixty-five. However, the data
do not as clearly support the prediction that n Achievement levels should be
highest in the middle years. This was not the case for men or women in the
1976 survey.
The virtue of Raynor's analysis is that it calls attention in a variety of ways
to the importance of the cognitive context in which the achievement motive in-
teracts with Performance. The motive has different effects, depending on what
the Performance means to the subject, particularly in relation to future goals, as
well as what Steps, with what success probabilities, are necessary to get to them.
COGNITIONS AFFECTING THE TRANSLATION OF

MOTIVATION INTO THE IMPULSE TO ACT
As pointed out in diflferent sections of this chapter, cognition affects the

motivation-action sequence in many different ways. Figure 12.15 attempts to
summarize these influences in schematic form. It represents an expanded version
of Figure 6.1, and each of the cognitive influences has been given a number to
make it easier to identify the relationship to be explained.
1. Cognitions influence the extent to which social or physiological cues con-
tact incentives, which link up with motive dispositions leading to motive arousal.
Several empirical demonstrations of this fact were reviewed in the first part of
the chapter. A typical example is provided by the finding that references to
leadership and ability do not link up with achievement incentives for certain
types of women and thus arouse achievement motivation for them, whereas the
same cues do arouse achievement motivation in comparable groups of men. The
term demands implies a cognitive definition of which actions will obtain the
goal implied in the incentive. That is, in mentioning the achievement incentive
of doing well in life, the experimenter also states that a person who wants to do
well in life should do well on these tasks. Normally, a person accepts this defini-
tion of the Situation, but it is possible for a person's achievement motivation to
be aroused by the incentive of doing well in life and for the person to still not
believe that doing well on these tasks is related to that goal. In this case, a cog-
Situation Person Opportunity (Situation)
(2) (3) \
Cues (Social, Motive Arousal x Ps (Probability x K(Values, = Impulse •• Actions ••Cognitive
siologlcal) (Motivation) of Success), Skill Importance to A c r Interpretations
of Goal) Causal
Attributions,
Perceived Ps (on Other Values Future Goals
This Task and Shaping Acts
in General) Chosen
(1) i
Cognitions, (6) | (5) ] (4)

(7)
Demands
Incentives — • Motive Dispositions
f<5>
"Motivation," as Atkinson, Weiner, and others use the term.
Figure 12.15.
Role of Cognitive Variables in the Motivation-Action Sequence.
nitive variable does not affect motive arousal (1) but affects instead the relevance
of probability of success at this task (2) for the achievement incentive. So the
impulse to do well on the task is lowered.
2 and 3. The position taken throughout this book is that an aroused motive
combines with two other types of person variables to produce excitatory poten-
tial or the impulse to act, labeled motivation somewhat confusingly by Atkinson,
Weiner, and others. Initially, Variable 2, the probability of success, was defined
by the skill a person showed as represented by the number of practice trials in
the behaviorist tradition (see Chapter 3). However, in the Atkinson model the
variable got redefined in cognitive terms as the perceived probability of success.
And in its most general form it became self-confidence, social norm ability, or
self-efficacy, as we shall see. Variable 3, the value placed on actions related to
the motive, is also cognitive. Atkinson's (1957) application of expectancy-value
theory to achievement motivation simplified the equation by making the proba-
bility of success (Ps) and the value of success (V) completely dependent on each
other (that is, V = 1 — Ps). This limits the value placed on the activity to diffi-
culty, and whereas there is no question that performing a more difficult task is
more valued by most people, many theorists have pointed out that other values
also influence the impulse to act. We will consider some of these other values.
What it is important to realize here is that both Ps and V are essentially cogni-
tive variables that interact with an aroused motive to produce the impulse
to act.
4. In the model, the impulse to act combines with the opportunities to act
in the environment to yield a response, which is then interpreted by the person
in various ways. These interpretations feed back to influence various aspects of
the motivation-action sequence. If a person ascribes failure to lack of effort, it
may make the importance of striving harder more salient. Effortful striving is an
important value in our culture; according to the model it should therefore influ-
ence the impulse to act, but technically it is not part of the motive aspect of the
action sequence.
5. If an act is perceived as successful, it feeds back to influence the sequence
in two places. At the most direct level it increases the Ps variable, making the
act more likely, but it also influences the incentive value of the activity in ways
described in Chapter 7. If initial perceived probability of success is low, success
increases its incentive value, whereas the reverse is true if perceived probability
of success is high.
6. Causal attributions can affect motive arousal directly. Subjects who
attributed their enduring thirst to their own voluntary choice showed less thirst
than subjects who attributed their Situation to external reasons (Mansson, 1969).
In a different type of study showing the same relationship, Breit (1969) asked
subjects to write essays on how they did (or did not) really control what hap-
pens to them. Based on the findings that subjects high in n Achievement are
more likely to attribute Performance to internal factors like ability and effort, he
reasoned that subjects who thought about having control over things would have
more aroused achievement motivation than those who thought about not having
control over things. The results conformed to this hypothesis: the subjects who
had been thinking about having control behaved more like subjects high in
n Achievement in their liking for occupations of varying diflficulty as compared
to subjects who had been thinking about not being in control. The findings are
not direct confirmation of the relationship, because Breit did not measure
aroused n Achievement in stories written by the subjects and because the treat-
ment could have aflfected the Ps or V variables in the equation. However, the
findings are consistent with the theoretical expectation that cognitive interpreta-
tions can feed back to influence motive arousal directly. Furthermore, Patten
and White (1977) have demonstrated that asking a subject to give reasons for
failure (a cognitive task) seems to feed back and increase the achievement moti-
vation aroused.
7. The interpretation a person gives an act can also feed back and alter the
demands that gave rise to the motivation for Performance in the first place. As a
simple illustration, consider Raynor's example of a person nearing retirement,
which represents the end of a closed, contingent career path. At this point there
are no next Steps to be taken toward the goal, so the social demand or cue value
of the activity giving rise to the achievement incentive declines or disappears.
The research reviewed in this chapter has dealt mostly with cognitive feed-
back effects on motivation of the type just summarized. Another kind of re-
search deals with how variations in the Ps and V factors interact with motiva-
tion to produce diflferent types of Performance. Here the findings are not easy to
summarize, because they come from a broad field in which studies have not
been carried out in a very systematic way. Properly speaking, what is at issue
here is the dynamics of action (Atkinson & Birch, 1970), or the effort to explain
action in terms of all its determinants, only one of which is motivation. The
most systematic recent attempt to build a model of how action occurs has been
carried out by Atkinson and Birch (1978). To review that enterprise is beyond
the scope of this book, but it is worth calling attention to some of the main as-
pects of the perceived probability of success and value that have been found to
interact with motivation to produce behavior.
Self-confidence or Self-efficacy
As the Atkinson model demonstrates, perceived probability of success at a task
interacts with aroused motivation to influence what the subject does. The sub-
ject's perception of his or her likelihood of success in general in all types of
tasks also has an important influence on Performance. This important person
variable has been called a sense of internal control, feeling like an Origin
(deCharms, 1976), mindfulness (Chanowitz & Langer, 1982), self-efficacy
(Bandura, 1982), or a sense of personal responsibility. Whatever it is called, it
has been shown to facilitate Performance under a variety of circumstances.
Some typical findings are presented in Figure 12.16. The research on which
the figure is based involves raising the sense of self-efficacy among phobics by
enactive mastery of progressively more threatening activities (Bandura, 1982). A
typical study involved "severe agoraphobics, whose lives were markedly con-
stricted by profound coping inefficacy that makes common activity seem filled
Cognitive EJfects on Motivation 507
100 -
90 -
80 -
/
70 -
/
60 /
/
50
Intergroup
40
1 1
Low Medium High Low Medium High
Level of Self-efficacy
Figure 12.16.
Mean Performance Attainments as a Function of Differential Levels of Perceived Self-efficacy. (The
left panel shows the Performances of groups of subjects whose self-percepts of efficacy were raised to
either low, medium, or high levels; the right panel shows the Performances of the same subjects at
different levels of self-efficacy (after Bandura, 1982).
with danger. . . . Therapists, who accompanied the agoraphobics into commu-

nity settings, drew on whatever Performance induction aids were required to en-
able their clients to cope successfully with what they dreaded" (Bandura, 1982).
After the treatment, they were able to live successfully through previously
dreaded situations such as traveling by automobile, using an elevator, dining in
a restaurant, or Shopping. At the same time, the perception of self-efficacy in-
creased markedly, and the degree of perceived self-efficacy predicted future suc-
cessful Performance, as Figure 12.16 shows. In fact, the perceived self-efficacy
predicted Performance better than past Performance did, demonstrating that at
least under some circumstances the cognitive aspect of the probability of success
is more important than its skill aspect.
Heckhausen has increasingly in recent years tended to conceptualize the
achievement motive as a self-evaluative System (Heckhausen & Krug, 1982).
Those high in n Achievement think of themselves as capable (having high self-
efficacy or probability of success) and approach tasks with confidence; those low
in n Achievement (or high in / Failure) think of themselves as low in ability or
probability of success and act defensively with respect to tasks. This thesis is en-
tirely consistent with the model of the motivation-action sequence in Figure
12.15, in which the perceived probability of success is an important determinant
of the impulse to act, but the self-evaluative System is not the motive itself, al-
though it may influence and be influenced by the motive.
Langer and Rodin (1976) have demonstrated the importance of this person
variable in another setting. They arranged to have much greater personal re-
sponsibility assigned to residents on one floor of a nursing home for elderly peo-
ple than on another floor where, as is usual in such settings, all decisions were
made for the residents by the staff. It was explained to those receiving the Spe-
cial treatment that they really had the responsibility of caring for themselves
and of making decisions such as what movie to attend, how they wanted their
rooms arranged, how they spent their time, and how the procedures in the
home might be changed to suit their needs better. They were also each given a
plant to care for. Several measures of attitudes and activities of the residents on
both floors were taken one week before the Special treatment and three weeks
afterward. Those who had been given a greater sense of personal responsibility
reported themselves at the later period to be happier and more active than those
living in the normal decision-free environment, and they were also rated to be
more generally improved and alert by observers. Behavioral measures of time
spent visiting with other residents and talking to the staff supported these con-
clusions. In short, giving these elderly people a sense of personal responsibility
or efficacy facilitated actions of all sorts.
What is needed is more investigation of the way in which the sense of self-
confidence interacts with motives to produce behavior. One such study is de-
scribed in the next chapter.
Values
Certain goals are intrinsic to motives. People high in n Achievement want to do
better (which has many cognitive determinants). People high in n Power want
to have impact, which is also defined in increasingly complex ways as people
grow older and more sophisticated. People high in n Affiliation want to be with
people. But there are many other factors that combine with these goals to define
the V variable in the motivation-action sequence. One is simply the importance
of any of these goals, as it is consciously defined in the person's world view. In-
dividuais high in n Achievement are most "turned on" by tasks with moderate
probability of success, because Performance in such situations gives them the
best feedback on their improvement. However, that does not automatically
imply that doing better on such tasks has great importance for them.
As Raynor and Entin (1982b) point out, the importance of the goal of a
motive has no formal Status in the theory of achievement motivation. They had
subjects rate the importance of a future goal and also the probability of success
of attaining it. They found that more important goals were perceived, on the av-
erage, as having .7 Ps, as compared with .5 Ps for less important goals. This
demonstrates the need to distinguish between the goal intrinsic to a motive and
the importance of that goal as seen from an extrinsic point of view. A person
high in n Achievement might conceivably be more attracted to the goal of grad-
uating from law school if the perceived probability of success was moderate
(Ps = .5) if no other extrinsic values were involved. But of course they are in-
volved: graduating from law school may be important for all sorts of other rea-
sons, so the person chooses to go to law school (the action outcome of the
motivation-action sequence) only if the perceived probability of success is
considerably higher than the .5 Ps theoretically optimal for a person high in

n Achievement.
It is even possible that the goal State that gives satisfaction to a person's
motive is considered completely unimportant by that person. People high in
n Affiliation are satisfied by being with people, but in the next chapter we will
consider the case of individuals high in n Affiliation who do not value being
with people but prefer solitude. In this case the motive and value conflict and
produce an interesting compromise action—namely, writing letters to people.
Most of the research in this area has dealt with extrinsic values, which in-
crease the likelihood that achievement motivation will lead to a particular type
of achievement behavior. Heckhausen (1980) has stressed the importance of
what might be called the instrumentality value: if people see the connection be-
tween an outcome or an attribution and an act, they are more likely to perform
the act, although this depends also on achievement motive strength and per-
ceived probability of success (Kleinbeck & Schmidt, 1979). Such a result might
be explained on the grounds that people have learned that it is valuable to know
what leads to what. A feeling of personal control, of being able to do something
that leads to a result, is an analogous variable. It is highly valued and facilitates
Performance, as the nursing home experiment demonstrates. Perlmuter, Scharff,
Karsh, and Monty (1980) have even shown that the words people choose to
learn are learned faster than words they are simply given to learn.
Value is also placed on effort in our society, so when effort becomes salient,
Performance improves (see Figure 12.8). In fact, all the causal attribution litera-
ture can be reconceptualized in terms of value theory. The dimensions of per-
ceived causation—internal versus external locus of control, controllability, and
stability—can be thought of as values (or at least Schemata ordering the way
things work) that influence Performance. In general, the values of internality,
controllability, and stability promote successful Performance, whereas their op-
posites diminish it.
Many investigators have been critical of the achieving style, which interpre-
tations of the achievement motive research have suggested is intrinsic to the mo-
tive itself. It is sometimes assumed that people high in n Achievement energeti-
cally pursue their own goals without respect to others, in a kind of caricature of
the upward-aspiring, high-achieving Western male. Such an image is incorrect
on many counts. At the simplest level it confuses achievement behavior, or the
action Output variable in Figure 12.15, with a particular motivational input. Ob-
viously, from the model of the motivation-action sequence and from findings re-
ported in earlier chapters, not all achievement behavior (or successful perfor-
mance) is the result of achievement motivation. As Gallimore (1981) has
carefully documented, the n Affiliation score among Hawaiians is more closely
associated with many types of achievement behavior than the n Achievement
score is. Hawaiians value interpersonal relations and work hard primarily when
they are working together on a project. Thus, obviously a value placed on coop-
erative work in combination with an achievement motive ought to lead to differ-
ent behaviors than if the value were placed on individual accomplishment.
The same point has been made in connection with sex differences by
Lipman-Blumen, Leavitt, Patterson, Bies, and Handley-Isaksen (1980) and Par-

sons and Goff (1980). Women in our society tend to value relationships more
than men do, so for them, achievement motivation tends to be associated with a
relational achieving style, whereas for men it is associated more with a direct
achieving style. Thus, women are more likely to achieve in a vicarious way or in
a contributory or collaborative relationship than men are (Lipman-Blumen et
al., 1980). Parsons and Goff (1980) describe much the same difference in terms
of Bakan's model of agency (which characterizes the direct achieving style of
men) and communion (which characterizes the collaborative style of women).
Both groups of researchers have developed value questionnaires to determine the
extent to which individuals subscribe to direct or collaborative orientations in
the achievement area. This is obviously a Step in the right direction in accor-
dance with the motivation-action sequence as outlined in Figure 12.15, for it is
only as motives are studied in combination with other determinants of action
like self-confidence and values that psychologists will begin to account for the
various ways in which people behave.
Research in this area has not advanced very far, because people have con-
fused motives and values and not treated them as separate determinants of ac-
tion. Many more studies are needed that examine how important values interact
with motives to produce action—in particular, studies of how they interact with
the affiliative and power motives, which have tended to be neglected in favor of
an almost exclusive concern with achievement behavior. What is also needed is
a paradigmatic study of just how the three major determinants of action (mo-
tives, probability of success, and values) interact to produce Performance. That
is the topic of the next chapter.
NOTES AND QUERIES
1. Different incentives arouse achievement motivation in different groups

of people: References to money rewards arouse more n Achievement in
working-class than middle-class people (Douvan, 1956); references to social
acceptability arouse more n Achievement in some groups of women than
men. Does this mean that achievement motivation is solely determined by
people's understanding of a Situation? For example, it has been argued that
it is wrong to conclude that lower-class blacks have lower average
n Achievement than middle-class blacks or whites (B. C. Rosen, 1956), be-
cause lower-class blacks are just as achievement oriented, but their achieve-
ment motivation is aroused by different incentives. Evaluate this argument.
Be sure to distinguish between motive arousal and dispositional level of
n Achievement. Think about a person whose achievement motivation could
be aroused only by one peculiar Situation. Would you expect this person's
dispositional level of n Achievement to be high or low?
2. An extreme version of the Schachter and Singer or James-Lange theory
would seem to argue that moods associated with physiological arousal are
solely determined by the meaning people attribute to the arousal. Suppose

you had a headache. Can you imagine an understanding of the Situation
that would lead you to feel really happy about the headache? What limits
do you think may exist in our ability to feel what our understanding teils us
we should feel?
3. Design a cognitive dissonance experiment that would lower the level of
aroused achievement or power motivation the way Mansson's experiment
lowered the level of thirst.
4. What kind of cognitive dissonance, if continued over time, might lower dis-
positional n Achievement? (If necessary, consult Andrews, 1967).
5. Try explaining Shrable and Moulton's (1968) finding that picture cues relat-
ing to an individual's competencies have differential effects for arousing
achievement motivation for individuals of higher and lower intelligence. In
other words, could the results have anything to do with perceived probabil-
ity of success?
6. So far no one has studied the causal attributions of individuals high and low
in n Power or n Affiliation. Make some predictions as to how motive
strength in these areas might affect causal explanations for success or failure
in gaining power or being loved. Would different types of causes be involved
than those important in the achievement area?
7. Think of a successful experience you have had recently. Write out the rea-
sons why you think you were successful. Try classifying these reasons ac-
cording to the three dimensions discovered in previous research. Do the
same for a recent failure experience. Are the reasons you gave related to
your motive profile in ways that would be expected from previous research?
8. One can easily imagine how people high in n Achievement would develop a
positive proactive image of themselves with respect to Performance from ob-
serving how they behave over time. Similarly, people low in n Achievement
develop a defensive self-image and style where Performance is concerned
from observing how they behave. In these instances the self-picture and mo-
tive syndromes are consistent and logically related, so it may look as if one
is equivalent to the other. But sometimes motive Systems and self-images are
quite inconsistent with each other. Under what circumstances is such a State
of affairs likely to occur? How would you know it existed, and what conclu-
sions would you draw about the person?
9. Do you believe the contingent path model would apply to individuals vary-
ing in the strength of motives like n Power? Why or why not? Design an
experiment to test the hypothesis that subjects high in n Power work harder
to attain a goal when attaining it opens up the opportunity to attain a more
important goal.
10. Construct a theoretical explanation for why contingent paths might not af-
fect the behavior of individuals high in intimacy motivation.
11. If n Achievement is more related to time cognitions—that is, the interre-
latedness of acts over time—can you think of reasons why n Power might
be related to space cognitions (to how they and others are placed in space)?
Design an experiment to test this hypothesis.
12. Figure 12.15 has been constructed in a way that shows that relatively stable
motive dispositions in a person are an important determinant of the
motivation-action sequence. But it is possible to argue that such motive dis-
positions either do not exist or can be ignored. Such an approach would ex-
plain individual differences in aroused motivation solely in terms of cogni-
tive variables like probability of success, values, understandings of what
leads to what, and incentives. Try leaving motive dispositions out of the
model. Do you run into any difficulties in explaining certain types of behav-
ior when you do so?
13. Take each of the relationships labeled 1 through 7 in Figure 12.15 and try
to find an experimental result showing the existence of such a relationship
for the power motive or the affiliative motives. Do you run into any particu-
lar difficulties?
14. Heckhausen's value of instrumentality appears to be related to Raynor's
conception of contingent paths. Both suggest that perceived means-end rela-
tionships are an important determinant of action. Yet when people are
asked why they succeeded or failed, they seldom reply that it was because
doing well on the task was (or was not) important for something eise they
wanted. Why not? Does this throw doubt on the causal attribution method
of arriving at values (what is important to people) or on the importance of
the value of instrumentality?
15. Heider argued strongly for the importance of cognition in determining moti-
vation. Consider, for example, the following Statements:
If you want to persuade an individual (p) he should attack o (an "other" person),
try to convince him that o harmed him, that o is much better off, etc. O must be
portrayed as the unjustified aggressor, the frustrator, etc., and that he is determined
to härm p. It is a matter of beliefs; if you control beliefs, then you control emotions.
One Information, one change in belief, affects motivation, affect, etc., and these di-
rect effects have further effects on action. (Heider, cited in Benesh & Weiner, 1982)
In terms of the Schema presented in this chapter for the determination of

action, in what sense can it be said that beliefs control emotion or determine
motivation? Are beliefs the only determinants of the impulse to attack? If
so, how could one explain the greater tendency of the United States to at-
tack at some times in its history than at other times (see Chapter 11)? If it
is a question of the extent to which one nation believes another nation de-
serves to be attacked, for the reasons Heider gives, why is a nation more
likely to believe such cognitive information at one point in its history than
another?
16. Heider is also critical of the way in which psychologists like Murray think
in terms of aggressive or power motives, because they leave out the cogni-
tive elements involved in choosing a target or goal for the motive: "Othello
does not want just to kill anybody; he does not have a general need to kill,
only the need to kill Desdemona" (Heider, cited in Benesh & Weiner, 1982)
because of a lot of beliefs he holds about her. How is Heider's legitimate
concern for what determines the specific goals of a motive in a given Situa-
tion handled by the Schema for the multiple determination of action pre-
sented in this chapter? Make a model based on Figure 12.15 that shows
some of the key determinants in Othello's impulse to kill Desdemona.
13
How Motives Interact
with Values and Skills to
Determine What People Do
Drives, Habit Strength, and Incentives as Determinants of Response Strength

Motives, Expectancies of Success, and Values as Determinants of
Performance
Habit Strength as Expectancy of Success
The Achievement Incentive and Other Incentives Affecting Achievement
The Distinction Between Motives and Values
How the Achievement Motive, Skill, and Achievement Values Affect

Performance
How Motives, Skills, and Values Jointly Determine Success as a Naval
Officer
Factors Influencing Affiliative Acts and Choices
How the Situation Combines with the Affiliative Motive and Affiliative Values to
Predict Affiliative Choices
How Values Affect the Way the Affiliative Motive Expresses Itself
The Joint Effect of the Affiliative Motive, Social Skill, and Affiliative Values on
Affiliative Acts and Choices
How the Determinants of Affiliative Acts and Choices Interact with Each Other to
Produce Their Eflfects on Behavior
• The Distinction Between Motives and Intents
514
E /ver since psychologists observed that motivated people or animals learn
faster, they have been interested in how motives combine with other variables to
increase the probability that a response will occur. In more general terms the
question is, What factors in what combinations will best predict what response
will be made, or if made, how often and how strongly it will be made?
All psychologists except a few association theorists like Guthrie (1935) as-
sume that motives, rewards, or reinforcers are one of the determinants of re-
sponse strength, and that there are also several other determinants that need to
be taken into account. To begin with, the environmental Situation is obviously
important. A hungry rat will run faster through a maze or learn the correct
turns more quickly than a satiated rat, but only if there is food in the goal box,
and only if the rat can get into the maze. In other words, if one is interested in
predicting the strength of the maze-running response, it is helpful not only to
know how hungry the rat is, but also that the rat has access to the maze and
that there is food available at the end of it. Response strength is jointly deter-
mined by a motivational variable in the organism and certain environmental
variables.
In similar fashion it has been shown that individuals high in n Achievement
will perform better than those low in n Achievement, but only if the environ-
ment provides them with a moderately difficult task incentive (see Clark &
McClelland, 1956; French, 1955). If the task is too easy or too difficult, they do
not work harder than those with low n Achievement. The environment has to
provide them with an opportunity to do better in order for the motive to in-
crease response strength. As noted in Chapter 8, McKeachie (1961) reported
that men high in n Power will work harder (get better grades) than men low in
n Power in classes in which there are many opportunities to speak, as contrasted
with classes in which there is little opportunity to speak. An environmental
variable—the opportunity the instructor provides for speaking in class—
combines with a motive disposition to predict Performance. Several personality
theorists, notably Endler (1981) and Magnusson (1976), have stressed the impor-
tance of taking into account interactions between the person and the environ-
ment in explaining behavior.
• DRIVES, HABIT STRENGTH, AND INCENTIVES AS

DETERMINANTS OF RESPONSE STRENGTH
Chapter 3 summarized the attempt made by Clark Hüll (1943) and others (for
example, Spence, 1956) to predict response strength or Performance from two
variables in the organism, ignoring the environmental contribution for the mo-
ment. One determinant was motivation or drive. The other was habit strength.
Hull's formula was SER = D X SHR- I n other words, excitatory potential, or
the tendency to make a response (SER), is a function of drive strength (Z>) mul-
tiplied by habit strength (SHR). What Hüll meant by habit strength was the
amount of reinforced practice the animal had had in making the response previ-
ously, or in more general terms, the skill the animal had acquired in making the
response.
Obviously, both determinants of the tendency to make a response are impor-
515
tant, and Hüll argued—partly for empirical and partly for theoretical reasons—
that multiplying the two determinants gave the best prediction of response
strength. Empirically the two curves draw apart as practice or habit strength in-
creases, indicating a multiplicative relationship. Theoretically it seemed reason-
able to assume that if either drive or habit were reduced to zero, there would be
no response, which is the prediction the equation yields if the two variables are
multiplied by each other. In commonsense terms, people might be quite skilled
at playing the piano, but if they do not want to, they will not play. And no
matter how much they want to play, if they do not have the skill to play, they
also will not play.
As noted in Chapter 3, later Hüll (1952) added a third variable to his equa-
tion to take into account the effects of incentive value on Performance. Rats
run a maze faster for food that they like (for example, bran mash) than for
other foods, like sunflower seeds, that they do not particularly like. Or they
run faster for sixteen pellets of food in the goal box than for one pellet. So
the formula was expanded, again multiplicatively, to read as follows: SER —
D X SHRX K.
Spence (1956) agreed that incentive value (K) was important, but he feit it
should not be combined multiplicatively with the other determinants but instead
should be added to drive strength and the sum of these two multiplied by habit
strength. In his formulation, the presence of either drive or incentive would lead
to some behavior if habit strength existed, whereas in HulPs formulation, if ei-
ther incentive or drive were zero, there would be no tendency to act.
The intricacies of just what was meant by these terms and how various em-
pirical studies can be interpreted using the alternative formulas need not concern
us here. What is important is the identification of the types of variables that
need to be taken into account to explain response strength and their application
to human Performance. McClelland (1951), for example, applied the same set of
three variables to explain complex human behavior by arguing at the level of
personality structure that motive strength is equivalent to drive strength; traits
or skills, to habits; and Schemata or values, to incentives. This general model
has been used throughout this book in trying to explain how motives combine
with other variables to produce the impulse to act. (See Figure 6.1 and Fig-
ure 12.15.)
MOTIVES, EXPECTANCIES OF SUCCESS, AND VALUES

AS DETERMINANTS OF PERFORMANCE
As Atkinson (1964) and Weiner (1980a) pointed out, and as we noted in Chap-
ter 1, Lewin's model of motivated behavior identified three variables very similar
to those used by behaviorists like Hüll and Spence, although Lewin defined
them in phenomenological or cognitive terms. Thus, he spoke of need or ten-
sion, which is equivalent to drive strength; properties of the goal object, which
is equivalent to incentive value; and psychological distance, which is a kind of
phenomenological equivalent of the skill or habit variable in terms of how easy
or how hard it is to get from the need to the goal.
How Motives Interact with Values and Skills 517
These ideas were formalized first by Atkinson (1957) and later by Atkinson
and Birch (1978) into the best-known contemporary theory of the determinants
of action. The formula arrived at was very similar to HulFs except that the vari-
ables in it were defined in cognitive terms and operationalized in measures ob-
tained from human subjects rather than animals. As reviewed in Chapter 7, the
initial formula read as follows: Ts = Ms X Ps X Is. That is, the tendency to
achieve success (Ts) is a multiplicative function of the motive to achieve success
(Ms), expectancy or probability of success (Ps), and the incentive value of suc-
cess (Is). It was further assumed that the incentive value of success could be de-
fined as 1 — P5, meaning that the more difficult the task (the less the probabil-
ity of succeeding at it), the greater the reward value would be of succeeding at
it. As we noted in Chapter 7, the multiplicative model worked fairly well in ac-
counting for the general tendency of subjects to choose to work at tasks of mod-
erate difficulty, since P5 X (1 — Ps) is greatest when Ps = .50.
Later, Atkinson and Birch (1978) shifted from what they called the tradi-
tional episodic view of behavior to a stream of behavior paradigm in which the
central problem became not how an act is initiated, but how one act replaces
another. That is, they argued that the organism is always doing something, and
the problem is to explain how it happens to shift from doing one thing to doing
another thing. They identify a number of factors that are responsible for
changes in activity, such as instigating force (F), which is very like the old con-
cept of drive; consummatory force (the weakening of a tendency by expressing
it); and inhibitory force. Without going into the details of how the variables in
their new model are defined and relate to each other in determining action, it is
clear that they emphasize at least one variable that had not previously been part
of modeis of the determinants of action. For reconceptualizing the motivational
problem in terms of shifts in the stream of behavior calls attention to the great
importance of other instigating forces in the Situation. Thus, Atkinson (1980) ar-
gues that it is important not only to know the strength of n Achievement in a
Situation, but the strength of other motives as well. In its simplest form, this
equation can be stated as follows:
M
Percentage of time spent in Activity A =
Ma + Mb + . . . + Mn
That is, the tendency to spend time thinking about achieving is a function of the
ratio of the motive to achieve (Ma) to all motives (Ma + Mb + . . . + Mn)
present in the Situation. Thus, Atkinson (1979) points out the following:
. . . [T]wo individuals who differ greatly in absolute strength of achievement motive

(the numerator) might, nevertheless, obtain the same thematic apperceptive
n Achievement score if they also differ in a certain way in number and/or strength of
competing motives (the denominator). For example, the ratio of
is equivalent to . Both equal 1/3 and imply the same per-

M F P
1 + 1+ 1 3 + 3 + 3
centage time thinking about achievement, yet the strengths of the achievement mo-
tive are 1 and 3.
Since Atkinson and Birch's new model of the dynamics of action empha-
sizes shifts in the stream of behavior, it also focuses attention more on the per-
centage of time spent on various activities, as in these equations, rather than on
the initiation and acquisition of responses, choice, latency, resistance to extinc-
tion, or persistence, which are the more traditional indicators of response
strength. In this respect they follow Skinner (1966), who maintains that the
probability of an operant response's occurring represents what is commonly
called purpose. That is, the frequency with which a rat presses a bar to get food
when no Stimulus that elicits the response is easily identifiable could be seen as
representing the strength of the rat's purpose in seeking food as opposed to
doing other things (scratching itself, sniffing, and so on).
The Atkinson-Birch model developed as a highly formalized attempt to ex-
press the ideas in what has come to be called loosely expectancy-value theory,
defined as follows:
A general class of theories that relate tendencies either to perform particular actions
or not to perform them to the strength of expectancies that these actions will lead to
specific outcomes and the valences or subjectively perceived values of those outcomes
for the person. . . . The expectancies and valences are usually assumed to combine
multiplicatively to determine the strength of tendencies. (Feather, 1975)
Habit Strength as Expectancy of Success

Atkinson translated the expectancy term in the equation to predict response
strength into probability of success, but probability of success has come to have
two somewhat different meanings. On the one hand, it is a function of the avail-
ability of the response: the probability of success is lower if a person cannot
make a response very well. In this sense it has the same meaning as the habit
strength or skill variable in Hull's (1943) or McClelland's (1951) equations. The
importance of the skill variable in predicting response strength subsequently has
tended to be ignored in much motivational research either by using very over-
learned responses, such as canceling e's and o's in a series of letters, or by sub-
tracting across successive trials to "correct" for individual differences in skill.
Thus, in a study that will be discussed at length later in the chapter, Patten and
White (1977) deal entirely with improvements in Performance from one trial to
the next. However, improvements get greater and greater the more practice the
subject has with a task, indicating that the subject is getting more skillful, as
well as that the better the Performance on one trial, the better it is likely to be
on the next. Skill itself—even in routine tasks—tends to be a product of motives
and expectancies plus practice, but no one doubts that skill is a major determi-
nant of the probability of success and hence contributes in a major way to pre-
dicting response strength.
On the other hand, perceived probability of success as a cognitive variable is
different. It is determined in a major way by beliefs about the efficacy of making
a response, which may be somewhat independent of a person's skill in making
it. As pointed out in Chapter 12, two types of such beliefs have been studied ex-
tensively. One type has to do with the efficacy of effort in bringing about a con-
sequence through a particular response in a given Situation. The other type has
to do with the generalized confidence people have that they can bring about out-
comes through activities of any kind. We have reviewed elsewhere (see Chapter
12) the evidence that a belief in the importance of effort facilitates Performance.
For example, W. U. Meyer (1973; see also Weiner, 1980a) found that those who
attributed their failure to lack of effort tended to show more improvement on
the next trial. And Dweck (1975) and Chapin and Dyck (1976) reported that
children with learning difficulties who were trained to believe that their failures
were due to lack of effort tended to do better subsequently. We can conceptual-
ize these studies as meaning that subjective probability of success was increased
through the expectation that greater effort would be more likely to produce a
desired outcome.
As also reviewed in Chapter 12, the importance for predicting response
strength of generalized beliefs about personal efficacy has been demonstrated in
a variety of studies. Perlmuter, Scharff, Karsh, and Monty (1980) have summa-
rized evidence showing that giving subjects a choice of what to do improves
their Performance over what it would be if they were made to do it. They argue
that giving subjects a choice gives them a perception of control over the Situa-
tion, which enhances Performance. Seligman (1975) has summarized a number
of studies showing that animals and human beings can "learn helplessness." If
they are put in a Situation where they are repeatedly exposed to inescapable
shock, they will learn to give up trying to do anything about it and will fail to
try to escape in a new Situation in which they actually could avoid the shock. In
other words, their expectancy of the probability of success from taking action
has been lowered in general, and a measure of their sense of helplessness should
predict Performance in a variety of situations.
In contrast, deCharms (1976) demonstrated that teaching junior high school
students to act like Origins—to believe that they could personally cause things
to happen—significantly improved their grade placement, test scores, and the
Proportion of boys receiving such training who graduated from high school (see
Chapter 14). And Bandura's (1982) self-efficacy training, reviewed in Chapter
12, is similarly effective in helping people get over phobias. Just as Seligman's
subjects learned that there was little or no relationship between doing something
and success, deCharms's and Bandura's subjects learned that there was a con-
nection between the two. In all these instances, perceived probability of suc-
cess—whether high or low—-is strongly related to subsequent response strength
or Performance.
The Achievement Incentive and Other Incentives Affecting Achievement

A major misunderstanding has occurred in the discussions of achievement moti-
vation theory because of the failure to distinguish between the incentives specific
to the motive and allied incentives or other values affecting the valence of suc-
cess. The misunderstanding has arisen in part because of Atkinson's terminolo-
gy. He defines the incentive value of success (Is) as 1 — Ps. But there are obvi-
ously many factors other than task difficulty that determine the incentive value
of success broadly conceived. He was writing within the context of achievement
motivation studies and was trying to explain the preference of subjects high in
n Achievement for tasks of moderate difficulty, but he used terminology that has
been taken to have a more general significance.
Maehr (1974) has been particularly vigorous in criticizing achievement moti-
vation theory for not taking into account cultural variations in the meaning of
achievement and success. He and his colleagues believe they have shown that
"the traditional conception of achievement motivation results in an ethnocentric
bias" (Maehr & Kleiber, 1981), because it has become associated with individu-
alistic, rather than cooperative, achievement; with the Protestant work ethic;
and generally with upward occupational striving (see also Duda, 1980).
But strictly speaking, the incentive for the achievement motive relates to
such outcomes—if at all—only indirectly and in combination with other values.
It is quite specific and limited and can be defined in general terms as the plea-
sure derived from doing something better. Many studies have demonstrated that
if a task is too easy or too difficult, the incentive is not present and subjects
with high n Achievement do not perform better (see Chapter 7). If something
like getting time off from work is the incentive for successful Performance, they
also do not do better than subjects low in n Achievement (French, 1955). Obvi-
ously, whether the pleasure derived from performing better gets connected with
upward occupational striving or individualistic versus collaborative achievement
depends on other values to which the person is exposed. As noted in Chapter
12, it is quite proper to ask what kinds of achievement are valued by older peo-
ple, by people from other cultures (Gallimore, Boggs, & Jordan, 1974), or by
women so long as it is not assumed that these values directly alter the incentive
for the achievement motive, which remains doing something better. What a per-
son decides to do better will obviously vary, depending on other values In moti-
vational expectancy-value theory, values are independent determinants of re-
sponse strength or the tendency to act.
Weiner (1980a) and others have created further terminological confusion by
referring to measures of achievement values as if they were measures of achieve-
ment motives. A generation ago deCharms, Morrison, Reitman, and McClelland
(1955) established that valuing achievement (v Achievement) as measured in an
attitude questionnaire not only does not relate to the n Achievement score as
obtained from coding thought samples, but also has quite different behavioral
correlates. Nevertheless, as noted in Chapter 6, ever since that time, a number
of investigators have tried to develop "objective" measures of n Achievement
through attitude questionnaires, although none of them has correlated consis-
tently with the n Achievement measure or predicts better Performance in a
moderate risk Situation when the achievement incentive is present.
The most populär and best validated of these v Achievement measures is
one developed by Mehrabian (1969). Unfortunately, Weiner (1980a) and others
refer to scores obtained from this instrument as measures of n Achievement,
with the result that the interpretation of much empirical research in this area in
recent years is confused and misleading. The disagreement here is not simply a
matter of arguing over which measuring instrument is better. Rather, using the
Mehrabian instrument as a measure of motives blurs the distinction between
motives and values, which leads to just the kind of objections Maehr and others
have made to achievement motive theory as it has been developed by some peo-
ple. Motives appear to be present and function in much the same way in men
and women (Stewart & ehester, 1982) and in various ethnic groups, but since
the extent to which these groups subscribe to achievement values differs, it may
well appear that only the groups that subscribe to achievement values have any
achievement motivation—which is patently untrue. For example, certain groups
of women do not value strongly the typical achievement values of agentic striv-
ing, but there is no evidence they are lower in n Achievement.
The Distinction Between Motives and Values

At several points throughout the book we have pointed to differences between
motives and values. Values are normally measured by asking people to indicate
what is important to them. In the measure of values developed by Rokeach
(1973), subjeets are simply asked to rank order the importance to them of termi-
nal values represented by words like freedom, equality, and true friendship or in-
strumental values like independent, intellectual, or ambitious (see Chapter 5).
There is a strong cognitive, self-conscious component in these judgments. Sub-
jeets know what they are doing. They are reporting what is important to them.
In contrast, motives are measured from coding operant thought content; subjeets
do not know what aspect of their thinking is of interest, so they normally can-
not consciously shape their thoughts to score high on various motives.
The correlation between these two types of measures is quite low. Beyond
this, Rokeach (1973) has reported that while some of his value rankings corre-
late with motive scores in a meaningful way, others do not. For example,
n Achievement scores correlated significantly with the value ranking for inde-
pendent and intellectual but not with sense of aecomplishment, ambitious, or ca-
pable. This is typical of results from a number of studies showing that value at-
titudes do not correlate with motive scores obtained from coding spontaneous
thought (see Child, Frank, & Storm, 1956). The most reasonable interpretation
of such findings is that these two types of measures are essentially independent,
as they ought to be on theoretical grounds, and that when occasional correla-
tions appear between them, they are the produet of a peculiar set of circum-
stances related to the particular group being tested.
In contrast, both motives and values can influence choiees or the valence of
one outcome over another, as Feather (1975) has pointed out: "Like motives,
values can influence valences. Money, for example, may be a potent incentive for
people who assign a very high priority to a comfortable life in their value Sys-
tems" (p. 302). And under certain circumstances, money can also have high va-
lence for those high in n Achievement—when it can be used as a measure of
how well the person is doing (see Chapter 7).
Motives and values are also both associated with afFect. A person who val-
ues freedom is happy when given freedom and unhappy when it is taken away.
Similarly, subjeets with high n Achievement feel good when thinking about
achievement goals and feel bad when thinking about failure (McClelland, Atkin-
son, Clark, & Lowell, 1953).
There appears to be a difference about when the affect occurs. Zajonc (1980)
has summarized evidence indicating that the affective response to a Stimulus
often occurs before the Stimulus is recognized. Before subjects understand con-
sciously what Stimulus is being presented, they react to it with positive or nega-
tive affect (see Chapter 2). The affect associated with motives appears to be of
this kind, since motives operate often without clear cognitive understanding of
what is going on. For example, as noted in Chapter 8, subjects high in n Power
react more sensitively to power cues—when sensitivity is measured in terms of
electrical responsivity of the brain—within a quarter of a second after presenta-
tion of the cue and before one might presume that the person had recognized or
understood the meaning of the Stimulus (Davidson, Saron, & McClelland, 1980).
In other words, subjects high in n Power are reacting very quickly, automatical-
ly, and strongly to power cues as contrasted with other cues, and as contrasted
with subjects low in n Power. In contrast, affect associated with values comes
after understanding of the Situation, since values are made up of such under-
standings. A man is angry after he understands his freedom has been taken
away.
Values are much more affected than motives by social norms and by societal
and institutional demands (Feather, 1975; Rokeach, 1973). Thus, values clearly
affect choices, serving to direct the energies of a motive into one Channel or an-
other. It is at this point that the complaints of Maehr (1974) and others about
the limitations of traditional achievement motivation theory make sense. The
point can be made very simply from a study carried out by French and Lesser
(1964) in which they discovered that the energies of women high in n Achieve-
ment went into different activities, depending on their values. The data are sum-
marized in Table 13.1.
If the women were intellectually oriented, those high in n Achievement did
better at an intellectual task, but not at a task involving social skills. In con-
trast, among women who valued the traditional female role, those high in
n Achievement tended to perform better at a social skill than those low in
n Achievement. Both groups of women high in n Achievement did better at
something, but what they did better at depended on their values. Note also that
what is involved in this experiment is not v Achievement—that is, the extent to
which the women valued achievement consciously—but quite a different set of
values related to what women should be doing. Thus, values quite outside the
achievement complex can influence the valence of various activities, shunting
motivational energies in one direction or another.
HOW THE ACHIEVEMENT MOTIVE, SKILL, AND

ACHIEVEMENT VALUES AFFECT PERFORMANCE
While it is clear that values affect choices and shunt motivational energies in
one direction or another, it is not yet clear that they energize behavior and lead
Table 13.1.
RELATIONSHIP OF n ACHIEVEMENT SCORES TO TWO TYPES OF
PERFORMANCE DEPENDING ON WOMEN'S VALUES (after French & Lesser,
1964)
Correlations of n Achievement with

Intellectual Social
Value Orientation of Women Skill"
Traditional .04 .59*
Career oriented .55* .32
Performance on an anagrams test.

^Performance on a number of questions such as, If you were a stranger in town, how would you make
friends?
*p < .05.
to faster learning of related activities the way motives do. The best evidence on
this point comes from a number of studies that have shown that subjects high in
v Achievement measured in various ways do not ordinarily perform better than
subjects low in v Achievement. Since the point is an important one, it is worth
examining in some detail through a recent study. Patten and White (1977) de-
signed an experiment patterned after an earlier one conducted in Germany by
W. U. Meyer (1973). Subjects were asked to complete four rows of digit-symbol
Substitution in one minute. On the second trial, the subjects in the failure condi-
tion were interrupted on the fourth row because time was up so that they failed
to complete the task in the time allotted. Then they were asked whether they
had failed because of lack of effort, lack of ability, luck, or task difficulty. The
measure of response strength was the improvement in time taken to complete
the first three rows of digit-symbol Substitution on the next trial. As a control,
some subjects were run through the experiment without experiencing failure. Al-
though Meyer had originally used a measure of n Achievement obtained in the
usual way, Patten and White (1977) switched to a measure of v Achievement
obtained from the Mehrabian questionnaire, although they refer to it as a mea-
sure of achievement motivation. They also ran some subjects under ego-involved
conditions (without failure) when the achievement motive has been shown to be
aroused (McClelland et al., 1953).
Some of the main results of this carefully designed experiment are shown in
Figure 13.1. The improvement score represents a decrease in the number of sec-
onds from the first trial to later trials. Note first that when the achievement mo-
tive had been aroused experimentally by failure, subjects gained significantly
more in the second block of trials than they did in the first block of trials. No
such improvement occurred for subjects in the neutral, no-failure condition. Fur-
thermore, W. U. Meyer (1973) had shown that for the same task, subjects who
20
n Achievement Aroused Condition

16 (Reasons Requested for Failure)
c
~ 12
E Neutral Control Condition

High v Achievement
Low v Achievement
Trial Blocks (Representing Increased Practice)
Figure 13.1.
Effect of Aroused n Achievement and Differences in v Achievement on Group Mean Block
Improvement Scores (after Patten & White, 1977).
scored higher in n Achievement, if success oriented, behaved like those in whom

the achievement motive had been experimentally aroused: they gained more
from the first to the second block of trials than those low in n Achievement
(Heckhausen, 1980). This is a crucial point, because it shows that the effect of
increased motive strength is the same whether it is the result of individual dif-
ferences or situational manipulations. To put it a different way, those who score
high in n Achievement behave chronically like those whose achievement motive
has been situationally aroused.
Technically, this confirms the following relationship in terms of HulPs equa-
tion or Atkinson's reformulation of it:
Hüll: SER = D X SHR.
Atkinson: Ts = Ms X Ps.
Freely translated, the tendency to work hard on the task (or Performance) is a
function of motive strength (represented by the n Achievement score) times skill
(represented by the number of practice trials).
Note that a similar relationship does not hold if the v Achievement measure
obtained from the Mehrabian questionnaire is used to estimate differences in
n Achievement levels. This is one more confirmation of many similar results
showing that the Mehrabian measure cannot be considered a measure of the
achievement motive.
Let us assume instead that it is a measure of the incentive value of success
(or Is, in Atkinson's terms). Then we can write the following: Ts (or SER ) =£ Is
(or K) X Ps (or SHR). Freely translated, the tendency to work hard on the task
is not a function of the incentive value of success (represented by the v Achieve-
ment measure) times skill (represented by the number of practice trials). There
is no greater improvement with practice for those high in v Achievement than
for those low in v Achievement.
The failure of value attitude measures to predict Performance the way mo-
tives do exists not only for laboratory tasks like this one, but for outcomes in
real life. For example, McClelland and Boyatzis (1982) have reported that the
leadership motive syndrome (high n Power higher than n Affiliation, and high
Activity Inhibition) is associated with people's greater eflfort and managerial suc-
cess in a large Company after sixteen years, whereas no value measures obtained
from questionnaires like the Edwards Personal Preference Schedule significantly
predict managerial success over the same period of time (see Bray, Campbell, &
Grant, 1974, as well as Chapters 8 and 12 of this book).
However, it makes good sense to regard v Achievement as a measure of in-
centive strength, for Hüll would also argue that if drive strength were zero, the
product of incentive and habit would not increase the tendency to make a re-
sponse. That is, no matter how much practice a rat has had in pressing a bar
to get food, nor how large the incentive, the rat will not press the bar if it is
not hungry. This corresponds in the Patten and White experiment to the neu-
tral, no-failure condition in which no achievement motive is aroused, although
there are variations in the extent to which people in that condition value
achievement.
In contrast, as the Hüll formula would predict, if the achievement motive is
aroused, then incentive value does make a difference. As Figure 13.2 shows
(Patten & White, 1977), when the achievement motive was aroused through ego-
involving instructions, subjects for whom the incentive value of achievement was
high (high v Achievement) performed much better than those for whom the in-
centive value of achievement was low (low v Achievement). This is analogous to
a hungry rat's responding more when the reward is large than when it is small.
It is as if valuing achievement increases the push to do better in this Situation,
once the subject's achievement motivation has been aroused. Figure 13.2 also
confirms the fact that the difference in v Achievement has no eflfect on perfor-
mance in the neutral condition, when the achievement motive has not been
aroused.
Furthermore, there is evidence in the Patten and White experiment for the
Joint eflfect of all three variables on response strength or the impulse to work
hard at the task. If the top curve in Figure 13.1 is broken down and plotted
separately for those high and low in v Achievement, it appears that those high
in v Achievement gain significantly more with practice when the achievement
motive is aroused through failure than those low in v Achievement under similar
conditions. In other words, Ts = Ms X Ps X /5, or the tendency to work hard
equals n Achievement times practice times v Achievement. Patten and White
(1977) were not interested in making such a multivariate prediction of perfor-
mance, but their results support the inference that all of these variables should
contribute to predicting response strength in some combination or another.
28 - y
26 —
T3
High v Achievement,/'
c 24 -
t (lriSec
O
22 -
20 -
c 18
QJ
aj / ' Low v Achievement

IAO.
16 —
mpr
14 -
12 -
1 1
Neutral Ego-Involved
(n Achievement Aroused)
Instructions
Figure 13.2.
Total Group Mean Improvement Scores, as a Function of Achievement Classification and Instruc-
tional Conditions (after Patten & White, 1977).
Patten and White (1977) were also able to show that subjects who attrib-
uted their failures more to lack of effort showed a greater improvement in per-
formance by the fifth trial, even when the effects of skill on previous trials and
v Achievement scores were controlled. In contrast, those who attributed their
failures to a lack of ability or task difficulty did not show a greater improvement
on any of the trials after failure. In other words, the belief or expectancy that
effort would make a difference led to a greater improvement in Performance.
In terms of the model, this belief contributes to perceived probability of suc-
cess in the same way prior practice does. However, this contribution may occur
only under conditions of achievement motive arousal, because Patten and White
also found that simply asking people for reasons why they failed increased their
Performance, as contrasted to not asking subjects to give reasons for failure.
They argue that asking for causal attributions may simply be another way of
arousing achievement motivation. Thus, when they report a significant correla-
tion between v Achievement levels and improvement scores in an experiment in
which causal attributions were asked for, the result is essentially the same as
that shown on the right-hand side of Figure 13.2: when achievement motivation
is aroused, high v Achievement leads to better Performance.
Patten and White also argue that Weiner's theory cannot be correct that
v Achievement leads to people's attributing failure to lack of effort, which leads
to better Performance. For under conditions of achievement motive arousal, high
v Achievement is significantly associated with better Performance no matter
what causes people assign for failure. There is no indication that the effect of
high v Achievement is mediated by attributing failure to lack of effort. Once
again, the data support the model which states that the incentive value of suc-
cess, here represented by v Achievement, contributes to response strength inde-
pendently of the perceived probability of success (here represented by the belief
that effort will make a difference in the outcome).
• HOW MOTIVES, SKILLS, AND VALUES JOINTLY

DETERMINE SUCCESS AS A NAVAL OFFICER
What is needed is more research demonstrating how all these variables combine
to predict response strength. McClelland (1981a) reported one investigation in
which he was able to show that motive, Schema or value, and trait or skill vari-
ables all contributed independently in a multiple regression to predicting supe-
rior Performance among commanding or executive officers in the U.S. Navy (see
Table 13.2).
The variables tested were not chosen to check on the model being discussed
here, but it was possible to show that motive variables, Schema or value vari-
ables, and skill or trait variables each contributed to predicting officer success.
For example, the leadership motive pattern (see Chapter 8), the belief that
power must be used in the officer's job (a belief or value), and scheduling skill
all contributed significantly to the multiple correlation predicting a person's suc-
cess as a commanding or executive officer in the Navy. Furthermore, combina-
tions of any two of these types of predictors improved the multiple correlation
coefficient (R) still further, whereas using all three types of determinants pro-
duced the rather substantial multiple R of .68, which is considerably higher
than is usually found in studies of this type.
McClelland uses this argument to answer Mischers (1968) complaint that
relationships found in personality study are generally low, hovering around cor-
relations of .30. This is true, McClelland argues, primarily because researchers
typically focus on single variables rather than on the multivariate determination
of response strength. While Table 13.2 makes the point about the importance of
multiple determinants of an outcome in a general way, a much more carefully
designed investigation is needed to discover what variables in what combinations
best predict response strength or Performance.
• FACTORS INFLUENCING AFFILIATIVE ACTS

AND CHOICES
Constantian (1981) has carried out an experiment that has several advantages
for testing various ideas that have been put forward about the way motives, val-
ues, and skills determine response strength. First, she worked in the area of af-
filiation rather than achievement. This is an advantage, because previous modeis
proposed are supposed to be general, although they have all been derived from
achievement situations. Second, she obtained a measure of the frequency of op-
erant or spontaneous affiliative acts over time, thus adopting a procedure that is
a crucial part of the Atkinson and Birch model, although they have never em-
Table 13.2.
CHARACTERISTICS RELATED TO SUPERIOR PERFORMANCE AMONG COMMANDING OR
EXECUTIVE U.S. NAVY OFFICERS: N = 12 (after McClelland, 1981a)
Unique Variance
Tested Variable Contributed (Percent) p
Motives:
1. Leadership motive pattern (n Power > n Affiliation, high Activity 6 .01
inhibition)
2. Low entrepreneurial interesta 10 <.01
Schemata or values:
3. Ideal climate: responsibility organizational, not personal 5 .02
4. Ideal climate: reacting against bureaucratic inertia 9 <.01
5. Use of power required in job 10 < .01
6. Use of affiliation not required in job 2 .11
Traits:
7. Manages by optimizing 12 <.01
8. Low on manages by helping 5 .02
9. Has scheduling skill 7 <.01
Multiple R = .68, p < .001, for all nine variables. Regression stopped when additional variables
added less than .01 to R.
Motives Motives Schemata

Motives Schemata Traits Plus Plus Plus
Ahne Alone Alone Schemata Traits Traits
Multiple R .24 .40 .30 AI .39 .52
P .12 .02 .095 .01 .05 .001
a
From Strong-Campbell Vocational Interest Blank.
ployed it in their empirical work. To obtain this measure, Constantian adopted a

method reported earlier by Larson and Csikszentmihalyi (1978). Subjects who
were summer-school students were asked to wear electronic pagers, or beepers
(of the type doctors use), for a week. They were beeped randomly seven times a
day between the hours of 9:00 A.M. and 11:00 P.M., in sets of two-hour periods
(9:00 to 11:00 A.M., 11:00 A.M. to 1:00 P.M., and so forth). When they were
beeped, they were to fill out a brief checklist explaining what they were thinking
and doing at the time. The measure of operant affihative activity was the pro-
portion of times beeped when the person reported he or she was conversing with
someone or writing a letter to someone.
Constantian (1981) also obtained a more conventional measure of peoples'
preferences for doing things with people, which we will call the Affihative
Choice Measure. Subjects were asked to indicate on a scale of 1 to 7 how much
How Motives Internet with Values and Skills 529
they would like doing fifteen different types of things with friends, such as
working at a Job, doing errands, visiting a museum, living in an apartment,
and so on.
The key question is how the affiliative motive combines with other factors to
determine the strength of either operant affiliative acts or affiliative choiees. Con-
stantian obtained a measure of n Affiliation coded in the usual way (see Heyns,
Veroff, & Atkinson, 1958) from stories written to six pictures. To obtain a mea-
sure of the value subjeets placed on affiliation, she added up the number of posi-
tive reactions and subtracted the number of negative reactions to being with
people. This will be referred to as the v Affiliation measure. Finally, a measure
of the pereeived probability of success in social situations was obtained in the
following way. Subjeets rated themselves on a scale of 1 to 7 on a number of
items such as "How often do you feel that you have handled yourself well at a
social gathering?" (on a scale of almost never to almost always) and "How sure
of yourself do you feel among strangers?" (on a scale of not at all sure to ex-
treme ly sure). The mean rating on these items represents subjeets' estimates of
their degree of social skill, or their pereeived probability of success in social
situations.
How the Situation Combines with the Affiliative Motive and

Affiliative Values to Predict Affiliative Choiees
The subjeets were also asked to rate how much they liked doing various activi-
ties alone and with friends (Constantian, 1981). Indirectly this gives an indica-
tion of how often activities are in fact carried out alone or with others, as Table
13.3 demonstrates. People report that they like reading for pleasure much more
by themselves than with others and that they like dining out with others rather
than by themselves. This simply confirms what all theorists have argued—
namely, that the environmental Situation is a very important determinant of be-
havior. Whatever peoples' motives, values, and social skills in the affiliative area
are, some activities like Shopping and dining out provide more opportunities for
affiliative behaviors than other activities like reading for pleasure or working on
a hobby. Any observer wanting to predict whether a person will be with a friend
should first take into account the Situation in which the behavior is supposed to
oecur. People in restaurants are much more likely to be with others than people
reading a book.
Person variables also interact with situations differently, as Table 13.3 also
shows. Note that people high in n Affiliation are significantly more likely to like
doing errands, exploring a new city, or Shopping with friends, but they are not
more likely to like to read, do a hobby, or cook with friends. Those who value
affiliation strongly like doing almost any activity more with friends except run-
ning errands or cooking. The contrast in the results for partieipating in a hobby
or doing errands with friends is particularly interesting. For n Affiliation is posi-
tively associated with doing errands with friends and v Affiliation is not, whereas
the reverse is true for partieipating in a hobby with friends.
The difference between the two activities is that doing errands is compulso-
Table 13.3.
MEAN LIKING FOR DOING VARIOUS ACTIVITIES ALONE OR WITH FRIENDS: N = 111 (after
Constantian, 1981)
Mean Liking (Scale of

1 to 7)
With Correlations with
Activity Ahne Friends n Affiliation v Affiliation
Reading for pleasure 5.77 3.43 .07 .19*
Doing art or hobby you enjoy 5.17 4.79 .08 .29**
Doing errands 4.19 4.56 .24* .09
Cookiing 4.07 4.77 .06 .10
Dining out in a restaurant 2.02 6.41 .14 .21*
Exploring a new city 3.47 6.10 .21* .20*
Shopping 3.96 5.02 .20* .37***
*p < .05.
**p < .01.
***p < .001.
ry, whereas participating in a hobby is completely voluntary. Thus, the results

suggest that high v Affiliation, but not high n Affiliation, leads to liking to do
things with people that are normally done alone, such as reading for pleasure or
participating in a hobby. One may doubt that in fact they will read more often
with friends and suspect they are just responding in a stereotyped way to the
idea of doing things with friends. In contrast, subjects high in n Affiliation ap-
pear to like to do things with friends that they have to do anyway, such as run-
ning errands. Of course, some activities are liked more by those both high in
n Affiliation and v Affiliation, like Shopping or exploring a new city. The point
of this analysis is that personal characteristics interact with different situations
in different ways. Thus, it is not enough to know whether a Situation calls for
more or less affiliative activity. We must also know whether it is a required or
completely voluntary activity. If it is the former, people with high n Affiliation
are more likely to do it with friends, whereas if it is the latter, people high in
v Affiliation are more likely to do it with people.
The relationship of Situation and person variables to behavior in combina-
tion is illustrated by Table 13.4. In this analysis, the Situation is represented
symbolically when subjects are asked how much they would like doing some-
thing either alone or with friends. Note that in Section A of the table, the Situa-
tion is very important in predicting how much the students will like working. If
they are with friends, they will like working much more than if they are alone.
The person variable is also important: those with high n Affiliation like work
Table 13.4.
RELATIONSHIP OF A PERSON AND A SITUATION VARIABLE ON LIKING (SCALE OF 1
TO 7) VARIOUS ACTIVITIES: N = 111 (after Constantian, 1981)
A. Mean Liking for Working at a Job

Person Variable
Low High
n Affiliation n Affiliation
Situation Variable (N = 54) (N = 57) Combined
Alone 3.7 2.8 32
' (Difference = 2.3; p <.001)
With friends 11 1£
Combined 4.5 4.2
Difference = 0..3; p = .02
Interaction of motive and Situation variables: p = .03.
B. Mean Liking for Doing Errands

Person Variable
Low High
Situation Variable (N = 54) (N = 57) Combined
Alone 4.7 3.7 4 2
(Difference = 0.3; n.s.)
With friends iii 4,9 4.5
Combined 4.4 4.4
Difference = 0; n.s.
Interaction of motive and Situation variables: p = .001.
less than those with low n Affiliation, presumably because the "highs" would
prefer socializing with people to working. The person and Situation variables
also interact: High n Affiliation has opposite effects on liking for work, depend-
ing on the Situation. If they are working alone, those high in n Affiliation like
work less than those low in n Affiliation; if they are with friends, they like work
more than those with low n Affiliation.
In Section B of Table 13.4, neither the person variable nor the Situation
variable by itself is associated with liking to do errands more or less. But the in-
teraction is very significant: those high in n Affiliation like errands less if they
are alone and more if they are with friends, in contrast to those with low
n Affiliation. To predict successfully how much a person will like running er-
rands, it is necessary to know the Situation variable (whether he or she runs
them alone or with friends) and the person variable (n Affiliation score), as well
as how they interact. Those who are high in n Affiliation and doing errands
with friends will like doing errands most (mean = 4.9), whereas those who will
like running errands least are those high in n Affiliation and doing them alone
(mean = 3.7).
How Values Affect the Way the Affiliative Motive Expresses Itself
Constantian's study was designed to get at the interactive affects of the affiliative
motive and affiliative values. According to theory, the affiliative motives involve
affectively toned associative networks built on natural incentives connected with
early contact gratifications. In contrast, as people grow up, they form concepts
relating to their experiences with people, which become conscious values (see
Chapters 5 and 9). These values are much influenced by parental and societal
inputs as to the importance of affiliative activities; therefore, people's affiliative
values need bear no necessary relationship to their affiliative motive dispositions,
that is, to the affective experiences they have actually had with people, especially
early in life. In Constantian's (1981) study, for example, the correlation between
the n Affiliation score, which is not under conscious control, and the v Affilia-
tion score, which is, is .21, iV = 111, p < .05. In other words, there are almost
as many people high in n Affiliation and low in v Affiliation as there are those
who are high in both measures. This means that motives and incentives could
turn out to be relatively independent determinants of response strength, as ex-
pected according to theory.
How do motives and values interact? Constantian (1981) also obtained a
measure of the extent to which people valued solitude using an exactly parallel
method except that she substituted the phrase being alone for being with people.
Liking for solitude and liking for being with people were not highly correlated
(r = .14; not significant). So there are about as many people who feel positive
both about being alone and being with people as who prefer one State to the
other. However, it proved possible to get two groups of people nearly equal in
size who either valued solitude more than affiliation or vice versa. Table 13.5
shows that liking to do things with friends (that is, the summary score for all
activities) is jointly increased by the affiliative motive and affiliative values acting
in concert. Those who are high in both characteristics have the highest average
score (mean = 52.0), much higher than those who are low in both characteris-
tics (mean = 47.0). Both independent variables contribute significantly to liking
for doing things with friends, but there is no interaction between them showing
that they combine additively.
In contrast, when Constantian (1981) asked the subjects in another part of
the questionnaire to indicate what they thought the good effects of a prolonged
period of being alone would be, the results were quite different. Those high in
n Affiliation mentioned increased peace and self-awareness significantly more
often than those low in n Affiliation, and so did those valuing solitude more
than affiliation. The result here is that high n Affiliation combines with the
value for solitude (not for affiliation) to produce a stronger response of this type.
Table 13.5.
DIFFERENT EFFECTS OF VALUING AFFILIATION AND SOLITUDE ON CORRELATES OF
n AFFILIATION SCORES (after Constantian, 1981)
A. Mean Liking for Doing Things with Friends:"

Values
Affiliation
Less,
Affiliation Solitude
Motive Classification More More Mean
High n Affiliation 52.0 50.2 51.0
Low n Affiliation 50.9 47.0 48.9
Mean 51.4 48.6
Analysis of variance
Motive source P = .05
Value source p = .01 Interaction insignificant
B. Mean Frequency of Mentioning Peace and Self-awareness as Values of Prolonged Solitude.>a

Values
Affiliation Solitude
Motive Classification More More Mean
High n Affiliation 49.6 53.2 51.4
Low n Affiliation 47.0 50.2 48.6
Mean 48.3 51.7
Analysis of variance
Motive source P = .03
Value source P = .01 Interaction insignificant
a
T-scores (mean = 50; SD = 10).
What this finding illustrates is how values can shift the way a motive expresses
itself. That is, if people high in n Affiliation value affiliation, they seek being
with people more; if they value solitude, they may seek the peace and self-
awareness that comes from being alone more.
The point is made even more dramatically in Figure 13.3. Note first that
valuing affiliation more than solitude had little affect on liking for taking a
country walk with friends for those low in n Affiliation (left side of the figure).
However, for subjects high in n Affiliation, if they also valued affiliation, they
reported that they liked much more taking a country walk with friends. This is
the same result as is shown in Section A of Table 13.5 for doing all types of ac-
tivities with friends.
Q Affiliation Valued More Than Solitude

7.00 [ j Solitude Valued More Than Affiliation -70
I 6.00 -
*
-60
-
I 5.00 -50
4.00 -- 4 0
3.00 - 1—|
-- 3 0
I
2.00 1— -- 2 0 O
I * - 10 o
o
Low High Low High

*Difference significam at p< .05.
Figure 13.3.
Effects of Differing Values on the Outlets of High n Affiliation (data from Constantian, 1981).
The right side of Figure 13.3, however, shows quite a different result, this
time for the proportions of persons found to be writing letters when they were
paged. In general, those who valued solitude were more likely to be found writ-
ing letters, but the difference did not even appioach significance for those low in
n Affiliation. For those high in n Affiliation, however, a significantly higher pro-
portion of those valuing solitude more than affiliation were found to be writing
letters than those valuing affiliation more than solitude. To put the question in a
somewhat different perspective, we might ask what it is that a person high in
n Affiliation might do to establish contact with others if he or she valued soli-
tude more than being with people. One answer clearly is writing letters, because
it is an activity that involves contact with people but that a person can do
alone.
In short, values can readily shift the way the affiliative motive expresses
itself in various activities. It is precisely this point that the critics of the
n Achievement literature have failed to recognize in arguing that n Achievement
is too individualistically defined. It may be defined that way in some cultures
but not in others, just as for some people, valuing being with people is more im-
portant than valuing solitude. Such people, if they have high n Affiliation, will
be more likely to do things with friends. But there are other people (just as
there are other cultures) who value solitude more, and in those instances the
outlet for relating to people, if people have high n Affiliation, will be by contact-
ing them in some way like writing letters that does not involve joining with
them in activities.
The Joint Effect of the Affiliative Motive, Social Skill,

and Affiliative Values on Affiliative Acts and Choices
As theory would predict, the presumed personality determinants of affiliative be-
haviors do not correlate very highly with each other. As Table 13.6 shows, the
n Affiliation score is correlated with the v Affiliation score only at a barely sig-
nificant level, and not at all with people's estimates of their social skill. The
v Affiliation and social skill scores are much more highly correlated, because
both are influenced by the subject's self-concept. In the one case, people are
asked how much they like being with people, and in the other, how successful
they are when interacting with people. It seems reasonable to assume that judg-
Table 13.6.
CORRELATIONS OF PERSONALITY DETERMINANTS WITH EACH OTHER AND WITH
AFFILIATIVE ACTS AND CHOICES: N = 111, EXCEPT FOR AFFILIATIVE ACTS COLUMN
(data from Constantian, 1981)
Operant Respondent
Personality Affiliative Affiliative
Determinant Motive Value Skill Acts (N = 48)a Choicesb
Motive:
n Affiliation .21* .05 .42** .21*
n Affiliation ratioc .17 -.04 .45*** .19*
Value:
v Affiliation — .45*** .17 .41*
Skill:
Social skill _ -.14 .06
Operant affiliative .33*
acts (N = 48)
a
Percentage of times found conversing with others or writing letters when beeped.
b
Mean liking for carrying out fifteen activities with friends.
c
T-score n Affiliation/T-score n Affiliation + T-score n Power + T-score n Achievement + T-score Activity Inhibition.
*p < .05.
**p < .01.
***p < .001.
ments about social skill would be related to liking for being with people. Indi-
viduais who feel they are shy and awkward are unlikely to respond by saying
that they like being with people. Ideas about skill and liking are part of people's
general understanding of their relations to people.
In this light, it seems surprising that liking to do things with friends (affilia-
tive choices) is not related to perceived social skill (r — .06). The reason may
be that the social skill items have more to do with dealing with strangers or un-
familiar situations, as well as that some people who think they have poor social
skill prefer doing things with friends precisely for that reason. This would re-
duce the expected positive association between the two variables.
The strength of the affiliative motive, as Table 13.6 shows, is strongly re-
lated to the frequency of operant affiliative acts reported by the person beeped.
Neither the value placed on affiliation nor the estimate of one's social skill is re-
lated significantly to the operant affiliative act measure. In contrast, the v Affilia-
tion measure is much more strongly related to the affiliative choices score than
the n Affiliation score is. The social skill measure does not seem to be a deter-
minant of affiliative choices, either.
Why should the n Affiliation score be more related to the operant affiliative
acts score and the v Affiliation measure be more related to the affiliative choices
score? The key to understanding the difference lies in the realization that the
acts score does not involve the subjects' conscious perception or judgment. That
is, it is not based on the subjects' report of what they did, which is influenced
by all sorts of cognitive factors such as memory distortions, beliefs about the
seif, or ideas as to what they should be doing. The affiliative choices score does
reflect just such cognitive variables. Motives affect the frequency with which
people do things that are not passed through conscious cognitive filters. Values,
on the other hand, represent the usually conscious conceptions in terms of
which people organize their experience and preferences. If people answer in one
part of the questionnaire that they like being with others, they should be more
likely to answer in another part of the questionnaire that they would prefer
doing things with people. Both answers are determined by the value the person
places on affiliation.
In contrast, the affiliative motive score is not conscious and therefore does
not automatically elicit the value people consciously place on affiliation. It leads
instead to more operant interactions with people, because pleasure has been ob-
tained from that type of interaction in the past. Many of the relationships of
motive dispositions to operant activities have been missed by psychologists, be-
cause they typically use self-report measures in which there is a large cognitive
element, which is not part of the motive disposition. As Table 13.6 makes clear,
motives are more important for predicting what people will spontaneously do,
whereas values are more important for determining what they will cognitively
decide should be done.
Having emphasized the main result, we may still wonder whether the deter-
minants of operant action do not influence each other. Does it not seem likely
that consciously valuing affiliation might strengthen the motive to affiliate (or
vice versa) and increase the tendency to interact with others? The data do not
permit an unequivocal answer to the question, for there is a slight but signifi-
cant positive correlation between n Affiliation and v Affiliation scores, which
could mean either that the value promotes the motive or the motive promotes
the value. Only longitudinal studies or attempts to change values or motives
through education could answer definitely what influences what. And, whereas
valuing affiliation is not associated with more operant affiliative acts, we might
argue that the affiliative choices measure is an alternative way to get at the
value people place on being with people—at least on being with friends. This
value indicator does correlate significantly with the affiliative acts score (r =
.33), perhaps because in College students can often chat with roommate friends.
Thus, it is not surprising to find that there is some relationship between subjects'
saying they like to do things with friends and finding that they are more often
conversing with people (often roommate friends).
Even this relationship between a cognitive judgment of one's interests and
actual act frequency is thrown into doubt by the fact that the choices measure is
related to the n Affiliation measure, which is in turn related to the affiliative acts
frequency. If a partial correlation is calculated to remove the influence of the
Joint relationship to n Affiliation, the correlation of the choice indicator of af-
filiative values with affiliative acts frequency is reduced to insignificance. The
overall conclusion remains the same: estimates of the cognitive value students
place on affiliation do not relate to operant affiliative acts as well as operant mo-
tive thoughts do. This may be because neither operant thoughts nor acts are di-
rectly influenced by the kind of cognitive factors that determine choices.
Table 13.6 also presents the results for the n Affiliation ratio to check the
prediction from the Atkinson-Birch model that it is not so much the absolute
strength of the motive as its strength relative to other motives in the Situation
that will predict the frequency of acts related to it. Constantian (1981) also
scored the TAT stories for n Power; n Achievement; and Activity Inhibition, a
measure reflecting a concern for Controlling one's behavior (see McClelland,
1975). The motive ratio is the strength of n Affiliation relative to all motiva-
tional tendencies present in the person. Its correlation with the affiliative
acts score is slightly higher in line with prediction, although not signifi-
cantly so.
How the Determinants of Affiliative Acts and Choices Interact

with Each Other to Produce Their Effects on Behavior
By using the technique of multiple correlation, we can determine the combined

effects of motives, values, and perceived skill on predicting affiliative acts or
choices. In other words, how much of the Variation in frequency of affiliative
acts can be predicted by taking all three of its possible determinants and their
combinations into account? Table 13.7 gives the relevant results. The affiliative
acts measure was available on only forty-eight subjects, but the correlations
based on the total number of subjects were used in the analysis wherever possi-
ble. This procedure seemed better than limiting the sample just to the forty-
eight subjects for each correlation, because the beeper sample was chosen by
Constantian (1981) to represent extremes in attitudes toward being alone or
with others. This exaggerated correlations between variables in some cases. The
partial correlations represent the "pure" contribution of a given determinant to
affiliative behavior, with the influence of other variables removed.
The results shown at the top of Table 13.7 simply confirm in a purer form
what was previously shown in Table 13.6—namely, that the n Affiliation score is
the sole significant contributor to predicting the affiliative acts score, and the
v Affiliation score is the sole significant predictor of the affiliative choices score,
once the eflfect of other variables has been removed. The effect of all three vari-
ables taken together is to increase slightly the ability to predict affiliative behav-
ior over and above a prediction obtained from any one of the personality deter-
minants. The multiple R's in Table 13.7 are slightly higher than the single-
variable r's in Table 13.6.
Atkinson (1957), Hüll (1943), and Spence (1956), have made specific predic-
tions as to how these determinants will interact in affecting behavior, so the in-
teraction terms in Table 13.7 are of particular interest. They show that only the
motive times the perceived skill interaction contributes significantly to predicting
the affiliative acts score, as the theoretical modeis would predict. In fact, the ad-
Table 13.7.
PREDICTING FREQUENCY OF AFFILIATIVE ACTS AND CHOICES FROM PERSONALITY
DETERMINANTS
Affiliative Acts Affiliative Choice

(N = 48)a (N = \Ul)b
Personality Determinant Partial r P Partial r P
Motive (M)\
n Affiliation .39 < .001 .14 n.s.
Value (V):
v Affiliation .19 n.s. .40 <.001
Skill (S):
Social skill -.24 n.s. -.14 n.s.
Multiple R = .48 < .01 .45
Interactions:
M X V -.19 n.s. -.07 n.s.
M X S •41 < .01 .09 n.s.
V X S -.12 n.s. .10 n.s.
Multiple R = .60 .46
Gain in R by adding in interactions <.O5 n.s.
a
Percentage of times found conversing with others or writing letters when beeped.
^Mean liking for carrying out fifteen activities with friends.
dition of this interaction raises the multiple correlation from .48 to .60, a signifi-
cant increase.
Its meaning is more clearly illustrated by Figure 13.4, which shows that
perceived social skill contributes to predicting affiliative acts frequency only if
n Affiliation is high. The subjects with low n Affiliation showed no increase in
the frequency with which they were found conversing with another person the
greater they perceived their success to be in social situations. Figure 13.4 plots
the relationship between perceived social skill and affiliative acts when n Affilia-
tion is assumed to be high (half a Standard deviation above the mean) or low
(half a Standard deviation below the mean). To put it a different way, people's
believing that they are likely to be successful in social interaction does not lead
to more interaction unless they are motivated to use the skill. This is the same
type of conclusion Hüll came to in observing rats in mazes. Even though the
rats knew how to run in the maze from previous practice, they would not do so
unless motivated by hunger. But if the rats were motivated, their previous expe-
rience in running the maze (conceptualized as probability of success) would
greatly increase the speed and success with which they could now traverse the
maze. Likewise, if people want to affiliate, perceived skill in affiliating will
greatly increase the likelihood that they will be found conversing with others.
This result is the same as that obtained in the achievement area in the Pat-
50
&s
.§ ? 40 High n Affiliation (T-Score = 55)
M 30
•S 20
£5
10
< §
(L» O
5 ••5' ° Low n Affiliation (T-Score = 45)

< 3
1 1 1 1 1
30 40 50 60 70
Perceived Social Skill (x), in Standard Scores
Figure 13.4.
Interaction of Affiliative Motive Strength and Social Skill in Predicting Affiliative Act Frequency.
ten and White (1977) experiment, where a different motive (n Achievement) and
measure of skill (number of practice trials) were employed. That is, the top line
in Figure 13.1 shows that people in whom n Achievement is higher gain more
from practice. In that instance and in the Constantian experiment, motive and
skill combine to increase response strength, defined in the affiliation experiment
as response probability.
In the Constantian study, too, there is no significant interaction of the value
and skill measures. Subjects high in v Amliation do not engage in more affilia-
tive acts the greater their perceived social skill. This is analogous to the finding
in the Patten and White (1977) experiment that subjects high in v Achievement
with increased practice (or probability of success) do not perform better. (See
the bottom two lines in Figure 13.1.) This finding may seem contrary to com-
mon sense, because we often assume that if we think something is important—
like getting enough exercise—we would be more likely to do it if we were more
skillful at doing it, that is, at getting exercise by jogging, playing tennis, and so
on. But the findings in these two experiments suggest this is not the case. Exer-
cise would have to satisfy some motive for us to do it more often, and then
greater skill would lead us to exercise more often.
The results for the third interaction—between motive (n Affiliation) and
value (v Affiliation)—are not the same as in the Patten and White experiment,
in which high v Achievement improved Performance if the motive was aroused
or high. Here the trend is in the opposite direction, although it is not signifi-
cant: it takes less v Affiliation to lead to an affiliative act if n Affiliation is high.
So we cannot draw a firm conclusion as to whether valuing something and also
being motivated for it combine to increase the likelihood that a person will be
found doing it. Perhaps it makes a difference how well someone can do some-
thing, as in the Patten and White experiment, versus how often he or she will
do it, as in the study of affiliation acts frequency (Constantian, 1981). Perhaps
values combine with motives to improve Performance but not to increase the
frequency with which the Performance is undertaken—which is largely a func-
tion of motive level and the combination of motive and skill level.
Where M = motives, S = perceived social skill, and V = incentive value,
the M X S combination does not contribute to predicting the affiliative choice
measure, nor do the two other interactions (M X V or V X S) contribute sig-
nificantly to predicting that measure. In other words, values do not combine
with perceived social skills the way motives do, nor is the effect of multiplying
motives times values significant.
If Atkinson's (1980) motive ratio variable is substituted for the n Affiliation
score, the effect is striking. The multiple correlation goes from .49 for the affilia-
tive acts score without the interactions to .76 including the interactions, a highly
significant gain. Again, there is no effect of using a motive ratio for predicting
the affiliative choices score. And once more, the effect of the motive times skill
interaction makes a highly significant contribution to the multiple R for the af-
filiative acts measure, as does the motive times value interaction. That is, valu-
ing affiliation has less effect in increasing affiliative acts for those with high
n Affiliation relative to other motives than for those with low n Affiliation rela-
tive to other motives. Since such an interaction is the opposite of what would be
theoretically predicted and is not significant when the n Affiliation ratio is not
used, it needs further confirmation.
Next, we should examine the third-order interactions directly, since both
Hull's and Atkinson's modeis assume that the three determinants of response
strength multiply with each other. Thus, we would like to see the effect for pre-
dicting affiliative behavior of multiplying M X S X V, but it is not possible to
do so meaningfully with these data because of what statisticians call multicolin-
earity. Roughly speaking, this means that multiplying the variables two at a
time has taken up so much of the explanatory power of the determinants that
there is none left over to be accounted for by multiplying all three together.
There is an approximate way to test the applicability of the Hüll and
Spence modeis. It will be recalled that Spence argued that if either incentive
(here called value) or drive (here called motive) were present, some behavior
would result, whereas in the Hüll model, if either of these variables were re-
duced to zero, the equation would predict no response. Thus, Spence suggested
that M and V should be added rather than multiplied, as in the Hüll model.
Table 13.8 shows the effects of adding or multiplying the M and V variables.
That is, we can treat either the sum or the product of these two determinants as
independent variables in a multivariate prediction of affiliative behavior. Then
we can see whether either the product (M X K) or the sum (M + V) times
perceived social skill (S) contributes significantly to predicting affiliative behavior.
As Table 13.8 shows, in this instance the Hüll formula works and the
Spence formula does not. That is, both the sum and the product of M and V
contribute significantly to predicting affiliative acts or affiliative choices, but only
the (M X V) X S interaction adds nearly significantly to the multiple R, and
contributes significantly to predicting the affiliative acts score. The (M + V)
X S interaction does not contribute significantly to predicting the affiliative acts
score or to increasing the multiple R.
At first glance the results for the M X V interaction term in Tables 13.7
and 13.8 may seem contradictory, for it contributes negatively to predicting af-
filiative acts in Table 13.7 and positively in Table 13.8. However, the difference
lies in the fact that the contribution of M X V in Table 13.7 is assessed after
the contribution of M to the multiple correlation has been taken into account,
whereas in Table 13.8 the contribution of M is included in the M X V term.
This means that once the effect of M by itself has been removed, the remaining
interaction of M X V may contribute negatively to the multiple correlation.
The Hull-Atkinson formula, however, is not stated this way: it is more
nearly represented by the formula (M X V) X S, which predicts fairly well
the affiliative acts score (R = .57). To this extent, the results may be taken as
confirmation that a multiplicative relationship among these three determinants
represents a reasonable way of combining them to predict response strength.
However, as good a prediction of affiliative acts can be obtained from the deter-
minants listed in Table 13.7, consisting largely of M and M X S (R = .60).
So multiplying all three determinants is not the only, or even the best, way of
predicting response strength in this Situation.
Table 13.8.
COMPARISON OF HÜLL AND SPENCE FORMULAS FOR PREDICTING AFFILIATIVE ACTS
AND CHOICES; MOTIVE = n AFFILIATION, VALUE = v AFFILIATION
Alternative
Formulas Affiliative Actsa Affiliative Choicesb
for Predicting
Behavior Partial r p Partial r P
Hüll formula (M X V X S):
Social skill -.29 n.s. -.06 n.s.
Motive X value (affiliation) .46 <.01 .39 <.01
R == AI .39
(M X V) X S .37 <.01 .12 n.s.
R =-- .57 <.01 .41 <.01
Gain in R <.10 n.s.
Spence formula (M + V) X S:
Social skill -.30 .05 -.06 n.s.
Motive + value (affiliation) .48 <.01 .39 <.01
R == .49 <.01 .39 <.01
(M + V) X S .13 n.s. .11 n.s.
R == .50 <.01 .40 <.01
Gain in R n.s. n.s.
a
Percentage of times found conversing with others or ikvriting letters when beeped.
b
Mean liking for carrying out fifteen activities with friends.
Neither Hull's nor Spence's model for interactions of the variables improves
the multiple R for the affiliative choices measure of response strength (see Table
13.8). What this analysis shows above all is that it is vitally important to distin-
guish between operant and respondent indicators of response strength and, in
trying to predict these indicators, to include measures of motives, skill, and val-
ues as independent determinants of action tendencies. Personality theorists have
repeatedly confused these variables and talked about them as if they were inter-
changeable, just as Weiner (1980a) refers to v Achievement as if it were a mea-
sure of the achievement motive. Or what is worse, they have argued that per-
sonality is inconsistent because the measures do not correlate highly with each
other (Mischel, 1968), when, according to the theory outlined here, they should
be independent and therefore uncorrelated with each other. If we accept the fact
that motives, skill, and values are three nearly independent personal determi-
nants of response strength, they can, taken together, account for a surprising
amount of the Variation in spontaneous or operant behavior.
In the Constantian study, if the n Affiliation ratio is used as a determinant,
it together with affiliative values and skill and their interactions accounts for
over 75 percent of the Variation with which people engage in affiliative acts over
time (see D'Andrade & Dart, 1982, for a discussion of the importance of ac-
counting for Variation rather than variance in behavior). This is quite impressive
in view of the error undoubtedly involved in sampling people's behavior at dif-
ferent times during a week. The remaining 25 percent of the Variation could eas-
ily be accounted for by situational factors Controlling the opportunity for interac-
tion. For example, it is practically impossible for a person to be conversing with
someone if he or she is sleeping, in class, or riding the subway when beeped.
In contrast, the determinants do not combine as expected to increase the
predictability of affiliative choices. This confirms the importance of the distinc-
tion Skinner (1966) made between operants (strength measured by response
probability) and respondents (strength measured by amplitude, latency, persis-
tence, and so on), as well as his inference that operants more readily show the
influence of "purpose," or motivational factors. Affiliative choices as responses
to contrived Stimulus situations represented by questionnaire items are deter-
mined almost entirely by values deriving from the same cognitive source as the
choices. People carry around with them cognitive Schemata that organize their
feelings, attitudes, and choices in a particular area such as affiliation or achieve-
ment. These Schemata are often part of their self-image, although they are not
coextensive with self-relevant values. They influence behavior, particularly when
the nature of the Situation in which the behavior occurs is clearly defined in
cognitive terms—that is, when a Situation is seen to be affiliation or achievement
related.
When people are asked whether they would like doing something "with
friends," the question taps a value associated with liking for people, which de-
termines how they answer the question. In contrast, the frequency with which
people converse with someone eise is not determined by the value placed on af-
filiative activity, but on the pleasure they unconsciously get from such interac-
tions, as reflected in the strength of the affiliative motive.
The distinction between conscious cognitive values and relatively uncon-
scious affectively toned associative networks or motives can also be invoked to
explain the differences Silverman (1976) has found in the reactions of people
with various personality disorders to emotionally toned phrases or pictures rele-
vant to their pathology when the Stimuli could or could not be consciously rec-
ognized and processed (see Chapter 2). That is, if a phrase like go shit was pre-
sented to stutterers too fast to recognize, it touched off affective associations
that produced more stuttering. But if the same phrase was presented slowly
enough to be recognized, it was processed in terms of cognitive Schemata (here
called values) that filtered out its effect on stuttering.
Just as HulPs and Atkinson's theoretical modeis have stated, social skill, or
a person's perceived probability of success in an interaction, will contribute to
the frequency of the interactive behavior only if the person is motivated to affili-
ate. People's being good at doing something does not lead to their doing it more
often unless they are really interested in the activity—an obvious point perhaps,
but one that has often been overlooked. Even people's valuing a type of activity
or its end result does not lead to their doing it more often unless they are moti-
vated to do it. What is needed are more studies of this type that show the Joint
effects of several determinants on what people do and that also make the dis-
tinction between predicting operant acts and cognitively based choices.
• THE DISTINCTION BETWEEN MOTIVES AND INTENTS
At the very beginning of our consideration of what the term motivation means
in Chapter 1, we distinguished between conscious intents, such as are contained
in Statements like "I want to go talk to my friend," and unconscious intents, de-
duced from the fact that I am often found talking to my friend or from a be-
havioral fact such as a person seeks out a friend in a crowd of people even
though the person reports having had no conscious experience of wanting to
talk to a friend. Freud deduced unconscious intents from unusual or symptom-
atic acts; we have just deduced an unconscious intent to affiliate with people
from the operant frequency with which people are found talking with others
when beeped at random times throughout a week. In other words, a motive be-
haves like a dispositional unconscious intent, since it is unlikely that the stu-
dents beeped had made a lot of conscious choices to talk to people when they
were found to be doing so. deCharms (1982) is so impressed by the difference
between thinking of people as being moved by motivational forces and as intend-
ing to do things that he believes the two ways of looking at people involve
wholly different perspectives: "Objects are moved by causes, organisms behave
from motives, and persons act with intention." He argues that talking about mo-
tives as if they were biological forces moving behavior neglects the all-important
experience of choice and intention.
From the perspective of this and other chapters, the distinction between a
motive and an intent is clear and obvious, and it does not require a major shift
in perspective. A motive is a largely unconscious determinant of spontaneously
generated behavior, and an intent is a largely conscious preference for doing
something that should be determined jointly by motive strength, probability of
success, and values associated with doing it. The conscious-unconscious aspect
of this distinction is not essential to it: what is essential is the difference between
choice of what to do—what Atkinson calls motivation, or the tendency to
approach or avoid alternatives when they confront people in a respondent
Situation—and the operant frequency with which people act in certain ways.
For people may be conscious of their need for affiliation or may choose an
alternative without the choice's being conscious. Nevertheless, the conscious-
unconscious distinction catches what characterizes the two types of behavior
most often and helps explain why deCharms feels "organisms behave from mo-
tives"—for example, are moved unconsciously to do something without con-
sciously intending to do it.
In the Constantian study, conscious affiliative choices or intents turned out
to be determined largely by affiliative values, a little by the affiliative motive,
and not at all by perceived affiliative skill. However, the choices involved, like
going to a movie with a friend, did not require much affiliative skill. One can
safely assume that if the choices had required much skill—that is, if they had
dealt with making friends, for example—the subjects' perceived affiliative skill
would have played more of a role in affecting their intent to make friends.
It cannot be emphasized too often that a motivational intent or choice—
whether conscious or not—is a product of several determinants, including at
least two nonmotivational ones: probability of success and values. Operant

trends in behaviors, in contrast, appear to be the product of largely unconscious
motives and skill, or probability of success if it is combined with a strong mo-
tive. Conscious values play less of a role in determining operant behaviors.
NOTES AND QUERIES

1. In this chapter the similarity is noted between skill and perceived probabil-
ity of success. Normally we would expect these variables to be closely relat-
ed: if people know they can do something well, they are more likely to esti-
mate they can do well at it. But think of a Situation where this would not
be true. Which of the two variables would be a better predictor of behavior?
Would it depend on the kind of behavior—whether it was respondent (as in
solving a problem) or operant (as in the frequency with which a person
turned to solving problems)?
2. The French and Lesser experiment (Table 13.1) shows the Joint effect on be-
havior of motives and values. Redesign the experiment in a way that would
also show the effect of skill or the probability of success on behavior.
3. Think through the implications of Patten and White's (1977) conclusion
that asking people the reasons for their failure arouses achievement motiva-
tion. Does this mean that attributing failure to lack of effort improves per-
formance only when achievement motivation is aroused? Is there any way to
examine this question by getting causal attributions without arousing
achievement motivation? Individuais high in n Achievement might believe
that in general effort is important, but would you expect that belief to im-
prove Performance in a particular Situation if their motivation to achieve
was not aroused? Is this analogous to asking if probability of success or
self-confidence will improve Performance if there is no motivation to
perform?
4. Weiner (1980a) would appear to argue at times that a motive has its effect
on Performance only through causal ascriptions and at other times that the
pattern of causal ascriptions is the motive. Evaluate each of these hypothe-
ses in terms of existing findings and design research that would shed further
light on the issues involved.
5. One problem with the approach examined in this chapter is that so many
different aspects of people's understanding of a Situation or the values they
bring to it have been identified that it seems impossible to reduce them to a
manageable number in trying to predict different kinds of behavior. Cogni-
tive consistency is important, but so is a sense of control, internality, instru-
mentality, and a flock of other values like cooperativeness, autonomy, and
social approval. Rokeach (1973) has listed eighteen instrumental and eigh-
teen goal-state values that might influence behavior. In the study of Navy
officers summarized in Table 13.2, four values were measured that do not
relate particularly well to any of these values. Do you see any way to deter-
mine what the most important values are that must be taken into account,
or will the relevant ones be specific to each Situation?
6. In Constantian's beeper study, what do you think people high in n Achieve-

ment and n Power would be most likely to be thinking and doing when
paged at random intervals throughout a week? Would you expect sex differ-
ences? If so, why?
7. Design a treatment study that would determine whether increasing the value
placed on being friendly would increase the affiliative motive.
8. Our culture values people's being friendly to others, so there is congruence
between n Affiliation and v Affiliation scores. This is less true in the power
area, since our culture is ambivalent about people's consciously wanting to
be dominant or powerful. So there should be many people high in n Power
who are low in v Power. How do you think their power motive would be
most likely to express itself? The analogy is with people discussed in the
chapter who are high in n Affiliation and low in v Affiliation.
9. It is suggested in the text that the reason the M X V interaction has differ-
ent effects in the achievement and affiliation studies may be because in the
former, improvements in Performance are being predicted and in the latter,
increased frequencies of an act. Can you think of other possible explana-
tions? Do you think the findings in the affiliation area would be the same as
in the achievement area if the behavior to be predicted were shifted to per-
formance on a task of learning relationships among people (as in Figure
9.1)? That is, in that Situation, would increases in v Affiliation in combina-
tion with high n Affiliation lead to better Performance, as opposed to the re-
sult shown in Table 13.7, where the reverse result was obtained? That is, in
Table 13.7, the higher the n Affiliation, the less v Affiliation is needed to get
an increase in the frequency of affiliative acts.
10. In the end, the frequency over time of affiliative acts turns out to be much
more predictable (multiple R as high as .76) than the average liking for fif-
teen types of affiliative activity (multiple R = .46). Why do you think this
turned out to be the case? Might it be due to the fact that a larger number
of responses was being recorded in the first than the second instance (fifty
versus fifteen), to the fact that some appropriate determinants were left out
in the second instance, or to the fact that operants are easier to predict than
respondents? Design experiments that would answer such questions.
11. Apply the model for predicting behavior explained in the chapter to na-
tional behavior. Under what circumstances would a country be most likely
to go to war or commit frequent aggressive acts? Chapter 11 summarizes
some of the evidence on motivational determinants of such behavior. What
other key determinant of this behavior is suggested by the equations tested
in this chapter, and how would you measure it?
14
Motivation Training
Applying Expectancy-Value Theory to Improving Academic Performance

The Pygmalion Effect
Increasing Perceived Probability of Success
Increasing Understanding of and Control over One's Own Performance
Achievement Motivation Training for Entrepreneurs
Training Inputs in Courses for Small Businessmen
Changes in Business Activity After Motivation Training in India
Factors Affecting the Success of Entrepreneurial Training in India
Entrepreneurial Training for Minority Businesspeople in the United States
Achievement Motivation Training in Schools
Origin Training in the Classroom
Power Motivation Training

Developing Courage in Outward Bound
Socializing the Power Motive Among Alcoholics
Empowering the Staff of Community Action Agencies
Leadership Training for Managers
547
Downloaded from https:/www.cambridge.org/core. University of Exeter, on 08 Jul 2017 at 12:55:59, subject to the Cambridge Core terms of use, available at
• APPLYING EXPECTANCY-VALUE THEORY TO
IMPROVING ACADEMIC PERFORMANCE
As it became evident in the 1960s that human motives were related to important
human endeavors like entrepreneurship and management, investigators turned
their attention to methods of changing motives to improve Performance. Since
the emphasis was on Performance Output rather than on motive change in itself,
these efforts to produce change can best be understood in terms of the formula
for predicting response Output presented in previous chapters. According to that
formula, response Output, given environmental opportunity, is a function of mo-
tive strength (M) times probability of success (Ps) times incentive value (V).
Technically, response probability can be increased by changing environmental
opportunity or any one of the three person variables in the equation. Early ef-
forts to introduce change focused on affecting the probability of success variable.
School learning or skill acquisition affects this variable. If people learn how to
do something better, it by definition increases the probability of their succeeding
at that activity and makes it more likely that they will carry out the activity if
they are also motivated to do it and they value it. But motive development
courses approached probability of success in a different way. They manipulated
the perceived probability of success without teaching skills directly.
The Pygmalion Effect

Impetus was given to this approach by Rosenthal's (1966) demonstration that an
experimenter's expectation or bias strongly influenced the way subjects behaved
in a variety of different types of situations. Of particular importance was the evi-
dence summarized in Rosenthal and Jacobson's book Pygmalion in the Class-
room (1968) leading to the conclusion that simply giving a teacher the expecta-
tion that certain randomly picked pupils in their classes were of high ability was
sufficient to improve greatly the academic Performance of those pupils. A large
number of studies on this Pygmalion effect were carried out, and while some of
them were open to criticism (see Elashoff & Snow, 1970), there seems little
doubt that the phenomenon exists (Rosenthal & Rubin, 1978).
A typical result is illustrated in Figure 14.1. This study was carried out on
inner city black boys aged seven to eleven in second to fifth grade, most of
whom were not doing very well in school. The teacher for each grade was given
the real ability test scores for a random half of the class and ability test scores
that were one Standard deviation higher than their actual scores for the other
half. In other words, the teachers were led to believe that half of their pupils
were much brighter than their test scores showed they actually were. At the end
of the school year these students showed a much larger gain in reading and
arithmetic achievement test scores than the pupils whose true ability scores had
been given the teachers. The difference in score gain for the two types of pupils
occurred for each grade, but obviously there were also differences from grade to
grade. Thus, in third grade the pupils in the control group also gained greatly in
test scores, although those the teachers thought were particularly bright gained
548
Motivation Training 549
Control Pupils
.1 + 30 ! lExperimental Pupils
£ Ü
+ 20
+ 10
5 S
I
-10
Grade 2 Grade 3 Grade 5
Figure 14.1.
Effects on Pupil Achievement Test Scores of Telling Teachers Their Actual Ability Scores (Control
Pupils) or Scores Considerably Higher Than Actually Attained (Experimental Pupils) (after
Keshock, 1970).
even more. The pupils in the control group in fifth grade, in contrast, actually
lost ground, whereas those from whom the teachers expected better work only
managed to hold their own. So it looks as if there are important teacher effects
over and beyond what they expect from their students, but however good the
teachers are, if they think certain pupils are smarter, those pupils tend to per-
form better.
Increasing Perceived Probability of Success

What is the mechanism by which teacher expectations aflfect pupil Performance?
Heckhausen (1974, 1980) reasoned that the mechanism involves the pupils' per-
ceived probability of success as it is influenced by the reasons assigned by teach-
ers and pupils for success and failure. Teachers who know their pupils lack abil-
ity will attribute failure to lack of ability rather than to other causal factors
such as lack of eflfort, luck, or task difficulty (see Chapter 12). This leads the
teachers to act very diflferently toward the pupils than they would if they
thought they were of high ability, as Brophy and Good (1970) have demonstrat-
ed. For example, they found that if a child gives a wrong answer, the teacher is
much more likely to rephrase the question or give the child a clue to the right
answer if the teacher regards the child as having high ability rather than low
ability. Children of low ability much more often get no feedback from the teach-
er—whatever they do or say—as contrasted with those of high ability, to whom
the teacher nearly always responds somehow. In time, certain children begin to
respond in ways that confirm the teacher's expectation that they will perform
poorly. For instance, they are much less likely to raise their hands in class than
those considered to be of high ability. In short, the teacher conveys to the child
that he or she is of low ability and is not expected to perform well; this, in time,
creates exactly the same impression in the child's mind, which lowers the child's
perceived probability of success and, according to the general formula for pre-
dicting response Output, decreases the likelihood of successful Performance.
Raising teachers' expectations about children's ability reverses this trend.
Heckhausen (1974) checked this possibility directly by instructing teachers
of underachieving students to attribute failure to lack of effort rather than lack
of ability. There were three to eight underachieving target students in each of
several classes of fourth-graders, and their teachers were simply told that they
should say to these students when they did not do very well, "You could do
better if you try harder." There were also other underachieving pupils in each
classroom who served as controls. The pupils were given several tests before the
experiment started, as well as four and a half months later. All the students in
the classes irnproved on a number of these measures, perhaps because the teach-
ers did not limit their comments about effort only to the target pupils.
However, it was possible to get another control group from the classrooms
of two teachers who had not attended the Session in which they were instructed
to attribute failure to lack of effort. In comparison with the pupils from these
classes, the targeted pupils from the experimentally treated classes were signifi-
cantly more likely to attribute failure to lack of effort rather than to lack of
ability; their Test Anxiety scores decreased significantly; and they scored higher
on primary mental ability tests involving speed of Performance. However, they
did not show significant gains in arithmetic Performance or the expected gain in
the h Success aspect of n Achievement.
These results are consistent with those obtained by other investigators. For
example, Dweck (1975) had children perform a number of tasks during training,
on most of which they succeeded but on some of which they failed. Whenever
they failed they were told, "You got only right. That means you should
have tried harder." Other children were given only success feedback. Those who
had received training in attributing failure to lack of effort performed better
after training and were less disrupted by a failure experience than those who
failed and had not had the training.
It took Dweck twenty-five training sessions to produce this effect, but An-
drews and Debus (1978) found they could produce a similar change after only
one hour of training. As subjects performed a task, they were asked after each
trial to give causal attributions for the outcome. Whenever they gave an effort
attribution—that is, attributing success to extra effort or failure to lack of ef-
fort—they were strongly reinforced by the experimenter by Statements like
"OK!" or "Good!" The investigators found that this type of training increased
persistence in the face of failure afterward and also increased the likelihood that
the subjects would make effort attributions. However, when the subjects were re-
tested sixteen to seventeen weeks later, while those who had been trained contin-
ued to make more effort attributions, the effect on persistence had disappeared.
In terms of the formula for predicting response strength, what these experiments
are doing is increasing the subjects' perceived probability of success by training
them to believe that more eifort leads to greater success. To put it the other way
around, attributing success to greater effort decreases the expectation of failure
that comes from attributing it to stable factors like lack of ability or task diffi-
culty. (See Chapter 12.)
Stamps (1973) tested the direct effect on fear of failure of increasing the per-
ceived probability of success. There were three groups of subjects, two of which
received treatment of different types and one of which served as a no-treatment
control group. In one treatment group, the subjects worked on various arithme-
tic problems for two weeks, rewarding themselves with tokens exchangeable at
the end for prizes, in accordance with how well they were doing. This was
called the Self-Reinforcement Group. It was designed to increase self-confidence
in doing work of this type or, in terms of the formula, to increase perceived
probability of success. The other treatment group received a kind of group ther-
apy in which the subjects played various games as the experimenter commented
on their Performance. The goal of these training sessions was to make the sub-
jects feel positive about their Performance, so they were greatly encouraged for
effort, and whenever they failed the experimenter either gave a neutral response
or encouraged them to do something eise.
Both types of treatment significantly lowered fear of failure as measured by
the Hostile Press score discussed in Chapter 10. One might suppose that lower-
ing / Failure would have some long-term effects on Performance of a similar
type, although this was not tested in this experiment. In fact, all any of these
experiments show is that it is relatively easy to change perceived probability of
success in an experimental Situation so far as that particular Situation is con-
cerned. This change may have effects that are long-term and generalize to other
situations (see the Bandura research on self-efficacy mentioned in Chapter 12),
but it is not yet known exactly under what conditions such major effects occur.
Increasing Understanding of and Control over One's Own Performance

Heckhausen (1974, 1980; Heckhausen & Krug, 1982) reports a number of more
ambitious attempts to change academic Performance through exercises derived
from achievement motivation theory. By and large, they focus on the cognitive
aspects of performing better, because Heckhausen was thinking largely in terms
of how people high and low in n Achievement evaluate themselves differently.
So he designed the training to help those low in n Achievement to think, plan,
and evaluate the seif like a person high in n Achievement, that is, to set moder-
ate rather than unrealistic goals for Performance and to attribute success to abil-
ity and failure to lack of effort rather than low ability.
Typically, the approach involved teaching students to set moderate goals in

tasks varying from ringtoss to school subjects, giving students feedback on the
extent to which they had achieved their goals, teaching them to monitor their
own Performance and reward themselves for doing well, training them in causal
attribution so that the importance of effort could be clearly seen, and trying to
generalize what they had learned in the training sessions to everyday life. Fur-
thermore, the pupils observed the experimenter as he displayed the characteris-
tics of a person high in n Achievement: "He spoke out loud what went through
his mind while setting a Standard, planning his actions, calculating effort expen-
diture, monitoring Performance, evaluating Performance outcome, weighing
causal attributions and administering seif-reward" (Heckhausen, 1974). Then the
students took turns doing the same thing, first speaking aloud and then silently
to themselves as they went through the achievement motive thought sequence.
There were two control groups of students, one of which received no training
and one, another kind of training that was announced as a "program for im-
proving motivation to learn," although the games and activities in it were irrele-
vant to motive change. This group was included to see whether any effects ob-
tained were due purely to Suggestion or expectation that something was being
done to improve the students' Performance.
After fourteen weeks, those trained in achievement motive practices had
gained in a number of ways over those in the other two groups. They showed a
larger gain in their h Success score as contrasted with their / Failure score, as
measured by a semiprojective test (Schmält, 1976) in which a person picks out
of several alternatives the one most likely to describe what is going on in a
picture. Their goal setting had become more realistic, and their level of aspira-
tion tended realistically to go up after success and down after failure. Also,
"seif-reward after success had gone up whereas there was no difference in self-
punishment after failure . . . last but not least, the gain on the scholastic
achievement test exceeded the other two groups." (Heckhausen, 1974). The only
area in which an expected shift did not occur involved causal attributions after
success and after failure, for which there were no differences among the groups
or from before to after the training.
Improvements in goal setting, causal attributions, and h Success were found
after such training in several studies, although none of these changes had any
long-term effects on school grades (Heckhausen & Krug, 1982). Heckhausen
(1980) interprets the results in terms of the cognitive characteristics of the
achievement motive discussed in Chapter 12. That is, he believes that what such
training does is help individuals define their own goals for Performance, which
leads to a realistic level of aspiration, particularly if the teacher aids in the pro-
cess by assigning tasks of appropriate difficulty for a particular Student. Of Spe-
cial importance is orienting the pupils to judging their Performance in terms of
their own past accomplishments rather than the accomplishments of others (that
is, using a social norm Standard for evaluating their achievements), for people
can improve over their past Performance even though they are still doing poorly
relative to others. All these procedures increase the student's sense of personal
responsibility for setting the goals in the first place, for the Performance that fol-
lows, for evaluating the Performance, and for deciding what to do next. So the
Student develops internal, as opposed to external, reasons for success and failure,
which in turn leads to increased valuing of the seif for having been the key fac-
tor in the whole achievement process.
All these factors involve cognitive processes that increase the perceived
probability of success. That is, if students understand that they are picking the
Standards for Performance in terms of which they will be judged; if, when they
learn how well they have done, they can see the outcome as a result of their
personal efforts; if they have control over what level of difficulty they will per-
form at next; and if they are in charge of rewarding themselves for performing
well, they will have higher self-esteem. And higher self-esteem translates into a
greater perceived probability of success in general, and also a greater per-
ceived probability of success in particular with respect to the tasks they are
working on.
Most of the effects of the courses Heckhausen (1980) reviews have been in
the cognitive area—that is, the changes observed have been in goal setting,
causal attribution, and self-confidence. Longer-range effects on Performance in
school or in life have seldom been obtained. Heckhausen and Krug (1982) con-
clude that the reason for the lack of effects of these changes on school perfor-
mance lies in the way school learning is structured:
Although children are more achievement motivated after the training program, they
are still academic low achievers. They would like to be successful, but they now find
that the school Situation does not lend itself to satisfying their need to achieve. It
may be possible to assume self-responsibility, but setting one's own goals is hardly
feasible. Furthermore, assignments are too difficult compared to one's Performance
Standard. Success, if any, is therefore still more likely to be attributed to chance than
effort, and there are few opportunities for self-reinforcement or for anticipating posi-
tive effects.
So the conclusion might be drawn that cognitive inputs produce cognitive

Outputs, but that the ultimate effect of changing the cognitions associated with
the achievement motive is not well documented. This type of motivation training
can be viewed as focusing primarily on the cognitive variables that influence
probability of success in the general formula governing the likelihood that a re-
sponse will occur (summarized in Figure 12.15).
ACHIEVEMENT MOTIVATION TRAINING

FOR ENTREPRENEURS
In the 1960s a motivation training program was developed that was aimed at in-
fluencing all the theoretical determinants of Performance—namely, motive
strength, perceived probability of success, and the incentive value of success. It
grew out of the research on the achievement motive (reviewed in Chapters 7
and 11), which showed that people high in n Achievement had many of the
characteristics needed for successful entrepreneurship and that successful entre-

preneurs or small businesspeople were in fact higher in n Achievement. Since it
had also been demonstrated that countries more oriented around n Achievement
in populär literature developed more rapidly economically, it seemed reasonable
to infer that this came about in part because there were more energetic entrepre-
neurs in such countries and that therefore one of the ways to promote economic
development in a country might be to strengthen the achievement motive among
its business entrepreneurs.
So training courses in achievement motivation were provided for small busi-
nessmen, particularly in India at the outset, because it was a country that was
underdeveloped and needed more successful entrepreneurs outside its great met-
ropolitan centers to create more Jobs for unemployed or underemployed workers
in smaller cities throughout the country (McClelland & Winter, 1969/1971).
After the courses had been successfully offered in a number of cities in India
and elsewhere, it was decided to try to have a major impact by this means on
economic development in a limited area.
Business activities in two small cities in southeast India were carefully moni-
tored for several years before, during, and after the business leaders in one of
the cities received achievement motivation training. The cities were chosen to be
as comparable as possible. They were about the same size, with populations of
around 100,000; the distribution of the work force in various categories was sim-
ilar; they consumed equivalent amounts of electricity per capita, indicating ap-
proximately equal levels of industrial development; and they were located in the
same State, Andhra Pradesh, so the people belonged to the same culture and
spoke the same language. One was a seaport and the other, a river port on a
major inland waterway. It was hoped that providing achievement motivation
training for the businessmen in one city, Kakinada, would stimulate them to
outperform the businessmen in the neighboring control city, Rajahmundry, pro-
ducing in the long run a significant difference in the level of economic develop-
ment of the two urban areas.
Through Speeches at local businessmen's clubs and other recruitment efforts,
some fifty-two businessmen from Kakinada were persuaded to leave their busi-
nesses and spend two weeks being trained in achievement motivation at the
Small Industries Extension Training Institute at Hyderabad, over two hundred
miles away. They were trained in four batches of ten to fourteen participants
each. The training was intensive, lasting ten or more hours a day, and it was
based on all that was known at the time about how to introduce personality
change. It was generally believed that it was difficult, if not impossible, to
change adults in any very fundamental way such as might be involved in in-
creasing their need to achieve. Certainly the research findings on the effects of
prolonged psychotherapy were not very encouraging. It typically involved many
more hours of person-to-person contact than was available in a short group
training course, and its effects in producing fundamental personality change
were in some doubt (Eysenck, 1952). However, religious groups like the Mor-
mons or the Jesuits seemed to be effective in producing personal change through
intensive training, so it was decided to make an all-out effort to produce motive
change in any way possible, drawing on the experience of religious groups and
on whatever scientific Information was available.
Training Inputs in Courses for Small Businessmen
Twelve different types of training inputs were devised (McClelland, 1965;

McClelland & Winter, 1969/1971). They can be classified according to which
determinant of Performance Output they should affect, as shown in Table 14.1.
Great emphasis was placed on increasing motive strength directly by learning
how to think, talk, and act like a person high in n Achievement. So they wrote
Table 14A.
TRAINING INPUTS IN COURSES FOR BUSINESS ENTREPRENEURS (after McClelland & Winter,
1969/1971)
Purpose Is to Increase
Incentive Value
Motive Strenth Perceived Probability of Being More
(n Achievement) of Success En trepreneurial
M X p. V
1. Learning the achievement 1. Prestige Suggestion that they 1. Realization that chosen
associative network (scoring could change and become occupation requires it
System) more entrepreneurial
2. Practicing such things as 2. Commitment to concrete 2. Realization that chosen
moderate goal setting in goals and plans (increased occupation fits in with life
action; getting turned on by effort) goals
Performance
3. Following behavior of 3. Keeping track of progress 3. Clarification of how this
attractive entrepreneurial toward goals; using feedback value fits in with other
modeis (vicarious to increase self-confidence values that might conflict
satisfaction) with it
4. Receiving warmth and 4. Retreat setting to dramatize
respect from trainers during importance of life change
confusion and failure
Gaining confidence from
experiences in course
exercises (see Item 2 in M
column)
Gaining confidence from 5. New reference group of
self-knowledge and from trainees to give reminders
knowing how to change and reinforcement after
training
imaginative stories to pictures, learned to score them for n Achievement, and

practiced writing stories that scored as high as possible in n Achievement. Thus,
they learned to think and talk in achievement terms all day long. They learned
to act like a person high in n Achievement by participating in various games in
which they had to set goals for themselves and observe whether they reached
them or not. In particular, they participated in a business Simulation game in
which they had to set production goals for themselves and either made or lost
money, depending on whether they attained the goals or not. They observed and
discussed their own behavior as compared with the behavior of others in the
same Situation and compared their Performance with the model of the person
high in n Achievement, who sets moderate goals, takes personal responsibility,
uses feedback to modify goals, and takes initiative to find new ways of achieving
goals. (See the summary in Chapter 7 of the characteristic behaviors of people
high in n Achievement.)
The purpose of these games was not only to give people experience in
action of behaving in particular ways, but also to provide the positive affect
from effective Performance that by definition is the basis on which the achieve-
ment motive is built. The excitement and enthusiasm generated by the various
group activities was considerable and could have been strong enough to help
forge new motivational associative networks. Finally, the participants were
exposed to attractive modeis of successful entrepreneurial behavior in the form
of visitors who had done well in business and who talked to the group about
what they had done. Thus, the trainees could observe for themselves that these
successful men in fact did think, talk, and act like the model of the person
with high n Achievement that they were attempting to emulate in their own
behavior.
Several training inputs should have affected the perceived probability of suc-
cess determinant of Performance outcome. Every effort was made to convince
the businessmen that they could and should change and become more entrepre-
neurial and successful. Prestige Suggestion has been found by psychologists over
and over again to be an important means of introducing attitude change (see
Hovland, Janis, & Kelley, 1953). It lies behind such diverse phenomena as the
placebo effect (Benson & Epstein, 1975), in which a patient gets better just from
being given a "sugar pill" by a medical authority, and the Hawthorne effect, in
which workers in the Hawthorne plant of Western Electric performed better
whenever management introduced change, even if the change was back to a pre-
vious way of doing things (Roethlisberger & Dickson, 1947). So efforts were
made to build the prestige of those offering the training by stressing the fact
that they represented a world-famous university, that the training was based on
carefully collected scientific evidence, and that it had the backing of an impor-
tant government-sponsored institution and the Ford Foundation. All these in-
puts were designed to counteract the prevailing, rather low level of perceived
probability of success, since the businessmen were gloomy about their own pros-
pects and about the possibility of economic development in India as a whole.
Of equal importance was the emphasis at the end of the program on getting
the businessmen to commit themselves to concrete goals for business activities
that were so specific that they could e'asily keep track over the coming months
of how well they were achieving them. Most of them were used to setting
goals in general terms like "When I get back I am certainly going to work
harder." But they learned that such a goal is not acceptable to a person high in
n Achievement, because it is not specific: there would be no way to teil whether
they had reached it or not, and it is not specific as to when it would be
achieved. They learned to set more appropriate goals like "Monday after I get
back I am going to work out a careful plan showing how the purchase of an
air conditioner for my sari shop should increase turnover and profits over a six-
month period so that I can pay off the loan to purchase the air conditioner. I
am going to take this plan to the Bank of India the next day to get a loan of so
many rupees. I expect to get an answer on the loan within one week, to pur-
chase the air conditioner and install it within another week, and to track the in-
creased business over the next six months so that I can calculate at the end of
the time whether or not the investment has been worth it." Such plans not only
improve business practice; they also create more confidence in people that they
know exactly what they are doing and are therefore more likely to succeed.
In the course of the businessmen's learning what their motives were, as re-
vealed in the stories they wrote and as shown in their behavior in discussion or
in the business games, many received unpleasant surprises as they realized they
had not been behaving in ways consistent with the model of successful entrepre-
neurial behavior. Self-knowledge is an obvious first Step to change, but it can in-
volve strong feelings of failure and confusion. Thus, it was considered very im-
portant to generate an atmosphere of warmth and respect for the individual
during the training, which helped people remain confident as they underwent
the necessary restructuring of their thinking about themselves and their lives. In
other words, the trainers were behaving like the experimenters in the studies re-
ported earlier by Heckhausen (1980) and Stamps (1973): they were diminishing
fear of failure by reacting to failure either neutrally or as if it were normal
under the circumstances, stressing the importance of renewed effort to change
the Situation, and giving all kinds of positive reinforcement for efforts to under-
stand oneself and to change. As noted at the bottom of the middle column in
Table 14.1, self-confidence or the perceived probability of success was also pro-
moted by the experience the men were gaining in performing more successfully
in various games throughout the course, as well as by the growing realization
that they were understanding themselves better all the time and knew how to do
specific things that would change their business behavior for the better. For it
increases self-confidence to realize that you know something other people do not
know about how to get what you want.
The third set of training inputs affected primarily the incentive value of
being more entrepreneurial. Most of the men in the course were already in busi-
ness, so it was easy to explain to them that if they meant to do well in their
chosen occupation, they needed to behave more along the specific lines outlined
in the course. The incentive value of being entrepreneurial was bolstered by the
simple argument that their occupation required it. If they wanted to do some-
thing eise, such as be a teacher, an artist, or a sadhu (a religious renunciate),
being more entrepreneurial would not be so important; since they had decided
to be businessmen, however, it obviously was very important to them.
Raising the question of their occupational goals inevitably stimulated them

to think about what their general goals in life were. In one exercise, for exam-
ple, they were asked to go off by themselves, meditate for an hour, and prepare
an inscription they would like to see on their tombstone. This led to a discus-
sion of what each man valued most in life and how that fitted in with the gen-
eral values held by others around him. The purpose of considering these matters
was to see how the incentive value of being more entrepreneurial fitted or failed
to fit in with other goals the person might have. For example, a man might
have written on his tombstone that he wanted to be remembered as "a generous
family man." Discussion might reveal that he often feit a conflict between
spending time at work and with his family. Thus, what he might do in the
course of discussing this conflict was figure out how his work might contribute
to his desire to be viewed as a generous family man. In another instance, a man
might realize that he had failed to take some initiative in business, because in
India many families operate under the Joint family System, in which the oldest
competent male—usually the father—is definitely in charge. Thus, it might have
seemed completely inappropriate to the man under the prevailing System of In-
dian values to take an initiative that clearly belonged to another in the family.
The purpose of these training inputs was to embed the value of being more en-
trepreneurial firmly within a network of supporting—or at least nonconflicting
—values relating to life goals and the norms of Indian culture.
Requiring that the men leave their home towns for training in a dormitory
setting in another place helped to dramatize the importance of the life change
the courses were supposed to produce. The effectiveness of religious retreats is
based in part on the fact that people withdraw from everyday pursuits and sur-
roundings and seek to create a new self-image or identity, which can be carried
back to the old setting. Furthermore, being away with others who are undergo-
ing the same emotional experiences creates a new reference group, and there is
considerable evidence that perceived membership in a new reference group helps
attitude change to persist (see Berelson & Steiner, 1964). The men in these
courses came to know each other very well, since they spent over one hundred
hours together discussing their most intimate thoughts and plans. They came to
realize that the future economic progress of their Community depended on them
and, more specifically, on the plans each of them had made for increased busi-
ness activity. They also knew what each member of the group had committed
himself to do when he returned, so they would obviously be in a position to
know whether he had done what he said he was going to do or not. All of this
served to underline for every participant the great importance of being more en-
trepreneurial when he returned to contribute to Community development and to
earn the approval of his many newfound friends, who were part of the same for-
ward movement.
Changes in Business Activity After Motivation Training in India

What were the eifects of these intensive efforts to increase achievement motiva-
tion and improve entrepreneurial Performance? Since the men had been taught
the scoring System for n Achievement, it was not possible to determine directly
in the usual way whether their achievement motivation had increased, as in the
Heckhausen training courses reviewed earlier, which had not employed the TAT
as a training device. So attention was focused on developing an index of entre-
preneurial behavior or business activity that could apply equally well to people
in any line of work (McClelland & Winter, 1969/1971). The men were ques-
tioned about all aspects of their business activity during the previous two years
at the time they started the training and again six, twelve, eighteen, and twenty-
four months after the training was completed. Efforts were made to get Informa-
tion on capital invested, monthly business income, profits, and number of em-
ployees, although it was very difficult to get accurate economic figures from
many of them, partly because of their reluctance to provide it for fear that tax
collectors might find out about it, and partly because many of the businessmen
were not very good at keeping such records.
In the end, a very simple coding System was devised for assessing the level
of business activity. If, by all signs, business activity had proceeded at about the
same pace for the past two years, the subject received a 0. If his business had
noticeably declined in income or profitability or if he had gone out of business,
he received a — 1. He received a -hl if he had undertaken some specific plans to
improve his business, such as taking a course in accounting, planning a new
plant, joining a voluntary business organization, or going to work more often or
earlier. A score of + 2 was given if he had actually started a new business, been
responsible for an unusual increase in the firm's income or profitability level, or
received an unusual promotion in the firm. To receive a + 2 , a man had to have
done something that was clearly visible and obvious to all as representing an im-
provement in business activity, and that involved a clear financial change. Every
attempt was made to check carefully on Statements made in interviews about
changes. Thus, if a man said he had built a new warehouse, the interviewer
would go to make sure it was really there and had been built since the course; if
a man said he had hired three new employees to handle increased business, the
interviewer would make sure the employees were really present when he visited
the man's place of business.
Achievement motivation training was effective in increasing the level of
business activity measured in this way, but the effect was much larger for those
who had the opportunity to change their behavior. The men were divided into
those who were in charge of their businesses and those who worked for someone
eise. For seventy men in charge who were trained in achievement motivation
from various cities in India (including especially Kakinada), the average business
activity score in the two years after the training was 1.39, in contrast to an av-
erage score of 0.57 for sixty businessmen in charge from the same or similar
cities in India (including especially Rajahmundry) who were not trained in
achievement motivation. The difference is highly significant. However, the com-
parable average score for men not in charge who were trained was 0.71, in con-
trast to an average score of 0.38 for those not trained and not in charge, a dif-
ference that does not reach accepted levels of significance. In other words, a
change in motivational intent is unlikely to lead to a change in actual behavior
unless the opportunity is present, as Figures 6.1 and 12.15 make clear.
For those who were in charge, an effort was made to see whether any par-
ticular combination of training inputs was more effective than any other combi-
nation. That is, some of the training inputs listed in Table 14.1 were not present
in some training courses given in other cities in India. For example, some were
given without having a retreat setting or without the extensive follow-up every
six months for two years that was carried out for the Kakinada trainees. A total
training input score for each course was obtained by assigning a value of 0 for
not present, 1 for partly present, or 2 for fully present for each of the twelve
training inputs, so the maximum total score would be 24.
Table 14.2 shows the average business activity level scores in the two years
before and the two years after training for a representative sample of courses
varying in strength of training inputs. The average level of business activity in
the two years before the course was about the same for all groups compared.
Any kind of achievement motivation training increased business activity levels in
the two years after the course as compared with the average level for untrained
Indian businessmen in the same two-year period. Beyond this, it appears that
the more training inputs that were present in the course, the greater was the in-
crease in activity level afterward. Clearly, the Hyderabad course was less effec-
tive than the Kakinada courses, and the Bombay and Barcelona courses feil
somewhere in between. The more training inputs that were fully present, the
greater was the increase in business activity levels after the training.
Beyond this, we cannot draw any conclusions from the data as to whether
particular inputs were essential or more effective than others. No courses were
run without the training inputs designed to increase motive strength and focus-
ing, for example, just on the cognitive training inputs designed to increase per-
ceived probability of success, as in the Heckhausen studies previously referred
Table 14.2.
BUSINESS ACTIVITY LEVELS OF BUSINESSMEN WITH AND WITHOUT ACHIEVEMENT
MOTIVATION TRAINING, FOR MEN IN CHARGE ONLY (after McClelland & Winter, 1969/1971)
Mean Activity Scores

for Two- Year Period
Percent rated +2
Amount of Before After After
Group Training Input N Training Training Training
Untrained Indian businessmen (including 0.0 60 0.58 0.57 28
those in Rajahmundry)
Training courses in:
Hyderabad, India 10.0 10 0.60 1.06 40
Bombay, India 15.5 18 0.65 1.39 67
Barcelona, Spain 20.5 19 0.53 1.37 58
Kakinada, India 22.5 21 0.35 1.61 81
to. However, these data also show that the result is not just dependent on some
general prestige Suggestion effect, for if it were, the men in the Hyderabad
course should have gained as much in activity level as those receiving a fuller
set of training inputs in the Kakinada courses.
Other measures of the effectiveness of the training were employed. One of
the most significant ones was the number of new Jobs created per entrepreneur
before and after the training, because it could be determined with accuracy and
because it represented the contribution the training was making to employment,
which was a matter of considerable interest to the Indian government. Figure
14.2 shows that in the two-year period before the training in the two cities di-
rectly compared (Kakinada and Rajahmundry), on the average an entrepreneur
had created employment for between one and two persons. In the two years
after the training had taken place for businessmen in Kakinada, the picture was
quite different. The entrepreneurs trained in achievement motivation in Kaki-
nada had hired, on the average, twice as many new employees as the untrained
entrepreneurs in the control city of Rajahmundry. The proportion of firms em-
ploying more people afterward than before is significantly higher for the trained
than the untrained businessmen.
However, we could argue that this effect is not contributing to overall em-
Kakinada
8.0 {City Where Training Took Place)
Rajahmundry (Control City)
6.0
4.0
2.0
2 Years Before 2 Years After

Training Training in Kakinada
Figure 14.2.
Increased Jobs Created by Entrepreneurs Trained and Untrained in Achievement Motivation (after
McClelland & Winter, 1969/1971).
ployment or economic development. Rather, is it not possible that the trained

businessmen are simply taking business away from their less-motivated competi-
tors? If a man adds an air conditioner to his sari shop, he may do more busi-
ness, but shop owners without air conditioners may do less business, so there is
no net gain in employment. Case studies suggested that this was not all that
happened, because a number of new firms were created where none had existed
before. For example, one man created a trucking Company, and it got fresh veg-
etables to market that had simply spoiled previously.
Furthermore, the statistics on overall employment in the two cities show
that the efforts of the trained entrepreneurs affected the economy of the area as
a whole, as Figure 14.3 shows. When the training was started, the two cities
were closely matched in the total number of employees in firms employing ten
or more people. Shortly thereafter, there was a general economic depression that
affected the east coast of India, which resulted in a significant drop in employ-
ment in both cities by the spring of 1967. However, the drop was much larger
in Rajahmundry, where no training had taken place, and no recovery in overall
21,000 Kakinada (Where Training Took Place)

o
o
•Q. 19,000
17,000
Q.
6
Rajahmundry (No Training)
fc 15,000
X>
3
Z
Training Introduced
I ! J_
March 1964 March 1967 March 1970
Time
Figure 14.3.
Total Employment over Time in Two Indian Cities Where Entrepreneurial Training Did and Did
Not Take Place (after McClelland & Winter, 1969/1971).
employment had occurred in that city by the spring of 1970. However, in Kaki-
nada, where fifty-two key businessmen had been given achievement motivation
training, employment rose again considerably so that by 1970 it was somewhat
higher than six years earlier. Case studies documented just how and why this in-
crease in employment had taken place as a result of the increased efforts of the
businessmen who had undergone motivation training.
McClelland and Winter (1969/1971) also make the point that the cost per
new job created through entrepreneurial training is one twentieth of the cost of
creating a new job by other methods the government of India had employed.
For example, in a government loan program to aid small industry in Andhra
Pradesh, actual figures show that the cost per new job created was over 3500 ru-
pees, whereas the cost of motivation training to produce a new job was esti-
mated to be around 183 rupees. Since in terms of local purchasing, a rupee was
roughly the equivalent of a dollar at that time, it is clear that motivation train-
ing is a comparatively much less expensive way to create employment. Obvious-
ly, the application of a little carefully validated psychological knowledge can
have far-reaching economic effects.
Factors Affecting the Success of Entrepreneurial Training in India

McClelland and Winter (1969/1971) were interested in finding out what psycho-
logical factors might be responsible for the fact that the training was effective
for some participants and not others, so they divided the businessmen into those
who became active after the course and those who were inactive and compared
them on a number of variables. Their flrst hypothesis was that traditional Indian
values might have prevented some men from believing that they could or should
be more active and successful. However, they found no significant differences in
the way "actives" and "inactives" responded to an extensive questionnaire cov-
ering measures of Cautious Fatalism, Respect for Powerful Others, Conformity
to Caste Rules, and Submissive Conflict Avoidance. For example, it was thought
that inactives might be less willing to take initiative because they subscribed
more to sentiments like "It is better to follow the leadership of wise eiders than
to do what one thinks is best." However, all the businessmen tended to agree
somewhat with this Statement, and those who changed after the course agreed
with it as much as those who did not change. On the whole, the businessmen
were not particularly fatalistic or tradition bound, but within these limits the ac-
tives were as traditional as the inactives. For example, less than half of the men
said they observed Rahukal (a period during the day when they did not conduct
any business), but 50 percent of the actives said they followed this tradition as
compared with only 20 percent of the inactives. This is even the opposite of
what might be expected, although overall there was no significant difference in
how much the actives and inactives followed tradition.
One psychological variable that predicted who would not change was the
n Power score, as obtained from TATs administered before and after the course
and two to three years later. For those in charge who remained inactive, the av-
erage n Power score increased from 3.2 before training to 6.8 two to three years
later, a highly significant difference. Among the men who were also in charge
but became active, the increase in n Power score was small and insignificant. A
careful analysis showed that those in charge who remained inactive gained more
in n Power than they did in n Achievement, whereas the reverse was true of
those in charge who became active. In other words, it looks as if the achieve-
ment motivation training had led to an increased desire to have impact or to be
recognized as a success for some men, which does not translate into taking the
entrepreneurial, moderate-risk Steps necessary to improve business Performance.
It had been expected that the n Achievement scores might also remain
higher two to three years after the course for those who were active as com-
pared to those who were inactive, on the grounds that the achievement-related
associative network should have remained more salient for those whose behavior
showed they were more guided by it. However, results did not confirm this hy-
pothesis. The average n Achievement score increased greatly right after training,
as would be expected, since the businessmen had been trained specifically to
write stories high in n Achievement. When the men were retested two to three
years later, the average score had declined considerably, showing forgetting, but
there were no differences at any testing point in the average n Achievement
scores of the actives and inactives.
Later Heckhausen (1971, 1980) scored the protocols for his measures of
h Success and/Failure, in which the men had not received specific instruction.
He found that for the men in charge, those who were active had declined in
/ Failure from before the course to two or three years later, whereas those who
were inactive had increased in / Failure. The result is shown most clearly in the
Net Hope score, which subtracts the / Failure score from the h Success score,
as shown in Figure 14.4. On the left side of the figure the results for the trained
Indian businessmen show that the Net Hope score rose more steeply for those
who became active than it did for those who remained inactive. On the right
side of the figure are some comparable results obtained by Varga (1977) in a
similar evaluation of the effects of achievement motivation training courses in
Indonesia, Pakistan, Iran, and Poland. Once again, those who became active
after training gained much more in Net Hope than those who stayed inactive.
Heckhausen points out that the difficulty with the Indian results is that it is
impossible to know whether lower Net Hope causes inactivity or inactivity
causes lower Net Hope, because the retesting was done so much later that the
participants would by then have known whether they were going to be success-
ful or not in improving their business. On the other hand, Varga's data are not
as open to the same criticism, since the retesting occurred earlier, at a time
when the businesspeople would be less certain about how well they were doing.
Thus, the most likely conclusion is that achievement motivation training is most
effective for increasing business Performance for those in whom it succeeds in
increasing the Net Hope score.
A somewhat similar result was obtained by applying the scoring System for
"efficacy" developed by Pizer (in McClelland & Winter, 1969/1971) to differen-
tiate the story content of those who were active or inactive after an achievement
Z 2 -
Before 2 to 3 Years Before 1 Year

Training Later Training Later
Time Time
Figure 14.4.
Changes in Values for Net Hope Score Between the First Measure Taken Before Training and Later
Measures for Participants Who Were Active or Inactive After Training. (Left) Reanalysis of data
from the Indian training (McClelland & Winter, 1969/1971) by Heckhausen (1971). (Right)
Evaluation of courses sponsored by the United Nations Industrial Development Organization
(UNIDO) in Indonesia, Pakistan, Iran, and Poland by Varga (1977) (after Heckhausen & Krug,
1982).
training course in Bombay. It scored for whether the main character in the
story was doing something or being passive, for whether the hero relied on in-
ternal or external resources, and for whether he or she solved or avoided prob-
lems. In the stories obtained two to three years after training in Kakinada, it
was found that those who had become active had gained much more in efficacy
scores as contrasted with those who had remained inactive. In other words, a
positive problem-solving approach in the stories after the training—whether it is
defined as Net Hope or efficacy—is significantly associated with becoming more
active. What is also of interest is that the Pizer efficacy score is very similar in
conception to the deCharms Origin-Pawn score, which, as we shall see in a mo-
ment, has been found to be related to better Performance in school after motiva-
tion training.
Entrepreneurial Training for Minority Businesspeople in the United States

Similar achievement motivation training courses have been offered to a number
of different groups of businesspeople in the United States, particularly those in
underdeveloped areas or from minority groups. As compared with untrained
people, a much higher percentage of trained white businesspeople in McAlester,
Oklahoma, or trained black businesspeople in Washington, D.C., showed six or

more concrete signs of increased business activity six months after training.
Also, the number of new employees hired among the trained businesspeople was
approximately double what it was among the untrained businesspeople, in a pat-
tern very similar to what was discovered in India (McClelland & Winter,
1969/1971).
Figure 14.5 summarizes the effects of achievement motivation training for
black and Hispanic businesspeople in nine different cities across the United
States. In this study it was possible to get better estimates of the financial condi-
tion of the firms than in previous research. As the comparisons in Figure 14.5
show, training increased significantly monthly sales, personal income, and prof-
its. In presenting these data, Miron and McClelland (1979) point out that the
gains are much larger than could be attributed to general improvement in busi-
ness conditions at the time, as measured by similar statistics for the country as a
whole or for the regions in which the training took place. Overall, the evidence
is persuasive that achievement motivation training improves the Performance of
small businesspeople, as it ought to according to theory.
Before Training, January 1971-June 1972

After Training, January 1973-June 1974
e $7000
c
c
? 6000
| 5000
s
| 4000
3000
2000 -
1000 -
Monthly Sales
i£L
Personal Income Profits
Figure 14.5.
Eflfects of Achievement Motivation Training for Sixty-seven Minority Entrepreneurs in Nine U.S.
Cities on Monthly Sales, Personal Income, and Profits (after Miron & McClelland, 1979).
• ACHIEVEMENT MOTIVATION TRAINING IN SCHOOLS
It was natural to wonder whether achievement motivation training would im-

prove academic Performance, although on theoretical grounds it was not alto-
gether clear why it should. For higher n Achievement should lead to more en-
trepreneurial activities—setting moderate goals for oneself, taking initiative, and
so on—but it was not at all clear that such activity would lead to better grades.
On the contrary, most teachers set the goals for their students and require that
the students do the lessons as they are assigned rather than take initiative to do
things in their own way, as Heckhausen and Krug (1982) point out. Further-
more, n Achievement scores have seldom been shown to be related to getting
better grades in school (see Chapters 6 and 7). The Heckhausen studies referred
to earlier made more sense theoretically, since they were designed primarily to
increase self-confidence or perceived probability of success rather than achieve-
ment motive strength per se.
Nevertheless, attempts were made to increase achievement motivation in
schoolchildren in the United States (Alschuler, 1973) and India (Mehta, 1969)
by adapting the courses used to train businesspeople for application to the
school Situation. In the United States, intensive courses in n Achievement were
given in Special classes for children run by outside experts rather than their
usual teachers (McClelland, 1972). The goal, as in the case of the entrepreneur-
ial training for businesspeople, was to create the maximum pressure for personal
change, using most of the training inputs outlined in Table 14.1. That is, the
children were taught the scoring System for n Achievement and practiced vari-
ous goal-setting games so that they could learn to think, talk, and act like a per-
son high in n Achievement. Efforts were made to dramatize the training by seg-
regating it from regulär schoolwork so that they could really believe they could
perform better in all sorts of ways, including getting better grades in school.
This kind of training has not generally improved academic Performance, al-
though it does appear to increase entrepreneurial activities outside of school
(Alschuler, 1973). Mehta (1969) reports that such training in Indian schools was
sometimes effective in improving academic Performance. Whenever effects were
found, they tended to be larger for boys and for Performance in quantitative
studies like math and science.
Some typical results are summarized in Figure 14.6. In this study, some
eighth-graders in San Mateo County, California, received achievement motiva-
tion training on four Weekends either at school or at a camp in the mountains
(Ryals, cited in McClelland, 1972). The question of interest was whether stu-
dents' grades had gained more or less than would be expected based on their
grades in the same subject the previous year. In the measure utilized, students
who got just as high a grade as would have been expected from their grade the
previous year would end up with a score of 50; if they did better than would
have been expected, they would get a score above 50; and if they did worse,
they would get a score below 50. As Figure 14.6 shows, those who received
achievement motivation training either at school or in the camp did not get bet-
JControl Students
\n Achievement Training at School on Weekends

n Achievement Training
D in a Camp on Weekends
60.0
QQ ™
'S °°
5 a 55.0
st 50.0
45.0
Grade Point Average Math Grades
Figure 14.6.
Effects of Achievement Motivation Training on Grades Relative to What Is Expected Based on Past
Performance for Eighth-Graders (after Ryals, cited in McClelland, 1972).
ter grades overall than those who had received no such training. However, they
did get significantly higher math grades, and the gains were greater for those
who had had achievement motivation training in a retreat setting designed to in-
crease the incentive value of change, as shown in Table 14.1. The results could
not be due to the fact that teachers would assign higher grades to those they
knew had received motivation training, because the same result appeared for an
objective test of science knowledge. These courses also differed from earlier ones
that were less effective in that they were given by the students' own teachers
rather than by Outsiders.
Using the students' own teachers to provide the motivation training has the
advantage of capitalizing on the Pygmalion effect. The teachers believed the
training would increase motivation, and they should therefore have begun to
react to the students as if they were better motivated and more capable. This, in
turn, should have increased the students' perceived probability of success, which
should have facilitated their Performance.
As one might expect, this effect is magnified if the motivation training is
spread out through the school year and integrated with regulär classroom in-
struction. In Berea, Ohio, nearly all the teachers in the Smith elementary school
were given instruction in how to train students in achievement motivation, so
the whole school was oriented more toward motivating pupils (McClelland,
1979a). The teachers implemented the general principles of achievement motiva-

tion training in various ways in their classrooms, using curricular materials de-
signed to teach the children the achievement problem-solving sequence in
thought and in action through games.
All the children in the sixth grade of Smith received achievement motiva-
tion training and then graduated to the Roehm junior high school in another
part of the city. There they entered the seventh grade with graduates of a com-
parable elementary school (Fairwood) who had not received achievement moti-
vation training. It was unlikely that the teachers in the junior high school paid
much attention to which elementary school their pupils had come from or that
they knew that some of their students had received achievement motivation
training, so the grades they assigned the seventh-grade students should be rela-
tively unbiased reports of how well the students were performing. All the pupils
had taken the Ohio Academic Ability Test, so it was possible to develop an
equation to predict from their scores on this test the grade point average they
should be receiving in the seventh grade. It was determined on the basis of this
prediction equation whether a Student was doing better or worse than expected.
As the formula for predicting behavior states, Performance should be a function
of M X Ps X V. In this instance, the Ps variable is represented by the aca-
demic skills measured by the Ohio Academic Ability Test. To test for the effects
of increases in motive strength, the influence of ability on Performance is re-
moved by simply noting whether students are performing better or worse than
would be expected based on their ability.
As Figure 14.7 makes clear, a much higher percentage of the students who
had received achievement motivation training were academic overachievers in
seventh grade. As compared with pupils from Fairwood, motivation training at
Smith increased the proportion of overachievers among students of high, medi-
um, or low academic ability. Overall, 67 percent of the Smith children who had
received achievement motivation training were overachievers in the seventh
grade, as compared with less than half (44 percent) of the Fairwood children
who had not received achievement motivation training (McClelland, 1979a). The
difference is highly significant statistically. What is particularly interesting about
this result is that the training changed the children in ways that carried over to
the next school year in a different school with teachers who had received no
Special instruction in the field of motivation. So their better Performance could
no longer be attnbuted to teacher expectations that they were abler and would
do better.
ORIGIN TRAINING IN THE CLASSROOM

deCharms developed a variant of achievement motivation training that included
his Origin-Pawn concept. In his terms,
An Origin is a person who feels that he is in control of his fate; he feels that the
cause for his behavior is within himself. A Pawn feels that he is pushed around, that
someone eise pulls the strings and he is the puppet. He feels the locus of causality
100
D Untrained (Did Not Receive n Achievement Training

in Sixth Grade)
80
D Received n Achievement Training in Sixth Grade
6©
•S 60
§1
'S -o 40
S 20
/V= 23 11 24 19 19 16
Ability: High Medium Low
(80-100 Percentile) (50-79 Percentile) (1-49 Percentile)
Ohio Academic Ability Test
Figure 14.7.
Percentage of Academic Overachievers in Seventh Grade as a Function of Receiving or Not
Receiving Achievement Motivation Training in Sixth Grade (after McClelland, 1979a).
for his behavior is external to himself. The motivational effects of these two personal
states are extremely important. The Origin is positively motivated, optimistic,
confident, accepting of challenge. The Pawn is negatively motivated, defensive,
irresolute, avoidant of challenge. The Origin feels potent; the Pawn feels powerless.
(deCharms, 1976)
He relates the Origin concept to the feeling of personal causation, which is re-
lated to Robert White's concept of effectance. As Chapters 5 and 8 pointed out,
the power motive probably derives from a sense of effectance, from having im-
pact, or from feeling potent. So one could argue that deCharms introduced some
aspects of power motivation into what he called Origin training.
The training was tried out in a black inner city school district in St. Louis,
where many of the children were underachievers and feit like Pawns. A number
of sixth-grade teachers from the district participated in a five-day residential
training Workshop very much like one used to train entrepreneurs (see Table
14.1). They learned about n Achievement, n Power, and n Affiliation both in
theory and in practice; they also were introduced to the Origin-Pawn concept.
They experienced Origin-Pawn feelings by playing a game in which they first
could construct anything they liked out of Tinker Toys and then were forced to
use the same materials in ways precisely and assertively ordered by the instruc-
tor. The instructions were purposely exasperating and even contradictory: 'Take
a spool, connect it to a red rod on the round side of the spool. I did not say to
put the assembly down. Put the assembly down now. Pick up the assembly.
Connect another round spool to the other side of the red rod" (deCharms,
1976). The teachers got some feeling of what it was like to be a Pawn and real-
ized that they could readily create these feelings in their pupils through overly
conscientious instruction. So the overall goal of the Workshop was not only to
inform the teachers on how to develop achievement motivation, but also to help
them learn how to treat their pupils like Origins.
deCharms also feit that the teachers should be treated as Origins, so they
decided themselves at the end of a Workshop how to utilize the information they
had been given to set specific goals for changes in what they did in their class-
rooms in the Coming academic year. They decided to focus their efforts in four
areas. The first was achievement thinking. For example, the children wrote ten
weekly essays entitled "My Stories of Achievement." The second area was self-
concept. The students examined themselves by responding to promptings such as
"My favorite daydream . . ." or 4Tf I had three wishes . . ."
Goal setting was the third area. The students were taught how to get
pleasure from moderate goal setting by participating in a version of the old-
fashioned spelling bee modified to take achievement motivation concepts into ac-
count. Modern teachers and students do not like the spelling bee very much, be-
cause poor spellers consistently fail in public and drag their team's score down.
The students' Performance is evaluated in reference to a social norm—how well
others spell. But deCharms altered the procedure so that Performance could be
evaluated in terms of a self-experienced norm—how well the Student had spelled
previously, as in some of Heckhausen's studies. The teacher determined which
words each child could spell at the beginning of each week by testing him or
her for the words to be learned that week. Then, as children's turns came in the
spelling bee, they were asked whether they wanted to try to spell an easy, mod-
erately difficult, or very difficult word for them, for which they would receive
one, two, or three points for success, depending on the word's difficulty for
them. This put the poor spellers and the good spellers on the same basis in
terms of potentially contributing points to their team's score. Most of the chil-
dren learned the advantages of picking moderately difficult words to spell, and
spelling improved greatly as a result.
The fourth area in which the teachers decided to focus their efforts was
learning to act like an Origin:
The children were told that an Origin is someone who takes personal responsibility,
prepares his work carefully, plans his life to help him reach his goals, practices his
skills, persists in his work, has patience, for he knows that some goals take time in
reaching, performs—he knows he has to do things in order to reach his goals—
checks his progress—i.e., uses feedback—and moves toward perfecting his skills, pay-
ing Special attention to improvement. (deCharms, 1976)
The teachers practiced behaving like Origins by setting goals for schoolwork and
for the weekend, checking on progress, analyzing what they could or could not
do, examining the importance of goals in terms of their self-concept, and so on.
Much of this training overlaps with training in the achievement motivation
problem-solving thought sequence, as summarized in Figure 6.6.
To determine the effects of ten weeks of Origin training, deCharms em-
ployed the usual measures of motive strength and moderate goal setting; he also
designed a Special coding System for stories, which measured whether the chil-
dren were thinking like an Origin. He was led to do this partly because more
traditional self-report measures of whether the children feit they were internally
or externally controlled did not sensitively pick up the effects of Origin training.
Despite the repeated emphasis in the training on taking personal control of
what they did, those children who had received the training did not score any
higher on Rotter's Measure of Internal Control (1966) than those who had not
received the training, nor, for that matter, was there a consistent increase in av-
erage n Achievement scores among the trained as compared with the untrained
students. Boys did show a significant increase in n Achievement scores after
training in the sixth grade, but girls did not, and by the seventh grade the boys'
average n Achievement score had dropped back to about what it was at the end
of the fifth grade. The training did produce an increase in moderate goal set-
ting—in picking what level of difficulty to work at among arithmetic tasks—but
the effect was significant only for "externals" (those who, on the Rotter scale,
said that the way things turned out depended on outside factors beyond their
control). They had tended before the training to pick the most difficult items to
work on, and after the training they shifted much more often to choosing mod-
erately difficult tasks.
However, the main effects of Origin training were on the Origin thinking
score and on Performance on achievement tests. As the graphs in Figure 14.8
show, the mean Origin score rose directly after motivation training, whether it
was in the sixth or seventh grade. The higher average Origin score of the
trained students feil off somewhat in the eighth grade and still further by the
eleventh grade, when no training had occurred since the seventh grade, but it
was still significantly higher than for the untrained students (deCharms, 1980).
The Origin scoring System gives points for references in the stories to internal
control, internal goal setting and choice of acts to pursue the goal, reality per-
ception, personal responsibility, and self-confidence.
For example, the Origin score of the following story is very high:
The boy is thinking of a present to get his mother.
The boy got three job no matter how long and hard to get the money.
he wants her the best present for her.
He will get her something very meanfull and lovely to her. (deCharms, 1980)
o
c
'S)
o
c
Grade Grade
Sixth- and Seventh-Grade Training Seventh-Grade Training Only
o o
C/3
c
6
c
5 6 7
Grade Grade
Sixth-Grade Training Only No Training (Controls)
« Training
•— — — • No Training
Mean Origin Score Before and After Motivational Training (after deCharms, 1976).
The story is scored for internal control, because there is a completely internally
motivated decision, approach, and solution to the problem. The boy internally
sets his own goal, takes the instrumental activities necessary to achieve it ("three
Job"), shows perception of reality in that he has no money, shows personal re-
sponsibility for persistence ("no matter how long and hard to get the money"),
and shows concern for others ("wants her the best present"); self-confidence is
expressed in the last sentence.
The boy acts as an Origin throughout and ends up doing something that
has an emotional impact on his mother, which means the story also contains
power imagery. But the power thinking is definitely of the socialized type, and
one might expect, because of the emphasis on personal responsibility for others,
that the Origin score would correlate with the socialized power (s Power) score
(see Chapter 8). In fact, this is the case. Wittcoif (1980) has shown that Origin
training increases the s Power score and decreases the personalized power
(p Power) score from the fifth to the sixth grade. The s Power score, it will
be recalled from Chapter 8, is associated with high Activity Inhibition and
p Power, with low Activity Inhibition. In a sample of twenty-one pupils in
deCharms's study who were high in Activity Inhibition, the correlation between
the n Power and Origin scores is .36 (p < .10). Among an equal number of pu-
pils low in Activity Inhibition, the correlation between these two measures is
only .12. Thus, there is reason to believe that the Origin score taps something of
the same characteristic measured by the s Power score. Both emphasize having a
responsible impact on others.
The effect of Origin training on academic achievement is marked, as Figure 14.9
[
0 -
[^\Trained (N= 83)
-5 ~ Trained •
o
o ^MJntrained (/V=»64)
-10
I
1
5
1 l
Grade
Figure 14.9.
Mean Discrepancy from Grade Placement (Iowa Test of Basic Skills) for Pupils Given or Not Given
Origin Training in Sixth and Seventh Grades (after deCharms, 1980).
makes evident. The measure is based on the Iowa Test of Basic Skills, which is
designed to show whether children are achieving at the level expected for their
grade placement. The untrained students in this district feil farther and farther
behind, until by the seventh grade they were, on the average, ten months below
the achievement expected at that grade. In contrast, those who received Origin
training in the sixth and seventh grade reversed this trend (deCharms, 1980). By
the end of the eighth grade—one year after they stopped receiving training
—they were, on the average, achieving at the level expected of that grade.
deCharms (1980) also followed up the trained and untrained students some
years later to determine what proportion graduated from high school. As Figure
14.10 makes clear, the percentage of pupils graduating from high school tends
to be quite low in this district—particularly among boys, only a little over 20
percent of whom graduate from high school if they have not had Origin training
(0 years of Origin training in Figure 14.10). Higher percentages of girls graduate
from high school, and Origin training does not significantly increase this figure.
However, significantly more of the trained boys went on to high-school gradua-
tion. Again, this effect is notable because it occurred some five years after the
training intervention had taken place. It is also cost effective, since an inexpen-
sive, short-term training input resulted in a gain in the proportion of high-school
graduates, which should have a significant impact on the life-long earning
streams of those who graduate.
| [Girls
70
^jßoys
1 60
50 - —
o
1
40 -
X
O 30 —
5 20 -
Du
10 -
1
Years of Origin Training
Figure 14.10.
Effects of Amount of Origin Training on High-School Graduation for Girls and Boys (after
deCharms, 1980).
As would be expected from these results, those whose Origin scores were
high after training did significantly better on the Iowa Test of Basic Skills for
their grade than those who resisted the training and whose Origin scores stayed
low. The "natural highs"—those whose Origin scores were high even before the
training—performed the best, those whose scores changed from low to high
scored next best, and those who showed no change performed worst. Even
among untrained groups, those whose Origin score was naturally high per-
formed significantly better on the achievement tests than those whose Origin
score was naturally low. This seems to establish the fact that Origin think-
ing mediated the connection between the training and improved academic
Performance.
deCharms also developed a questionnaire that enabled him to find out
whether the pupils in a class perceived the teacher as creating an Origin climate.
For example, an item like "The teacher lets us try new ways of doing things" is
scored positively for Origin climate; one like "The teacher gets upset when we
don't do things her way" is scored negatively. He found that teachers who had
received motivation training were perceived by their pupils as having created
classrooms with significantly higher Origin climate scores. Finally, there was a
significant correlation between the perceived Origin climate scores and the aca-
demic Performance of the children in the classrooms (deCharms, 1976). All of
these relationships that deCharms empirically established are summarized in
Figure 14.11. Origin motivation training for teachers led them to Sponsor Origin
training for their pupils, which improved school learning through getting the
children to think more like Origins. Origin training for teachers also helped
create classroom climates that were perceived by pupils as encouraging Origin
behavior, and this perception was also associated with improved school learning.
Some teachers created Origin climates without having had Special training, and
in these classrooms, too, the children performed better.
While Origin training included elements of achievement thinking and acting,
it is doubtful both on theoretical and empirical grounds that it served to im-
prove academic Performance through increasing n Achievement. As noted at the
Origin Training for Teachers *- Origin Classroom Climate

as Perceived by Pupils
(Whether
or Not
Teachers
Pupils Think Like Origins Were
Trained)
Origin Training for Pupils • Improved School Learning
Figure 14.11.
Effects of Origin Training for Teachers and Students on School Learning (after deCharms, 1976).
outset, high n Achievement should increase entrepreneurial behavior, which

is not usually associated with better academic Performance. Furthermore,
n Achievement scores did not rise consistently in deCharms's studies. Rather,
the effectiveness of the training may lie in its influence on the Ps, or self-
confidence, variable in the Performance equation and also through its effect on
the value placed on Controlling one's own life through careful planning, goal set-
ting, and evaluation of Performance.
Finally, there is the hint that Origin training increases socialized power con-
cerns, which could facilitate better academic Performance, since a lot of prestige
and recognition goes with getting better grades in school. Remember that one
study cited in Chapter 8 found that high n Power is associated with getting bet-
ter grades in school; reference to another such study occurs in the next section.
Origin training, as it has so far been delivered, seems to be somewhat more ef-
fective for boys, perhaps because the assertive achievement-power complex is
more associated at the present time in the United States with male, rather than
female, role behavior. It is probable that motivation training for girls would be
more effective if it were designed with their needs, realities, and perceptions
taken more into account.
Another possible explanation for the effectiveness of Origin training in im-
proving academic achievement lies in the fact that it emphasized the problem-
solving aspects of the scoring System for n Achievement, which Lasker (1978)
found to be related to higher levels of ego development, as measured by Loev-
inger (1966). (See Figure 7.14.) Thus, it could have improved Performance by
teaching the children to cope with the world in a more mature way. The only
difficulty with this interpretation is that the deCharms group failed to find an in-
crease in Loevinger's measure of ego development after Origin training over and
above what occurred naturally between the seventh and ninth grades. Since
maximum changes in ego development scores occur at this age, it is possible
that the measure is not sensitive enough to pick up additional improvements in
maturity possibly introduced by Origin training.
• POWER MOTIVATION TRAINING

Developing Courage in Outward Bound
Fersch (1971) reasoned that Out ward Bound training should develop power mo-
tivation, because the participants spend time in a wilderness setting carrying out
demanding and dangerous physical exploits. They go for long hikes, climb
mountains, walk tree trunks high above the ground, escape from being tied and
thrown into the water, and spend three days all alone in the woods with no
food and very little equipment. The emphasis is on being tough, courageous, and
capable of meeting all types of physical challenges (Newman, 1980). The ideol-
ogy of the Outward Bound movement is captured by the titles of populär arti-
cles describing it like "You Will Never Be Afraid to Try" or "Metamorphosis
of a Marshmallow."
Fersch studied a group of ninth-graders about fourteen years of age, most of

whom were underachieving in school and economically underprivileged, who
went for a month's Outward Bound training one summer (mostly on scholarship
money). They also participated in a month's Upward Bound training in achieve-
ment motivation. He tested them in May before the summer, again when they
returned from Outward Bound, and a year later. He had expected that Outward
Bound would increase average n Power scores but found that this was not the
case either for those who stayed for the program or for the substantial number
who dropped out after a few days. However, those who scored high in n Power
before going were significantly more likely to stay rather than drop out of the
program, demonstrating that the program provided opportunities for displaying
power that would be more appealing to those high in n Power.
Furthermore, for those who stayed in the program, the n Power score ob-
tained afterward significantly correlated with social competence displayed during
the following school year. The social competence measure was made up of gains
in grade point average over the previous year plus the number of socialized
power and achievement activities the Student participated in during the school
year. Socialized power activities included assuming leadership positions of vari-
ous types either in school or outside of school.
Among those who stayed in the Outward Bound program who were from
impoverished backgrounds, the gain in the n Power score from before to after
the training also significantly predicted who would do well academically and so-
cially in the Coming year. No other variable significantly predicted increased
social competence. That is, neither the n Power scores beforehand nor the
n Achievement scores beforehand or afterward, nor staff ratings of how much
they learned from Outward Bound, nor self-image scores predicted improved
Performance in school. Somehow, the Outward Bound training in meeting physi-
cal challenges capably made the n Power scores after the course more diagnostic
of who would be more likely to behave assertively in a responsible way later.
There was an overall significant drop in n Power scores for those who stayed for
the Outward Bound program. This suggests that so many demands for being
powerful and assertive had led through catharsis to a decrease in n Power,
much as eating satiates the hunger drive. Thus, those whose n Power score re-
mained high despite this "satiation effect" would be more likely to be those
whose power motive was strongest and most likely to continue to express itself
throughout the school year.
Socializing the Power Motive Among Alcoholics

Once it was realized that heavy drinking was a response to a strong impulsive
need for power (see Chapter 8), it seemed reasonable to try to help alcoholics
through a training program that explained to them why they were drinking and
how they might satisfy their power needs in other, less destructive ways. So a
course was designed, patterned after the motivation training courses for entre-
preneurs, except that it placed primary emphasis on understanding how the
power motive affected thought and action (Boyatzis, 1974; Cutter, Boyatzis, &
Clancy, 1977).
Participants learned the n Power scoring System and participated in games

like arm wrestling or leading a person around blindfolded that gave them con-
crete affective experiences of being powerful or powerless. They learned, in par-
ticular, the difference between the drive for personalized power over someone
eise and socialized power, which involves more self-control and acting power-
fully on behalf of others, as when a mother cares for her baby or a father looks
after his child. They reviewed the evidence that alcohol increases feelings of
power, which is particularly appealing to those with a strong impulsive personal-
ized power need that is blocked from finding satisfaction in other ways.
Toward the end of the course they set concrete goals for how they planned
to satisfy their power needs in other ways than through drinking. Thus, they
might want to try to improve the respect with which they were regarded by
dressing better, by doing better at their Jobs, or by helping others, using a tech-
nique that Alcoholics Anonymous has found to be useful in rehabilitating alco-
holics. In that organization, reformed alcoholics spend a lot of time helping
each other whenever they feel like drinking or get in trouble. Concrete goal set-
ting and the realization that they understand their drinking problem both con-
tribute to the perceived probability of success variable, which should lead to im-
proved Performance after the course (see Table 14.1). Furthermore, efforts were
made to affect the third, or incentive, variable in the response Output prediction
equation by increasing the value to themselves and to their friends of overcom-
ing alcohol addiction, although most of the participants understood the value
quite well but feit they were incapable of doing anything about their problem.
One systematic attempt has been made to study the effectiveness of this type
of power motivation training (PMT). The participants were hard-core alcoholics,
nearly all male, who were patients living at a Veterans Administration clinic
near Boston. All of them were receiving Standard treatment for alcoholism, in-
cluding detoxification, ingestion of Antabuse (which causes nausea and vomiting
when alcohol is swallowed), group therapy, participation in Alcoholics Anony-
mous, and occupational therapy, if desired. Nearly everyone had been an alco-
holic for years and been in and out of the hospital many times. Most of them
were unable to work consistently because of their drinking problem.
Fifty of the alcoholics were randomly assigned to PMT in groups of eight to
ten members meeting approximately three and a half hours a day for ten days,
for a total of thirty-five hours of training. They were followed up six and twelve
months after the courses to see how they were doing in comparison with a
matched group of fifty alcoholics who had received only the Standard treatment.
The criterion of improved Status included both decreased drunkenness and abil-
ity to work, except for the few who were retired or physically incapacitated.
Total abstinence was not used as a measure of improvement, although it is the
most widely used measure, because it did not seem reasonable to classify people
as improved who are abstinent but not working or abstinent because they are in
jail or back in the hospital undergoing further treatment. So people were re-
garded as reasonably rehabilitated if they were drunk less than once a month
and working five out of the past six months, unless they were unable to work
because of age or physical incapacity (McClelland, 1977b).
By this criterion, the Standard treatment plus PMT was significantly more
effective than the Standard treatment alone, as Figure 14.12 shows. After the
Standard treatment, only about one quarter of the alcoholics could be considered
improved a year later, but the percentage doubled for those who had received
PMT in addition. The greater effectiveness of PMT in the rehabilitation of alco-
holics is not exceeded by any of the other intensive therapies for alcoholics that
have been evaluated by Armour, Polich, and Stamboul (1978), although direct
comparisons are difficult, because they used a much less rigorous criterion of re-
habilitation than the one adopted here.
There were other small signs that the PMT had been helpful. Of those who
were not improved in the first six months after discharge, a larger percentage
(32 percent) of those receiving PMT showed up as improved after twelve
months than was true among those who had received the Standard treatment (12
percent). The trend is interesting, because it is the opposite of the general belief
in the alcoholism field that once people have started to drink again, they inevita-
bly get worse. Furthermore, live men from among those receiving Standard
treatment died in the year after discharge, whereas none from the PMT group
died (McClelland, 1977b).
[Standard Treatment
50 1 Standard Treatment Plus PMT

ics Impro
40
o
Alcor
30
20
10
After 6 Months After 12 Months

Time
Figure 14.12.
Eflfects of Power Motivation Training on Improved Functioning of Alcoholics (after McClelland,
1977b).
Empowering the Staff of Community Action Agencies
During the late 1960s and early 1970s, the Office of Economic Opportunity au-
thorized the creation of Community Action Agencies (CAAs), which were to be
"developed, conducted, and administered with the maximum, feasible participa-
tion of the residents of the areas and members of the groups served" (McClel-
land et al., 1975). The agencies, which mounted programs to serve the poor,
were to be administered as much as possible by the poor themselves. However,
it was soon apparent that poor people had had little experience in managing
others or in acting powerfully and effectively to serve their interests in the com-
munity. So a form of power motivation training was provided for 194 staff
members or neighborhood workers from ten different CAAs in Kentucky be-
tween January and April of 1970 (McClelland, Rhinesmith, & Kristensen,
1975). The objectives of the Workshop were to "explain how socialized power
needs have to be developed to promote Community change" (McClelland et al.,
1975) and to help participants experience ways of strengthening the organization
by making others feel strong, by cooperating rather than competing, by con-
fronting and resolving conflicts and difficulties rather than denying them, and by
stimulating others to take proactive, strong action rather than be passive.
To implement these objectives, the participants learned the n Power scoring
System, participated in a negotiation exercise and the blindfold game, to experi-
ence feelings of being powerful or powerless, and also studied cases in which
Community action had been successful or not. All three types of inputs listed in
Table 14.1 were included—namely, increasing power motive strength, improving
the perceived probability of success, and emphasizing the great importance of
becoming more effective influencers if they were to improve Services to the dis-
advantaged in their communities. All staff members were rated for the degree of
their participation in the Workshop on a scale of 1 to 4. In two of the agencies,
the average level of participation of staff members was significantly lower than
for the other eight agencies, indicating that the training had been ineffective, ei-
ther because the staff members were not interested or because the trainers had
failed to put the training across.
Approximately six months after the training, each staff member was con-
tacted and interviewed intensively by a graduate Student from the University of
Kentucky who had had no contact with the training. The staff members' perfor-
mance was then rated by the graduate students on a scale of — 1 to + 2 in
terms of how effective they had been since the training. The points in the scale
were defined as they were for evaluating entrepreneurial behavior, — 1 indicating
that the person had definitely been performing less well than usual; 0 meaning
that the person was performing pretty much as he or she had been all along;
-f 1 indicating a definite improvement in Performance, including particularly
planning better and involving more people in programs; and + 2 meaning the
person had undertaken a new project, had been promoted, or had shown un-
usual initiative.
According to this method of evaluation, 52 percent of those who had been
trained were judged to have improved in Performance over how they had be-
haved previously. But how can such a figure be compared with the proportion
who might have improved without training? A base line figure can be estimated
from calculating the percentage who improved from the two agencies where the
power motivation training was judged at the time to be ineffective or poor. Only
25 percent of the staff members from these two agencies were judged by the in-
terviewers to have improved their Performance, although the interviewers, of
course, had no way of knowing that the power motivation training in these in-
stances had been judged ineffective. In contrast, 60 percent of the staff members
from the remaining agencies were judged by the interviewers to have improved
their Performance some months after training, a difference that is highly signifi-
cant statistically. Although the measure of evaluation is crude, it suggests that
power motivation training was effective in making staff members of CAAs be-
come more effective Organizers and influencers in their communities.
Leadership Training for Managers

The leadership motive pattern (high n Power higher than n Affiliation, and high
self-control or Activity Inhibition) is associated with greater managerial success,
as studies reviewed in Chapter 8 demonstrate. So it seemed reasonable to as-
sume that trying to create this motivational pattern in managers would improve
their Performance. Accordingly, courses have been offered to business managers
that provide training in power motivation as a key element in a program de-
signed to improve managerial Performance. The courses provide training in the
thought characteristics of all three social motives, power- and achievement-
related games, self-study, goal setting and planning, and all of the other charac-
teristics usually involved in motivation training courses, as summarized in Table
14.1. More emphasis is given on exercising power in a socialized way to influ-
ence others on behalf of the institution, as well as to cases and goal setting spe-
cifically related to managerial responsibilities.
An evaluation of the effectiveness of this type of training in a large U.S.
corporation was carried out by McClelland and Burnham (1976). In this in-
stance, the men trained were managers of sixteen sales districts in different parts
of the country. The first measure of the effectiveness of the training involved a
comparison of how the salesmen perceived the organizational climate in their
districts before and after their managers were trained. As Figure 14.13 shows,
the salesmen, after their managers had been trained, feit a greater sense of re-
sponsibility, more organizational clarity, and greater team spirit. The results fit
expectation, because the emphasis in the training of the sales managers was on
the fact that they were no longer salesmen—individual achievers—but managers
of others, whose primary responsibility was to empower others to take more ini-
tiative and to create a climate that strengthened and supported the individual
salesmen.
The second measure of the effectiveness of the training involved whether it
was associated with an increase in sales for the Company. The training for the
sales managers took place late in 1972; sales at the Company were up substan-
tially in 1973 over 1972, and profitability shifted from a $15 million loss in 1972
j JBefore PMT for Salesmen's Managers
60 - Q j After PMT for Salesmen's Managers

jsmen
w 50
tional Norn
;e Scores by
- —1
o E
^ -£ 20 -
1
°- 10
iense of Org ani;!ati(;>nal Team Spirit

Responsibility Clarity
Figure 14.13.
Effects of Power Motivation Training for Managers on Organizational Climate Perceptions of the
Salesmen They Managed (after McClelland & Burnham, 1976).
to a $3 million profit in 1973. Of course, many factors other than managerial

training could, and probably did, play a role in increasing sales and profitability,
but it is significant that increases in sales were most closely associated with in-
creases in the organizational climate dimensions, which had been affected by
managerial training. The salesmen filled out the organizational climate question-
naire in the fall of 1972, before their managers were trained, and again in April
1973, some months after their managers had been back on the Job. Figure 14.14
shows that it was precisely those sales districts whose climate morale scores
were highest in April 1973 that turned in the best gains in sales for the whole
year of 1973. It seems unlikely that the higher morale scores in April could be
attributed entirely to sales gains already made early in the year, so it seems rea-
sonable to infer that power motivation training for managers was associated
with higher morale scores among those who worked for them, who in turn pro-
duced higher gains in sales. The managers were using their knowledge of their
own and others' motives to behave in ways that created a climate that empow-
ered their salesmen to do better.
30
Os
O 25 a = 6 Districts
<N b = 4 Districts
Os c = 4 Districts
d = 2 Districts
•g 20
5
.£ 15
Ü
ä 10
_L _L J_ _L
'' a b c d
30 40 50 60 70
Climate Morale Score (April 1973): Organizational Clarity Plus Team Spirit
Figure 14.14.
Relationship of Perceived Organizational Climate to Subsequent Gains in Sales (after McClelland «
Burnham, 1976).
Research on power motivation training has not been nearly as extensive as

research on achievement motivation training, but such as it is, it appears to be
reasonably effective and for much the same reasons. That is, it affects motive
strength, probability of success, and the value placed on the particular kind of
successful outcome being emphasized in the training. Much more research is
needed to find out which training inputs are most efFective in what combina-
tions. But what is known is that power and achievement motivation training
courses are effective in changing the way people think, talk, and act, which in
turn leads to socially important outcomes, such as better Performance in school,
small business success, more frequent recovery from alcoholism, and more effec-
tive organizational Performance for managers.
NOTES AND QUERIES
1. Why do you think the teacher in the fifth grade in Figure 14.1 had so
much less of an effect on pupil test Performance than the teacher in the
third grade? Formulate a hypothesis to explain this difference based on

deCharms's research on the Origin-Pawn dimension, reviewed later in
the chapter, and explain how you would test the hypothesis.
2. Explain the relative failure of increasing perceived probability of success in
improving academic Performance in terms of the findings presented in
Chapter 13 on the interactive effects on Performance of the M, Ps, and V
variables.
3. In Chapter 8, some findings are reported that suggest that perceived
probability of success in performing is not as important in determining the
behavior of individuals high in n Power as it is for individuals high in
n Achievement. If the power motive turns out to be the major motive be-
hind academic success, could this explain why increasing perceived probabil-
ity of success by training has weak long-term effects on school grades?
4. Do you believe training programs for small businesspeople would increase
their Performance if training omitted teaching n Achievement but included
inputs related to the probability of success (Ps) and the incentive value of
success (V). Justify your answer in terms of theory and research evidence.
Review the literature on the effects of skills training on the Performance of
small businesspeople (see Miron & McClelland, 1979).
5. In terms of the model for predicting operant behavior reviewed in Chapter
13, list as many reasons as you can that might explain why blacks and
women in the United States have until recently seldom been successful
entrepreneurs.
6. Indians outside of India (for example, in East Africa) often appear to be
more successful businesspeople than Indians in India. Give several alterna-
tive theoretical explanations for this phenomenon. How would you test
each one?
7. In terms of findings reported in Chapters 6 and 7, make a list of the kinds
of classroom situations in which pupils receiving training in achievement
motivation would not be expected to do better.
8. If Origin training improves school Performance, why is it not being used in
school Systems today? Try to frame your answer in terms of the dominant
motives and goals of the main actors in the school System—teachers, pupils,
parents, and administrators.
9. Suppose achievement motivation training were really effective in the schools.
Can you think of any possible negative consequences of such an eventuality?
10. There is some evidence (Ellingstad & Struckman-Johnson, 1978) that power
motivation training for individuals picked up for drunken driving is not par-
ticularly effective in reducing drunken driving behavior. Try to explain this
result in theoretical terms.
11. In Chapter 7 it is reported that a strong emphasis on toilet training in
childhood is associated with high n Achievement in adulthood. Try to ac-
count for this result in terms of the model used in this chapter to explain
the effectiveness of various training inputs. Use the model also to explain
the association between permissiveness about sex and aggression in early

childhood and high n Power in adulthood.
12. According to the theory presented in this book (see especially Chapters 4
and 5), motives are built on the affect associated with natural incentives.
Explain in terms of such a theory how motives could be acquired or
strengthened in the types of training programs described in this chapter.
What natural incentives are experienced? How have they been modified by
cognitive growth? What affect is experienced at what points? Does this help
explain why nearly all training programs of this type employ games?
13. Is there any solid evidence in the chapter that a motive disposition has been
strengthened by training, or do you believe all the results obtained can be
explained by changes in the other determinants of response Output or by
changes in the way an existing motive expresses itself?
14. Design an experiment to show that training definitely increased motive
strength rather than changed some other variable. If, after achievement mo-
tivation training, students responded differently to homogeneous and hetero-
geneous classrooms (see Chapter 7), would that be firm evidence that the
overall n Achievement level had been raised?
15
Milestones in the Progress
Toward a Sdentific
Understanding of Human
Motivation
Measuring Motives
Definition of a Motive
Accumulated Knowledge About Three Important Human Motive Systems
The Power Motive
Understanding How Motives Combine with Other Characteristics to
Determine Action
Some Issues Needing Further Clarification
Examining the Biological Basis of Major Motive Systems
Defining and Measuring Other Major Motive Systems
Determining What Collective Motive Levels Mean
Improving the Theoretical and Practical Understanding of How to Develop Motives
The Relationship of Progress in Psychology to Its Role in Society
587
M . otivation has always fascinated people and will continue to fascinate
them so long as there are people around to wonder why human beings and ani-
mals behave as they do. Nearly everyone develops an explicit or implicit theory
of motivation. We think we know why our parents are sometimes disagreeable
and try to continue to control what we do: they want the pleasure of continuing
to dominate us. Or we think we know why our girlfriend or boyfriend has aban-
doned us: he or she prefers to be with someone with greater prestige or more
possessions than we have. Or we think we know why we cannot seem to study
very hard: we have a low need to achieve. Authors, philosophers, economists,
politicians, and the people next door have all operated in terms of theories of
motivation. Shakespeare vividly portrayed the lust for power in MacBeth and
the longings of love in his Sonnets. Plato explored the nature of love in the
Symposium. Economists think in terms of the desire to acquire possessions, and
they write of the importance of the profit motive. Political observers from Ma-
chiavelli to the present have stressed the importance of the desire for power in
human affairs.
What have psychologists added to this parade of observations and theories
about human motivation? It is time to review what has been accomplished and
what remains to be discovered. The mass of detail in individual chapters may
obscure the broad outline of progress that has been made. So let us take a look
at the big picture without bothering to cite evidence to support the generaliza-
tions made. It can be found in previous chapters. What follows is a brief natural
history of developments in the field of motivation, with an emphasis on psychol-
ogy's accomplishments and on matters that need further clarification.
• MEASURING MOTIVES
To begin with, psychologists have tried to be more systematic than others inter-
ested in motivation. They have sought to define what constitutes a motive care-
fully and to arrive at lists of the most important motives, as McDougall and
Murray did for normal behavior, or Freud and the psychoanalysts did for ab-
normal behavior. But the key to progress in science lies not only in theoretical
clarification, but also in adequate measurement. Consequently, Murray and Cat-
tell devised methods of measuring the various motives that psychologists and
others had found to be necessary to explain behavioral outcomes as diverse as
neurosis, creativity, or animal learning.
They succeeded only to a limited extent, in part because the measures they
developed were composites that did not distinguish clearly enough between mo-
tives and other personal characteristics. In Murray's studies, judges pooled their
observations of many different behaviors into a final estimate of the strength of a
particular motive, but such judgments were strongly influenced by other charac-
teristics the person was thought to have, so the motive measure was impure. Al-
though Cattell pooled many different measures statistically, the same difficulty
arose: the items making up the factor score for a particular motive contained re-
sponses that reflected other personal characteristics like traits as well as motives.
588
Milestones in the Progress 589
And inventories of self-reports of desires, wishes, and needs developed by many

psychologists proved to have the same drawback: they too reflected important
nonmotivational characteristics, such as the response bias to present oneself in a
favorable light.
Another attempt to measure motives grew out of the behaviorist tradition
represented by Hüll, which considered tension, or strong Stimulation, as the ulti-
mate wellspring of human and animal action. Organisms were conceived as seek-
ing to reduce the tension caused by deprivation of substances like food and
water needed for survival. An elaborate theory was constructed that explained
how secondary motives and rewards were acquired through association with a
primary drive like hunger and its relief through eating.
Since the theory was applied deductively to explain human behavior, it did
not result in further investigation of complex human motives until Spence, Tay-
lor, and their associates realized that individual differences in drive strength
might be measured by a Manifest Anxiety questionnaire, as explained in Chapter
3. That is, according to behaviorist theory, an organism would soon learn, in the
course of trying to reduce the tension of biological needs, that certain cues
would be associated with a lack of the rewards needed to reduce the tension.
The discomfort or upset caused by the anticipation of the lack of something
needed could readily be conceptualized as anxiety. So people manifesting more
anxiety could be understood as being under greater tension or as having greater
drive strength. The use of various questionnaires to measure the strength of the
anxiety motive led to the accumulation of much information about how this mo-
tive functions to facilitate some behaviors and to inhibit others, as outlined in
Chapters 3 and 12. Research using this approach to measuring anxiety led to
important developments, not only in the understanding of factors governing effi-
ciency of Performance, but also those underlying neurosis (Eysenck, 1957a).
However, the approach has important limitations in extending our under-
standing of human motivation. It is based on a greatly oversimplified theoretical
notion that all motives can be traced to a Single master motive, namely, anxiety.
Furthermore, the measure of anxiety obtained from self-reports on question-
naires is impure, because some of the items refer more to traits or styles of re-
sponding than to motives in themselves, and because such self-reports are influ-
enced by response biases—like the desire to gain sympathy or avoid disapproval
—that have nothing to do with the strength of the anxiety itself. Finally, the ap-
proach is weak because it does not deal carefully with what it is a person is anx-
ious about or with the negative State the person is trying to avoid. From clinical
studies of individuals it is clear that a distinction must be made between differ-
ent types of anxiety, such as anxiety over failure and anxiety over possible rejec-
tion. Later work with another method of measurement has differentiated a vari-
ety of avoidance motives, including fear of failure, fear of rejection, fear of
success, fear of power, fear of intimacy, and so forth (see Chapter 10). Empirical
progress in this area has been limited, but it has been sufficient to show that
these different fears have very different effects on behavior, so it is not possible
to think in terms of one master anxiety motive. For example, whereas those
high in fear of failure generally avoid competition (Table 10.5), there is no evi-
dence that those high in / Power do; in fact, they report that they get into argu-
ments more frequently than those low i n / Power (see Table 10.9).
An important step in the methodology of measuring human motives was
taken when it was decided to arouse motives experimentally and examine their
effects on all types of behavior (Chapter 6). Arousing the hunger motive in
human subjects established that its effects were most sensitively reflected in fan-
tasy or associative thought, thus confirming an important insight Freud had
reached fifty years earlier. He too found that motivational factors could be most
readily identified in dreams and free associations. Other types of behavior, such
as self-reports of motive states, percepts, or action trends, reflect motive arousal
less sensitively, because they are more influenced by nonmotivational factors
such as response bias, objective Stimulus properties of the external environment,
or opportunities for action in the environment.
Because of these drawbacks, the effects of arousing various motives on asso-
ciative thought were carefully identified and worked into coding Systems that
could be objectively and reliably scored. It was then reasoned that if a person
produced under neutral testing conditions many of the thoughts characteristic of
people in whom the motive was aroused, then that person could be presumed to
be thinking characteristically as if he or she were regularly in a State of arousal
so far as that motive was concerned. For example, if a man thinks like hungry
people when he is not hungry (that is, when he has just eaten), then it could be
presumed that he has a characteristically high level of concern about hunger or
eating (that is, he has a strong hunger motive).
One important advantage of this method is that motive definitions do not
depend on a priori conceptions of how a motive should show itself in thought or
on dictionary definitions of characteristics supposedly typical of a given motive.
Rather, the coding System that makes up the definition of a motive is empiri-
cally determined by the effects of arousing the motive on associative thought.
This method of measuring human motives has been successful in producing cod-
ing Systems that identified several major motive Systems, the effects of which
have been extensively studied for the past generation.
DEFINITION OF A MOTIVE
This method also led to a working definition of a motive as a recurrent concern
for a goal State based on a natural incentive—a concern that energizes, Orients,
and selects behavior. Explaining the key terms in this definition should help clar-
ify and summarize what psychologists have learned about human motivation as
covered in this book. Basically, a motive disposition refers to thinking about a
goal State frequently—to a recurrent concern. It does not refer to fleeting or oc-
casional thoughts, since nearly every eventuality occurs to everyone once in a
while. A person who has just eaten may once in a while think about being with-
out food, but it is only the person who thinks about being without food fre-
quently when he or she is not hungry whom we would want to characterize as
having a strong hunger motive. Similarly, it is people who think about doing
things well when there is no Stimulus to do so whom we would characterize as

high in n Achievement.
The fact that the concern is about a goal State has the important implica-
tion that the means of getting to the goal is not part of the definition of a mo-
tive. The goal State may be defined in terms of the outcome of certain acts, such
as doing something better (for the achievement motive), or having impact (for
the power motive), but the particular acts that lead to such outcomes are not
part of the definition. This is an important distinction, because many theorists
identify motives in terms of certain acts or behavioral trends that define them.
For example, psychologists often speak of an aggressive drive. But aggression is
a type of activity that is highly controlled in human society, so it would be haz-
ardous to attempt to infer the strength of the aggressive urge or drive from the
number of aggressive acts a person commits. It is better to define the motive be-
hind aggressive acts in terms of their intended effect as determined from
thoughts about goal states, which research suggests can be more broadly defined
as "having impact." Then intended injury to someone, which defines aggression
for many psychologists, becomes one type of having impact, and attaining
higher Status becomes another way of having impact. Or aggression can be seen
as one means of having impact and accumulating prestige as another. This is
why the concern to have impact is defined as the power motive rather than
the aggression motive, since the latter defines the means of having impact too
narrowly.
The goal states involved in motivational concerns are presumably based on
and derived from natural incentives, which innately give rise to various emo-
tions. The chief theoretical advantage of basing motives on natural incentives is
that it explains why there should be relatively few major motive Systems, why
they have such a pervasive effect on behavior, and why motives are so intimately
connected with emotional states. If it were not for the grounding of major mo-
tive Systems in something, it would be hard to differentiate a purely cognitive
wish like "I want to know how this story turns out" from pervasive major mo-
tive Systems. We may think of such cognitive desires as being for a subgoal in
the Service of some ultimate goal, such as the need to achieve, which has been
shown to lead people to try to finish incompleted tasks (Chapter 7); however,
what is it that makes a goal State like wanting to do something better (as in the
achievement motive) more basic or pervasive than any simple cognitive desire?
If a natural incentive is not responsible for this difference, then some other, as
yet undiscovered, factor must be responsible.
The insistence on the importance of natural incentives has another advan-
tage. It accounts for the phenomenon of intrinsic motivation, or the tendency of
people to engage in activities that seem to be satisfying in themselves like whis-
tling, chewing gum, dancing, jogging, or playing with Rubik's cube. It stretched
belief to figure out how such activities could be in the Service of major motive
Systems like n Achievement or n Power. Furthermore, arousing these other mo-
tive Systems interferes with carrying out such intrinsically satisfying activities
(Deci, 1975). Thus, it seems better conceptually to think of such activities as in-
centive driven. That is, the natural incentives continue to produce the primitive
behaviors that define them, quite apart from whether they have become orga-
nized into major motive Systems for an individual. Thus, chewing gum or hitting
a nail with a hammer may continue to give impact satisfaction if occasion pro-
vides the opportunity, whether or not such activities have become organized into
a major power motive System for a person.
The use of the word concern in the motive definition refers directly to the
fact that motives are best measured by coding concerns in associative thought or
fantasy. The measurement implication of the definition is important, because
without it, the definition would not represent much of an advance over previous
definitions provided by McDougall, Cattell, Freud, or Maslow. The term con-
cern is used because it does not imply the conscious goal-directed striving that
is part of many definitions of motives, for all the empirical work on motives
identified through coding associative thought indicates that motives are not nec-
essarily part of the person's conscious self-image. Thus, another of Freud's in-
sights has been confirmed—namely, that some important motives are not con-
scious. In contrast, not all people are not conscious of the strength of their
major motive Systems, as Freud seems to suggest at times. Some are and some
are not. This is the inference to be drawn from the empirical fact that the con-
scious value placed on achievement, affiliation, and power is essentially uncor-
related with the n Achievement, n Power, and n Affiliation scores as determined
by coding the content of associative thought. For this reason it is essential to
distinguish between values, which are part of the self-picture, and motives,
which are not. The ideas people have about what is important in life or to them
have a strong influence on what they choose to do, but in the scheme presented
here, they are determinants of behavior that are essentially independent of mo-
tive Systems. Thus, choiees in many situations, particularly those in which the
alternatives are cognitively understood, are determined primarily by values, and
motives play little part in determining what people do in such situations (see
Chapter 13).
The influence of values, skills, and opportunities on conscious choiees and
actions explains why it is so difficult to assess motive strength from what people
say and do. They may say over and over again in a questionnaire that they want
to achieve, but that is because the items explicitly tap the value they and their
eulture place on achievement. It is not evidence that they have a recurrent con-
cern about the goal State of doing things better. Or they may greet a number of
guests effusively at a cocktail party, but that is not a sure indication that they
have high n Affiliation. They may be acting that way because of social norms
governing the behavior of hosts at such a party and because they have an image
of themselves as good hosts. Or even if they are high in n Affiliation, they may
fail to show affiliative behavior in the presence of superiors, because social
norms they have interiorized dietate that they should not be affiliative to such
people.
So motivational concerns are best measured in associative thought where the
influence of values, skills, and opportunities is less marked. Technically, of
course, it should be possible to measure the strength of a motivational concern
from its behavioral effects, if those effects are relatively uninfluenced by values,
skills, and opportunities or if the effect of these other determinants of behavior

is controlled or standardized. A typical example is the "doodling" index of
n Achievement devised by Aronson (1958) and described in Chapter 7. People
high in n Achievement doodle differently from those low in n Achievement.
Since there are no conscious values affecting how a person should doodle, since
it requires no particular skill, and since situational influences are slight, the doo-
dling index of n Achievement is reasonably valid. That is, it is influenced pri-
marily by the motivational concern and not by other determinants of behavior.
In similar fashion, we could attempt to devise an index of the strength of a per-
son's power motive from its behavioral correlates as summarized in Chapter 8,
such as, in the case of men, the frequency of getting into arguments, playing
competitive sports, drinking alcohol, and accumulating prestige supplies. Such
attempts have been made, but they have limited value because the outlets for the
power drive vary so much as a function of sex, age, level of maturity, and social
class.
The only motivational concerns that can be estimated with reasonable valid-
ity from conscious choices are those that involve negative incentives. Thus, an
avoidance motive like fear of failure can be measured fairly adequately by a Test
Anxiety questionnaire, although even here the direct measure of the concern
over failure in associative thought appears to be somewhat more valid (Chapter
10). The explanation for the greater validity of self-report measures of an avoid-
ance motive like the fear of failure may lie in the fact that the source of the
negative incentive is often external, in contrast to a positive incentive like doing
better or having impact, which can only be observed internally. Thus, fear of
failure develops out of threats from external authorities such as teachers or par-
ents, which people can readily identify. Furthermore, there is no strong value
against admitting test anxiety, at least among most students, so conscious re-
ports can reflect the avoidance motive reasonably accurately. In contrast, values
may interfere with admitting to fear of weakness or fear of success, so it is a
matter for further investigation to discover whether self-report measures can
substitute for the more direct measures of avoidance motives in associative
thought.
The term concern also signifles that a motive is more general than an intent.
An intent is, by definition, more specific and limited in time. Thus, Freud un-
covered unconscious intents that led to the forgetting of a name or dreaming
that one could not give a dinner party. But the intents were in the Service of
more general motivational concerns. Conscious intents, in contrast, are the prod-
uct not only of motives, but also of other determinants of behavior such as the
value people place on performing an act or their estimate of the probability that
it can be performed.
The measures of motivational concerns derived from coding associative
thought have proved to satisfy for the most part the requirements for good mea-
surement. They reflect sensitively known differences in motivational states when
aroused experimentally. They signify differences in motive strength more
uniquely than other behaviors that are more influenced by nonmotivational de-
terminants such as values, skills, and opportunities. They are valid in the sense
that they satisfy the requirements of the functional definition of a motive, which
states that it energizes, Orients, and selects behavior. That is, people who score
high on one of these motive measures, such as n Affiliation, behave as they
should if they were in fact more motivated for affiliation. They behave more en-
ergetically in an affiliative manner; that is, they perform affiliative acts more fre-
quently. They perceive affiliative cues in the environment more readily, and they
learn affiliative networks more quickly. Other reported measures of motive dis-
positions have not been shown to satisfy these requirements for distinguishing
between a more and less motivated person in a particular area.
The chief difficulty with the associative thought measures of motive strength
has been their presumed unreliability. The motive scores people obtained on one
occasion have generally not agreed very well with the scores they obtained on a
second occasion, nor have the scores obtained from one group of stories agreed
very well with the scores obtained from another group of stories, even when all
the stories are obtained on the same occasion. In psychometric terms, the inter-
nal consistency and test-retest reliability of the measures have been so low that
many psychologists have feit the measures could not be taken seriously.
What can a person's n Achievement level really be if estimates of its
strength vary so much from one occasion to another? Two recent develop-
ments—one theoretical and the other practical—suggest that the reliability of
the measures is not nearly as low as it has been estimated to be. Both develop-
ments rely on the fact that once a thought has been expressed spontaneously, it
tends to be used up, so to speak, so different thoughts are more likely to be ex-
pressed next. If a person teils an achievement-related story first, he or she is less
likely to teil an achievement-related story next. Thus, there is little consistency
in the amount of achievement imagery expressed on two occasions.
Atkinson and Birch (1978) have built a theoretical model questioning
the common assumption that variations in scores represent "error" in estimating
a true score and demonstrating that there can be consistency in levels of
n Achievement expressed over long periods of time even though there is
moment-to-moment inconsistency, which is in fact predicted by their theory and
therefore cannot be considered "error" (see Chapter 6). At a practical level,
Winter and Stewart (1977) have demonstrated that test-retest reliability increases
greatly if the subjects' set to be creative and to vary the type of story told is
broken by telling them that they can teil the same or a different story, depend-
ing on how they feel at the moment. Under these instructions, the estimates of
motive strength from coding associative thought are satisfactorily stable. Now
that substantial progress has been made in establishing the reliability of these
measures, investigation of other motive Systems should proceed more rapidly.
Nevertheless, associative measures of motive strength are easily influenced
by situational factors that constrain the spontaneity with which a person thinks
and writes. If subjects feel threatened or anxious, their thought processes are
no longer spontaneous, and motive scores derived from them are no longer
valid. Lundy (1981a) has developed a method for scoring stories that estimates
the extent to which they are spontaneous so that the investigator can deter-
mine whether the motive scores derived from them are likely to be valid (see
Chapter 6).
ACCUMULATED KNOWLEDGE ABOUT THREE

IMPORTANT HUMAN MOTIVE SYSTEMS

Gains in knowledge about what motives are and how they can be measured
have led to substantial progress in the accumulation of knowledge about three
important motive Systems that govern human behavior. Historically, the first of
these to be intensively investigated was the achievement motive, or n Achieve-
ment. As work on it progressed, it became apparent that it might better have
been named the efficiency motive, because it represents a recurrent concern
about the goal State of doing something better. Doing something better implies
some Standard of comparison—either internal or external—and it is perhaps best
conceived in terms of efficiency or an input/output ratio. Doing better or im-
proving means getting the same Output for less work; getting greater Output for
the same work; or, best of all, getting greater Output for less work. So people
high in n Achievement are attracted primarily to situations where there is some
possibility for improvement of this kind. They are not attracted to, and do not
work harder in, situations where there is no possibility of improvement—that is,
where tasks are very easy or very difficult—or where external rewards such as
money or recognition are provided. In order to know whether they are doing
better, they prefer situations in which they have personal responsibility for the
outcome and that give them feedback on how well they are doing.
What proved particularly important was the realization that people with
these characteristics should make good entrepreneurs, as well as the demonstra-
tion that successful entrepreneurs did, in fact, have higher n Achievement. This,
in turn, led to an extensive investigation linking high levels of n Achievement in
individuals and in countries to increases in entrepreneurial behavior and more
rapid rates of economic growth. It turned out that the achievement motive has
been a major factor in the economic rise and fall of ancient and modern civiliza-
tions. In fact, since efficient business activity is a key element in the economic
success of individuals and nations, it is not stretching the evidence too far to
suggest that the achievement motive has a lot to do with wealth and poverty or
the Standards of living people enjoy.
Early studies indicated that high n Achievement might be fostered in middle
childhood by parents who set higher Standards of Performance for their chil-
dren, often because they belonged to reform groups that stressed that their be-
liefs or ways of doing things were better than traditional ways. In this manner,
higher n Achievement levels could be explained among certain minority groups,
among Protestants in the early days of the Reformation, or in countries where
the Communist reformation first occurred (see Chapter 11). More recent re-
search has found that parental insistence on high Standards of Performance in
the first years of life for Controlling involuntary functions like eating or elimina-
tion is associated with higher levels of adult n Achievement. The finding is of
great theoretical importance, because it Supports the notion that motivational
concerns can be formed very early in life, before infants have developed much
self-consciousness or the cognitive capacity to evaluate their own Performance.
This also helps explain why conscious values, which are formed much later in
life, after cognitive development, can be quite different from motivational con-
cerns based on very early affective learnings.
The Power Motive

The need for power as coded in associative thought represents a recurrent con-
cern to have impact certainly on people, and perhaps on things as well. The lat-
ter possibility has not been extensively investigated, although McAdams (1982a)
has shown that peak experiences, involving feelings of physical or psychological
strength, were more likely to be recollected by individuals high in n Power.
High n Power is associated with many competitive and assertive activities and
with an interest in attaining and preserving prestige and reputation.
However, since competitive and particularly aggressive activities are highly
controlled by society because of their potentially disruptive effects, the outlets
for the power motive vary greatly according to the norms the person has in-
teriorized as to what is acceptable behavior. Thus, sex typing has an important
effect. The behavioral outlets of high n Power in men and women are different.
The power motive leads to more openly competitive and assertive behavior
among men than women, and also more often to accumulating resources as a
means of exerting influence among traditional women than among men. Social
class also makes a difference. Lower-class men high in n Power tend to be more
openly aggressive than middle-class men at the same level of n Power. Level of
maturity also shunts the power drive in one direction or another. If people high
in n Power are at the oral intake stage (Stage I), they tend to engage more in
power-oriented reading; if they are at the stage of autonomy or self-control
(Stage II), they tend to bottle up much anger; if they are at the assertive stage
(Stage III), they express their anger more openly in assertive and competitive
activities; and if they are at the stage of mutuality (Stage IV), they tend to share
more secret information with their intimates and to assume responsible positions
in voluntary organizations.
From reviewing this research, one gets the impression that the power motive
is something like a hydra-headed monster that shows very different faces de-
pending on other variables. Sometimes it leads to dominance, as in Stage III
people who are oriented around assertiveness. And sometimes it leads to submis-
siveness, as in Stage I people who are oriented around gaining power from fol-
lowing leaders from whom they take in strength.
Nowhere is this variability more obvious than when high n Power is either
associated with a high or low degree of internal inhibition. If men are high in
n Power and low in inhibition, they tend to think in terms of personally domi-
nating others. They drink too much. They are Don Juans trying to seduce as
many women as possible. They lie, trick, and deceive. They are socially irre-
sponsible. In short, they have some of the characteristics of the image of Satan.
In contrast, men high in n Power and high in inhibition have some of the char-
acteristics of the image of God. They think in terms of exercising their power
on behalf of others; they engage in less impulsive self-indulgent behavior like
drinking; they assume positions of leadership in voluntary organizations; they

believe in centralized authority, hard work, self-sacrifice, and charity. In short,
they are responsible Citizens. They make good husbands and good managers in
business organizations, particularly if they are also low in n Affiliation. People
with this motivational profile (high n Power higher than n Affiliation, and high
in Activity Inhibition), variously called the imperial power motive syndrome or
the leadership motive syndrome, tend to be empire builders, and countries with
the same profile in their populär literature tend also to be imperial: they collect
more of their gross national income in taxes; spend more of it on military prepa-
rations; and, at least in the case of the United States, tend more often to go to
war. Just as the achievement motive appears to lie at the base of economic
growth, the power motive, if it is disciplined, lies behind the effective governing
of people, which can lead to the organizational success of great civilizations like
the Roman Empire.
The chief weakness of people with the leadership motive syndrome is that
they are particularly prone to illness if they are stressed or frustrated. They are
more likely to develop high blood pressure under such conditions and to fall
prey to illnesses resulting from a weakening of the body's immune defense sys-
tem, for the inhibition of a strong power motive in combination with low inter-
est in affiliation appears to result in hormonal changes that damage the lympho-
cytes that function to maintain the body's immune defenses.
Less is known about the sources of the power motive in society and in
childhood, but what is known follows the pattern discovered for the achieve-
ment motive, which suggests that how parents treat their children makes a dif-
ference. Parents who have lost power and been oppressed, like the Jews in Nazi
Germany or unemployed black males in the United States, tend to develop a
strong power motive, perhaps in retaliation. They are also probably more per-
missive in allowing their children to express power. The one solid piece of evi-
dence we have is that parents who are permissive about expressions of sex and
aggression in early childhood tend to have children who as adults have a higher
n Power. It is also true that children of Jewish parents who were particularly
persecuted in Germany had higher average levels of n Power than children of
other Jewish parents who had left Germany but who had not been so oppressed.
And there is evidence that the particularly oppressed and stressed parents placed
more emphasis on the assertion of Jewish identity in order to insure survival,
from which it may be inferred that they were also more permissive about early
assertiveness in their children.

Less is known about the affiliative motives than about the achievement or power
motives. The recurrent concern involved is for the goal State of being with an-
other, but it is not certain what the natural incentive for this goal State is. Does
love, or the desire to be with others, develop primarily out of sexual contact
gratifications, as suggested by Freud and Plato, or is it connected with the
"good vibes" identified by Condon as characterizing harmonious relations be-
tween two persons? (See Chapter 9.) Are the sexual motive and the affiliative
motives connected at all, for that matter? No one knows for sure, because it has
proved very difficult to develop a measure of the strength of the sexual motive
by arousing and coding fantasy changes in the usual way. The reason is that
anxiety, inhibition, or some other factor blocks the direct expression of sexual
thoughts following sexual arousal, just as Freud discovered years ago. Some ten-
tative attempts have been made to derive an indirect or symbolic measure of the
strength of the sexual motive, but they need much further elaboration.
Coding Systems for two affiliative motives have been developed and investi-
gated. One, the n Affiliation score, is associated somewhat with fear of rejection
and with more active efforts to be with people as if one fears being left alone.
The other, the intimacy motive, is more of a "being" as opposed to a "doing"
motive. Those who score high in it take pleasure in intimate sharing with others
and are perceived to be natural, warm, sincere, appreciative, and loving. They
are interested in establishing and maintaining warm interpersonal relationships,
but they are not anxious in the absence of others. In contrast, those high in
n Affiliation, while having some of the same interest in interpersonal relation-
ships, are more often anxious about whether they are liked or not, so that often
they are not very populär.
The affiliative motives are of central importance for mental and physical
health and a sense of well-being, just as Plato argued centuries ago. College men
who scored high in the intimacy motive were judged to have made a happier,
better adjustment to life by a psychiatrist who interviewed them regularly over a
period of years without knowledge of their scores. And several studies have
shown that a strong n Affiliation or intimacy motive, particularly if it is accom-
panied by low inhibition, is associated with a stronger immune defense System
against disease (see Chapter 9). The reason for this relationship is not known,
although presumably it involves some reduction in the stress hormones that
damage lymphocyte function. At any rate, the data, so far as they go, support
the folk psychological notion that love is good for your health.
Little is known about the origins of the affiliative motives in society or in
childhood. In an extensive study, no parent training practices in the early years
of life turned out to be significantly and consistently associated with adult
n Affiliation scores. The reason may be that the investigators focused more on
what the parents did than on the nature of the relationship between parent and
child, which may have had more to do with the formation of the affiliative mo-
tives. Certainly the matter cries out for further investigation using modern tech-
niques for coding the relationships between parents and children.
UNDERSTANDING HOW MOTIVES COMBINE WITH

OTHER CHARACTERISTICS TO DETERMINE ACTION
Psychologists have had a great and abiding interest in building theories or mod-
eis to predict choices or action tendencies. They have nearly always in recent
years assumed that motive strength was one of the determinants of choice. A
significant Step forward was taken by Atkinson when, building on earlier work
of Lewin and Hüll, he proposed that an approach tendency was jointly deter-
mined by motive strength (Ms), probability of success (Ps), and the incentive
value of success (Is). He assumed further that the incentive value of success was
a simple linear function of difficulty or the reciprocal of the probability of suc-
cess (1 — Ps)9 and that multiplying the three determinants by each other pro-
duced the best prediction of an approach tendency. The model proved to be
quite useful in predicting the level of difficulty at which people would choose to
work as a function of the strength of their n Achievement, the probability of
success, and the incentive value of success defined as 1 — Ps. Subsequently, psy-
chologists have spent a great deal of time trying to clarify and expand the mean-
ing of the model. Atkinson himself, in collaboration with Birch (Atkinson &
Birch, 1978), developed a more abstract definition of the variables in the model
and their interrelationships that, when formulated in terms of a Computer
model, permitted the deduction of what behaviors one could expect under vari-
ous circumstances.
Others worked to clarify the meaning of the original model. As a first Step,
it seems desirable not to use the term motivation as Atkinson has to describe the
impulse or tendency to act that is a product of all the determinants of action,
because it is confusing to think of the cognitive and skill determinants of the im-
pulse to act as motivational in nature. Furthermore, we must distinguish be-
tween motive as a disposition and the arousal of the motive at a particular time
and place, which is best conceived as a State of motivation. But if the term mo-
tivation is used to describe an aroused motive, it cannot also be used to describe
the outcome of all the determinants of action. Much confusion in the interpreta-
tion of experiments has arisen from using the term motivation in these two dif-
ferent senses.
Also, the meaning of the probability of success term in the model has come
in for considerable expansion. It Covers the perceived probability of success at a
task, and also the demonstrated skill at the task, which affects the probability of
success in actuality. It refers to one's general level of self-confidence or feeling of
efficacy, and also to the confidence one feels in reference to a particular task. It
includes the feeling of whether one has voluntarily chosen to do something or
whether one has control over the Situation or not. And much work has been
done on the causal explanations for success or failure, which either increase or
decrease the perceived probability of success. For example, if people ascribe fail-
ure to lack of effort, they are likely to have a higher perceived probability of
success on the next trial than if they perceive failure as having been due to lack
of ability. All these variables contribute to the perceived probability of success,
which is an important determinant of the impulse to act.
As for the incentive value in the model, attention has shifted from difficulty
as the only or the major determinant of the outcome to other values that influ-
ence the impulse to act in different ways. The importance of the action or
whether it is necessary to get to subsequent goals are both important contribu-
tors to its incentive value. Other values relating to whether the action is appro-
priate for one's sex, age, or culture also influence the strength of the incentive to
perform the task in one way or another. For example, the incentive value of
working in groups is apparently much greater for native Hawaiians than it is for
other U.S. Citizens. What is needed is a much more systematic classification of

the values that enter into influencing the outcome of various equations designed
to predict one kind of Performance or another from motives, probability of suc-
cess, and incentive values. At the present time psychologists have pretty much
limited their efforts to demonstrating that different values, in combination with
different motives and probabilities of success, can produce different behavioral
outcomes.
Finally, the model has been in great need of empirical testing outside the
achievement area in connection with motives and outcomes other than approach
tendencies to tasks of different difficulty levels. In Chapter 13 we demonstrated
that the model had considerable Utility in predicting Performance success on a
laboratory task and also the frequency with which College students were found
to be interacting with another person. In both cases, motive strength (as repre-
sented by either n Achievement or n Affiliation scores) predicted the outcome,
but the incentive value placed on achievement or affiliation did not. And in both
instances, the motive level multiplied by the skill level (either actual skill or per-
ceived social skill) predicted the action outcomes over and beyond what could
be predicted from either variable alone. Skill contributed to more of a certain
kind of Performance only when a person was highly motivated. In contrast,
multiplying aroused motive strength (n Achievement) times the incentive value
placed on achievement (v Achievement) improved Performance, whereas multi-
plying n Affiliation times the value placed on affiliation did not increase the fre-
quency with which students performed affiliative acts. What accounts for this
difference? Under what circumstances does incentive value multiply with motive
strength to increase the probability of a response, and under what circumstances
does it not? Many more multivariate studies of this type need to be carried out
to test the general applicability of the model of the multiple determinants of the
impulse to act.
What is suggested by these preliminary investigations is that (1) the three
variables do appear to give the best prediction of an action tendency if they are
multiplied as the model assumes; (2) motive dispositions are more powerful pre-
dictors of operant or spontaneous acts, whereas incentive values are better pre-
dictors of cognitive choices; and (3) a better prediction of operant outcomes can
be made if competing motivational dispositions are taken into account. All of
these conclusions have great significance for further research in the field of moti-
vation, for psychologists need to test the assumption that the determinants of
action multiply with each other outside the area of choosing tasks of different
difficulty.
Furthermore, psychologists have not usually made the distinction between
trying to predict cognitively based choices as compared with operant actions.
And even though Atkinson (1980) has insisted on the theoretical importance of
taking motive ratios into account, he has apparently never done so in an empiri-
cal test as to whether this gives a better prediction of outcomes. In fact, if the
ratio of n Affiliation to other motives present in the individual is combined with
perceived social skill, and with the value placed on affiliation and their interac-
tions, over 75 percent of the Variation in the frequency with which students are
found to be relating to others at random moments can be accounted for from

the three predictors. The remaining Variation appears largely due to the environ-
mental conditions that made interaction impossible. Such a result may be a
fluke, since it suggests a much higher degree of predictability for human behav-
ior than has been thought possible, but it serves to underline the importance of
doing many more such multivariate studies if one is seriously interested in ac-
counting for the variability in human behavior. Psychologists have expended far
too much of their energies studying one variable at a time.
What these studies also indicate is that whereas the values people espouse
may be slightly influenced by motive dispositions, they are essentially indepen-
dent determinants of action outcomes, just as probability of success is. Value
and motive measures are not highly correlated, and they predict different action
outcomes singly or in combination with other variables.
• SOME ISSUES NEEDING FURTHER CLARIFICATION

While some theoretical and factual progress has been made in our understanding
of human motivation, much more remains to be made. As we have reviewed
milestones in what has been learned about various aspects of motivation, we
have often concluded by observing that we need to know more about this mat-
ter. Certainly there will continue to be refinements in the way motives are de-
fined and measured and in our understanding of how they influence behavior,
but it is also useful to call attention to major gaps in our knowledge and to
speculate about matters that should engage the attention of psychologists in-
creasingly in the future. What are some of the major issues that especially need
clarification?
Examining the Biological Basis of Major Motive Systems

We have repeatedly argued on theoretical and empirical grounds that motives
are based on natural incentives, specific to the human species, which innately
arouse or "turn on" the individual in different ways, as reflected in a limited
number of types of emotions. Yet the evidence for such a proposition is not as
direct or conclusive as it ought to be. Most of it is circumstantial or indirect,
based on inferences from such diverse phenomena as language learning in chil-
dren, intrinsically satisfying activities, what produces smiling in infants, the uni-
versality of a few human emotions, and the involvement of older parts of the
brain in emotion and motivation, as summarized in Chapters 4 and 5.
Furthermore, in a few cases the evidence for innate biological involvement
in a motive System is quite direct. This is true for the hunger motive, in which
hypothalamic arousal occurs directly in response to lowered available blood
sugar or to the sight or smell of tasty food. It is also true for the sexual motive,
in which the production of sex hormones at puberty clearly provides the biologi-
cal basis for a type of affective arousal in connection with sexual Stimuli that
makes learnings associated with it much stronger and more basic than they
would be if they had a purely cognitive basis. And there is some beginning evi-
dence, as noted in Chapter 8, that the catecholamine hormonal System, particu-
larly brain norepinephrine turnover, provides the biological Substrate that pro-
duces the affect associated with the impact or power incentive.
What is exciting about this type of evidence is that it suggests that the bio-
logical aspect of different motive Systems may be quite specific. Nearly all earlier
theories of motivation involved some type of physiological or cortical arousal,
but it was conceived in very general terms. Thus, drive—whether as tension or
anxiety—was thought to result in general psychophysiological arousal, and dif-
ferent emotions seemed to have more or less the same physiological effects. But
newer, more sensitive techniques of biochemical analysis have begun to produce
evidence that the effects of diflferent motivational or emotional states on hor-
mones may be quite diflferent. So it could turn out that each of the major motive
Systems is based on a diflferent natural incentive, as demonstrated by uniquely
diflferent hormonal eflfects of each natural incentive.
Clearly, what is needed is a much better way of determining what natural
incentives exist for the human species, how they produce diflferent types of aflfec-
tive arousal represented by varying profiles of hormone release, and how specific
types of aflfective arousal form the basis for developing major motive Systems
through cognitive development and associative learning. Working out these rela-
tionships is a problem of the greatest importance to understanding the nature of
human motives and will require great ingenuity to solve.
Defining and Measuring Other Major Motive Systems

Progress in understanding the achievement, power, affiliative, and avoidance mo-
tives has been considerable, but it inevitably raises several questions. Why have
these motives been studied and not others? Are they the most important mo-
tives? What other motives need study? Such questions cannot be answered with
any certainty.
Work on the achievement motive began because of a historical accident:
psychologists at the time had had some experience in influencing Performance
through what they then called ego-involving instructions. It was soon realized
that these instructions set higher achievement goals for the subjects and pro-
duced changes in associative thought that came to be called the achievement mo-
tive. Yet if one were to imagine coding all the thoughts that all the people in
the world are having at this moment, or have had over time, the frequency of
achievement-related thoughts would certainly be lower than the frequency of
thoughts related to power or affiliation. Empirical evidence for this conclusion is
to be found in the scoring of hundreds of pages of fiction, children's textbooks,
and hymns throughout the history of the United States (McClelland, 1975).
Power-related thoughts were the most common; afüliation-related thoughts, next
in frequency; and achievement-related thoughts, least frequent. To come at it
from a diflferent angle, if one were to code the imagery in great legends or sa-
cred texts like the Bhagavad Gita or the Bible, one would find that they are pri-
marily about power and love. Achievement concerns play a minor role. This is
Mliestones in the Progress 603
not to say that achievement concerns are unimportant: they seem to be a key
factor in technological and economical progress, but they are of lesser interest to
most people most of the time than power and affiliative concerns.
What about other motives? There are a number crying out for investigation,
such as the need to nurture or be nurtured, curiosity, or simply the need to
maintain consistency and avoid dissonance. In the last instance, psychologists
have identified what appears to be a strong natural incentive but have not devel-
oped a measure of individual differences in motive strength centered on this
incentive.
There is no lack of motives to be investigated. What has inhibited people
from investigating them by the methods proposed here are doubts about the reli-
ability of the measures, which have only recently been put to rest; the effort re-
quired to develop a motive scoring System through arousal experiments, which
must then be validated behaviorally; and the training and time needed to get re-
liable coding of associative thought, as compared to the minimal effort needed to
score responses on questionnaires, which can be coded instantly by machine. Yet
the great expenditure of time and effort that developing coding Systems for asso-
ciative thought requires appears to be fully justified in terms of the greater theo-
retical and practical value of the measures of motive strength obtained in this
way.
Determining What Collective Motive Levels Mean

A surprising virtue of measuring motive Systems by coding thought turned
out to be that applying the same codes to the populär literature (that is, the
thoughts produced by a society) appears to give a useful estimate of motive
levels among cultures or nations. Although questionnaires and other tests cannot
be administered to individuals from long past historical periods to measure the
strength of their motives, the cultures in which they lived have left behind
songs, stories, legends, literary documents, and even designs on vases that can
and have been scored for motivational Systems in the same way records from in-
dividuals living at the present time have been scored. What is surprising is that
nations whose records score high in a particular motive behave like individuals
whose records score high on the same motive. The parallels are striking. Indi-
viduals whose stories score high in n Achievement show the characteristics of
successful entrepreneurs, and small businesses headed by people higher in
n Achievement grow faster. A high level of n Achievement in the folk tales of
preliterate cultures signifies the presence of more entrepreneurs in those cultures
than in cultures whose folk tales score low in n Achievement. Furthermore, na-
tions whose children's stories score higher in n Achievement tend to grow more
rapidly economically, and in several countries, rates of economic growth were
greater after historical periods when n Achievement levels were high in populär
literature than they were when n Achievement levels were low.
Men who write stories high in n Power and low in Activity Inhibition tend
to drink more heavily, and drinking tends also to be heavier in cultures whose
folk tales score high in n Power and low in Activity Inhibition. Individuals scor-
ing high in n Power and high in Activity Inhibition tend to make good manag-
ers of large, complex organizations, and countries with the same motivational
pattern show signs of being better organized in several ways. People scoring
high in n Affiliation like people and show signs of being in better health. Coun-
tries scoring high in n Affiliation show signs of respecting the rights of individu-
als more, and their public health figures indicate a lower incidence of several
physical illnesses.
What do these parallels mean? Can we think in terms of a group mind that
has motivational characteristics analogous to those possessed by an individual?
Or are estimates of national motive levels obtained from group literary products
a useful approximation of average motive levels that would be obtained from na-
tional representative sample surveys of individuals in the countries concerned?
Or is the ideational content of populär literature somehow responsible for levels
of individual motivation in the country? These are all questions that need fur-
ther investigation.
The important point is that the method of measuring individual motive
strength is applicable to group products and therefore opens up the possibility of
all sorts of historical and cross-sectional studies of the influence of levels of mo-
tivational ideology on the behavior of a country or groups within a country.
Different types of literature are read by difFerent Segments of a society, so it
should be possible to estimate motive levels of these different Segments by cod-
ing literature that is populär in each segment. Would, for example, levels of mo-
tivation in men's and women's magazines in 1957 and again in 1976 parallel the
average levels of motive strength among men and women at the two time peri-
ods, as determined by scoring individual protocols in the national sample sur-
veys conducted at these two time periods? Would the motivational content of
songs populär for different generations in the United States predict what that
generation would spend its money on, how many children it would have on the
average, or how susceptible it would prove to be to death from various illnesses?
The possibilities of this type of analysis appear almost limitless. Psychology has
made available to historians and sociologists a new tool that can measure moti-
vational levels in the past or in various Segments of society, which can be used
to interpret what happens in a country in much the same way as economic sta-
tistics were used by social scientists when they became available to explain social
developments.
Improving the Theoretical and Practical Understanding

of How to Develop Motives
Motives are learned, although they are based on natural incentives, which have
an innate component. If they are learned, what kinds of teaching or learning
promote their development? Some evidence points to the importance of the em-
phasis parents place on certain behaviors in early childhood or, in the case of
the achievement motive, in middle childhood. In that instance, if the parents set
high Standards for Performance and the father does not direct too closely the
child's effort to achieve these goals, the development of the achievement motive
is facilitated.
This only raises two further questions. If parents are in effect developing
motives or teaching them unconsciously by what they do, why do some parents
behave that way and others do not? The answer to the question is by no means
well worked out, but it seems to involve Special experiences the parents have
had as members of minority groups or reformist groups that have provided
them with an ideology that leads them to emphasize one aspect of child training
over another. However, the connection is by no means obvious or straightfor-
ward. Because parents want their child to have achievement motivation, they
may insist that the child do well at all tasks early, but this does not automati-
cally result in a higher level of adult n Achievement. The child training of im-
portance may appear to be quite unrelated to achieving and involve scheduling
feeding and severe toilet training. If so—if the development of a motive depends
on such chancy matters as the exposure of a parent to reformist ideology or the
unavailability of disposable diapers, which makes severe toilet training more de-
sirable, why leave the matter to chance? Why not teach the motives directly?
Can they be taught?
Many efforts have been made to teach motivation, with varying degrees of
success. Generally speaking, German psychologists have concentrated on the
cognitive aspects of achievement motivation, teaching students to set moderate
goals and to attribute failure to lack of effort, so as generally to build self-
confidence or the perceived probability of success variable in the formula for
predicting Performance outcomes. They did not include the associative thought
measure of n Achievement as part of the instruction, partly because they wanted
to reserve it to determine whether motive strength had in fact been increased,
and partly because the cognitive aspects of behavior change were easier to intro-
duce into the schoolroom. In general, they found significant changes in the cog-
nitive variables they were trying to influence, and at times they also produced
an increase in achievement motive strength. However, the long-term effects of
the motivation training on Performance in school were seldom significant, and
they attributed this to the fact that school does not provide opportunities for the
proper exercise of the achievement motive.
Another possible explanation for the result is that by concentrating on cog-
nitive changes, the German psychologists were influencing situation-dependent,
or respondent, behavior, whereas if they had concentrated more on changing the
motive itself, they might have influenced spontaneous, nonsituation-dependent
(or operant) behaviors, which in the long run might have made more of a differ-
ence in school Performance. To put the matter simply, if pupils have learned to
set moderate goals, it is necessary that their teacher provide them with opportu-
nities to choose moderate goals if the acquired behavior is to result in an im-
provement in Performance. But since teachers do not regularly do this, learning
to set moderate goals may have little effect. In contrast, if pupils are taught
achievement thinking, or Origin thinking, they may spontaneously think of mod-
erate goals they can set without being given the opportunity by the teacher. The
evidence for such an inference is by no means conclusive, since much achieve-
ment motivation training in the United States, as well as in Germany, has been
unsuccessful in improving academic Performance. The matter cries out for fur-
ther investigation.
For the most part, U.S. psychologists started motivation training courses by
teaching participants the scoring System for the various motives so that they
could learn to think in appropriate ways. The cognitive and value aspects of the
motivation-action sequence were also taught, as in Germany. The courses in
achievement motivation were successful in improving entrepreneurial behavior
for small businesspeople in a number of different countries around the world
and among minority businesspeople in the United States. The improvement
seems to be due to a shift from the fear of failure to the hope of success aspect
of achievement motivation, and also to strengthening the other determinants of
an action outcome—namely, the perceived probability of success and the value
placed on business improvement. Effects of achievement motivation training on
academic Performance have sometimes been successful and sometimes not, per-
haps because schools do not provide opportunities to exercise the achievement
motive, as the German psychologists argue, or perhaps because doing well in
school is more in the service of the power motive than the achievement motive.
The most successful type of motivation training for improving academic perfor-
mance is Origin training, as developed by deCharms (1976), and it contains an
element of power motivation training.
Power motivation training has also shown some preliminary success in
changing individuals for whom the power motive is a key element in their per-
formance. Since high n Power and low inhibition is associated with heavy drink-
ing in men, it was reasoned that training emphasizing inhibition or socializing
the power motive would help male alcoholics recover, and this turned out to be
the case. And since people high in n Power and high in Activity Inhibition tend
to make better managers, motivation training emphasizing this motive syndrome
ought to improve managerial behavior. This also turned out to be the case. But
the findings are preliminary and greatly in need of further confirmation.
The fact is that despite a great deal of effort and experience, we still do not
know whether psychologists have succeeded in developing motives, although
they ha/e certainly changed behavior. Nor do we know precisely how motives
are most easily developed, probably because we do not yet have a clear grasp of
the natural course of development of motives as children grow up. Further re-
search in the area of motive development is greatly needed.
• THE RELATIONSHIP OF PROGRESS IN PSYCHOLOGY

TO ITS ROLE IN SOCIETY
Psychologists who have attempted to develop motives have encountered many
difficulties, both of a theoretical and a practical nature. Their experience high-
lights the relationship of progress in psychology to its role in society. On the
theoretical side, the pressure, for example, to produce improved Performance in
school has at times been so great that it has clouded the issue of whether the
training is developing motives. Or the interest in getting results in improved per-
formance has diverted attention from the theoretical issues of just what educa-
tional inputs are producing the changes. On the practical side, ethical questions
have been raised as to whether it is proper to try to influence children's motives,

and there has been some confusion as to who should be in charge of motivation
training programs. Furthermore, psychologists may be interested in developing
training programs but not be particularly skilled at or interested in continuing
to deliver them. Even when training has shown promise, it is not clear what
agents or institutions should provide the training. If power motivation training
continues to show promise in rehabilitating alcoholics, who will provide it? Al-
coholism in our society is defined as a disease, and therefore doctors usually
provide treatment for alcoholism. But doctors do not think of themselves as ed-
ucators or trainers and are rarely willing to prepare themselves to offer motiva-
tion training.
The problem has been most evident in the case of achievement motivation
training for small business entrepreneurs, which has been demonstrated over and
over again in different countries to be effective so far as the participants are con-
cerned, and also very probably so far as the Community is concerned in which
they work, and ultimately so far as the nation is concerned of which the com-
munity is a part. Despite all this evidence, achievement motivation training for
entrepreneurs is seldom offered on a regulär basis anywhere in the world today.
Dozens of individuals in the public and private sector have been trained to give
the courses and have given them successfully in many different countries. Typi-
cally, they continue to offer the courses only so long as a Special research grant
is available to pay for their Services.
The training is of obvious benefit to the entrepreneurs, but not necessarily to
an educational institution that provides it, particularly since the training courses
are short and not part of the regulär curriculum the institution provides. So the
courses are discontinued. Or the trainers have become upwardly mobile because
of their greater achievement orientation, and they get promoted to better Jobs
and stop training. The small businesspeople themselves cannot afford to pay the
amount of money it would take to provide the courses without a research subsi-
dy. Banks or public agencies supposedly fostering small business have failed to
provide the training, partly because they do not think of themselves as educators
and partly because psychological interventions simply do not seem as important
or as easily made as economic changes brought about by the infusion of more
capital, a decrease in the tax rate, an improvement in the terms of trade, or the
exploitation of a new natural resource.
This story provides a case study in the relationship of psychology to society.
Ultimately, progress in psychology depends on the extent to which it is valued
and used by society. It has not always been regarded as a worthwhile field of
scientific endeavor and in fact is so regarded in only a few countries in the
world today, including the English-speaking countries, Scandinavia, The Nether-
lands, West Germany, Japan, and India. With the support it has received in
these countries, the psychology of motivation has made real progress in the last
generation. People in this field have discovered much new knowledge, some of it
of practical use to society. If society can accept such contributions as valuable,
the promise of psychology will appear even greater, and the field will get the fi-
nancial and institutional support and the commitment from individual investiga-
tors it needs to study the many questions about human motivation that still
need answers.
For, while some progress has been made in measuring a few human motives
and in understanding their origins and effects on behavior, much remains to be
discovered and clarified. The scientific study of human motives is still at an
early stage.
Bibliography
Abelson, R. P. Psychological Status of the Script concept. American Psychologist, 1981, 36, 715-729.
Ach, N. Über die Willensakt und das Temperament. Leipzig: Quelle and Meyer, 1910.
Adams, W. I., Lorens, S. A., & Mitchell, C. L. Morphine enhances lateral hypothalamic
self-stimulation in the rat. Proceedings of the Society for Experimental Biology, 1972, 140,
770-77'1.
Adler, A. Neurotic constitution. New York: Moffat, Ward, 1917.
Adorno, T. W., Frenkel-Brunswik, E., Levinson, D. J., & Sanford, R. N. The authoritarian
Personality. New York: Harper, 1950.
Ainsworth, M. D. S. The development of the infant-mother attachment. In B. M. Caldwell & H. N.
Ricciuti (Eds.), Review of child development research (Vol. 3). Chicago: University of Chicago
Press, 1973.
Alexander, C. N. Ego development, personality, and behavioral change in inmates practicing the
transcendental meditation technique: A cross-sectional and longitudinal study. Unpublished
doctoral dissertation, Harvard University, 1982.
Allen, G. J., & Condon, T. J. Whither subliminal psychodynamic activation studies? Journal of
Abnormal Psychology, 1982,97, 126-130.
Allport, G. W. Personality: A psychological Interpretation. New York: Holt, 1937.
Allport, G. W. Effect: A secondary principle of learning. Psychological Review, 1946, 53, 335-347.
Allport, G. W. The trend in motivational theory. Journal of Orthopsychiatry, 1953, 23, 107-119.
Alschuler, A. S. Developing achievement motivation in adolescents. Englewood Cliffs, N.J.:
Educational Technology Publications, 1973.
Amoroso, D. M., Brown, M., Pruesse, M., Ware, E. E., & Pilkey, D. W. An investigation of
behavioral and psychological reactions to pornographic Stimuli. In Technical report of the
Commission on Obscenity and Pornography (Vol. 8). Washington, D.C.: U.S. Government
Printing Office, 1971.
Andrews, G. R., & Debus, R. L. Persistence and the causal perception of failure: Modifying
cognitive attributions. Journal of Educational Psychology, 1978, 70, 154-166.
Andrews, J. D. W. The achievement motive in lifestyle among Harvard freshmen. Unpublished
doctoral dissertation, Harvard University, Department of Social Relations, 1966.
Andrews, J. D. W. The achievement motive in two types of organizations. Journal of Personality
and Social Psychology, 1967,6, 163-168.
609
610 Bibliography
Angelini, A. L. Studies in projective measurement of achievement motivation of Brazilian students,

males and females: Proceedings of the 15th International Congress of Psychology, Brüssels,
1957. Acta Psychologica, 1959, 15, 359-360.
Argyle, M., & Cook, M. Gaze and mutual gaze. Cambridge, England: Cambridge University Press,
1976.
Armour, D. J., Polich, J. M., & Stamboul, H. B. Alcoholism and treatment. New York: Wiley, 1978.
Aronoff, J. Psychological needs and cultural Systems. New York: Van Nostrand, 1967.
Aronoff, J. The cane-cutters of St. Kitts. Psychology Today, August 1971, pp. 53-55.
Aronson, E. The need for achievement as measured in graphic expression. In J. W. Atkinson (Ed.),
Motives in fantasy, action, and society. Princeton, N.J.: Van Nostrand, 1958.
Asch, S. E. Effects of group pressure upon the modification and distortion of judgments. In H.
Guetzkow (Ed.), Groups, leadership, and men. Pittsburgh: Carnegie Press, 1951.
Atkinson, J. W. The achievement motive and recall of interrupted and completed tasks. In D. C.
McClelland (Ed.), Studies in motivation. New York: Appleton-Century-Crofts, 1955.
Atkinson, J. W. Motivational determinants of risk-taking behavior. Psychological Review, 1957, 64,
359-372.
Atkinson, J. W. Towards experimental analysis of human motivation in terms of motives,
expectancies, and incentives. In J. W. Atkinson (Ed.), Motives in fantasy, action, and society.
Princeton, N.J.: Van Nostrand, 1958.
Atkinson, J. W. Introduction to motivation. New York: Van Nostrand, 1964.
Atkinson, J. W. Thematic apperceptive measurement of motivation in 1950 and 1980. In G.
d'Ydevalle & W. Lens (Eds.), Cognition in human motivation and learning. Hillsdale, N.J.:
Erlbaum, 1980.
Atkinson, J. W., & Birch, D. The dynamics of action. New York: Wiley, 1970.
Atkinson, J. W., & Birch, D. Introduction to motivation (2nd ed.). New York: Van Nostrand, 1978.
Atkinson, J. W., Bongort, K., & Price, L. H. Explorations using Computer Simulation to
comprehend TAT measurement of motivation. Motivation and emotion, 1977, 1, 1-27.
Atkinson, J. W., Heyns, R. W., & Veroff, J. The effect of experimental arousal of the affiliation
motive on thematic apperception. Journal of Abnormal and Social Psychology, 1954, 49,
405-410.
Atkinson, J. W., & Litwin, G. H. Achievement motive and test anxiety conceived as motive to
approach success and motive to avoid failure. Journal of Abnormal and Social Psychology,
1960, 60, 52-63.
Atkinson, J. W., & McClelland, D. C. The effects of different intensities of the hunger drive on
thematic apperception. Journal of Experimental Psychology, 1948, 38, 643-658.
Atkinson, J. W., & O'Connor, P. Neglected factors in studies of achievement-oriented Performance:
Social approval as an incentive and Performance decrement. In J. W. Atkinson & N. T.
Feather (Eds.), A theory of achievement motivation. New York: Wiley, 1966.
Atkinson, J. W., & Raphelson, A. C. Individual differences in motivation and behavior in particular
situations. Journal of Personality, 1956,24, 351-363.
Atkinson, J. W., & Reitman, W. R. Performance as a function of motive strength and expectancy of
goal attainment. Journal of Abnormal and Social Psychology, 1956, 53, 361-366.
Atkinson, J. W., & Walker, E. L. The affiliation motive in perceptual sensitivity to faces. Journal of
Abnormal and Social Psychology, 1956, 53, 38-41.
Axelrod, J. Neurotransmitters. Scientific American, 1974, 230(6), 59-71.
Bakan, D. Sigmund Freud and the Jewish mystical tradition. New York: Van Nostrand, 1958.
Bibliography 611
Bakan, D. The duality ofhuman existence. Boston: Beacon Press, 1966.

Bakan, P. Two streams of consciousness: A typological approach. In K. S. Pope & J. L. Singer
(Eds.), The stream of consciousness. New York: Plenum, 1978.
Bales, R. F., & Cohen, S. P. SYMLOG: A System for the multiple level observation of groups. New
York: Free Press, 1979.
Bandura, A. Self-efficacy mechanism in human agency. American Psychologist, 1982, 37, 122-147.
Barclay, A. M. The effect of hostility on physiological and fantasy responses. Journal of Personality,
1969, 37, 651-667.
Barclay, A. M. Linking of sexual and aggressive motives: Contributions of "irrelevant" arousals.
Journal of Personality, 1971,59, 481-492.
Barclay, A. M., & Haber, R. N. The relation of aggression to sexual motivation. Journal of
Personality, 1965, 33, 462-475.
Baron, J. Menstrual hormone changes and instinctual tendencies in dreams. Motivation and Emotion,
1977, 1, 273-282.
Barrett, G. V., & Franke, R. H. Psychogenic death: A reappraisal. Science, 1970, 167, 304-306.
Bartmann, T. Denkerziehung im Programmierten Unterricht. Munich: Manz, 1965.
Beardslee, D. C , & Fogelson, R. Sex differences in sexual imagery aroused by musical Stimulation.
In J. W. Atkinson (Ed.), Motives in fantasy, action, and society. Princeton, N. J.: Van
Nostrand, 1958.
Bell, A. P., Weinberg, M. S., & Hammersmith, S. K. Sexual preferenee: Its development in men and
women. Bloomington, Ind.: Indiana University Press, 1981.
Benesh, M., & Weiner, B. On emotion and motivation: From the notebooks of Fritz Heider.
American Psychologist, 1982, 37, 887-895.
Benson, H. The relaxation response. New York: Morrow, 1975.
Benson, H., & Epstein, M. D. The placebo effect: A neglected asset in the care of patients. Journal
of the American Medical Association, 1975, 232, 1225-1227.
Berelson, B., & Steiner, G. A. Human behavior: An inventory of scientificßndings. New York:
Harcourt, Brace and World, 1964.
Berlew, D. E. The achievement motive and the growth of Greek civilization. Unpublished honors
thesis, Wesleyan University, 1956.
Berlew, D. E. A study of interpersonal sensitivity. Unpublished doctoral dissertation, Harvard
University, 1959.
Berlew, D. E., & Williams, A. F. Interpersonal sensitivity under motive arousing conditions. Journal
of Abnormal and Social Psychology 1964,65, 150-159.
Berlyne, D. E. Arousal and reinforcement. In D. Levine (Ed.), Nebraska Symposium on Motivation:
1967. Lincoln: University of Nebraska Press, 1967.
Berlyne, D. E., & Madsen, K. B. Pleasure, reward, preference. New York: Academic Press, 1973.
Beswick, D. G. Theory and measurement of human curiosity. Unpublished doctoral dissertation,
Harvard University, 1965.
Bieber, I., Dain, H. J., Dince, P. R., Dreilich, M. G., Grand, H. G., Gundlach, R. H., Kremer, M.
W., Rifkin, A. H., Wilbur, C. B., & Bieber, T. B. Homosexuality: A psychoanalytic study. New
York: Basic, 1962.
Birch, H. G. The role of motivational factors in insightful problem-solving. Journal of Comparativc
Psychology, 1945, 38, 295-317.
Birk, L., Huddieston, W., Millers, E., & Cohler, B. Avoidance conditioning for homosexuality.
Archives of General Psychiatry, 1971,25, 314—323.
612 Bibliography
Birney, R. C , Burdick, H., & Teevan, R. C. Fear of failure motivation. New York: Van Nostrand,
1969.
Blehar, M. C , Lieberman, A. F., & Ainsworth, M. D. S. Early face-to-face interaction and its
relationship to later mother-infant attachment. Child Development, 1977, 48, 182-194.
Blodgett, H. C. The effect of the introduction of reward upon the maze Performance of rats.
University of California Publications in Psychology, 1929, 4(8), 113-134.
Borg, S., Kvande, H., & Sedvall, G. Central norepinephrine metabolism during alcohol intoxication
in addicts and healthy volunteers. Science, 1981, 213, 1135-1137.
Bowlby, J. Attachment and loss (Vol. 1), Attachment. New York: Basic, 1969.
Boyatzis, R. E. A two factor theory of affiliation motivation. Unpublished doctoral dissertation,
Boyatzis, R. E. Affiliation motivation. In D. C. McClelland & R. S. Steele (Eds.), Human
motivation: A book of readings. Morristown, N.J.: General Learning Press, 1973.
Boyatzis, R. E. Power motivation training workbook. Boston: McBer and Company, 1974.
Boyatzis, R. E. The competent manager: A model for ejfective Performance. New York: Wiley,
1982.
Bradburn, N. M , & Berlew, D. E. Need for achievement in English economic growth. Economic
Development and Cultural Change, 1961, 10, 8-20.
Bramante, M. Sex differences in fantasy patterns: A replication and elaboration. Unpublished
doctoral dissertation, City University of New York, 1970.
Bray, D. W., Campbell, R. J., & Grant, D. L. Formative years in business: A long term study of
managerial lives. New York: Wiley, 1974.
Brazelton, T. B., Koslowski, B., & Main, M. The origins of reciprocity: The early mother-infant
interaction. In M. Lewis & L. A. Rosenblum (Eds.), The effect of the infant on the caregiver.
New York: Wiley, 1974.
Breit, S. Arousal of achievement motivation with causal attributions. Psychological Reports, 1969, 25,
539-542.
Bremer, J. Asexualization. New York: Macmillan, 1959.
Briddell, D. W„ Rimm, D. C , Caddy, G. R., Krawitz, G., Sholis, D., & Wunderlin, R. J. Effects of
alcohol and cognitive set on sexual arousal. Journal of Abnormal Psychology, 1978, 87,
418-430.
Brophy, J. E., & Good, T. L. Teacher's communication of differential expressions for children's
classroom Performance: Some behavioral data. Journal of Educational Psychology, 1970, 61,
365-374.
Broverman, D., Klaiber, E. L., Kobayashi, Y., & Vogel, W. Roles of activation and inhibition in sex
differences in cognitive abilities. Psychological Review, 1968, 75, 23-50.
Broverman, D. M., Klaiber, E. L., Vogel, W., & Kobayashi, Y. Short-term versus long-term effects
of adrenal hormones on behaviors. Psychological Bulletin, 1974, 81, 672-694.
Brown, J. S. Gradients of approach and avoidance responses and their relation to level of
motivation. Journal of Comparative and Physiological Psychology, 1948, 41, 450-465.
Brown, J. S. The motivation of behavior. New York: McGraw-Hill, 1961.
Brown, N. O. Life against death. Middletown, Conn.: Wesleyan University Press, 1959.
Brown, R. A first language: The early stages. Cambridge, Mass.: Harvard University Press, 1973.
Bruner, J. S., & Goodman, C. C. Value and need as organizing factors in perception. Journal of
Abnormal and Social Psychology, 1947, 42, 33—44.
Bruner, J. S., Matter, J., & Papanek, M. Breadth of learning as function of drive level and
mechanization. Psychological Review, 1955, 62, 1-10.
Bibliography 613
Brunswik, E., & Reiter, L. Eindruckscharaktere schematisierter Gesichter. Zeitschrift für Psychologie,
1938, 142, 67-134.
Buber, M. Between man and man. New York: Macmillan, 1965.
Buber, M. / and thou. New York: Scribner, 1970.
Burdick, H. A., & Burnes, A. J. A test of "strain towards symmetry" theories. Journal of Abnormal
Bursztajn, H., Feinbloom, R. I., Hamm, R. M., & Brodsky, A. Medical choices, medical chances:
How patients, families, and physicians can cope with uncertainty. New York: Delacorte, 1981.
Burton, M. J., Mora, F., & Rolls, E. T. Visual and taste neurons in the lateral hypothalamus and
substantia innominata: Modulation of responsiveness by hunger. Journal of Physiology, 1975,
252, 50-51.
Burwen, L. S., & Campbell, D. T. The generality of attitudes toward authority and non-authority
figures. Journal of Abnormal and Social Psychology, 1951,54, 24-31.
Butler, R. A. Incentive conditions which influence visual exploration. Journal of Experimental
Psychology, 1954, 48, 19-23.
Byrne, D. Anxiety and the experimental arousal of affiliation need. Journal of Abnormal and Social
Psychology, 1961, 63, 660-662. (a)
Byrne, D. Interpersonal attraction as a function of affiliation need and attitude similarity. Human
Relations, 1961, 14, 283-289. (b)
Byrne, D. Response to attitude similarity-dissimilarity as a function of affiliation need. Journal of
Personality, 1962, 30, 164-177.
Byrne, D., McDonald, R. D., & Mikawa, J. Approach and avoidance affiliation motives. Journal of
Personality, 1963, 31, 21-37.
Cannon, W. B. Bodily changes in pain, hunger, fear, and rage: An account of recent researches into
the function of emotional excitement. New York: Appleton-Century-Crofts, 1915.
Cannon, W. B., & Washburn, A. L. An explanation of hunger. American Journal of Psychology,
1912, 29, 441-154.
Capote, T. In cold blood: A true account of a multiple murder and its consequences. New York:
Random, 1965.
Carlson, N. R. Physiology of behavior. Boston: Allyn and Bacon, 1977.
Cattell, R. B. Personality and motivation: Structure and measurement. Yonkers, N.Y.: World Book,
1957.
Cattell, R. B. The scientific analysis of personality. Baltimore, Md.: Penguin, 1965.
Cattell, R. B., Kawash, G. F., & DeYoung, G. E. Validation of objective measures of ergic tension:
Response of the sex urge to visual Stimulation. Journal of Experimental Research on
Personality, 1972, 6, 76-83.
Ceranski, D. S., Teevan, R., & Kalle, R. J. A comparison of three measures of the motive to avoid
failure: Hostile press, test anxiety, and resultant achievement motivation. Motivation and
Emotion, 1979, 3, 395-104.
Chanowitz, B., & Langer, E. Knowing more (or less) than you can show: Understanding control
through the mindlessness/mindfulness distinction. In M. E. P. Seligman & J. Garber (Eds.),
Human helplessness. New York: Academic Press, 1982.
Chapin, M., & Dyck, D. G. Persistence in children's reading behavior as a function of N length and
attribution retraining. Journal of Abnormal Psychology, 1976, 85, 511-515.
Child, I. L., Frank, K. F., & Storm, T. Self-ratings and TAT: Their relations to each other and to
adulthood background. Journal of Personality, 1956,25, 96-114.
Cicchetti, D., & Sroufe, L. A. The relationship between affective and cognitive development in
Down's syndrome infants. Child Development, 1976, 47, 920-929.
614 Bibliography
Clark, R. A. The projective measurement of experimentally induced levels of sexual motivation.

Journal of Experimental Psychology, 1952,44, 391-399.
Clark, R. A. The effects of sexual motivation on phantasy. In D. C. McClelland (Ed.), Studies in
motivation. New York: Appleton-Century-Crofts, 1955.
Clark, R. A., & McClelland, D. C. A factor analytic integration of imaginative and Performance
measures of the need for Achievement. Journal of General Psychology, 1956, 55, 73-83.
Cofer, C. N. Motivation and emotion. Glenview, 111.: Scott, Foresman, 1972.
Combs, A. W. A comparative study of motivations as revealed in Thematic Apperception stories
and autobiography. Journal of Clinical Psychology, 1947, 3, 65-75.
Comfort, A. The joy of sex. New York: Simon and Schuster, 1974.
Condon, W. S. An analysis of behavioral organization. In S. Weitz (Ed.), Non-verbal communication:
Readings with commentary (2nd ed.). New York: Oxford University Press, 1979.
Conners, C. K. Needs for affiliation and response to social influence. Unpublished doctoral
dissertation, Harvard University, 1961.
Constantian, C. A. Motives and dimensions as predictors of behavior in a small group. Unpublished
manuscript, Harvard University, Department of Psychology and Social Relations, 1978.
Constantian, C. A. Attitudes, beliefs, and behavior in regard to spending time ahne. Unpublished
doctoral dissertation, Harvard University, Department of Psychology and Social Relations,
1981.
Corbett, D., La Fernere, A., & Miliver, P. M. Plasticity of the medial prefrontal cortex: Facilitated
acquisition of intracranial self-stimulation by pretraining self-stimulation. Physiology and
Behavior, 1982, 28, 531-534.
Cortes, J. B. The achievement motive in the Spanish economy between the 13th and 18th centuries.
Economic Development and Cultural Change, 1960, 9, 144-163.
Cortes, J. B., & Gatti, F. M. Delinquency and crime: A biopsychosocial approach. New York:
Seminar Press, 1972.
Costa, P., & McClelland, D. C. Predicting rank in class from motivational, social class, and
intelligence measures. Unpublished manuscript, Harvard University, Department of Psychology
and Social Relations, 1971.
Cottrell, N. B. Social facilitation. In C. G. McClintock (Ed.), Experimental social psychology. New
York: Holt, Rinehart and Winston, 1972.
Cottrell, N. B., Wack, D. L., Sekerak, G. J., & Rittle, R. Social facilitation of dominant responses
by the presence of an audience and the mere presence of others. Journal of Personality and
Social Psychology, 1968, 9, 245-250.
Couch, A. S., & Keniston, K. Yea-sayers and nay-sayers: Agreeing response set as a personality
variable. Journal of Abnormal and Social Psychology, 1960, 60, 151-174.
Courts, F. A. Relations between experimentally induced muscular tension and memorization.
Cramer, P., & Hogan, K. Sex differences in verbal and play fantasy. Developmental Psychology,
1975, 11, 145-154.
Crespi, L. P. Quantitative Variation of incentive and Performance in the white rat. American Journal
of Psychology, 1942, 55, 467-517.
Crockett, H. J., Jr. The achievement motive and differential occupational mobility in the United
States. American Sociological Review, 1962, 27, 191-204.
Cronbach, L. J., & Meehl, P. E. Construct validity in psychological tests. Psychological Bulletin,
1955, 52, 281-302.
Crowne, D. P., & Marlowe, D. The approval motive. New York: Wiley, 1964.
Csikszentmihalyi, M. Beyond boredom and anxiety. San Francisco: Jossey-Bass, 1975.
Bibliography 615
Cutter, H. S. G., Boyatzis, R. E., & Clancy, D. D. The effectiveness of power motivation training
for rehabilitating alcoholics. Journal of Studies in Alcohol, 1977, 38(\), 131-141.
Cytowic, R. E. The long ordeal of James Brady. New York Times Magazine, September 27,
1981.
D'Andrade, R. G., Dart, J. The interpretation of r versus r2. Unpublished manuscript, University of
California at Irvine, Department of Anthropology, 1982.
Darley, J. M., & Batson, C. D. "From Jerusalem to Jericho": A study of situational and
dispositional variables in helping behavior. Journal of Abnormal and Social Psychology, 1973,
27, 100-108.
Davidson, R. J., Saron, C , & McClelland, D. C. Effects of personality and semantic content of
Stimuli on augmenting and reducing in the event-related potential. Biological Psychology, 1980,
11, 249-255.
Davies, E. This is the way that Crete went-Not with a bang but a simper. Psychology Today, June
1969, pp. 4 3 ^ 7 .
deCharms, R. Affiliation motivation and productivity in small groups. Journal of Abnormal and
Social Psychology, 1957,55, 222-226.
deCharms, R. Personal causation. New York: Academic Press, 1968.
deCharms, R. Enhancing motivation: Change in the classroom. New York: Irvington, 1976.
deCharms, R. The origins of competence and achievement motivation in personal causation. In L. J.
Fyans, Jr. (Ed.), Achievement motivation: Recent trends in theory and research. New York:
Plenum, 1980.
deCharms, R. That's not psychology! The implications of McClelland's approach to motivation. In
A. J. Stewart (Ed.), Motivation and society. San Francisco: Jossey-Bass, 1982.
deCharms, R., & Carpenter, V. Measuring motivation in culturally disadvantaged school children. In
H. J. Klausmeirer & G. T. O'Hearn (Eds.), Research and development toward the improvement
of education. Madison, Wis.: Educational Research Services, 1968.
deCharms, R., & Dave, P. N. Hope of success, fear of failure, subjective probability, and risk-taking
behavior. Journal of Personality and Social Psychology, 1965, /, 558-568.
deCharms, R., & Moeller, G. H. Values expressed in American children's readers: 1800-1950.
Journal of Abnormal and Social Psychology, 1962,64, 136-142.
deCharms, R., Morrison, H. W., Reitman, W. R., & McClelland, D. C. Behavioral correlates of
directly and indirectly measured achievement motivation. In D. C. McClelland (Ed.), Studies in
motivation. New York: Appleton-Century-Crofts, 1955.
deCharms, R., & Muir, M. S. Motivation: Social approaches. Annual Review of Psychology, 1978,
29, 91-113.
Deci, E. L. Intrinsic motivation. New York: Plenum, 1975.
Decke, E. Effects of taste on the eating behavior of obese and normal persons. In S. Schachter (Ed.),
Emotion, obesity and crime. New York: Academic Press, 1971.
Dember, W. N., Earl, R. W., & Paradise, N. Response of rats to differential Stimulus complexity.
Journal of Comparative and Physiological Psychology, 1957, 50, 514—518.
Diaz, A. J. An empirical study of the effect of CEO motives on intra-industry Performance with
examples drawn from US and Japanese auto manufacturers. Unpublished Bachelor of Arts
thesis, Harvard College, Department of Psychology and Social Relations and Economics, 1982.
Dienstbier, R. A. Emotion-attribution theory: Establishing roots and exploring future perspectives.
In H. E. Howe, Jr. (Ed.), Nebraska Symposium on Motivation: 1978. Lincoln: University of
Nebraska Press, 1979.
Doherty, M. A., & Walker, R. E. The relationship of personality characteristics, awareness and
attitude in a verbal conditioning Situation. Journal of Personality, 1966, 34, 504-516.
616 Bibliography
Dollard, J., Doob, L. W., Miller, N. E., Mowrer, O. H., & Sears, R. R. Frustration and aggression.
New Haven, Conn.: Yale University Press, 1939.
Donley, R. E., & Winter, D. G. Measuring the motives of public officials at a distance: An
exploratory study of American presidents. Behavioral Science, 1970, 15, 227-236.
Douvan, E. Social Status and success strivings. Journal of Abnormal and Social Psychology, 1956, 52,
219-223.
Duda, J. L. Achievement motivation among Navajo students: A conceptual analysis with
preliminary data. Ethos, 1980, 316-331.
Duffy, E. Activation and behavior. New York: Wiley, 1962.
Duncan, O. D., Featherman, D. L., & Duncan, B. Socio-economic background and achievement.
New York: Harcourt Brace, 1972.
Dutton, D. G., & Aron, A. P. Some evidence for heightened sexual attraction under conditions of
high anxiety. Journal of Personality and Social Psychology, 1974, 30, 510-517.
Dweck, C. S. The role of expectations and attributions in the alleviation of learned helplessness.
Journal of Personality and Social Psychology, 1975, 31, 674—685.
Dweck, C. S., & Wortman, C. B. Learned helplessness, anxiety and achievement motivation:
Neglected parallels in cognitive, affective and coping responses. In H. W. Krohne & L. Laux
(Eds.), Achievement, stress, and anxiety. Washington, D.C.: Hemisphere Publishing, 1982.
Easterlin, R. A. Population, labor force and long Swings in economic growth: The American
experience. New York: Columbia University Press, 1968.
Edwards, A. L. The social desirability variable in personality assessment and research. New York:
Dryden Press, 1957.
Eibl-Eibesfeldt, I. Ethology: The biology of behavior. New York: Holt, Rinehart and Winston,
1970.
Ekman, P. Universal and cultural differences in facial expressions of emotion. In J. K. Cole (Ed.),
Nebraska Symposium on Motivation: 1971. Lincoln: University of Nebraska Press, 1972.
Elashoff, J. D., & Snow, R. E. A case study in statistical inference: Reconstruction of the
Rosenthal-Jacobson data on teacher expectancy. (Tech. Rep. 15). Stanford, Calif.: Stanford
Center for Research and Development in Teaching, School of Education, Stanford University,
1970.
Ellingstad, U. S., & Struckman-Johnson, D. L. Short term rehabilitation study: 12 month analyses.
Vermillion: University of South Dakota, Human Factors Laboratory, 1978.
Ellis, H. The psychology ofsex. New York: New American Library, 1954.
Endler, N. S. Persons, situations, and their interactions. In A. I. Rabin, J. Aronoff, A. M. Barclay,
& R. A. Zucker (Eds.), Further explorations in personality. New York: Wiley, 1981.
Entin, E. E. The relationship between the theory of achievement motivation and Performance on a
simple and a complex task. Unpublished doctoral dissertation, University of Michigan, 1968.
Entin, E. E., & Feather, N. T. Attribution to success and failure in contingent and noncontingent
paths. In J. O. Raynor & E. E. Entin (Eds.), Motivation, career striving, and aging. New York:
Hemisphere Publishing Corp., 1982.
Entin, E. E., & Raynor, J. O. Effects of contingent future orientation and achievement motivation
on Performance in two kinds of tasks. Journal of Experimental Research in Personality, 1973,
6, 134-320.
Entin, E. E., & Raynor, J. O. Success versus moving on in contingent paths. In J. O. Raynor and E.
E. Entin (Eds.), Motivation, career striving, and aging. New York: Hemisphere Publishing
Corp., 1982.
Entwisle, D. R. To dispel fantasies about fantasy-based measures of achievement motivation.
Psychological Bulletin, 1972, 77, 377-391.
Bibliography 617
Epstein, S. The measurement of drive and conflict in humans: Theory and experiment. In M. R.
Jones (Ed.), Nebraska Symposium on Motivation: 1962. Lincoln: University of Nebraska Press,
1962.
Erikson, E. H. Childhood and society (2nd ed.). New York: Norton, 1963. (Originally published,
1950).
Erikson, E. H. Sex differences—The play configuration of pre-adolescents. American Journal of
Orthopsychiatry, 1951, 21, 667-692.
Erikson, E. H. Insight and responsibility. New York: Norton, 1964.
Estes, W. K. Generalization of secondary reinforcement from the primary drive. Journal of
Comparative and Physiological Psychology, 1949, 42, 286-295.
Estes, W. K. Stimulus response theory of drive. In M. R. Jones (Ed.), Nebraska Symposium on
Motivation: 1958. Lincoln: University of Nebraska Press, 1958.
Ewing, J. A., Rouse, B. A., & Pellizzari, E. D. Alcohol sensitivity and ethnic background. American
Journal of Psychiatry, 1914,131, 206-210.
Exline, R. V. Need affiliation and initial communication behavior in problem solving groups
characterized by low interpersonal visibility. Psychological Reports, 1962, 10, 79-89.
Eysenck, H. J. Dimensions of personality. London: Routledge and Kegan Paul, 1947.
Eysenck, H. J. The effects of psychotherapy: An evaluation. Journal of Consulting Psychology, 1952,
16, 319-324.
Eysenck, H. J. The dynamics of anxiety and hysteria. New York: Praeger, 1957. (a)
Eysenck, H. J. Sense and nonsense in psychology. Harmondsworth, England: Penguin, 1957. (b)
Eysenck, H. J. Fact and fiction in psychology. Harmondsworth, England: Penguin, 1965.
Eysenck, H. J. Eysenck on extroversion. New York: Wiley/Halsted, 1973.
Falbo, T., & Beck, R. C. Naive psychology and the attributional model of achievement. Journal of
Personality, 1979, 47, 185-195.
Farber, I. E. Anxiety as a drive State. In M. R. Jones (Ed.), Nebraska Symposium on Motivation:
Feather, N. T. The relationship of persistence at a task to expectation of success and achievement
related motives. Journal of Abnormal and Social Psychology, 1961, 63, 552-561.
Feather, N. T. Persistence at a difficult task with alternative tasks of intermediate difficulty. Journal
of Abnormal and Social Psychology, 1963,66, 604-609.
Feather, N. T. Values in education and society. New York: Free Press, 1975.
Feather, N. T., & Simon, J. G. Fear of success and causal attribution for outcome. Journal of
Personality, 1973, 41(4), 525-542.
Feij, J. A. Debilitating anxiety, facilitating anxiety, and general anxiety. In C. D. Spielberger & I. G.
Sarason (Eds.), Stress and anxiety (Vol. 1). New York: Wiley, 1975.
Feierabend, I. K., & Feierabend, R. L. Aggressive behaviors within politics, 1947-62: A
cross-national study. Journal of Conflict Resolution, 1966, 10, 249-271.
Fersch, E. A., Jr. Inward bound: The motivational impact of a combined Outward Bound-Upward
Bound program on adolescents from poverty families. Unpublished doctoral dissertation,
Feshbach, S. The drive-reducing function of fantasy behavior. Journal of Abnormal and Social
Psychology, 1955,50, 3-11.
Feshbach, S. The stimulating versus cathartic effects of a vicarious aggressive activity. Journal of
A bnormal and Social Psychology, 1961, 63, 381-385.
Feshbach, S. Aggression. In P. H. Müssen (Ed.), CarmichaeVs manual of child psychology. New
York: Wiley, 1970.
618 Bibliography
Festinger, L. Development of differential appetite in the rat. Journal of Experimental Psychology,

1943, 32, 226-234.
Festinger, L., & Carlsmith, J. M. Cognitive consequences of forced compliance. Journal of Abnormal
Fishman, D. B. Need and expectancy as determinants of affiliative behavior in small groups. Journal
of Personality and Social Psychology, 1966,4, 155-164.
Fiske, D. W., & Maddi, S. R. The functions of varied experience. Homewood, 111.: Dorsey, 1961.
Fleming, J. Approach and avoidance motivation in interpersonal competition: A study of black male
and female College students. Unpublished doctoral dissertation, Harvard University,
Department of Psychology and Social Relations, 1974.
Fleming, J. Fear of success as a motivational construct: A functional interpretation. Unpublished
manuscript, Radcliffe Institute, 1975. (a)
Fleming, J. Fear of success in black College students: Its assessment and functional significance.
Unpublished manuscript, Radcliffe Institute, 1975. (b)
Fleming, J. The impact of College environments on black students. San Francisco: Jossey-Bass, 1983.
Fodor, E. M., & Farrow, D. L. The power motive as an influence on the use of power. Journal of
Personality and Social Psychology, 1979,37, 2091-2097.
Fodor, E. M , & Smith, T. The power motive as an influence on group decision making. Journal of
Personality and Social Psychology, 1982, 42, 178-185.
Foss, G. The influence of androgens on sexuality in women. Lancet, 1951, 260, 667-669.
Frankenhaeuser, M. Psychoneuroendocrine approaches to the study of emotion as related to stress
and coping. In H. E. Howe, Jr. (Ed.), Nebraska Symposium on Motivation: 1978. Lincoln:
University of Nebraska Press, 1978.
Frankenhaeuser, M., Dünne, E., & Lundberg, U. Sex differences in sympathetic-adrenal medullary
reactions induced by different Stressors. Psychopharmacology, 1976, 47, 1-5.
French, E. G. Some characteristics of achievement motivation. Journal of Experimental Psychology,
1955, 50, 232-236.
French, E. G. Motivation as a variable and work partner selection. Journal of Abnormal and Social
Psychology, 1956, 53, 96-99.
French, E. G. Development of a measure of complex motivation. In J. W. Atkinson (Ed.), Motives
in fantasy, action, and society. Princeton, N.J.: Van Nostrand, 1958. (a)
French, E. G. Effects of the interaction of motivation and feedback on task Performance. In J. W.
Atkinson (Ed.), Motives in fantasy, action, and society. Princeton, N.J.: Van Nostrand, 1958.
(b)
French, E. G., & Chadwick, I. Some characteristics of affiliation motivation. Journal of Abnormal
and Social Psychology, 1956, 52, 296-300.
French, E. G., & Lesser, G. S. Some characteristics of the achievement motive in women. Journal of
Abnormal and Social Psychology, 1964,65, 119-128.
French, E. G., & Thomas, F. H. The relationship of achievement motivation to problem-solving
effectiveness. Journal of Abnormal and Social Psychology, 1958, 56, 45-48.
Freud, S. [Interpretation of dreams.] In A. A. Brill (Ed. and trans.), The basic writings of Sigmund
Freud. New York: Random House, 1938. (Originally published, 1900.)
Freud, S. [Psychopathology of everyday life.] In A. A. Brill (Ed. and trans.), The basic writings of
Sigmund Freud. New York: Random House, 1938. (Originally published, 1901.)
Freud, S. [Three contributions to the theory of sex.] In A. A. Brill (Ed. and trans.), The basic
writings of Sigmund Freud. New York: Random House, 1938. (Originally published, 1905.)
Freud, S. [The history of the psychoanalytic movement.] In A. A. Brill (Ed. and trans.), The basic
writings of Sigmund Freud. New York: Random House, 1938. (Originally published, 1910.)
Freud, S. [Totem and taboo.] In A. A. Brill (Ed. and trans.), The basic writings of Sigmund Freud.
New York: Random House, 1938. (Originally published, 1918.)
Bibliography 619
Freud, S. [The ego and the id] (J. Riviere, trans.). London: Hogarth, 1927. (Originally published,
1923.)
Freud, S. [The future of an illusion] (W. D. Robson-Scott, trans.). New York: Doubleday, 1957.
(Originally published, 1927.)
Freud, S. [Civilization and its discontents] (J. Riviere, trans.). New York: Doubleday, 1958.
(Originally published, 1930.)
Friedman, M., & Rosenman, R. H. Type A behavior and your heart. New York: Fawcett, 1974.
Frieze, I. H., & Weiner, B. Cue utilization and attributional judgments for success and failure.
Journal of Personality, 1971,39, 591-606.
Fromm, E. Man for himself. New York: Rinehart, 1947.
Galbraith, J. K. The new industrial State. Boston: Houghton Mifflin, 1967.
Gallimore, R. Affiliation, social context, industriousness and achievement. In R. H. Munroe, R. L.
Munroe, & B. B. Whiting (Eds.), Handbook of cross-cultural human development. New York:
Garland STPM, 1981.
Gallimore, R., Boggs, J. W., & Jordan, C. Culture, behavior and Education: A study of
Hawaiian-Americans. Beverly Hills, Calif.: Sage, 1974.
Garcia, J., Hankins, W. G., & Rusiniak, K. W. Behavioral regulation of the milieu interne in man
and rat. Science, 1974, 185, 824-831.
Garcia, J., & Koelling, R. A. Relation of cue to consequence in avoidance learning. Psychonomic
Science, 1966, 4, 123-124.
Geen, R. G., & Gange, J. J. Drive theory of social facilitation: Twelve years of theory and research.
Psychological Bulletin, 1977, 84, 1267-1288.
Gesell, A., & Ilg, F. L. Infant and child in the culture of today. New York: Harper, 1943.
Giese, H., & Schmidt, A. Studenten Sexualität. Hamburg: Rowohlt, 1968.
Giliberto, S. M. Motivation and the Methodist revival period. Unpublished honors thesis, Harvard
University, Department of Psychology and Social Relations, 1972.
Gilligan, C , & Pollak, S. The vulnerable and invulnerable physician. Unpublished manuscript,
Bunting Institute of Radcliffe College, 1982.
Glass, D. C , & Wood, J. D. The control of aggression by self-esteem and dissonance. In P. G.
Zimbardo (Ed.), The cognitive control of motivation. Glenview, 111.: Scott, Foresman,
1969.
Goleman, D. Jason and Medea's love story. Psychology Today, April 1976.
Gombrich, R. F. Precept and practice: Traditional Buddhism in the rural highlands of Ceylon.
London: Oxford University Press, 1971.
Gough, H. G., & Heilbrun, A. B., Jr. The adjective checklist manual. Palo Alto, Calif: Consulting
Psychologists Press, 1975.
Grant, E. C. G., & Myers, E. Mental effects of oral contraceptives. Lancet, 1967, 2, 945-946.
Greene, D. L., & Winter, D. G. Motives, involvements in leadership among black College students.
Joumal of Personality, 1971,39, 319-332.
Greenstein, F. I. Personality in politics. New York: Markham, 1969.
Grinker, J. The control of classical conditioning by cognitive manipulation. In P. G. Zimbardo
(Ed.), The cognitive control of motivation. Glenview, 111.: Scott, Foresman, 1969.
Grossman, S. P. Eating or drinking elicited by direct adrenergic or cholinergic Stimulation of
hypothalamus. Science, 1960, 132, 301-302.
Grünt, J. A., & Young, W. C. Differential reactivity of individuals and the response of the male
guinea pig to testosterone proprionate. Endocrinolcgy, 1952, 51, 237-248.
Gur, R. C , & Gur, R. E., Handedness, sex, and eyedness as moderating variables in the relation
between hypnotic susceptibility and functional brain asymmetry. Journal of Abnormal
Psychology, 1974, 83, 635-643.
620 Bibliography
Gutheil, E. A. The language of the dream. New York: Macmillan, 1939.

Guthrie, E. R. The psychology of learning. New York: Harper & Row, 1935.
Haber, R. N. Discrepancy from adaptation level as a source of affect. Journal of Experimental
Psychology, 1958, 56, 370-375.
Hagen, E. E. On the theory of social change. New York: Dorsey, 1962.
Haiisch, F., & Heckhausen, H. Search for feedback Information and effort regulation during task
Performance. Journal of Personality and Social Psychology, 1977, 35, 724-733.
Hall, C. S. A cognitive theory of dream Symbols. Journal of General Psychology, 1953, 48, 169-186.
Hall, C. S., & Lindzey, G. Theories of personality. New York: Wiley, 1957.
Hamburg, D. A., Hamburg, B. A., & Barchas, J. D. Anger and depression in perspective of
behavioral biology. In L. Levi (Ed.), Emotions: Their parameters and measurement. New York:
Raven, 1975.
Hancock, J. G., & Teevan, R. C. Fear of failure and risk-taking behavior. Journal of Personality,
1964, 32, 200-209.
Haner, C. F., & Brown, J. S. Clarification of the instigation to action concept in the
frustration-aggression hypothesis. Journal of Abnormal and Social Psychology, 1955, 51,
204-206.
Hardy, K. R. Determinants of conformity and attitude change. Journal of Abnormal and Social
Psychology, 1957, 54, 289-294.
Harlow, H. F. Learning to love. San Francisco: Albion, 1971.
Harlow, H. F., Harlow, M. K., & Meyer, D. R. Learning motivated by a manipulation drive.
Harlow, H. F., Harlow, M. K., & Suomi, J. J. From thought to therapy: Lessons from a primate
laboratory. American Scientist, 1971, 59, 538-549.
Harlow, H. F., & Zimmerman, R. R. Affectual responses in the infant monkey. Science, 1959, 130,
421^32.
Harris, H. An experimental model of the effectiveness of project management offices. Unpublished
master's dissertation, Massachusetts Institute of Technology, 1969.
Hassan, M. K., Enayatullah, & Khalique, A. A study of anxiety in school children as related to
child-rearing attitude and some personality traits of parents. Indian Journal of Psychiatrie
Social Work, July 1977, pp. 1-7.
Hawke, C. C. Castration and sex crimes. American Journal of Mental Deficiency, 1950, 55, 220-226.
Hayashi, T., & Habu, K. A research on achievement motive: An experimental test of the "thought
sampling" method by using Japanese students. Japanese Psychological Research, 1962, 4, 30-42.
Hayashi, T., & Yamaushi, K. The relation of children's need for achievement to their parents' home
diseipline in regard to independence and mastery. Bulletin Kyoto Gakugei University (Series A).
1964, 25, 31-40.
Heathers, G. L. The avoidance of repetition of a maze reaction in the rat as a funetion of the time
interval between trials. Journal of Psychology, 1940, 10, 350-380.
Heckhausen, H. Hoffnung und Furcht in der Leistungsmotivation. Meisenheim, West Germany:
Hain, 1963.
Heckhausen, H. The anatomy of achievement motivation. New York: Academic Press, 1967.
Heckhausen, H. Trainings-kurse zur Erhöhung der Leistungsmotivation und der unternehmerischen
Aktivität in einem Entwicklungsland: Eine nachträgliche Analyse des erzielten Motivwandels.
Zeitschrift für und Entwicklungspsychologie und Pädagogische Psychologie, 1971, 3, 253-268.
Heckhausen, H. Intervening cognitions in motivation. In D. E. Berlyne & K. B. Madsen (Eds.),
Pleasure, reward, preference. New York: Academic Press, 1973.
Bibliography 621
Heckhausen, H. How to improve poor motivation in students. Paper presented at the 18th
International Congress of Applied Psychology, Montreal, 1974.
Heckhausen, H. Effort expenditure, aspiration level and self-evaluation before and after unexpected
Performance shifts. Mimeographed lectures at the Universities of Bergen and Oslo, Norway,
1975. (a)
Heckhausen, H. Perceived ability, achievement motive and Information choice: A study by Meyer
reanalyzed and supplemented. Unpublished manuscript, Ruhr University at Bochum, West
Germany, 1975. (b)
Heckhausen, H. Motivation und Handeln. New York: Springer-Verlag, 1980.
Heckhausen, H. Task-irrelevant cognitions during an exam: Incidence and effects. In H. W. Krohne
& L. Laux (Eds.), Achievement, stress, and anxiety. Washington, D.C.: Hemisphere Publishing,
1982.
Heckhausen, H., & Krug, S. Motive modification. In A. J. Stewart (Ed.), Motivation and society San
Francisco: Jossey-Bass, 1982.
Heider, F. The psychology of interpersonal relations. New York: Wiley, 1958.
Heilbrun, K. S. Reply to Silverman. Journal of Abnormal Psychology, 1982, 91, 134-135.
Heiman, J. R. The psychology of erotica: Women's sexual arousal. Psychology Today, August 1975,
pp. 90-94.
Heller, D. Children of holocaust survivors: The second generation effect. Unpublished Bachelor of
Arts thesis, Harvard College, Department of Psychology and Social Relations, 1979.
Henley, N. M. Power, sex, and nonverbal communication. In Z. Rubin (Ed.), Doing unto others.
Englewood Cliffs, N.J.: Prentice-Hall, 1974.
Hermann, M. G. Assessing the personalities of Soviet politburo members. Personality and Social
Psychology Bulletin, 1980, 6, 332-352.
Heyns, R. W., Veroff, J., & Atkinson, J. W. A scoring manual for the affiliation motive. In J. W.
Atkinson (Ed.), Motives in fantasy, action, and society. Princeton, N.J.: Van Nostrand, 1958.
Hill, C. T., Rubin, Z., & Peplau, L. Breakups before marriage: The end of 103 affairs. Journal of
Social Issues, 1976,32, 147-168.
Hoffman, L. W. Fear of success in 1965 and 1972. Journal of Consulting and Clinical Psychology,
1974, 42, 353-358.
Hopkins, J. R., Zelazo, P. R., Jacobson, S. W., & Kagan, J. Infant reactivity to stimulus-schema
discrepancy. Genetic Psychology Monographs, 1976, 93, 27-62.
Homer, M. S. Sex differences in achievement motivation and Performance in competitive and
non-competitive situations. Unpublished doctoral dissertation, University of Michigan, 1968.
Horner, M. S. A psychological barrier to achievement in women: The motive to avoid success. In D.
C. McClelland & R. S. Steele (Eds.), Human motivation: A book of readings. Morristown, N.J.:
General Learning Press, 1973.
Horner, M. S., Tresemer, D. W., Berens, A. F., & Watson, R. I., Jr. Scoring manual for an
empirically-derived scoring System for motive to avoid success. Unpublished manuscript, Harvard
Horney, K. Our inner conflicts. New York: Norton, 1945.
Horowitz, R. A. N Achievement correlates and the executive role. Unpublished Bachelor of Arts
thesis, Harvard University, Department of Social Relations, 1961.
Hovland, C. I., Janis, I. L., & Kelley, H. H. Communication and persuasion: Psychological studies of
opinion change. New Haven, Conn.: Yale University Press, 1953.
Hoyos, C. Motivationspsychologie Untersuchungen von Kraftfahrern mit dem TAT nach
McClelland, Archiv für die gesamte Psychologie, 1965. (Supplementary Vol. 7)
Hudson, L. Contrary imaginations. London: Methuen, 1966.
622 Bibliograph)?
Hüll, C. L. Principles of behavior. New York: Appleton-Century-Crofts, 1943.

Hüll, C. L. A behavior System. New Haven, Conn.: Yale University Press, 1952.
Hurley, J. R. The Iowa Picture Interpretation Test: A multiple-choice Variation of the TAT. Journal
of Consulting Psychology, 1955,79, 372-376.
Inkeles, A., & Levinson, D. J. National character: The study of modal personality and sociocultural
Systems. In Handbook of social psychology (Vol. 2), Special fields in applications. Cambridge,
Mass.: Addison-Wesley, 1954.
Izard, C. E. Emotions as motivations: An evolutionary-developmental perspective. In H. E. Howe,
Jr. (Ed.), Nebraska Symposium on Motivation: 1978. Lincoln: University of Nebraska Press,
1979.
Jackaway, R., & Teevan, R. Fear of failure and fear of success: Two dimensions of the same molive.
Sex Roles, 1976, 2, 283-293.
Jackson, D. N. A modern strategy for personality assessment: The personality research form.
Research Bulletin No. 33c. London, Canada: University of Western Ontario, Department of
Psychology, 1966.
Jemmott, J. B., III. Psychosocial stress, social motives and disease susceptibility. Unpublished doctoral
Jenkins, J. G. Instruction as a factor in "incidental" learning. American Journal of Psychology, 1933,
45, 471-477.
Johnston, R. A. A methodological analysis of several revised forms of the Iowa Picture
Interpretation Test. Journal of Personality, 1957,25, 283-293.
Jones, D. F. The need for power as a predictor of leadership and exploitation in a variety of small
group settings. Unpublished Bachelor of Arts thesis, Wesleyan University, 1969.
Jones, E. E., & Davis, K. E. From acts to dispositions: The attribution process in person perception.
In L. Berkowitz (Ed.), Advances in experimental social psychology (Vol. 2). New York:
Academic Press, 1965.
Jones, M. C. The elimination of children's fears. Journal of Experimental Psychology, 1924, 7,
383-390.
Jones, M. R. (Ed.), Nebraska Symposium on Motivation: 1955. Lincoln: University of Nebraska
Press, 1955.
Jopt, U. J., & Ermshaus, W. Wie generalisiert ist das Selbstkonzept eigener Fähigkeit? Zeitschrift für
experimentelle und angewandte Psychologie, 1977, 24, 578-601.
Jopt, U. J., & Ermshaus, W. Motivspezifizität des Zusammenhangs Zwischen Selbstbelohnung und
Kausalattribierung nach Erfolg und Misserfolg. Archiv für Psychologie, 1978, 130, 53-68.
Jung, C. G. [Memories, dreams, reßections.] (A. Jaffe, Ed.; R. Winston & C. Winston, trans.). New
York: Random, 1961.
Kagan, J., Kearsley, R. B., & Zelazo, P. R. Infancy: Its place in human development. Cambridge,
Mass.: Harvard University Press, 1978.
Kagan, J., & Moss, H. A. Birth to maturity. New York: Wiley, 1962.
Kaiin, R. Social drinking in different settings. In D. C. McClelland, W. N. Davis, R. Kaiin, & E.
Wanner, The drinking man: Alcohol and human motivation. New York: Free Press, 1972.
Kaiin, R., Kahn, M., & McClelland, D. C. The effects of male social drinking on fantasy. Journal of
Kaltenbach, J. E., & McClelland, D. C. Achievement and social Status in three small communities.
In D. C. McClelland, A. L. Baldwin, U. Bronfenbrenner, & F. L. Strodtbeck (Eds.), Talent
and society. Princeton, N.J.: Van Nostrand, 1958.
Karabenick, S. Fear of success, achievement and affiliation dispositions, and the Performance of men
and of women under individual competitive conditions. Journal of Personality, 1977, 45,
117-149.
Bibliography 623
Karabenick, S. A., & Yousseff, Z. I. Performance as a function of achievement level and perceived
difficulty. Journal of Personality and Social Psychology, 1968, 10, 414-419.
Kardiner, A. (Ed.). Psychological frontiers of society. New York: Columbia University Press, 1945.
Kawamura-Reynolds, M. Motivational effects of an audience in the content of imaginative thought.
Journal of Personality and Social Psychology, 1911,35, 912-919.
Keister, M. Behavior of young children in failure. In R. Updegraff (Ed.), Studies in pre-school
education. Iowa City, Iowa: University of Iowa Studies in Child Weifare, 1938.
Kelly, G. The psychology of personal constructs. New York: Norton, 1955.
Keshock, J. D. An investigation of the effects of the expectancy phenomenon upon the intelligence,
achievement and motivation of inner city elementary school children. Unpublished doctoral
dissertation, Case Western Reserve University, 1970.
Kinsey, A., Pomeroy, W. B., & Martin, C. E. Sexual behavior in the human male. Philadelphia:
Saunders, 1948.
Kintsch, W. Memory and cognition. New York: Wiley, 1977.
Kirshnit, C. E., & McClelland, D. C. Short term shifts in immunocompetence in response to
emotion-arousing films. Unpublished manuscript, Harvard University, Department of
Psychology and Social Relations, 1983.
Klaiber, E. L., Broverman, D. M., & Kobayashi, Y. The automatization cognitive style, androgens,
and monoamine oxidase. Psychopharmacologia, 1967, 7/(4), 320-336.
Klein, S. B. Motivation: Biosocial approaches. New York: McGraw-Hill, 1982.
Kleinbeck, U., & Schmidt, K. H. Aufgabenwahl im Ernstfall einer betrieblichen Ausbildung:
Instrumentalitäts-theoretische Ergänzung aum Risikowahlmodell. Zeitschrift für
Entwicklungspsychologie und Pädagogische Psychologie, 1979, 11, 1-11.
Kleinginna, P. R., Jr., & Kleinginna, A. M. A categorized list of motivation definitions, with a
Suggestion for a consensual definition. Motivation and Emotion, 1981, 5, 263-291.
Kline, P. Fact and fantasy in Freudian theory (2nd ed.). London: Methuen, 1981.
Klinger, E. Short-term stability and concurrent validity of TAT need scores: Achievement, affiliation
and hostile press. Proceedings of the 76th Annual Convention of the American Psychological
Association, American Psychologist, 1968, 23, 878.
Klinger, E. Meaning and void. Minneapolis: University of Minnesota Press, 1977.
Klinger, E., Barta, S. G., & Maxeiner, M. E. Motivational correlates of thought content frequency
and commitment. Journal of Personality and Social Psychology, 1980, 39, 1222-1237.
Kock, S. W. [Management and motivation. ] English summary of a doctoral dissertation presented at
the Swedish School of Economics, Helsinki, Finland, 1965.
Kohlberg, L. Stage and sequence: The cognitive-developmental approach to socialization. In D.
Goslin (Ed.), Handbook of socialization theory and research. Chicago: Rand McNally, 1969.
Kolb, D. A. Achievement motivation training for underachieving high school boys. Journal of
Kolb, D. A., & Boyatzis, R. E. On the dynamics of the helping relationship. Journal of Applied
Behavioral Science, 1970,6, 267-289.
Kolp, P. Navaho economic change. Unpublished doctoral dissertation, Massachusetts Institute of
Technology, 1965.
Konecni, V. J. The mediation of aggressive behavior: Arousal level versus anger and cognitive
labelling. Journal of Personality and Social Psychology, 1975, 32, 706-712.
Korman, A. The psychology of motivation. New York: Prentice-Hall, 1974.
Lacey, J. I. Somatic response patterning and stress: Some revisions of activation theory. In M. H.
Appley & R. Trumbull (Eds.), Psychological stress. New York: McGraw-Hill, 1967.
624 Bibliography
Langer, E. J., & Rodin, J. The effects of choice and enhanced personal responsibility for the aged: A
field experiment in an institutional setting. Journal of Personality and Social Psychology, 1976,
34, 191-198.
Lansing, J. B., & Heyns, R. W. Need affiliation and frequency of four types of communication.
Journal of Abnormal and Social Psychology, 1959, 58, 365-372.
Lansky, D., & Wilson, G. T. Alcohol, expectations, and sexual arousal in males: An information
processing analysis. Journal of Abnormal Psychology, 1981,90, 35-45.
Larson, R., & Csikszentmihalyi, M. Experiential correlates of time alone in adolescence. Journal of
Personality, 1978, 46, 677-693.
Lasker, H. M. Ego development and motivation: A cross-cultural cognitive-developmental analysis of n
Achievement. Unpublished doctoral dissertation, University of Chicago, Department of
Sociology, 1978.
Laufen, A. Validierungsschlüssel des TAT Verfahrens zur Erhebung der Anschlussmotivation.
Unpublished diploma dissertation, Psychological Institute of the Ruhr University at Bochum,
West Germany, 1967.
Lawrence, P. R., & Lorsch, J. W. New management Job: The integrator. Harvard Business Review,
1967, 45(6), 142-151.
Lazarus, R. S., & Alfert, E. The short-circuiting of threat by experimentally altering cognitive
appraisal. Journal of Abnormal Psychology and Social Psychology, 1964, 69, 195-205.
Le Magnen, J. Hyperphagie provoquee chez le rat blanc par alteration du mecanisme de satiete
peripherique. Comptes Rendus des Seances de la Societe de Biologie, 1956, 150, 32.
LeVine, R. Dreams and deeds: Achievement motivation in Nigeria. Chicago: University of Chicago
Press, 1966.
Levy, D. M. Oppositional syndromes and oppositional behavior. In P. H. Hoch and J. Zubin (Eds.),
Psychopathology of childhood. New York: Grüne and Stratton, 1955.
Lewin, K. A dynamic theory of personality. New York: McGraw-Hill, 1935.
Lewin, K., Dembo, T., Festinger, L., & Sears, P. S. Level of aspiration. In J. M. Hunt (Ed.),
Personality and the behavior disorders (Vol. 1). New York: Ronald Press, 1944.
Ley, R. G., & Bryden, M. P. A dissociation of right and left hemispheric effects for recognizing
emotional tone and verbal content. Brain and Cognition, 1982, 1, 3-9.
Lichter, S. R., & Rothman, S. Jewish ethnicity and radical culture: A social psychological study of
political activists. Political Psychology, 1982, 3(1-2), 116-157.
Linton, R. The cultural background of personality. New York: Appleton-Century, 1945.
Lipman-Blumen, J., Leavitt, H. J., Patterson, K. J., Bies, R. J., & Handley-Isaksen, A. A model of
direct and relational achieving styles. In L. J. Fyans, Jr. (Ed.), Achievement motivation: Recent
trends in theory and research. New York: Plenum, 1980.
Littig, L. W. Effects of motivation on probability preferences. Journal of Personality, 1963, 31,
417-427.
Littig, L. W., & Williams, C. E. Need for affiliation, self-esteem, and social distance of black
Americans. Motivation and Emotion, 1978, 2, 369-373.
Litwin, G. H. Motives and expectancy as determinants of preference for degrees of risk. Unpublished
Bachelor of Arts thesis, University of Michigan, Department of Psychology, 1958.
Litwin, G. H., & Siebrecht, A. Integrators and entrepreneurs: Their motivation and effect on
management. St. Louis: Hospital Progress, 1967.
Locke, E. A. Toward a theory of task motivation and incentives. Organizational Behavioral and
Human Performance, 1968, 3, 157-189.
Locke, E. A. Personnel attitudes and motivation. Annual Review of Psychology, 1975, 26, 457-
480.
Bibliography 625
Locke, E. A., & Bryan, J. F. Grade goals as determinants of academic achievement. Journal of
General Psychology, 1968, 79, 219-228.
Locke, E. A., Shaw, K. N., Saari, L. M., & Latham, G. P. Goal setting and task Performance:
1969-1980 (Tech. Rep. for the Organizational Effectiveness Research Program). Washington,
D.C.: Office of Naval Research, 1980.
Loehlin, J. C , & Nichols, R. C. Heredity, environment and personality. Austin: University of Texas
Press, 1976.
Loevinger, J. The meaning and measurement of ego development. American Psychologist, 1966, 21,
195-206.
Loevinger, J. Ego development: Conceptions and theories. San Francisco: Jossey-Bass, 1976.
Lohr, S. The efficiency of Japanese management. New York Times, February 27, 1982, p. 29.
Lorenz, K. The past twelve years in the comparative study of behavior. In C. H. Schiller (Ed.),
Instinctive behavior. New York: International Universities Press, 1952.
Lorenz, K. On aggression. New York: Harcourt Brace Jovanovich, 1966.
Luborsky, L., & Sargent, H. D. The psychotherapy research project, V: Sample use of method.
Bulletin of the Menninger Clinic, 1956,20, 263-276.
Lundy, A. C. Conditions of lest administration and TAT validity. Unpublished manuscript, Harvard
Lundy, A. C. Situationally evoked Schema sets in TA T-based research: A cognitive explanation for
dijferential TAT validity. Unpublished doctoral dissertation, Harvard University, 1981. (a)
Lundy, A. C. The reliability of the Thematic Apperception Test. Unpublished manuscript, Harvard
University, Department of Psychology and Social Relations, 1981. (b)
Lynn, R. An achievement motivation questionnaire. British Journal of Psychology, 1969, 60,
529-534.
Maas, J. W., Hattox, S. E., Greene, N. M., & Landis, P. H. 3-methoxy-4-hydroxy-phenylglycol
production by human brain in vivo. Science, 1979, 205, 1025-1027.
Maccoby, E. E., & Jacklin, C. N. The psychology of sex differences. Stanford, Calif.: Stanford
University Press, 1974.
Mace, C. A. Incentives: Some experimental studies (Rep. No. 72). London: Industrial Health
Research Board (Great Britain), 1935.
MacKintosh, J. H., & Grant, E. C. The effect of olfactory Stimuli on agonistic behavior of
laboratory mice. Zeitschrift für Tierpsychologie, 1966, 23, 584-587.
MacLean, P. D. Sensory and perceptive factors in emotional functions of the triune brain. In L.
Levi (Ed.), Emotions: Their parameters and measurement. New York: Raven, 1975.
Maddi, S. R. Affective tone during environmental regularity and change. Journal of Abnormal and
Social Psychology, 1961,(52, 338-345.
Maddi, S. R., Charlens, A. M., Maddi, D., & Smith, A. J. Effects of monotony and novelty on
imaginative productions. Journal of Personality, 1962,50, 513-527.
Maehr, M. L. Culture and achievement motivation. American Psychologist, 1974, 29, 887-896.
Maehr, M. L., & Kleiber, D. A. The graying of achievement motivation. American Psychologist,
1981, 36, 787-793.
Magnusson, D. The person and the Situation in an interactional model of behavior. Scandinavian
Journal of Psychology, 1976, 17, 253-271.
Mahone, C. H. Fear of failure and unrealistic vocational aspiration. Journal of Abnormal and Social
Psychology, 1960, 60, 253-261.
Mandler, G., & Sarason, S. B. A study of anxiety and learning. Journal of Abnormal and Social
Psychology, 1952, 47, 166-173.
626 Bibliography
Mandler, J. M. Effect of early food deprivation on adult behavior in the rat. Journal of Comparative
and Physiological Psychology, 1958,5/, 513-517.
Mann, J., Berkowitz, L., Sidman, J., Starr, S., & West, S. Satiation of the transient stimulating effect
of erotic films. Journal of Personality and Social Psychology, 1974, 30, 729-735.
Mansson, H. H. The relation of dissonance reduction to cognitive, perceptual, consummatory, and
learning measures of thirst. In P. G. Zimbardo (Ed.), The cognitive control of motivation.
Glenview, 111.: Scott, Foresman, 1969.
Maranell, G. M. The evaluation of presidents: An extension of the Schlesinger poll. Journal of
American History, 1970,57, 104-113.
Marquis, D. P. Learning in the neonate: The modification of behavior under three feeding schedules.
Marshall, G. D., & Zimbardo, P. G. Affective consequences of inadequately explained physiological
arousal. Journal of Personality and Social Psychology, 1979, 37, 970-988.
Martindale, C. Romantic progression: The psychology of literary history. Washington, D.C.:
Hemisphere Publishing, 1975.
Martindale, C. The wie ofpoet, styles, and sinusoidal patterns of regression. Unpublished manuscript,
University of Maine, Department of Psychology, 1977.
Maslow, A. Motivation and personality. New York: Harper & Row, 1954.
Maslow, A. H. A theory of metamotivation: The biological rooting of the value life. Journal of
Humanistic Psychology, 1967, 7, 93-127.
Maslow, A. Toward a psychology of being. New York: Van Nostrand, 1968.
Mason, J. W. Emotion as reflected in patterns of endocrine integration. In L. Levi (Ed.), Emotions:
Their parameters and measurement, New York: Raven, 1975.
Masters, W., & Johnson, V. Human sexual response. Boston: Little, Brown, 1966.
Matthews, K. A., & Saal, F. E. Relationship of the type A coronary-prone behavior pattern to
achievement, power, and affiliation motives. Psychosomatic Medicine, 1978, 40, 631-636.
May, R. Sex and fantasy. New York: Norton, 1980.
Mayer, J. Regulation of energy intake and body weight: The glucostatic theory and the lipostatic
hypothesis. Annais of the New York Academy of Science, 1955, 63, 15-43.
Mayer, J., & Marshall, N. B. Specificity of gold-thio glucose for ventromedial hypothalamic lesions
and obesity. Nature, 1956, 178, 1399-1400.
McAdams, D. P. Validation of a thematic coding System for the intimacy motive. Unpublished
doctoral dissertation, Harvard University, 1979.
McAdams, D. P. A thematic coding System for the intimacy motive. Journal of Research in
Personality, 1980, 14, 413^32.
McAdams, D. P. Intimacy motivation. In A. J. Stewart (Ed.), Motivation and society. San Francisco:
Jossey-Bass, 1982. (a)
McAdams, D. P. Experiences of intimacy and power: Relationships between social motives and
autobiographical memory. Journal of Personality and Social Psychology, 1982, 42, 292-302. (b)
McAdams, D. P., & Constantian, C. A. Intimacy and affiliation motives in daily living: An
experience sampling analysis. Unpublished manuscript, Loyola University, 1982.
McAdams, D. P., Healey, S., & Krause, S. Relationships between social motives and patterns of
friendship. Unpublished manuscript, Loyola University, Department of Psychology, 1982.
McAdams, D. P., & Losoff, M. Friendship motivation in fourth and sixth graders: A thematic
analysis. Unpublished manuscript, Loyola University, Department of Psychology, 1982.
McAdams, D. P., & McClelland, D. C. Social motives and memory. Unpublished manuscript,
Harvard University, Department of Psychology and Social Relations, 1983.
Bibliography 627
McAdams, D. P., & Powers, J. Themes of intimacy in behavior and thought. Journal of Personality
McAdams, D. P., & Vaillant, G. E. Intimacy motivation and psychosocial adaptation: A
longitudinal study. Journal of Personality Assessment, 1982, 46, 586-593.
McClelland, D. C. Functional autonomy of motives as an extinction phenomenon. Psychological
Review, 1942, 49, 272-283.
McClelland, D. C. Personality. New York: Sloane, 1951.
McClelland, D. C. Some social consequences of achievement motivation. In M. R. Jones
(Ed.), Nebraska Symposium on Motivation: 1955. Lincoln: University of Nebraska Press,
1955.
McClelland, D. C. Methods of measuring human motivation. In J. W. Atkinson (Ed.), Motives in
fantasy, action, and society. Princeton, N.J.: Van Nostrand, 1958. (a)
McClelland, D. C. The use of measures of human motivation in the study of society. In J. W.
Atkinson (Ed.), Motives in fantasy, action, and society. Princeton, N. J.: Van Nostrand,
1958. (b)
McClelland, D. C. The achieving society. Princeton, N. J.: Van Nostrand, 1961. (Reissued with a
new preface, New York: Irvington, 1976.)
McClelland, D. C. The harlequin complex. In R. W. White (Ed.), The study of lives. New York:
Atherton Press, 1963.
McClelland, D. C. Toward a theory of motive acquisition. American Psychologist, 1965, 20,
321-333. (a)
McClelland, D. C. TV Achievement and entrepreneurship: A longitudinal study. Journal of
Personality and Social Psychology, 1965, 1, 389-392. (b)
McClelland, D. C. Does education accelerate economic growth? Economic Development and Cultural
Change, 1966, 14, 257-278.
McClelland, D. C. Money as a motivator: Some research insights. McKinsey Quarterly, 1967, 4,
10-21.
McClelland, D. C. What is the effect of achievement motivation training in the schools? Teachers
College Record, 1972, 74, 129-145.
McClelland, D. C. Power: The inner experience. New York: Irvington, 1975.
McClelland, D. C. Sources of stress in the drive for power. In G. Serban (Ed.), Psychopathology of
human adaptation. New York: Plenum Press, 1976.
McClelland, D. C. The psychological causes and consequences of modernization: An Ethiopian case
study. Economic Development and Cultural Change, 1977, 25 (Suppl.), 43-66. (a)
McClelland, D. C. The impact of power motivation training on alcoholics. Journal of Studies on
Alcohol 1977, 38(\), 142-144. (b)
McClelland, D. C. Inhibited power motivation and high blood pressure in men. Journal of Abnormal
Psychology, 1979, 88, 182-190. (a)
McClelland, D. C. Increasing achievement motivation (Mimeographed report). Boston: McBer and
Company, 1979. (b)
McClelland, D. C. Motive dispositions: The merits of operant and respondent measures. In L.
Wheeler (Ed.), Review of personality and social psychology (Vol. 1). Beverly Hills, Calif.: Sage,
1980.
McClelland, D. C. Is personality consistent? In A. I. Rabin, J. Aronoff, A. M. Barclay, & R. A.
Zucker (Eds.), Further explorations in personality. New York: Wiley, 1981. (a)
McClelland, D. C. Comment on Lord Vaizey's paper on psychological factors in economic growth.
In H. Giersch (Ed.), Towards an explanation of economic growth. Tübingen, West Germany:
Mohr, 1981. (b)
628 Bibliography
McClelland, D. C. The need for power, sympathetic activation and illness. Motivation and Emotion,
1982, 6(1), 31-41.
McClelland, D. C , Alexander, C , & Marks, E. The need for power, stress, immune function and
illness among male prisoners. Journal of Abnormal Psychology, 1982, 9/(1), 61-70.
McClelland, D. C , & Apicella, F. S. A functional classification of verbal reactions to experimentally
induced failure. Journal of Abnormal and Social Psychology, 1945, 40, 376-390.
McClelland, D. C , & Atkinson, J. W. The projective expression of needs, I: The effect of different
intensities of the hunger drive on perception. Journal of Psychology, 1948, 25, 205-222.
McClelland, D. C , Atkinson, J. W., Clark, R. A., & Lowell, E. L. The achievement motive. New
York: Appleton-Century-Crofts, 1953.
McClelland, D. C , Baldwin, A. L., Bronfenbrenner, U., & Strodtbeck, F. L. Talent and society
Princeton, N. J.: Van Nostrand, 1958.
McClelland, D. C , & Boyatzis, R. E. The leadership motive pattern and long term success in
management. Journal of Applied Psychology, 1982, 67(6), 737-743.
McClelland, D. C , & Burnham, D. H. Power is the great motivator. Harvard Business Review,
March-April 1976, pp. 100-110; 159-166.
McClelland, D. C , Colman, C , Finn, K., & Winter, D. G. Motivation and maturity patterns in
marital success. Social Behavior and Personality, 1978, 6, 163-171.
McClelland, D. C , Constantian, C , Pilon, D., & Stone, C. Effects of child-rearing practices on adult
maturity. In D. C. McClelland (Ed.), The development of social maturity. New York:
Irvington, 1982.
McClelland, D. C , Davidson, R. J., Floor, E., & Saron, C. Stressed power motivation, sympathetic
activation, immune function and illness. Journal of Human Stress, 1980, 6(2), 11-19.
McClelland, D. C, Davidson, R. J., & Saron, C. Evoked potential indicators of the impact of the
need for power on perception and learning. Unpublished manuscript, Harvard University,
1979.
McClelland, D. C , Davidson, R., Saron, C , & Floor, E. The need for power, brain norepinephrine
turnover and learning. Biological Psychology, 1980, 10, 93-102.
McClelland, D. C , Davis, W. B., Kaiin, R., & Wanner, E. The drinking man: Alcohol and human
motivation. New York: Free Press, 1972.
McClelland, D. C , & Friedman, G. A. Child-rearing practices and the achievement motivation
appearing in folktales. In T. M. Newcomb, E. L. Hartley, & G. E. Swanson (Eds.), Readings in
social psychology (2nd ed.). New York: Holt, Rinehart and Winston, 1952.
McClelland, D. C , & Jemmott, J. B., III. Power motivation, stress and physical illness. Journal of
Human Stress, 1980, 6(4), 6-15.
McClelland, D. C, & Kirshnit, C. Effects of motivational arousal on immune function. Unpublished
manuscript, Harvard University, Department of Psychology and Social Relations, 1982.
McClelland, D. C , & Klemp, G. Validation of the human Service worker test. Unpublished
manuscript, 1974. (Available from McBer and Co., Boston, Mass.)
McClelland, D. C , & Liberman, A. M. The effect of need for achievement on recognition of
need-related words. Journal of Personality, 1949,75, 236-251.
McClelland, D. C , Locke, S. E., Williams, R. M., & Hurst, M. W. Power motivation, distress and
immune function. Unpublished manuscript, Harvard University, 1982.
McClelland, D. C, & Maddocks, A. The effects offilms on arousing motives in individuals with
different motive dispositions. Unpublished manuscript, Harvard University, Department of
Psychology and Social Relations, 1983.
McClelland, D. C. & Maddocks, A. The Joint effect of n Power and brain norepinephrine turnover on
recall: A replication. Unpublished manuscript, Harvard University, Department of Psychology
and Social Relations, 1984.
Bibliography 629
McClelland, D. C , & Pilon, D. A. Sources of adult motives in patterns of parent behavior in early
childhood. Journal of Personality and Social Psychology, 1983, 44(3), 564-574.
McClelland, D. C , Rhinesmith, S., & Kristensen, R. The effects of power training on Community
action agencies. Journal of Applied Behavioral Science, 1975, 11, 92-115.
McClelland, D. C , Rindiisbacher, A., and deCharms, R. C. Religious and other sources of parental
attitudes toward independence training. In D. C. McClelland (Ed.), Studies in motivation. New
McClelland, D. C , & Teague, G. Predicting risk preferences among power-related tasks. Journal of
Personality, 1975, 43, 266-285.
McClelland, D. C , & Watson, R. I., Jr. Power motivation and risk-taking behavior. Journal of
Personality, 1973, 41, 121-139.
McClelland, D. C , & Watt, N. F. Sex-role alienation in schizophrenia. Journal of Abnormal
Psychology, 1968, 73, 226-239.
McClelland, D. C , & Winter, D. G. Motivating economic achievement. New York: Free Press, 1969.
(Afterword to the paperback edition, 1971.)
McDougall, W. An introduction to social psychology. London: Methuen, 1908.
McDougall, W. The energies of men. London: Methuen, 1932.
McKeachie, W. J. Motivation, teaching methods, and College learning. In M. R. Jones (Ed.),
McKenna, R. J. Some effects of anxiety level and food cues on the eating behavior of obese and
normal subjects. Journal of Personality and Social Psychology, 1972, 22, 311-319.
McMurray, G. A. Experimental study of a case of insensitivity to pain. Reprinted with additions
from the Archives of Neurology and Psychiatry, 1950, 64, 650-667.
Mehrabian, A. Measures of achieving tendency. Educational and Psychological Measurement, 1969,
29, 445^51.
Mehrabian, A. The development and validation of measures of affiliative tendency and sensitivity to
rejection. Educational and Psychological Measurement, 1970, 30, 417—428.
Mehrabian, A., & Ksionzky, S. A theory of affiliation. Lexington, Mass.: Heath, 1974.
Mehta, P. The achievement motive in high school boys. New Delhi: National Council of Educational
Research and Training, 1969.
Melton, A. W. Learning. In W. S. Monroe (Ed.), Encyclopedia of Educational Research. New York:
Macmillan, 1941.
Meyer, J. P. Causal attributions for success and failure: A multivariate investigation of
dimensionality, formation, and consequences. Journal of Personality and Social Psychology,
1980, 38, 704-718.
Meyer, W. U. Leistungsmotiv und Ursachenerklärung von Erfolg und Misserfolg. Stuttgart: Klett,
1973.
Meyer, W. U. [Perceived ability and informational choice. ] Unpublished manuscript, Psychological
Institute of the Ruhr University at Bochum, West Germany, 1975.
Meyer-Bahlburg, H. F. L. Sex hormones in male homosexuality in comparative perspective. Archives
of Sexual Behavior, 1977,6, 297-325.
Milgram, S. Behavioral study of obedience. Journal of Abnormal and Social Psychology, 1963, 67,
371-378.
Milgram, S. Some conditions of obedience and disobedience to authority. Human Relations, 1965,
18, 57-76.
Miller, N. E. Fear as motivation and fear-reduction as reinforcement in the learning of new
responses. Journal of Experimental Psychology, 1948, 38, 89-101.
Miller, N. E. Comments on theoretical modeis illustrated by the development of the theory of
conflict. Journal of Personality, 1951,20, 82-100.
630 Bibliography
Miller, N., & Dollard, J. Social learning and Imitation, New Haven, Conn.: Yale University Press,
1941.
Miron, D., & McClelland, D. C. The impact of achievement motivation training on small business
Performance. California Management Review, 1979, 27(4), 13-28.
Mischel, W. Personality and assessment. New York: Wiley, 1968.
Mischel, W., & Gilligan, C. Delay of gratification, motivation for the prohibited gratification
and responses to temptation. Journal of Abnormal and Social Psychology, 1964, 69, 411-
417.
Money, J., & Ehrhardt, A. Man and woman, boy and girl. Baltimore: Johns Hopkins University
Press, 1972.
Montgomery, K. S., & Segall, M. Discrimination learning based on exploratory drive. Journal of
Comparative and Physiological Psychology, 1955, 48, 225-228.
Morris, D. The naked ape. London: Constable, 1967.
Moulton, R. W. Notes for a projective measure of fear of failure. In J. W. Atkinson (Ed.), Motives
in fantasy, action, and society. Princeton, N. J.: Van Nostrand, 1958.
Moulton, R. W., Raphelson, A. C , Kristofferson, A_. B. & Atkinson, J. W. The achievement motive
and perceptual sensitivity under two conditions of motive-arousal. In J. W. Atkinson (Ed.),
Motives in fantasy, action, and society. Princeton, N. J.: Van Nostrand, 1958.
Mowrer, O. H. Learning theory and personality dynamics. New York: Ronald Press, 1950.
Moyer, K. E. The psychobiology of aggression. New York: Harper & Row, 1976.
Mücher, H., & Heckhausen, H. Influence of mental activity and achievement motivation on skeletal
muscle toneness. Perceptual and Motor Skills, 1962, 14, 217-218.
Mueller, S. C. Motivation and reactions to the work wie among female performers and music
teachers. Unpublished doctoral dissertation. University of Michigan, 1975.
Murray, E. J. Sleep, dreams, and arousal. New York: Appleton-Century-Crofts, 1965.
Murray, H. A. Explorations in personality. New York: Oxford University Press, 1938.
Murray, H. A. American Icarus. In A. Burton, and R. E. Harris (Eds.), Clinical studies of
personality. New York: Harper and Row, 1955.
Müssen, P. H., & Jones, M. C. Self-conceptions, motivations and interpersonal attitudes of late- and
early-maturing boys. Child Development, 1957, 28, 243-256.
Müssen, P. H., & Scodel, A. The effects of sexual Stimulation under varying conditions on TAT
sexual responsiveness. Journal of Consulting Psychology, 1955, 19, 90.
Nakamura, J. Achievement motivation in developmental perspective. Unpublished manuscript,
University of Chicago, Department of Sociology, 1981.
Neihardt, J. C. Black Elk speaks: Being a life story of a holy man of the Oglala Sioux. New York:
Simon and Schuster, 1972. (Originally published, 1932.)
Newman, R. S. Alleviating learned helplessness in a wilderness setting: An application of attribution
theory to Outward Bound. In L. J. Fyans, Jr. (Ed), Achievement motivation: Recent trends in
theory and research. New York: Plenum, 1980.
Nicholls, J. G. The development of the concepts of effbrt and ability, perception of academic
attainment, and the understanding that dimcult tasks require more ability. Child Development,
1978, 49, 80O-814.
Noujaim, K. Some motivation determinants of effort allocation and Performance. Unpublished
doctoral dissertation, Massachusetts Institute of Technology, 1968.
Nuttall, R. L. Some correlates of high need for achievement among urban Negroes. Journal of
O'Connor, P. A., Atkinson, J. W., & Homer, M. Motivational implications of ability grouping in
schools. In J. W. Atkinson & N. T. Feather (Eds.), A theory of achievement motivation. New
York: Wiley, 1966.
Bibliography 631
Ogilvie, D. Psychodynamics of fantasized flight: A study ofpeople in folk tales. Unpublished doctoral
Oke, A., Keller, R., Mefford, I., & Adams, R. N. Lateralization of norepinephrine in human
thalamus. Science, 1978,200, 1411-1413.
Olds, J. A physiological study of reward. In D. C. McClelland (Ed.), Studies in motivation. New
Olds, J. Drives and reinforcements: Behavioral studies of hypothalamic functions. New York: Raven,
1977.
Olds, J., & Milner, P. Positive reinforcement produced by electrical Stimulation of the septal area
and other regions of the rat brain. Journal of Comparative and Physiological Psychology, 1954,
47,
Olson, G. A., Olson, R. D., Kastin, A. J., & Coy, D. H. Endogenous opiates: 1979. Peptides, 1980,
1, 365-379.
Orne, M. T. On the social psychology of the psychological experiment: With particular reference to
demand characteristics and their implications. American Psychologist, 1962, 17, 776-783.
Panksepp, J. Aggression elicited by electrical Stimulation of the hypothalamus in albino rats.
Physiology and Behavior, 1971,6, 321-329.
Papanek, G. S. The development of entrepreneurship. American Economic Review, 1962, 52, 46-58.
Parsons, J. E., & Goff, S. B. Achievement motivation and values: An alternative perspective. In L. J.
Fyans, Jr. (Ed.), Achievement motivation: Recent trends in theory and research. New York:
Plenum, 1980.
Passer, M. W. Perceiving the causes of success and failure revisited: A multidimensional scaling
approach. Unpublished doctoral dissertation, University of California, Los Angeles, 1977.
Patrick, J. R. Studies in rationa1 behavior and emotional excitement, II: The effect of emotional
excitement on rational behavior of human subjects. Journal of Comparative Psychology, 1934,
18, 153-195.
Patten, R. L., & White, L. A. Independent effects of achievement motivation and overt attribution
on achievement behavior. Motivation and Emotion, 1977, 1, 39-59.
Patterson, M. L. An arousal model of interpersonal intimacy. Psychological Review, 1976, 83,
235-245.
Pearlson, H. B. Effects of temporal distance from a goal and number of tasks required for goal
attainment on achievement-related behavior. In J. O. Raynor & E. E. Entin (Eds.), Motivation,
career striving, and aging. New York: Hemisphere Publishing Corp., 1982.
Pearlson, H. B., & Raynor, J. O. Motivational analysis of the future plans of College men: Imagery
used to describe future plans and goals. In J. O. Raynor & E. E. Entin (Eds.), Motivation,
Peplau, L. A., Rubin, Z., & Hill, C. The sexual balance of power. Psychology Today, November
1976, pp. 142-151.
Perlmuter, L. C , Scharff, K., Karsh, R., & Monty, R. A. Perceived control: A generalized State of
motivation. Motivation and Emotion, 1980, 4, 35-45.
Phillips, A. G., Carter, D. A., & Fibiger, H. C. Differential effects of para-chlorophenyl alanine on
self-stimulation in eaudate putamen and lateral hypothalamus. Psychopharmacology, 1976, 49,
23-27.
Piaget, J. The moral judgement of the child. New York: Harcourt Brace, 1932.
Pilon, D. A., Jr. The effect of personality needs and motives on political values and attitudes.
Unpublished doctoral dissertation, Harvard University, 1981.
Plutchik, R. Emotion: A psychoevolutionary synthesis. New York: Harper & Row, 1980.
Pottharst, B. C. The achievement motive and level of aspiration after experimentally induced success
and failure. Unpublished doctoral dissertation, University of Michigan, 1955.
632 Bibliography
Rabin, A. I., Aronoff, J., Barclay, A. M., & Zucker, R. A. Further explorations in personality. New
York: Wiley, 1981.
Rachman, S. Sexual fetishism: An experimental analogue. Psychological Record, 1966, 16, 293-296.
Radnitzky, G. Contemporary schools of metascience. Chicago: Regnery, 1973.
Raphelson, A. C. The relationships among imaginative, direct verbal, and physiological measures of
anxiety in an achievement Situation. Journal of Abnormal and Social Psychology, 1957, 54,
13-18.
Raynor, J. O. The relationship between distant future goals and achievement motivation. Unpublished
doctoral dissertation, University of Michigan, 1968.
Raynor, J. O. A theory of personality functioning and change. In J. O. Raynor & E. E. Entin
(Eds.), Motivation, career striving, and aging. New York: Hemisphere Publishing Corp., 1982.
Raynor, J. O., & Entin, E. E. Motivation, career striving, and aging. New York: Hemisphere
Publishing Corp., 1982. (a)
Raynor, J. O., & Entin, E. E. Future orientation and achievement motivation. In J. O. Raynor and
E. E. Entin (Eds.), Motivation, career striving, and aging. New York: Hemisphere NY, 1982.
(b)
Raynor, J. O., & Harris, V. A. Performance in decreasing and increasing probability contingent
paths. Unpublished data, 1973. Summarized in J. O. Raynor & E. E. Entin (Eds.), Motivation,
Raynor, J. O., & Smith, C. P. Achievement-related motives and risk-taking in games of skill and
chance. Journal of Personality, 1966, 34, 176-198.
Raynor, J. O., & Teitelbaum, R. C. Time-linked sense of seif and the n Achievement score. In J. O.
Raynor and E. E. Entin (Eds.), Motivation, career striving, and aging. New York: Hemisphere
Publishing Corp., 1982.
Richter, C. P. Animal behavior and internal drives. Quarterly Review of Biology, 1927, 2, 307-343.
Robinson, R. G. Differential behavioral and biochemical effects of right and left hemispheric cerebral
infarction in the rat. Science, 1979, 105, 707-710.
Roethlisberger, F. J., & Dickson, W. J. Management and the worker. Cambridge, Mass.: Harvard
Rogers, C. R. Counselling and psychotherapy. New York: Houghton Mifflin, 1942.
Rogers, C. R. Client-centered therapy: Its current practice, implications, and theory. Boston:
Houghton Mifflin, 1951.
Rogers, E. M., & Neill, R. E. Achievement motivation among Colombian peasants. East Lansing:
Michigan State University, 1966.
Rogers, E. M., & Svenning, L. Modernization among peasants: The impact of communication. New
York: Holt, Rinehart and Winston, 1969.
Rokeach, M. The nature of human values. New York: Free Press, 1973.
Rose, R. M., Holaday, J. W., & Bernstein, I. S. Plasma testosterone, dominance rank, and aggressive
behavior in male rhesus monkeys. Nature, 1971, 231, 366-368.
Rosen, B. C. The achievement syndrome. American Sociological Review, 1956, 21, 203-211.
Rosen, B. C. Race, ethnicity and the achievement syndrome. American Sociological Review, 1959,
24, 47-60.
Rosen, B. C , & D'Andrade, R. G. The psychological origins of achievement motivation. Sociometry,
1959, 22, 185-218.
Rosen, D. H. Lesbianism: A study offemale homosexuality. Springfield, 111.: Charles C Thomas,
1974.
Rosenbaum, G. Stimulus generalization as a function of level of experimentally induced anxiety.
Journal of Experimental Psychology, 1953,45, 35—43.
Bibliography 633
Rosenfeld, H. M., & Franklin, S. S. Arousal of need for affiliation in women. Journal of Personality
and Social Psychology, 1966, 3, 245-248.
Rosenthal, R. Experimenter effects in behavioral research. New York: Appleton-Century-Crofts,
1966.
Rosenthal, R. (Ed.), Skill in nonverbal communication. Cambridge, Mass.: Oelgeschlager, 1979.
Rosenthal, R., & Jacobson, L. Pygmalion in the classroom. New York: Holt, Rinehart and Winston,
1968.
Rosenthal, R., & Rubin, D. B. Interpersonal expectancy effects: The first 345 studies. The Behavioral
and Brain Sciences, 1978, 3, 377-415.
Ross, H. L., & Glaser, E. M. A study of successful persons from seriously disadvantaged backgrounds.
Los Angeles: Human Interaction Institute, 1970.
Rothman, S. Radical Christians, radical Jews. New York: Oxford University Press, 1981.
Rothman, S., et al. Ethnic variations in Student radicalism: Some new perspectives. In S. Bialer & S.
Sluzar (Eds.), Radicalism in the contemporary age: Sources of contemporary radicalism (Vol. 1).
Boulder, Colorado: Westview, 1977.
Rothman, S., & Lichter, S. R. Power, politics and personality in "Post-industrial society." Journal of
Politics, 1978, 40, 675-707.
Rotter, J. B. Generalized expectancies for internal versus external control of reinforcement.
Psychological Monographs, 1966, Whole No. 80.
Rowland, N. E., & Antelman, S. M. Stress-induced hyperphagia and obesity in rats: A possible
model for understanding human obesity. Science, 1976, 191, 310-312.
Rubin, Z. Doing unto others. New York: Prentice-Hall, 1974.
Rubinstein, E., & Sprafkin, J. N. Television: A Channel to social values. In D. C. McClelland (Ed.),
The development of social maturity. New York: Irvington, 1982.
Russett, B. M., et al. World handbook of political and social indicators. New Haven, Conn.: Yale
Ryan, T. A. Intentional behavior. New York: Ronald Press, 1970.
Salter, C. A., Meunier, J. L., & Triplett, W. M. Multiple measurement of anxiety and its effects on
complex verbal learning. Psychological Reports, 1976,55, 691-694.
Samelson, F. The relation of achievement and affiliation motives to conforming behavior in two
conditions of conflict with a majority. In J. W. Atkinson (Ed.), Motives in fantasy, action, and
society. Princeton, N.J.: Van Nostrand, 1958.
Sanford, R. N. The effects of abstinence from food upon imaginal processes: A further experiment.
Journal of Psychology, 1937, 3, 145-159.
Sarason, I. G. Experimental approaches to test anxiety: Attention and the uses of Information. In C.
D. Spielberger (Ed.), Anxiety and behavior (Vol. 2). New York: Academic Press, 1971.
Sarason, S. B., & Mandler, G. Some correlates of test anxiety. Journal of Abnormal and Social
Psychology, 1952, 47, 810-817.
Sarason, S. B., Mandler, G., & Craighill, P. G. The effect of differential instructions on anxiety and
learning. Journal of Abnormal and Social Psychology, 1952, 47, 561-565.
Sawchenko, P. E., & Swanson, L. W. Central noradrenergic pathways for the Integration of
hypothalamic neuroendocrine and autonomic responses. Science, 1981, 214, 685-687.
Schachter, S. The psychology of affiliation. Stanford, Calif.: Stanford University Press, 1959.
Schachter, S. Some extraordinary facts about obese humans and rats. American Psychologist, 1971,
26, 129-144. (a)
Schachter, S. Emotion, obesity, and crime. New York: Academic Press, 1971. (b)
Schachter, S., Goldman, R., & Gordon, A. Effects of fear, food deprivation, and obesity on eating.
Journal of Personality and Social Psychology, 1968, 10, 91-97.
634 Bibliography
Schachter, S., & Singer, J. E. Cognitive, social, and physiological determinants of emotional State.
Psychological Review, 1962, 69, 379-399.
Schein, M. W., & Haie, E. B. Stimuli eliciting sexual behavior. In F. A. Beach (Ed.), Sex and
behavior. New York: Wiley, 1965.
Schmält, H. D. Die Messung des Leistungsmotiv. Göttingen, West Germany: Hogrefe, 1976.
Schnackers, U., & Kleinbeck, U. Machtmotiv und machtthematisches Verhalten in einem
Verhandlungsspiel. Archiv für Psychologie, 1975,727, 300-319.
Schneider, K. Leistungsmotive, Kausalerklärungen für Erfolg und Misserfolg und erlebte Affekte
Nach Erfolg und Misserfolg. Zeitschrift für experimentelle und angewandte Psychologie, 1977,
24, 613-637.
Schneider, K. Atkinson's "risk preference" model: Should it be revised? Motivation and Emotion,
1978, 2, 333-344.
Schneider, K., & Kreuz, A. Die Effekte unterschiedlicher Anstrengung auf die Mengen und
Güteleistung bei einer einfachen und schweren Zahlensymbolaufgabe. Psychologie und Praxis,
1979, 23, 34-42.
Scholem, G. G. Major trends in Jewish mysticism. Jerusalem: Schocken, 1941.
Scholem, G. G. La kabbala et sa symbolique. Paris: Payot, 1966.
Sears, R. R. Survey of objective Status of psychoanalytic concepts. New York: Social Science Research
Council, 1943.
Sears, R. R., Maccoby, E. E., & Levin, H. Patterns of child rearing. Evanston, 111.: Row Peterson,
1957.
Sears, R. R., Rau, L., & Alpert, R. Identification and child rearing. Stanford, Calif.: Stanford
Seligman, M. E. P. Helplessness: On depression, development, and death. San Francisco: Freeman,
1975.
Selye, H. The stress of life. New York: McGraw-Hill, 1956.
Sheffield, F. D., & Roby, T. B. Reward value of a non-nutritive sweet taste. Journal of Comparative
and Physiological Psychology, 1950,45, 471^81.
Sheffield, F. D., Wolff, J. J., & Backer, R. Reward value of copulation without sex drive reduction.
Journal of Comparative and Physiological Psychology, 1951, 44, 3-8.
Sheldon, W. H. (with the collaboration of S. S. Stevens). The varieties of temperament: A psychology
of constitutional dijferences. New York: Harper, 1942.
Sheppard, H. L., & Belitsky, A. H. The job hunt. Baltimore: Johns Hopkins University Press, 1966.
Sherif, M. An outline of social psychology. New York: Harper, 1948.
Shinn, M. Secondary school coeducation and the fears of success and failure. Unpublished honors
thesis, Harvard University, Department of Psychology and Social Relations, 1973.
Shiomi, K. Relations of pain threshold and pain tolerance in cold water with scores on the
Maudsley Personality Inventory and Manifest Anxiety Scale. Perceptual and Motor Skills,
1978, 47, 1155-1158.
Shipley, T. E., Jr., & Veroff, J. A projective measure of need for affiliation. Journal of Experimental
Psychology, 1952, 43, 349-356.
Shrable, K., & Moulton, R. W. Achievement fantasy as a function of variations in self-rated
competence. Perceptual and Motor Skills, 1968,27, 515-528.
Silverman, L. H. Psychoanalytic theory. "The reports of my death are greatly exaggerated."
American Psychologist, 1976, 31, 621-637.
Silverman, L. H., Bronstein, A., & Mendelsohn, E. The further use of the subliminal psychodynamic
actuation method for the clinical theory of psychoanalysis. Psychotherapy: Theory, Research,
and Practice, 1976, 13, 2-16.
Bibliography 635
Silverman, L. H., Klinger, H., Lustbader, L., Farrell, J., & Martin, A. The effect of subliminal drive
Stimulation on the Speech of stutterers. Journal of Abnormal Psychology, 1972, 155, 14-21.
Silverman, L. H., Kwawer, J. S., Wolitzky, C , & Coran, M. An experimental study of aspects of
the psychoanalytic theory of male homosexuality. Journal of Abnormal Psychology, 1973, 82,
178-188.
Sinha, B. P., & Mehta, P. Farmers' need for achievement and change-proneness in acquisition of
information from a farm telecast. Rural Sociology, 1972, 37, A\l-\21.
Skinner, B. F. The behavior of organisms. New York: Appleton-Century-Crofts, 1938.
Skinner, B. F. Science and human behavior. New York: Macmillan, 1953.
Skinner, B. F. The behavior of organisms. New York: Prentice-Hall, 1966, preface to the 7th ed.
Smetana, J. G., & Adler, N. E. Fishbein's value X expectancy model: An examination of some
assumptions. Personality and Social Psychology Bulletin, 1980, 6, 89-96.
Smith, C. P. Achievement-related motives in children. New York: Russell Sage, 1969.
Solomon, R. L. The opponent-process theory of acquired motivation: The costs of pleasure and the
benefits of pain. American Psychologist, 1980, 35, 691-712.
Sonnenfeld, J. It's time to break for a Station identification: A motivational analysis of a College radio
Station as a voluntary organization. Unpublished manuscript, Harvard University, Department
of Psychology and Social Relations, 1975.
Sorrentino, R. M. Extending theory of achievement motivation to the study of group processes. In
J. W. Atkinson & J. O. Reynor (Eds.), Motivation and achievement. Washington, D.C.:
Hemisphere Publishing, 1974.
Sorrentino, R. M., & Sheppard, B. H. Effects of affiliation-related motives on swimmers in individual
versus group competition: A field experiment. Journal of Personality and Social Psychology,
1978, 36, 704-714.
Southwood, K. E. Some sources of political disorder: A cross-national analysis. Unpublished doctoral
dissertation, University of Michigan, 1969.
Spelke, E., Zelazo, P., Kagan, J., & Kotelchuck, M. Father interaction and Separation protest.
Developmental Psychology, 1973, 9(1), 83-90.
Spence, K. W. Behavior theory and conditioning. New Haven, Conn.: Yale University Press,
1956.
Spence, K. W. Behavior theory and selective learning. In M. R. Jones (Ed.), Nebraska Symposium
on Motivation: 1958. Lincoln: University of Nebraska Press, 1958. (a)
Spence, K. W. A theory of emotionally based drive (D) and its relation to Performance in simple
learning situations. American Psychologist, 1958, 13, 131-141. (b)
Spence, K. W., Farber, I. E., & McFann, H. H. The relation of anxiety (drive) level to Performance
in competitional and noncompetitional paired-associates learning. Journal of Experimental
Psychology 1956, 52, 296-305.
Spielberger, C. D. (Ed.). Anxiety and behavior. New York: Academic Press, 1966.
Spielberger, C. D., Gorsuch, R. L., & Lushene, R. E., STAI manual for the state-trait anxiety
inventory. Palo Alto, Calif.: Consulting Psychologists Press, 1970.
Spitz, R. A., & Wolf, K. M. The smiling response: A contribution to the ontogenesis of social
relations. Genetic Psychology Monographs, 1946, 34, 57-125.
Spitz, R. A., & Wolf, K. M. Autoerotism during the first year of life. Psychoanalytic study of the
child, 1949, 3/4, 85-120.
Sroufe, L. A., & Waters, A. The ontogenesis of smiling and laughter: A perspective on the
organization of development of infancy. Psychologie'al Review, 1976, 83, 173-183.
Sroufe, A., & Waters, E. Attachment as an organizational construet. Child Development, 1977, 48,
1184-1199.
636 Bibliography
Stamps, L. The effects of Intervention techniques on children's fear of failure behavior. Journal of
Genetic Psychology, 1973, 123, 85-97.
Steele, R. S. The physiological concomitants of psychogenic motive arousal in College males.
Unpublished doctoral dissertation, Harvard University, 1973.
Steele, R. S. Power motivation, activation, and inspirational Speeches. Journal of Personality, 1977,
45, 53-64.
Steele, R. S. Psychoanalysis and hermeneutics. International Review of Psychoanalysis, 1979, 6,
389-411.
Stein, L. Norepinephrine reward pathways: Role in self-stimulation, memory consolidation, and
schizophrenia. In J. K. Cole & T. B. Sonderegger (Eds.), Nebraska Symposium on Motivation:
Stennet, R. G. The relationship of Performance level to level of arousal. Journal of Experimental
Psychology, 1957, 54, 54-61.
Stewart, A. J. Scoring System for stages of psychological development. Unpublished manuscript,
Harvard University, Department of Psychology and Social Relations, 1973.
Stewart, A. J. Longitudinal prediction from personality to life outcomes among college-educated
women. Unpublished doctoral dissertation, Harvard University, Department of Psychology and
Social Relations, 1975.
Stewart, A. J. The study of experience-induced affective development. Unpublished manuscript,
Boston University, Department of Psychology, 1978.
Stewart, A. J. (Ed.). Motivation and society. San Francisco: Jossey-Bass, 1982.
Stewart, A. J., & ehester, N. L. Sex differences in human social motives: Achievement, affiliation
and power. In A. J. Stewart (Ed.), Motivation and society. San Francisco: Jossey-Bass, 1982.
Stewart, A. J., & Rubin, Z. Power motivation in the dating couple. Journal of Personality and Social
Psychology, 1976, 34, 305-309.
Stone, A. The effect of sanetioned overt aggression on total instigation to aggressive responses.
Unpublished honor's thesis, Harvard University, Department of Psychology and Social
Relations, 1950.
Storms, M. D. A theory of erotic orientation development. Psychological Review, 1981, 88, 340-
353.
Sullivan, H. S. The interpersonal theory of psychiatry. New York: Norton, 1953.
Sulloway, F. Freud: Biologist of the mind. New York: Basic, 1979.
Swanson, G. E. Religion and regime. Ann Arbor: University of Michigan Press, 1967.
Taylor, J. A. The relationship of anxiety to the conditioned eyelid response. Journal of Experimental
Psychology, 1951, 41, 81-92.
Taylor, J. A. A personality scale of manifest anxiety. Journal of Abnormal and Social Psychology,
1953, 48, 285-290.
Teevan, R. C , Burdick, H., & Stoddard, N. Relationship between incentive and expeetations of
success. Psychological Reports, 1976, 39, 411-419.
Telford, C. W. The refractory phase of voluntary and associative processes. Journal of Experimental
Psychology, 1931, 14, 1-36.
Tennyson, R. D., & Woolley, F. R. Interaction of anxiety with Performance on two levels of task
difficulty. Journal of Educational Psychology, 1971, 62, 463-467.
Terhune, K. W. Motives, Situation and interpersonal conflict within prisoners' dilemma. Journal of
Personality and Social Psychology, 1968, 8(3), Part 2. (Monograph Suppl.) (a)
Terhune, K. W. Studies of motives, cooperation and conflict within laboratory microcosms. In G. H.
Snyder (Ed.), Studies in international conflict. Buffalo, N.Y.: University of Buffalo Studies,
1968, 4(1), 29-58. (b)
Bibliography 637
Tessler, M. A., O'Barr, W. M., & Spain, D. H. Tradition and identity in changing Africa. New
York: Harper & Row, 1973.
Thayer, R. E. Measurement of activation through self-report. Psychological Reports, 1967, 20,
663-678.
Thompson, D. A., & Campbell, R. G. Hunger in humans induced by 2-deoxy-D-glucose:
Glucoprivic control of taste preference and food intake. Science, 1911, 198, 1065-1068.
Thorndike, E. L. The associative processes in animals. Boston: Ginn and Company, 1899.
Thorndike, E. L. Animal intelligence. New York: Macmillan, 1911.
Thurstone, L. L. Ability, motivation, and speed. Psychometrika, 1937, 2, 249-254.
Tinbergen, N. The study of instinct. London: Oxford University Press, 1951.
Tolman, E. C. Purposive behavior in animals and men. New York: Appleton-Century-Crofts,
1932.
Tomkins, S. S. Ajfect, imagery, consciousness: The positive affects (Vol. 1). New York: Springer,
1962.
Tomkins, S. S. Ajfect, imagery, consciousness: The negative affects (Vol. 2). New York: Springer,
1963.
Toynbee, A. J. A study of history. New York: Oxford University Press, 1947.
Trevelyan, G. M. English social history. New York: Longmans, Green, 1942.
Trope, Y. Seeking information about one's ability as a determinant of choice among tasks. Journal
of Personality and Social Psychology 1975,52, 1004-1013.
Udry, J. R., & Morris, N. M. Distribution of coitus in the menstrual cycle. Nature, 1968, 220,
593-596.
Uleman, J. S. A new TAT measure of the need for power. Unpublished doctoral dissertation, Harvard
University, 1966.
Uleman, J. S. The need for influence: Development and validation of a measure, in comparison with
a need for power. Genetic Psychology Monographs, 1972, 85, 157-214.
Underwood, B., Moore, B. S., & Rosenhan, D. Affect and self-gratification. Developmental
Psychology, 1972, 8, 209-214.
Valins, S. Cognitive effects of false heart-rate feedback. Journal of Personality and Social Psychology,
1966, 4, 400-408.
Vander, A. J., Sherman, J. H., & Luciano, D. S. Human physiology: The mechanisms of body
function. New York: McGraw-Hill, 1975.
Varga, K. Who gains from achievement motivation training? Vikalpa (The Journal for Decision
Makers), 1977, 2, 187-200. (Available from the Indian Institute of Management, Ahmadabad,
India.)
Veroff, J. Development and validation of a projective measure of power motivation. Journal of
Abnormal and Social Psychology, 1957, 54, 1-8.
Veroff, J. Social comparison and the development of achievement motivation. In C. P. Smith (Ed.),
Achievement related motives in children. New York: Russell Sage, 1969.
Veroff, J. Assertive motivations: Achievement versus power. In D. G. Winter & A. J. Stewart (Eds.),
Motivation and society San Francisco: Jossey-Bass, 1982.
Veroff, J., Atkinson, J. W., Feld, S. C , & Gurin, G. The use of thematic apperception to assess
motivation in a nationwide interview study. Psychological Monographs, 1960, 74(12, Whole No.
499).
Veroff, J., Depner, C , Kulka, R., & Douvan, E. Comparison of American motives: 1957 versus
1976. Journal of Personality and Social Psychology, 1980, 39, 1249-1262.
Veroff, J., Douvan, E., & Kulka, R. The American experience. Basic, 1982.
638 Bibliography
Veroff, J., & Veroff, J. P. B. Reconsideration of a measure of power motivation. Psychological

Bulletin, 1972, 78, 279-291.
Vogel, W., Baker, R. W., & Lazarus, R. S. The role of motivation in psychological stress. Journal of
Waelder, R. Basic theory of psychoanalysis. New York: International University Press, 1960.
Wainer, H. A., & Rubin, I. M. Motivation of research and development entrepreneurs. Journal of
Applied Psychology, 1969, 53, 178-184.
Walker, E. L. Psychological complexity as a basis for a theory of motivation and choice. In D.
Levine (Ed.), Nebraska Symposium on Motivation: 1964. Lincoln: University of Nebraska Press,
1964.
Walker, E. L. Psychological complexity and preference: A hedge-hog theory of behavior. In D. E.
Berlyne & K. B. Madsen (Eds.), Pleasure, reward, preference. New York: Academic Press,
1973.
Walker, E. L., & Atkinson, J. W. The expression of fear-related motivation in thematic apperception
as a function of proximity to an atomic explosion. In J. W. Atkinson (Ed.), Motives in fantasy,
action, and society. Princeton, N.J.: Van Nostrand, 1958.
Walker, E. L., & Heyns, R. W. An anatomy for conformity. Englewood Cliffs, N.J.: Prentice-Hall,
1962.
Walker, W. I. Escape, exploratory, and food seeking responses of rats in a novel Situation. Journal
of Comparative and Physiological Psychology, 1959,52, 106-111.
Walster, E., Walster, G. W., & Berscheid, E. Equity theory and research. Boston: Allyn and Bacon,
1978.
Ward, L., & Wilson, J. P. Motivation and moral judgment as determinants of behavioral
acquiescence and moral action. Journal of Social Psychology, 1980, 112(2), 271-286.
Warden, C. J. Animal motivation: Experimental studies on the albino rat. New York: Columbia
Watson, R. L, Jr. Motivational and sex differences in aggressive behavior. Unpublished doctoral
Weber, M. [The Protestant ethic and the spirit of capitalism] (T. Parsons, trans.). New York:
Scribner, 1930. (Originally published, 1904.)
Weiner, 3. Role of success and failure in the learning of easy and complex tasks. Journal of
Weiner, B. A theory of motivation for some classroom experiences. Journal of Educational
Psychology, 1979, 71, 3-25.
Weiner, B. Human motivation. New York: Holt, Rinehart and Winston, 1980. (a)
Weiner, B. A cognitive (attribution)-emotion-action model of motivated behavior: An analysis of
judgments of help-giving. Journal of Personality and Social Psychology, 1980, 39, 186—
200. (b)
Weiner, B. The emotional consequences of causal ascriptions. Unpublished manuscript, University of
California, Los Angeles, Department of Psychology, 1981.
Weiner, B., Frieze, I., Kukla, A., Reed, L., Rest, S., & Rosenbaum, R. M. Perceiving the causes of
success and failure. In E. E. Jones, D. E. Kanouse, H. H. Kelley, R. E. Nisbett, S. Valins, &
B. Weiner (Eds.), Attribution: Perceiving the causes of behavior. Morristown, N.J.: General
Learning Press, 1971.
Weiner, B., Heckhausen, H., Meyer, W. U., & Cook, R. E. Causal ascriptions and achievement
behavior. Journal of Personality and Social Behavior, 1972, 21, 239-248.
Weiner, B., & Kukla, A. An attributional analysis of achievement motivation. Journal of Personality
Bibliography 639
Weiner, B., & Potepan, P. A. Personality characteristics and affective reactions toward exams of
superior and failing College students. Journal of Educational Psychology, 1970, 61, 144-
151.
Weiner, B., Russell, D., & Lerman, D. The cognition-emotion process in achievement-related
context. Journal of Personality and Social Psychology, 1979, 37, 1211-1220.
Weiss, J. M , Stone, E. A., & Harrell, N. Coping behavior and brain norepinephrine in rats. Journal
of Comparative and Physiological Psychology, 1970,72, 153-160.
Weitzner, M. Manifest anxiety, amphetamine, and Performance. Journal of Psychology, 1965, 60,
71-79.
Wendt, H. W. Motivation, effort and Performance. In D. C. McClelland (Ed.), Studies in motivation.
New York: Appleton-Century-Crofts, 1955.
Wendt, H. W. Risk-taking as a function of pre-verbal 'imprinting'? Some data and speculations.
Archiv für die gesamte Psychologie, 1961,775, 325-350.
Wendt, H. W. Early circannual rhythms and adult human behavior: Components of a
chronobehavioural theory, and critique of persistent artifacts. International Journal of
Chronobiology, 1974, 2, 57-86.
White, R. W. Competence and the psychosexual stages of development. In M. R. Jones (Ed.),
Whiting, B. B. Sex identity conflict and physical violence: A comparative study. American
Anthropologist, 1965, 67, 123-140.
Whiting, B. B., & Whiting, J. W. M. Children of six cultures: A psycho-cultural analysis. Cambridge,
Mass.: Harvard University Press, 1975.
Wilsnack, S. C. The effects of social drinking on women's fantasy. Journal of Personality, 1974, 42,
43-61; 243-261.
Wilson, G. T., & Lawson, D. M. Expectancies, alcohol, and sexual arousal in male social drinkers.
Journal of Abnormal Psychology, 1976, 85, 587-594. (a)
Wilson, G. T., & Lawson, D. M. The effects of alcohol on sexual arousal in women. Journal of
Abnormal Psychology, 1976, 85, 489^97. (b)
Wilson, G. T., & Lawson, D. M. Expectancies, alcohol, and sexual arousal in women. Journal of
Abnormal Psychology, 1978,57, 358-367.
Winter, D. G. Power motivation in thought and action. UnpuMished doctoral dissertation, Harvard
University, 1967.
Winter, D. G. The power motive. New York: Free Press, 1973.
Winter, D. G. The power motive in women. Unpublished manuscript, Wesleyan University,
Department of Psychology, 1982.
Winter, D. G., & Healy, J. M., Jr. An integrated System for scoring motives in running text:
Reliability, validity, and convergence (Paper presented at the American Psychological
Association, Los Angeles, 1981). Department of Psychology, Wesleyan University, 1982.
Winter, D. G., McClelland, D. C , & Stewart, A. J. A new defense of the liberal arts. San Francisco:
Jossey-Bass, 1982.
Winter, D. G., & Stewart, A. J. Power motive reliability as a function of retest instructions. Journal
of Consulting and Clinical Psychology, 1977, 45, 436-440.
Winter, D. G., & Stewart, A. J. Power motivation. In H. London and J. Exner (Eds.), Dimensions
of personality. New York: Wiley, 1978.
Winter, D. G., Stewart, A. J., & McClelland, D. C. Husband's motives and wife's career level.
Journal of Personality and Social Psychology, 1977, 35, 159-166.
Winter, S. Characteristics of fantasy while nursing. Journal of Personality, 1969, 37, 58-72.
640 Bibliography
Winterbottom, M. R. The relation of need for achievement to learning experiences in independence

and mastery. In J. W. Atkinson (Ed.), Motives in fantasy, action, and society. Princeton, N. J.:
Van Nostrand, 1958.
Wise, R. A. Action of drugs of abuse on brain reward Systems. Pharmacology, Biochemistry and
Behavior, 1980, 13 (Suppl. 1), 213-223.
Witteoff, C. Power motivation in the schools. Unpublished manuscript, Washington University,
Graduate Institute of Education, 1980.
Wolfe, J. B. Effectiveness of token-rewards for chimpanzees. Comparative Psychology Monographs,
1936, 12 (No. 60).
Wolfenstein, M., & Leites, N. The movies: A psychological study. New York: Free Press, 1950.
Woodworth, R. S. Experimental psychology. New York: Holt, 1938.
Wormley, W. Portfolio manager preference in an investment decision-making Situation: A
psychophysical study. Unpublished doctoral dissertation, Harvard University, 1976.
Wortman, C. B., & Brehm, J. W. Responses to uncontrollable outcomes: An integration of reaetance
theory and the learned helplessness model. Advances in Experimental Social Psychology, 1975,
8, 277-336.
Wundt, W. Grundzüge der physiologischen Psychologie. Leipzig: Engelmann, 1874.
Yerkes, R. M., & Dodson, J. D. The relation of strength of Stimulus to rapidity of habit formation.
Journal of Comparative and Neurological Psychology, 1908, 18, 459-482.
Young, P. T. Motivation and emotion. New York: Wiley, 1961.
Youngleson, M. L. The need to affiliate and self-esteem in institutionalized children. Journal of
Zajonc, R. B. Social facilitation. Science, 1965, 149, 269-214.
Zajonc, R. B. Feeling and thinking: Preferences need no inferences. American Psychologist. 1980, 35,
151-175.
Zajonc, R. B., & Sales, S. M. Social facilitation of dominant and subordinate responses. Journal of
Experimental and Social Psychology, 1966,2, 160-168.
Zanna, M. P., & Cooper, J. Dissonance and the pill: An attribution approach to studying the
arousal properties of dissonance. Journal of Personality and Social Psychology, 1974, 29,
703-709.
Zeigarnik, B. Über das Behalten von erledigten und unerledigten Handlungen. Psychologische
Forschung, 1927, 9, 1-85.
Zimbardo, P. G. The human choiee: Individuation, reason, and order versus deindividuation,
impulse, and chaos. In W. J. Arnold & D. Levine (Eds.), Nebraska Symposium on Motivation:
Zimbardo, P. G. Psychology and life. Glenview, 111.: Scott, Foresman, 1979.
Zuckerman, M. The Sensation seeking motive. Progress in Experimental Personality Research, 1974,
7, 80-148.
Zuckerman, M. General and situation-speeifie traits and states: New approaches to assessment of
anxiety and other instinets. In M. Zuckerman & C. D. Spielberger (Eds.), Emotions and
anxiety. Hillsdale, N.J.: Erlbaum, 1976.
Zuckerman, M., & Wheeler, L. To dispel fantasies about the fantasy-based measure of fear of
success. Psychological Bulletin, 1975, 82, 932-946.
Zumkley, H. Aggression und Katharsis. Göttingen, West Germany: Hogrefe, 1978.

David C. McClelland - Human - Motivation PDF

Uploaded by

Copyright:

Available Formats

David C. McClelland - Human - Motivation PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

David C. McClelland - Human - Motivation PDF

Uploaded by

Copyright:

Available Formats

What are some of the major topics covered in the book?

What are some of the major topics covered in the book?

How does the book approach measuring individual differences in human motive strength?

How does the book approach measuring individual differences in human motive strength?

Preface

The psychology of motivation is a broad and loosely defined field. It covers

Part 2 The Nature of Human Motives 105

Part 3 Important Motive Systems 221

Outlets of the Need for Power 280

Part 4 Contextual Effects on Human Motives 413

13 How Motives Interact with Values and Skills to Determine What

• Motives as One of Three Major Determinants of Behavior

• Are Unconscious Motives Important?

Consciously wanting something is an everyday experience. It will be called a

Conflicts in Conscious Intents

Needs music (2) X Enjoyment from opera (5) 10

Avoidance-avoidance conflict: deciding whether to endure Jung's criticisms or the

Needs acceptance of psychoanalysis (10) X Anti-Semitic rejection ( — 5) —50

Approach-avoidance conflict: deciding whether to teil the truth or avoid Breuer's

Needs Breuer's friendship (5) X Breuer's disapproval ( — 6) —30

An approach-approach conflict is unstable and easily solved. The traditional

of this alternative. The formula determines the attractiveness of Freud's staying

o ——~^ ^-^___<>_Strong Approach

Another consequence of the difference in slopes is illustrated by the example

Conscious Goal Setting

The relation of effort to Performance has been confirmed in a different type

Is Conscious Intent Necessary for Learning?

Factors Underlying Conscious Intents

I sacrificed unhesitatingly my budding popularity as a physician and a growing prac-

Forgetting a Proper Name Unintentionally

Freud's Explanation Forgotten Name Our Exploitation

Guilt over suicide of his patient Substitute Names Recalled:

Unconscious Motives in Freud

Conclusions to Be Drawn from Freud's Approach to Motivational Analysis

EXPERIMENTAL STUDY OF UNCONSCIOUS MOTIVES

• ARE UNCONSCIOUS MOTIVES IMPORTANT?

tion with unconscious motives that developed in psychiatry following Freud

Table 1.2. (Continued).

NOTES AND QUERIES

3. If setting a conscious goal reflects motivational and nonmotivational deter-

• Motives as Reasons for What People Spend Their Time Doing

• Motives as Reasons for Abnormal Behavior

• Motives as Reasons for Creativity and Growth

• Measuring Human Motives

• Stages in Motivational Development

• Checking the Validity of the Motivational Stage Theory

• MOTIVES AS REASONS FOR

To a considerable extent there is a theory of motivation to go with every field of

M cDougall's Classification of Motives

1. Food-seeking propensity: to seek (and perhaps störe) food.

MOTIVES AS REASONS FOR ABNORMAL BEHAVIOR

The Freudian Motivational Trilogy: Sex, Aggression, and Anxiety

Bad Desires and Good Desires

Jung's View of Motivation

Basic Motives and Later Psychoanalytic Theorists

curate to describe an infant as primarily motivated by physiological needs such

Lower needs are characterized by deficits—such as the lack of food and

MEASURING HUMAN MOTIVES

An early attempt at measuring human motives was made by W. H. Sheldon