Antonovsky (1979)

:
I
DATE DUE
Aaron Antonovsky
Health, Stress,
and Coping
•• Jossey-Bass Publishers
dK San Francisco • Washington • London • 1982
HEALTH, STRESS, AND COPING
Sew Perspectives on Mental and Physical Well-Being
by Aaron Antonovsky
Copyright © 1979 by: Jossey-Bass Inc., Publishers

433 California Street
San Francisco, California 94104
&
Jossey-Bass limited
28 Banner Street
London EC1Y 8QE
Copyright under International, Pan American, and

Universal Copyright Conventions. All rights
reserved.No part of this book may be reproduced
in —
any form except for brief quotation (not to
exceed 1,000 words) in a review or professional
—
work without permission in writing from the publishers.
Library of Congress Catalogue Card Number LC 79-83566
International Standard Book Number ISBN 0-87589-412-7
Manufactured in the United States of America
JACKET DESIGN BY WILLI BAUM
FIRST EDITION
First printing: May 1979
Second printing: February 1980
Third printing: January 1981
Fourth printing: April 1982
Code 7917
The Jossey-Bass
Social and Behavioral Science Series
For Helen
With love and gratitude
Preface
This book poses the question of salutogenesis, of the origins (genesis)

of health (saluto). The very need for a neologism suggests that,
whatever the work that has been done in this area, little of it can be
called serious or systematic; the study of pathogenesis still over-
whelmingly dominates medical research, whether biological or so-
cial.The book also proposes an answer to the question of saluto-
genesis The origins of health are to be found in a sense of coherence.
:
Taken together, the question and the answer are called the saluto-
genic model. This model, proposed by a sociologist, is, of course,
only a partial conceptualization of one of the greatest mysteries in
the study of people : How
do we manage to stay healthy? I should
like to think that the approach has sufficient cogency for it to be
taken up by colleagues in other life sciences, so that some day more

of the mystery is unraveled.
The book is not intended to be popular. It offers no easy
solutions; as the Epilogue indicates, I tend to be a pessimist. I make
Vll
viii Preface
no attempt to simplify a highly complex problem or to avoid techni-

cal discussion where it is necessary. The book, however, is not
directed only to my major reference group, my colleagues in medical
sociology. It has something to say to all those who, professionally
and personally, are committed to understanding and enhancing the

adaptive capacities of human beings. The professions of students
in one of my seminars — sociologist, psychologist, psychiatric nurse,
physician, health care organizer, epidemiologist, architect, commu-
nity organizer —suggest a cross-section of my intended audience.
All advances are made by exploiting the past. I trust that
those whose work has been useful to me have not been neglected in
the bibliographic references. But since the book is a culmination of
some fifteen years of work, as reviewed in the Introduction,
I have
surelyborrowed from some studies without adequate acknowledg-

ment. To their authors, my apologies. As to the works with which I
should be familiar but am not, I can only plead guilty to being far
from an encyclopedist.
It is appropriate to express particular indebtedness to those
whose work has been of major significance for me. The responsibility
is fully mine, but much of whatever merit the book has is due to
what I have learned from Hans Selye, Rene* Dubos, George Engel,
and, above all, my friend and colleague Melvin Kohn.
As always, one's work is shaped and facilitated by personal
contacts as well as by those whose work one has read. In this con-
text, I must mention Abraham David Katz. Not only was David Katz
a most provocative research assistant and colleague from 1968 to

1973, when many of my ideas were taking shape (Antonovsky,
1975), but more than anyone I have ever known, he embodied
what I later came to call the sense of coherence. Unhappily, David
Katz will never know of my gratitude; he was killed at the Suez
Canal in the Yom Kippur War.
More recendy, liana Shoham worked as my research assis-
tant. Together with her I moved from the trees of generalized

resistance resources to the forest of the sense of coherence. Dana
Shoham's brightness and critical skepticism have certainly con-
tributed to my thinking through many problems.
The final draft of the book was written during one of the
most relaxed years I have ever known, a sabbatical leave at the
Preface ix
School of Public Health, University of California, Berkeley. My

thanks are due to Moshe Prywes, my dean at the Beersheba School
of Medicine, for approving a sabbatical prior to the graduation of
our first class; to Warren Winkelstein, dean of the School of Public
Health, and to William Reeves, director of the Program in Epi-
demiology, who were my formal hosts. S. Leonard Syme, chairman
of the Department of Biomedical and Environmental Health Sci-
ences, invited me to Berkeley and, with the departmental adminis-
trative officer, Ingalill Fivian, worried about my every comfort.
Because Len Syme is probably the medical sociologist whose think-
ing is my own, I found it most useful as well as most
closest to
pleasurable to have spent a year down the hall from him.

As anyone who has ever written a book knows, the cheer, in-
humor, and efficiency of those who type and
telligence, sense of
retype the manuscript are often decisive to the author's sanity. Hav-
ing Constance Long and Donna Kimmel work with me was a sheer
pleasure. I am truly grateful.
Among the wise decisions I have made was that to give a
seminar at Berkeley devoted wholly to the manuscript of this book.
These classes became a continual source of challenge for me. I did
not always accept my student-colleagues' (they quickly became
such) criticisms, but the need to explain why was as useful as the
ideas I did take from them. My sincerest thanks, then, to Louis
Allen, Margaret Boyd, Lynda Brodsky, Shoshanna Churgin, Ber-

nard Cordes, Michael Freeman, Leslie Grant, Mario Gutierrez,
Kathryn Johnson, Barbaraterry Kurtz, Eric Rosen, Barbara Sack-
off, William Satariano, Sandra Stein, and Claudine Torfs.
Over the years, I have been the recipient of a number of

grants, which have allowed me to pursue my work, not always as
spelled out in the grant request. For these I am grateful to the
United States National Institutes of Health, the Ford Foundation
through the Israel Foundations Trustees, and, most recently, the
Israeli Ministry of Health for its support of me as an established in-
vestigator, Chief Scientist's Bureau.
Beersheba, Israel Aaron Antonovsky

April 1979
Contents
Preface vii
The Author xiii
Prologue : Evolution of a New Perspective 1
One Studying Health Instead of Disease 12
Two Measuring Health on a Continuum 38
Three Stressors, Tension, and Stress 70
Four Tension Management and Resources for

Resistance 98
xi
xii Contents
Five Perceiving the World as Coherent 123
Six Relation of the Sense of Coherence to Health 160
Seven The Salutogenic Model of Health 182
Eight Implications for an Improved Health Care

System 198
Epilogue: Outlook for Human Health 220
References 229
Index 247
The Author
Aaron Antonovsky is Kunin-Luncnfeld Professor of Medical

Sociology and chairman of the Department of the Sociology of
Health in the Faculty of Health Services, Ben Gurion University of
the Negev, Beersheba, Israel.
Antonovsky was born in Brooklyn in 1923 and attended
Brooklyn College, where his undergraduate studies in history and
economics were interrupted by service in the United States Army
during World War II; he was awarded the B.A. degree in 1945.
He did his graduate work in the Department of Sociology at Yale
University, where he had his first contact with the new field of
medical sociology, and was awarded the M.A. and Ph.D. degrees
in sociology in 1952 and 1955, respectively. He taught evenings at
Brooklyn College from 1955 to 1959, and in 1956 he became re-
search director of the New York State Commission Against Dis-
crimination. He was a Fulbright Professor of Sociology at the
University of Teheran in 1959-1960.
xiii
xiv The Author
With his wife Helen, a developmental psychologist, he emi-

grated to Israel in 1960 and accepted a post at the Israel Institute
for Applied Social Research in Jerusalem, with which he is still
affiliated. In the early 1960s, he began to teach in the Department
of Social Medicine, and he started a research project with Judith

Shuval on the latent functions of health care institutions. This
project was followed by work on multiple sclerosis, coronary artery
disease, menopause, preventive dental-health behavior, and early
detection of cancer, as well as a series of studies on social class and
various aspects of health and illness. Antonovsky spent 1965-1966
as a research fellow Harvard School of Public Health. In
at the
up a school of
1972, he was invited to take a central part in setting
medicine at Ben Gurion University of the Negev, where he is also
chairman of the Admissions Committee and where he continues his
commitment to sociomedical research.
Health, Stress,
and Coping
New Perspectives on Mental

and Physical Well-Being
V
Prologue: Evolution
of a New Perspective
It is said that Oriental scholars find it appropriate to introduce their

work by giving a personalized account of how they came to work
on the problem and how they moved ahead, made mistakes, clari-
fied positions, made and abandoned intellectual commitments, and
finally reached the point at which they now tentatively stand. I find
this approach appealing. I would like to think that, by pointing out
how I came to face the issues with which I propose to deal, I can
contribute clarity to the presentation. This, then, is the justification
for introducing this study in somewhat autobiographical terms.
Although retrospection always raises the danger of distorting the
meaning experiences had when they occurred, let me nonetheless
seek to retrace my steps, searching for the origins and development
of my concern with what will be called salutogenesis and the sense
of coherence.
2 Health, Stress, and Coping
My doctoral dissertation (Antonovsky, 1955, 1956) was not

concerned explicidy with either health or psychosocial stressors. I
asked how members of (what at that time was called) a minority
group defined their marginal social situation. In fact, I was investi-
gating cognitive coping responses to a socially structured psycho-

social stressor. In the six years that followed, I was concerned with
the same issue, though the focus shifted. Writing a history of the
Jewish labor movement in the United States (Antonovsky, 1961),

I dealt with the organizational response on a group level to a major
nineteenth- and twentieth-century stressor, immigration. Then, as re-

search director of the New York Commission Against Dis-
State
crimination in an era when poverty and discrimination had not
quite yet been publicly discovered, I directed a number of studies
on the consequences of the stressors of low income and discrimina-
tion (Antonovsky and Lorwin, 1959).
Shortly after I migrated to Israel in 1960, 1 became involved
in a major study in the sociology of medicine, focused on the fit
between "nonmedical" needs and the social structure of the health
services (Shuval, Antonovsky, and Davies, 1970) At the same time,
.
I began teaching in a school of public health. Both these projects
were preliminaries to my direct entry into the field of stress research.

I began collaboration with neurologists on an epidemiological study
of multiple sclerosis.The literature indicated a direct correlation
between distance from the equator and incidence rates. Being an
anthropologically oriented sociologist, I hypothesized that societies
closer to the equator differed, among other ways, from the devel-
oped societies further away in that societies closer to the equator
both suffered fewer stressors and were socioculturally more com-
petent at coping with the stressors they did face. On the basis of
a relatively few items in a questionnaire, I committed myself to
defining stressors objectively as those experiences that anyone any-
where would agree were stressors — for example, going hungry for
extended periods. Our findings were suggestive though not defini-
tive (Antonovsky and others, 1965; Antonovsky and Kats, 1967).
In 1965, enjoying a sabbatical at the Harvard School of
Public Health, I started a series of studies that marked a continua-
tion ofmy commitment to conceptualizing stressors as objective
phenomena and hypothesizing a direct link between stressors and
Prologue: Evolution of a New Perspective 3
diseases. I reviewed all the empirical studies that related social class
and some measure of disease. My primary concern was to bring the
data together, rather than go behind the data and ask Why? (An-
tonovsky, 1967a, 1967b, 1968; Antonovsky and Bernstein, 1977).
During that year, however, the late John Kosa, Irving Zola,
and I embarked on an editing venture that resulted in Poverty and
Health (Kosa, Antonovsky, and Zola, 1969). In the book we asked
Why? and sought to answer the question by having those writing the
What are the stressors in the
papers address themselves to this issue :
lives of poor people that underlie the brute fact that, with regard to
everything related to health, illness, and patienthood, the poor are
screwed?
Ideas presented in several of the chapters, and particularly
in Marc paper on mental health, marked a turning point
Fried's
in my Not only were the stressors important, he pointed
thinking.
out; but the poor also ended up badly because they had less where-
withal to battle these stressors. This marked the germination of an
idea. If two people were confronted by an identical stressor, it
struck me, but one had the wherewithal to successfully meet the
challenge and the other did not, how could this situation best be
conceptualized? Stressors by definition place a load on people. It oc-
curred to me to call the strain incurred tension. The word stress
would then be reserved for the strain that remains when the tension
is not successfully overcome. The distinction compelled me to intro-
duce a further concept, that of tension management, that is, the
process of dealing with the tension.
I became so enthusiastic about this step forward that it came
to dominate the design of my next major study. In collaboration
with Ascher Segall, who had responsibility for the idea, for obtain-
ing funding, and for doing the most complex epidemiological and
medical work, I embarked on a study of the relationship between
migration to Israel and coronary heart disease among North Ameri-
cans (and their nonmigrant siblings). In attempting to determine
whether people who migrate from a more industrialized to a less
industrialized society —rare birds, indeed —have lower rates of heart

disease, I included many measures of stressors in the study; but my
real interest had come to be tension management and overcoming
stressors (Antonovsky, 1971).
There was, however, a complication. Having abandoned

stressors, so to speak, Iwas influenced by what has come to be
known as the transactional approach. Looking back, I find confu-
sion between saying "the impact of a given external situation upon
a person is mediated by the psychological, social, and cultural re-
sources at his disposal" (Antonovsky and Kats, 1967, p. 16) and
"what is important for [the] consequences [of a life crisis] is the sub-
jective perception of the meaning of the event rather than its ob-
jective character" (Antonovsky, 1974, p. 246). My failure to distin-
guish between the definition of an event as a nonstressor (suggested
in the second quote) and the employment of resources to overcome
an event that is defined as a stressor was not, I think, eliminated
until I came to write Chapter Three in this book.
But to return to 1968. I presented a paper on the design of
the coronary study at a seminar attended by the noted British epi-
demiologist J. N. Morris. After I had distinguished among stressor,
tension, tension-management resources, and stress, Morris cogendy
remarked that he saw no particular reason that justified the use of
the model in relation only to heart disease. It could just as reason-
ably (or unreasonably) be applied to any other disease. I seemed to
be, he noted, interested in breakdown rather than in heart disease
per se. All the way back to Jerusalem from Tel Aviv, the phrase
rang in my head. It should be noted that in addition to the multiple
sclerosis and coronary studies, I had also worked on a study of pre-
ventive dental-health behavior and was at the time engaged in a
large-scale study of ethnic differences in adjustment to problems of
climacterium. In addition, from my work on social class and health,
I was familiar with agood part of the epidemiological and sociologi-
cal literature on cancer, mental illness, and so on.
And then it struck me. By God, Morris is right. I am not
interested in heart disease or multiple sclerosis or cancer; I am inter-
ested in breakdown. This, then, is the origin of my first major de-
parture from the mainstream. Preparing what has come to be known
as my "breakdown" paper for the Social Science and Medicine
Conference in Aberdeen in 1968 (Antonovsky, 1972), I proposed a
fundamental distinction between two problems: First, there is the
classic problem of all of medicine: why someone gets a particular
disease (or, in epidemiological terms, why a given group has a high
rate of a particular disease) . But, second, there is the problem of why

someone gets dis-ease (as in John Donne's "This great hospital, this
sick, this dis-easeful world" ) , or the problem of breakdown, what-
ever particular form or diagnostic category the breakdown might
have.
Explicidy, I had committed myself to the idea that it was
important to study breakdown as a dependent variable. I proposed
an operational definition of the concept, later to be tested in the
field ( Antonovsky, 1973 ) . The concept rests on the notion that there
are common facets to all diseases. This definition inevitably led me
to the notion of generalized resistance resources. Influenced by Hans
Selye's concept of the general adaptation syndrome, I wrote ( 1972,
p. 541) : "Because the demands which are made on people are so
variegated and in good part so unpredictable, it seems imperative to
focus on developing a fuller understanding of those generalized
resistance resources which can be applied to meet all demands."
Further influenced by Rene Dubos, I began to explore the concept
of adaptability in the psychological, social, and cultural spheres as
one major key to successful tension management in coping with a
wide variety of stressors.
At this point, then, the dependent variable that concerned

me was breakdown; the independent variables were generalized
resistance resources. Note that the level of stressors, objectively or
subjectively defined, no longer interested me. It was quite natural for
me, as a sociologist, to conceptualize breakdown, or the overall
health state, as a continuum. Not being a clinician,
was not caughtI
in the bind of categorizing people as healthy or was clear to

sick. It
me that all of us, as long as we are alive, are in part healthy and in
part sick, that is, we are somewhere on the breakdown continuum.
But "overall health state" is a misnomer. Like everyone else, I had a
pathogenic orientation. (The very use of the word breakdown points
My question was: What explains the fact that
in this direction.)
peoplemove down on the breakdown continuum?
The next major step in my thinking came in the context of
work on the menopause-adaptation study mentioned above. Our
central interest was in the relationship between the traditionalism-
modernity cultural continuum and successful adaptation to problems
confronting women in the age of climacterium. Our quite large-
scale study (some 1,150 women y was based on representative sam-

ples of five ethnic groups in Israel that could reasonably be ranked
on a traditionalism-modernity continuum (Datan, 1971; Datan,
Antonovsky, and Maoz, forthcoming) We had good theoretical rea-
.
sons for predicting a direct correlation between traditionalism and

adaptation (measured in a wide variety of ways, including medical
examination and location on the breakdown continuum). That is,
we had reason to expect that the relatively most traditional Israeli
Arab village women would be best adapted. But we had just as good
reason to predict that the urban, middle-class Israeli Jewish women
of Central European origin would be best adapted. I shall not re-
veal who on our team (a psychiatrist, a developmental psychologist,
and a sociologist) took which position. But we were all right. The
relation was curvilinear. The Central European women were best
adapted, by and large, and close behind them came the Arab village
women. The Israeli Jewish women of Turkish, North African, and
Persian origin were most poorly adapted. Our post hoc analysis led
to what for me was a major insight. The crucial variable in success-
ful adaptation was not the content of culture and social structure
but its relative stability. The Central European and Arab women
were rooted in stable cultural contexts; the other three groups were
immigrants in transition, uprooted and not yet rerooted. As will be-
come clear, particularly in Chapter Four, this insight became a cen-

tral element in my thinking about stress.
The menopause was the stimulus for an even more
project
fundamental turn in my For some reason that I cannot
thinking.
recall, we had included this question "During World War II, were
:
—
you in a concentration camp yes or no?" Of the 287 Central Euro-
—
pean women a representative sample of this age-sex-ethnic group
taken from a middle-class Israeli community 77 said "yes." There—
is, in the scientific literature, almost no instance of a randomly
selected sample study group compared with a control group. And

we had some fourteen measures of adaptation. The results may not
be surprising (Antonovsky and others, 1971): Significantly fewer
of the camp survivors than of the control group were well adapted,
no matter how adaptation was measured.
But then came what for me was the revolutionary question
and the origin of my concern with what I only later began to call
saluto genesis. A statistically significant difference between groups

simply means that more of Group A than of Group B are high than
can be accounted for by chance. It does not mean that no one in
Group B is high. More than a few women among the concentration
camp survivors were well adapted, no matter how adaptation was
measured. Despite having lived through the most inconceivably in-
human by Displaced Persons camps, illegal

experience, followed
immigration to Palestine, internment in Cyprus by the British, the
Israeli War of Independence, a lengthy period of economic auster-
ity, the Sinai War of 1956, and the Six Day War of 1967 (to men-
tion only the highlights), some women were reasonably healthy and
happy, had raised families, worked, had friends, and were involved
in community activities.
Other strands in my history began to be woven into the
picture. Though this is meant to be an intellectual account, I cannot
refrain from mentioning one personal element. My parents, now
eighty-nine and eighty-three and well adapted by any standards,
have not had an easy life. My father, son of a poor shtetl family in
czarist Russia, left home at age fifteen for the city. He ran through
the streets, in the midst of a raging pogrom, to call the doctor to at-
tend my sister's birth. Illegal crossing of the border in 1921. Cross-
ing the Atlantic in steerage three times because he and my mother
did not have the right papers. Without much of a formal education,
unskilled except as a Hebrew teacher, he raised a family with the
"large" income, particularly during the 1930s, that came from a
New York hand laundry. Mother, it is true, came from a middle-
class family. But she, too, lived through all of the above and, as a
woman, faced even more stressors, perhaps, than those known by
a man.
Working on the second edition of Poverty and Health, I be-
gan groping toward the question that occurs to one when examining
lives such as those of my parents: Whence the strength? Despite the
fact that the poor are screwed at every step of the way, as I have
put it, they are not all sick and dying. And going even further back
to the 1950s, when I taught courses on "The Negroes in the United

States," I recalled asking the same question about the sources of
had previously become rather uninterested in
their strength. If I
asking people about stressors, surely doing so was superfluous for
concentration camp survivors, the poor, Afro-American slaves, or

free Negroes. (For an extremely penetrating analysis of the question
Whence Gutman, 1976. Gutman points out that
the strength? see
American Negro historiography has been dominated by the question
of what slavery and subsequent conditions did to the Negro and has
seldom asked how the Negro shaped a viable culture in adapting to
the environment.) The important question, the fundamental ques-
and philosophical terms, became:
tion in scientific, humanitarian,
How do some of these people manage to stay reasonably healthy? I
was beginning to be freed from the pathogenic orientation. The an-
swer, I thought, could be found by exploring generalized resistance
resources.
I am nearing the close of my story, but two major steps re-
main to be related. Since 1973, I have been heavily involved in the
creation of a new medical school in Israel oriented to training com-
munity primary-care physicians. I liked the idea of conceptualizing
my role in this venture, together with that of my colleagues, as
training a doctor who would be a generalized resistance resource.
While my energies were devoted largely to being a medical
school teacher, my research colleague in
Jerusalem, liana Shoham,
was carrying out a field study that we had designed. Our dependent
variable was the breakdown continuum. Our independent variables
were an extensive series of what we thought were measures of gen-
eralized resistance resources. In the spring of 1977, shortly before I
was to leave on sabbatical, we what is technically called a
received
smallest-space analysis —
(Guttman, 1974) a computer printout of
the structure of relationships among a large series of variables. Lo
and behold True, many
! of the resistance resources were related to
breakdown. But one resource, as we had called it, not only was more
highly correlated with our measure of overall health status, but also
seemed to be the intervening variable between the other resources
and health. For days we puzzled over the nature and meaning of
this variable, and then it clicked. This was not just another general-
ized resistance resource. It was a way of looking at the world. After
considering a number of alternatives,we came to call it the sense of
coherence.
One final step remained to be taken before I could begin
writing this book. What was to be done with stressors? As long as I
was dealing with concentration camp survivors or poor people, I

could make a reasonable assumption. But what of the rest of us?
Many of my colleagues in stress research are engaged in major
efforts in the conceptualization and measurement of stressors. Per-
haps because of my continuing interest in history since my under-
graduate days or my experience as a Jew and as an Israeli or my
personal peculiar combination of pessimism and optimism, it strikes
me forcefully that the human condition is stressful. The question
then becomes not how some concentration camp survivors or some
poor people manage to stay healthy, but how any of us manage to
stay healthy —
the question of salutogenesis. I was now ready to
write the book.
The central concerns of this book are, I trust, clear by now.

But before we enter the detailed terrain of each chapter, in which
the overall argument is inevitably obscured, let me try to provide an
aerial photograph.
Chapter One spells out why salutogenesis rather than patho-
genesis is the great intriguing mystery and important human con-
cern in the field of health. The thesis is not theoretical. The hard
data indicate that at any one time, at the least one third and quite
possibly a majority of the population of any modern industrial
society is characterized by some morbid, pathological condition, by
any reasonable definition of the term. The reader who is already
persuaded that illness is not at all deviant or who is willing to accept
the thesis that salutogenesis is indeed at least a major issue might be

content with skiniming the inevitably boring statistics of the chapter.
In contrast to Chapter One, which conservatively accepts
the traditional medical-model dichotomy of health and illness,
Chapter Two explores the dependent variable. After reviewing the

dichotomous model that is used in most health research, clinical and
epidemiological, and a number of continuum models, I discuss the
health ease / dis-ease continuum, which I see as the most appro-
priate tool for the study of salutogenesis.
Having we turn to an exploration of the deter-
set the stage,
minants of someone's (or some group's) location near the ease end
of the continuum. Or, to put it dynamically, what explains move-
ment toward the ease end of the continuum? Chapter Three con-
fronts the hypothesis that the absence of, or a low level of, stressors,
subjectively or objectively defined, provides the answer. This is not

at all a straw man. The hypothesis lies at the basis of an entire
school, in theory, in research, and in much of contemporary every-
day life, of stress reduction. The evidence is overwhelming, I argue,
that the hypothesis must be rejected. Stressors, I contend, are omni-

present in human existence. Further, it is important to distinguish
between tension, the response of the organism to stressors, and stress,
the state of the organism in response to the failure to manage ten-
sion well and to overcome stressors. Tension may even be salutary;
stress is related to dis-ease.
If it is conceded that the answer to the salutogenic question

lies management, then the question becomes:
in successful tension
What determines such success? Chapter Four suggests a tentative
answer. It is based on a systematic analysis of generalized resistance
resources, ranging from the physical and biochemical to the macro-
sociocultural. But the chapter leaves one with a nagging disquietude
because of the extensive array of resources. Is there a culling rule,
a theoretical basis for expecting that a given phenomenon can in-
deed serve as a generalized resistance resource?
Chapter Five, which answers this question, is the heart of
the book. What is common to all such resources is that they facili-
tate making which we are
sense out of the coundess stimuli with
constantly bombarded. Such repeated experience generates a strong
sense of coherence. This central concept of the book is defined as a
global orientation that expresses the extent to which one has a
pervasive, enduring though dynamic feeling of confidence that one's
internal and external environments are predictable and thai there
is a high probability that things will work out as well as can rea-
sonably be expected. The bulk of the chapter is devoted to a sys-

tematic exploration of the childrearing, social-structural, and cul-
tural sources of the generalized resistance resources that foster a
strong sense of coherence. This exploration provides a link to a sub-
stantial body of work in the field that has not often been related.
Finally, I review the cultural bias of locus-of-control theory, point-
ing up how the sense of coherence differs radically from "I am in
control."
Chapter Six is devoted to a consideration of the available
evidence for the relationship between the sense of coherence and the
health ease/dis-ease continuum. Since research has overwhelmingly
been conducted with a pathogenic orientation and with a given
disease as the dependent variable, there is a built-in limitation on
the persuasibility of the argument. I hope it is sufnciendy cogent to
act as a stimulus for research.
The issue confronted in Chapter Seven is that of the overall
structure of the salutogenic model, which is more complex than the
line of argument may have suggested. Stressors are related to health,
as are genetic and constitutional predispositions and weak links.
Generalized resistance resources have a complex relationship to each

other and to the sense of coherence. Health, in turn, is not simply a
dependent variable; it is linked to well-being in other areas of life.
Chapter Eight is not necessarily an integral part of the salu-

togenic model. But because I am a teacher of medical students and
have a deep concern for the health care of individuals and commu-
nities and because I am firmly convinced that there are implications
for health care in the model, I thought it important to add this
chapter. In it, I seek to spell out some of these implications, both for
the relationship between the doctor (or health team) and the
patient and for the organization of health care services.
Finally, in a brief Epilogue, venturing into a philosophical
vein, I record the expectations that flow from the salutogenic model
for the health ease of human beings in the foreseeable future. In
doing compare and contrast my views with the optimistic ones
so, I
of LewisThomas, the on-balance optimistic views of Thomas Mc-

Keown, the on-balance pessimistic views of Rene" Dubos, and the
pessimism of Ivan Illich.
Chapter One
Studying Health
Instead of Disease
The problem of salutogenesis is one of the most mysterious, intri-

guing, and meaningful challenges for philosophy and the biological
and the social sciences. Pathogenesis —the origins of disease X, dis-
ease Y, disease Z —has preoccupied us (to the extent that we have
focused on origins and not only on diagnosis and therapy). Even
immunology, the science perhaps most closely linked to salutary,
homeostasis-maintaining and homeostasis-restoring processes, has
posed its questions largely in terms palatable to pathogenesis: What
will prevent this or that disease? I hope it will become clear in due
course that my concern no mere semantic quibble and that here,
is
as in all of science, how one poses the question is crucial to the

direction one takes in looking for the answers.
I had considered using the term orthogenesis, the origins of
being straight or upright The term, however, has been preempted
12
— —
Studying Health Instead of Disease 13
by others and generally refers to treatment aimed at correcting

mental and nervous defects in children. Eugenics, too, which con-
cerns improvement of the human race by mating, is not an apt
term for the concept I propose to develop. Since my concern is not
primarily with issues of repairing or straightening out (though a
theory of salutogenesis certainly bears important implications for
orthogenics) nor with improvement through mating, I have opted
for a neologism salutogenesis.
To ask why someone gets viral pneumonia or hypertensive
heart disease or a fractured rib or schizophrenia or any one of the
1,040 major morbid conditions listed in the International Statistical
Classification of Diseases, Injuries, and Causes of Death is to ask
a and clinically
difficult scientific important question. There is no
end and subsequently applied research needed to eluci-
to the basic
date the complex hundreds of answers. Commitment to such activ-
ity, however, tends to lead to what Dubos (I960)' has argued is a
commitment to an illusion. To think that conquering one disease

after another brings us that much closer to conquering disease in
general has double-edged consequences. On the one hand, it pro-
vides the motivation, courage, and resources for societies and people
to devote themselves to the struggle of understanding and coping
with real problems and real human suffering. On the other hand, it
dulls our sensitivity to and concern with some painfully hard
morbidity data in the second half of the twentieth century. More-
over, it inevitably diverts attention and resources from the real
mystery and from what I shall contend is the most promising direc-
tion to take in attempting to make our life better.
Let me put it bluntly. Given the ubiquity of pathogens
microbiological, chemical, physical, psychological, social, and cul-
tural —it seems to me self-evident that everyone should succumb to
this bombardment and constandy be dying. Dubos (1965, p. 35)
has highlighted the solution to part of the problem as "the control
of the disease states caused by microbial agents which are ubiqui-
tous in our communities in the form of dormant infections." Before
the public health triumphs in controlling sanitation, food, and
water; before the generally improved standards of living, particu-
larly of nutrition; before the discoveries of medical microbiology
and pharmacology, acute and semiacute infectious processes were
caused largely by microorganisms acquired through exposure to an

exogenous source of infection. These pathogens have hardly dis-
appeared, even from the Western world, but they can largely be
controlled or, at least, their consequences can be contained. There
are problems of drug-resistant strains of pathogenic agents and
social problems of application of existing knowledge and techniques.
But the major infectious diseases in the Western world are now re-
lated to the omnipresent, endogenous microorganisms that make

the event of infection much less important than the smoldering
infectious process.
There are, then, the descendants of the exogenous "bugs"
that have been of such great significance in human history (Burnet,
1953). There are the endogenous bugs, whose virulence may be
held in abeyance but whose threat is constant. There are, third,
those agents, be they viruses, mutant cells, pollutants, or agents of
physical trauma such as guns, knives, and motor vehicles, that pose
a constant threat of damage, reparable or irreparable, immediately
evident or slowly unfolding, to the human organism. And, finally,
there are those bugs variously called psychosocial stressors, presses,

strains: alienation, rapid social change, identity crises, ends-means
gaps, discrimination, anxiety, frustration. I shall, in Chapter Three,
discuss the issue of psychosocial stressors in some detail. But for the
time being, and in full awareness of the complexity, softness, specu-
lativeness, and even problems of elementary definition in viewing
psychosocial stressors as pathogens, I would maintain that there is
sufficient ground for including them in this litany of bugs.
In sum, it seems to me that elementary knowledge and ob-

servation of the world around us must inevitably raise the question
of how anyone ever stays alive. The human capacity for conscious-
ness, for taking in this miserably hostile world —
its glories and
—
wonders notwithstanding must certainly lead us to ponder the
ultimate mystery of salutogenesis. Surely, the reader must respond,
I have exaggerated. Even casual observation suggests that most of
us, most of the time, are not on our deathbeds, are not in the hos-
pital, and are more or less healthy. The world around us cannot be
as virulent as I have suggested ; or at least we have learned to live in

some balance with the world so that we do not succumb that often
to the bugs. What are the facts?
I shall, in Chapter Two, deal in some detail with the ques-

tion of what sickness, disease, ill health, or health is. I shall there
argue that it makes the most sense on a com-
to picture the answers
plex, multidimensional continuum. I see little utility in a dichoto-
mous classification of sick-well, pathological-salutological. But for
the present, I shall consider data that deal with conditions that
would be considered pathological or morbid in the orthodox medi-
cal model.
Morbidity Hypothesis
My contention can be put in a straightforward manner. At

any one time, at least one third and quite possibly a majority of the
population of any modern industrial society is characterized by
some morbid condition, by any reasonable definition of the term.
Or, to put it another way, deviance, clinically or epidemiologically
defined, is "normal." That is, significant departures from the clinical

picture of health are, statistically, far from unusual.
One would think that, having formulated this hypothesis on
the basis of the theoretical consideration that bugs are ubiquitous,
one could easily test it. All one need do is turn to any textbook of
epidemiology and find a distribution of the population as having
no morbid condition; one morbid condition, with subclassification;
two or more morbid conditions, with subclassification. A search of
six standard texts revealed no such table or reference to the issue. I
then examined the vast repository of data put out by the U.S.
National Center for Health Statistics on the basis of the National
Health Survey. Again, my efforts were largely in vain. In sum, I
found no single source that could provide sufficiently detailed evi-
dence to test the hypothesis. Oral inquiries to epidemiologists con-
firmed this lack of data; moreover, they not only noted the methodo-
logical difficulties in obtaining such a set of data but were usually
surprised that the question was raised (see Morris, 1957, p. 11 ).
That this is the case should not have surprised me, for it
bears out the point I wish to make. Our dominant ideological

paradigm, which shapes our society's clinical practice and scientific
research, focuses on and responds to a particular disease or clinical
entity. Our health care system, or, as Winkelstein (1972) has co-
gently suggested, our disease care system, focuses on individual dis-
eases.But even those such as Winkelstein who urge more attention

to health care still set problems in terms of pathogenesis: how to
prevent lung cancer, motor vehicle accidents, or poliomyelitis.
Hence data are most often collected on cases of specific disease
rather than on the location of people on a health-disease continuum
or even in an overall category of morbidity. We somehow assume
that most people are healthy because we focus on individual dis-
eases. Even so revolutionary a thinker as Dubos, who perhaps more
than anyone else first set me on the path I now tread in teaching us
of the mirage of health, thinks largely of specific diseases. Only
when we begin to pose the problem of salutogenesis will we begin
to fully implement a full-scale search for those factors that promote
health rather than cause specific diseases.
Morbidity Data
The problem of salutogenesis, however, emerges in its full

significance only when we look at the data. It is to these that I now
turn. It is not my purpose, nor would it have been possible, to arrive
would adequately answer the ques-
at a definitive set of data that
tion: At any one time, how many people are (or what part of the
population is) characterized by explicit, significant morbid condi-
tions? I have set as my limited goal to submit the contention pre-
sented above to a rough but fair test of empirical data. I have tried
to erron the conservative side. I leave it to the reader, given the
data and their sources, to judge the extent to which the contention
is supported.
Since this book was written in the United States, the data
most easily available to me were those that refer to that country.
My purpose not being epidemiological, I have not considered it im-
portant to present relevant data from other Western countries. They
have been reviewed and unhappily show a similar picture. For the
most recent published survey available to me, see the work of
Bridges- Webb (1973, 1974) on a small Australian community. It
seems quite safe to say that data on nonindustrialized societies would
show a far bleaker picture.
Let us start from a 1961 statement by White, Williams, and
Greenberg: "Data from medical care studies in the United States

and Great a population of 1,000 adults
Britain suggest that in
(sixteen years of age and over) in an average month 750 will ex-
perience an episode of illness and 250 of these will consult a physi-
cian." No doubt many of the 250 who do come to the doctor are
not what, by medical consensus at least, would be called ill; there is
equally no doubt that many among the 500 who experience an
illness episode but do not go to the doctor are not what, again by
medical consensus, we would call ill. But we must add that there
may well be some among the 250 without an episode of illness who
can conservatively be called ill. (We are not here concerned with
the extremely important issue of going to the doctor an issue that —
these data raise.)These data led Zola (1966, p. 616) to conclude
that "the empirical reality may be that illness, defined as the pres-
ence of clinically serious symptoms, is the statistical norm."
Chronic Illness. To the best of my knowledge,
only one pub-
lished study comes close to adequately documenting the requisite
data, and we would do best to start from it, even though it is limited
to one urban community. Moreover, the Baltimore study of the
Commission on Chronic Illness (1957) focuses, as the name indi-
cates, on one element (chronic illness) of the picture, albeit the
most important one. But before considering the Baltimore study, we

should keep in mind one fact noted by the authors of that study.
Neither it nor other studies based on fieldwork include institutional-
ized persons. In 1970, some 1,670,100 persons were in long-term-
care institutions for clearly nonhealthy people, a prevalence rate of
828/100,000. This figure includes those in mental and tuberculosis
hospitals, nursing homes, and homes and schools for the mentally
and physically handicapped ( National Center for Health Statistics,
1976) (For data on the health status of nursing home residents, see
.
National Center for Health Statistics, 1977b.) This figure excludes

all persons in correctional institutions, detention homes, homes for
—
neglected children and unwed mothers settings that are likely to
contain a higher proportion of sick people (by age group) than do
noninstitutional settings. Moreover, the datashow a marked decline
in numbers compared with the data for the previous decade, largely
because of the substantial decline of people in mental institutions.
It is not my purpose here to go into details in a critical re-

view of the Baltimore study or any of the other data presented since
it is totally out of the question to do anything more than obtain
rough approximations of the proportions of the population who are

ill. Suffice it to say that the Baltimore data are based on a repre-
sentative citywide sample of 11,574 persons. The Baltimore of the

mid 1950s differed in one major respect from some other large cities
in its large (27 percent) non white population, a fact that would
tend to be reflected in somewhat more chronic illness than might be
found in other cities. Keeping this in mind, we turn to the data.
The prevalence rate of chronic diseases in the city of Balti-
more in 1954 was 156,650/100,000 persons, or almost 1.6 diseases
per person (p. 50). Almost two thirds (64.9 percent) of the popu-
lation had a chronic condition, a proportion that varied from al-
most 30 percent of those under fifteen to 95 percent of those over
sixty-five. (See pp. 393-399 of the study for a definition of chronic
tions
—"thoseWhen
condition.) reference
which
is made to substantial chronic condi-
interfered with or limited the patient's activi-
ties or were likely to do so in the future, or which required or were
likely to require care" — the proportion of the population so char-
acterized fell to 44.4 percent: 19 percent with one substantial con-
dition, 13 percent with two, and 12 percent with three or more
(pp. 55-56). A classification of the conditions as mild, moderate,
or severe in the present stage showed that only one eighth of the
conditions were severe (as indicated by an increasing degree of dis-
abling effects, a more advanced state of the disease, a greater likeli-
hood of fatality, a need for more care, complications, or increasing
pain). But this qualification would seem to be unduly sanguine in
light of the detailed classification found in Table 1.
A
second (albeit secondary and not recent) source of data
on chronic diseases in the United States is Blum and Keranen's
authoritative review (1966). Following the definition of chronic
disease formulated by the authors of the Baltimore study, they esti-
mate that 42 percent of the population suffers from one or more
chronic conditions. Blum and Keranen most often do not give rates.
Where they do, they are by and large comparable to the Baltimore
findings. Thus, for example, they estimate that about 2 percent of
the population has diabetes (p. 145) compared with Baltimore's
2.67. Similarly, 2 percent are estimated to have no functional hear-
ing (p. 30), the same proportion found for deafness and impaired
hearing in the Baltimore study. Where Blum and Keranen differ,
they tend to have somewhat higher estimates than the rates found in
Baltimore. Thus "asthma affects about 2J4 percent of the U.S.
population" (p. 35) compared with Baltimore's 1.24 percent.
Blum and Keranen "an estimated two million
also speak of
survivors of stroke" (p. 265) ; anemia in "4 to 15 percent of the
adult population, depending upon the hemoglobin level denned as
abnormal" (p. 71) and "hyperuricemia is estimated to affect ap-
;
proximately 3 percent of the U.S. population" (p. 108).

Finally, Blum and Keranen classify as chronic a number of
conditions that are not covered in the Baltimore study and that are
of widespread significance. Among these are diseases of the bron-
chopulmonary system, such as chronic bronchitis, "estimated to
affectbetween 15 and 20 percent of persons over forty in both the
United States and England" (p. 37), and pulmonary emphysema,
which affects "about 12 percent of adult males" (p. 40) ; and alco-
holism, found among "nearly 5 percent of adults" (p. 153).
A more recent and comprehensive review of the data is
found in a publication of an authoritative, respectable body (Na-

tional Health Education Committee, 1976). The data must, how-
ever, be taken with some caution since they are assembled mainly
from organizations in the medical-research and medical-service
establishment, whose vested interest is reflected in a possible over-
estimate of prevalence rates. This issue is dealt with later in this
chapter, but it is of sufficient import to be noted at this point. I
should add that the estimates presented do not differ greatly from
those from other sources.
As seen in which the data on the main causes of
Table 2, in
disability in the is a total of about
United States are presented, there
108 million cases of these morbid conditions. This is about 52 per-
cent of the total population. There is of course overlap. Individuals
who have multiple conditions are counted as two or possibly more
cases. Thus, the estimated 30.33 million cases of the four major
circulatory diseases included in the table (high blood pressure,
coronary heart disease, rheumatic heart disease, and strokes) are
found among 27.13 million persons, an overlap that is perhaps sur-
prisingly low. There is no way of estimating how many individuals,
in all, are afflicted by these 108 million conditions.
Included in this list are 6 million mentally retarded, 11.5
Table 1. Prevalence of Chronic Diseases, Baltimore, 1954

Table 1. Prevalence of Chronic Diseases, Baltimore, 1954 (cont'd)

Table 2. Main Causes of Disability, United States, Early 1970s

.
This is, however, surely an underestimation. According to the esti-
mate of the American Cancer Society (see Table 2), "more than
one million people were under medical care for cancer in 1975." In
1974, a total of 655,000 new cases were reported, excluding carci-
noma in situ of the uterine cervix and superficial skin cancers, which
are unlikely to be reported. Given the fact that malignant neoplasms
are the primary cause of about 18 percent of all deaths and that,
roughly, the overall five-year cancer survival rate is about 40 per-
cent today, I have calculated that the prevalence rate of cancer is
likely to be in the neighborhood of 800/100,000. This estimate is
consistent with the incidence data given in the second and third
National Cancer Study reports (Dorn and Cutler, 1959, p. 13;
Cutler and Young, 1975, pp. 307-342) and is lower than the rate
given in Levin and others ( 1974, p. 4)
Estimates of the prevalence of selected chronic circulatory
conditions in the United States in 1972 were arrived at on the basis
of a national sample of 44,000 households in the Health Interview
Survey (National Center for Health Statistics, 1974). Rates per
100,000 persons range from 6,010 for hypertensive disease, 5,040
for heart conditions, 4,770 for hemorrhoids, and 3,680 for varicose
veins to 750 for cerebrovascular disease, 440 for congenital anom-
alies of the circulatory system, and 160 for phlebitis and thrombo-
phlebitis. On the basis of a number of methodological studies com-
paring interview reports with examination results, the report

concludes — in accord with all studies on the subject — that interview
data provide considerable underestimation of the prevalence of
chronic conditions (see Feldman, 1960). The extent, of course,
varies considerably from condition to condition. Bearing in mind
this underestimation in interview data, we still find over forty-four
million chronic circulatory conditions estimated. Again, one must
take note of duplication. The report estimates that these conditions
were found among 36,492,000 persons, or 17.7 percent of the U.S.
population. Of these, over 26,000,000 were under age sixty-five.
Obviously, conditions with a wide range of severity and impact are
included here. But even a conservative approach cannot dismiss the
existence of varicose veins as a morbid condition even though in
only 3.9 percent of the cases were varicose veins reported as having
caused some limitation of activity during the year.
A
somewhat earlier report in the Health Interview Survey
indicates the prevalence of fifteen selected chronic respiratory con-

ditions in 1970 (National Center for Health Statistics, 1973b).
Again, the prevalence rates vary greatly: from 10,300/100,000 for
chronic sinusitis and 5,420 for hay fever to only 60 for pneumo-
coniosis. Also, chronic bronchitis (3,270) and emphysema (660)
suggest far different prognoses than does a deflected nasal septum
(400). Nonetheless, if we project these figures to the total U.S.
civilian, noninstitutionalized population, we can estimate a total of
53,754,000 conditions distributed among 46,884,000 people, of
whom almost 42,000,000 are under sixty-five.
The third in the series of National Center for Health Sta-
tistics (1973a) studies of chronic morbidity that merits reference
here reports on chronic digestive conditions. The ten condition
groups reviewed in this 1968 survey (for example, ulcer of the
stomach and duodenum, a rate of 1,720; gall bladder condition,
1,030; chronic enteritis and ulcerative colitis, 930) are projected to
almost twenty-one million conditions in the country, affecting a total
of seventeen million people, of whom 4,610,000 are sixty-five and
over.
Finally, we may take note of a 1971 survey of the prevalence
of impairments that have merited listing in the International Classi-
fication of Diseases. The impairments reviewed range from hearing
impairments, which have a rate of 7,160 (of which only 4 percent
cause limitation of activity), to complete or partial paralysis, with a
rate of 690 (with 62 percent of the conditions causing limitation of
activity). In all, some 47.4 million conditions included here are
found in the United States, affecting 37.3 million persons, of whom
9.8 percent are sixty-five and over (National Center for Health
Statistics, 1975).
Two major chronic conditions have largely been ignored.
The disease known as dental caries is, as Blum and Keranen put it
(1966, p. 76), universal. "Very few people," they add, "escape

some degree of gingivitis or periodontal disease." Whether it is be-
cause historically medicine and dentistry have gone their separate
ways or because dental disease is not a cause of mortality among
humans (in contrast to animals) or perhaps because of their near
universality, such problems are not often referred to in discussions of
chronic diseases. Yet if one's criterion is suffering, then surely this
—
must be one of the major areas demanding attention, and it cer-

tainly cannot be neglected in this review of the data.
The second area, that of mental retardation and mental ill-
ness, has been mentioned. My own preference is to exclude mental

retardation from a discussion of chronic disease. Although 3 per-
cent of the population scores below 70 on nationally standardized
intelligence tests, the overwhelming majority of such persons are, I
am persuaded, not to be included within the scope of a conserva-
tively oriented discussion of disease problems. No more than 100 per
100,000 persons —a far cry from the 2,883 rate cited in Table 2
are the legitimate concern of the disease care institution. I do not,
of course, in any way mean that mental retardation is not a major
problem of social concern. I am saying that it is not, in the main, a
medical problem. Many persons classified as mentally retarded,
appropriately or not, indeed come to the attention of the disease
care institution, but this is a totally different issue from the one that
concerns us at present.
Mental illness, however, cannot be excluded from this dis-
cussion. I am persuaded that Szasz and and the label-

his colleagues
ing theorists in medical sociology have made a major contribution
in compelling us to confront the meaning of mental illness (Szasz,
1963; Becker, 1963; Goffman, 1961; Scheff, 1966; see Gove, 1970,
for a critique) But I cannot accept their position that mental illness
.
is not a legitimate concern of the health care or disease care insti-
tution. It would take us too far afield to consider the various argu-
ments and to support my position here. Suffice it to say that I am
fully aware of and indebted to Szasz and the others and seek to err
on the side of caution in considering the data.
The data suggest that few indeed are the adults who have
always been free of significant emotional symptoms. But Susser
notes, "The fact that the presence of symptoms is the norm in a
population does not necessarily rule out their pathological signifi-
cance" (1968, pp. 208-214). Being somewhat less inclusive, we

take note of the Midtown Manhattan study, which found that 23
percent of the community sample suffered mental impairment from
symptoms classified as marked, severe, or incapacitating ( Srole and
others, 1962). Susser, judiciously reviewing the evidence available,
suggests — as I read him —a true prevalence rate of severe psychiat-
)
ric illness ranging from 1,000 to 3,000 per 100,000 adults below
age sixty, the rate being somewhat higher among older persons.
Finally, in this review of chronic-disease data, we turn to the
category of allergy. We do best by quoting from a popular but
highly reliable source (Roueche, 1978, p. 63) "Allergy differs from
:
most other diseases in that its victims (including even most asth-
matics) seldom die and almost as seldom recover. . . . Allergy is,
... in fact, the most common of all chronic complaints through-

out the industrially developed world. . . . The allergic population
of the United States is on the way to forty million. Almost fifteen
million of these are hay fever sufferers. Nine million are asthmatic,
and another several million are allergic to some food or drug or
drink. The rest . . . are victims of allergic eczematous contact
dermatitis."
Notifiable Diseases. We now turn to a different categoriza-
tion of diseases, which cuts across the above-considered chronic dis-
eases and the soon-to-be-considered acute conditions — that is, speci-
fied notifiable diseases that are under the egis of the Center for
Disease Control. In 1977, a total of eighteen cases of plague were
reported in the United States, four of cholera, and none of smallpox
or yellow fever. But unlike these diseases, which are covered by
international quarantine agreement and which seem, for the time
being, to have come close to having been eradicated in the United
States, there are other communicablewhose incidence is not
diseases
negligible. For the present purposes, these would not include such
diseases as anthrax, botulism, brucellosis, diphtheria, leprosy, lepto-
spirosis, malaria, meningococcal infections, poliomyelitis, psittacosis,
rabies in man, congenital rubella syndrome, tetanus, trichinosis, tula-
remia, typhoid fever, typhus fever, and venereal diseases other than
syphilis and gonorrhea. (I have quite intentionally included this
long list of diseases, none of which has yet been removed from the
list of notifiable diseases, as a reminder that secular trends in disease
prevalence are not necessarily, as liberal ideology would have it,
unidirectional.
Those have annual incidence rates of
notifiable diseases that
10 or higher per 100,000 are contained in Table 3. The rates for
all but gonorrhea seem to be quite modest. Clearly, compared with
the chronic diseases, these morbid conditions are of relatively minor

Table 3. Reported Cases of Specified Notifiable Diseases,

United States, 1977
. —
Health Survey, our major source of data, defines an acute morbid-

ity condition as a reported "departure from a state of physical or
mental well-being . . . which has lasted less than three months and
which has involved either medical attention or restricted activity"
(National Center for Health Statistics, 1978, p. 57) ; chronic con-
ditions are thus excluded. The data are classified in five condition
groups, following the International Classification of Diseases: infec-
tive and parasitic diseases, respiratory conditions, digestive-system
conditions, injuries, and all other acute conditions (ear disorders,
headaches, genitourinary disorders, skin disorders, and so forth)
In the year July 1975-June 1976, every man, woman, and
child in the United States had more than two episodes, on the average,
of an acute illness or injury requiring medical attention or restriction
of usual daily activity (an annual incidence of 214,300/100,000).
The on the incidence and duration of these acute con-
detailed data
ditions are contained in Table 4.
Clearly, acute conditions vary greatly in seriousness and per-
sonal or social implications. On the one hand, the list includes the
common cold (accounting for 20.7 percent of all acute conditions)
and sprains and strains (3.4 percent) on the other hand, it includes
;
pneumonia (0.8 percent) and genitourinary disorders (2.6 per-

cent). Bearing in mind that the evidence points to substantial un-
derreporting of acute conditions as well, even in a two-week recall
period as is the case here, we can clearly see that the 450 million
estimated episodes of acute conditions in one year form part of the
record. Using the data contained in Table 4, I have calculated that
on any given day, almost 2.6 percent of the population had an acute
illness or injury (average number of restricted-activity days multi-
plied by 365).
Disability and Limitation. Having raised the issue of the im-
pact of illness or injury on disability or activity limitation, I will
pursue the issue further in order to look at our basic question
How many people are ill at any given time? —from another point of
view. Again using the data of the National Health Survey (Na-
tional Center for Health Statistics, 1977a), we find that in 1976 the
average American had 18.2 restricted-activity days —days on which
he or she substantially reduced normal activity for the whole day
because of an illness, either chronic or acute, or an injury. This fig-
i
CQ
CO *
OS 8
5 *.-s
|
3
•2
S3
a
c
U
s
I
i—
a
(4
ure does not include persons who have permanently reduced their
usual activities because of a chronic condition; the data refer only
to those who had a reduced level of activity within the two-week
period prior to the interview compared with the level of activity be-
fore that period. The number of restricted-activity days averages

around 1 1 per year per person under age twenty-five, 15.5 for per-
sons twenty-five to forty-four, 25.4 for persons forty-five to sixty-
four, and 40 for persons sixty-five years and over.
The same source provides another way of looking at the
direct, reported consequences of chronic conditions in the data on
limitation of activity. Setting up criteria appropriate to children,
housekeepers, workers, and others, the Health Interview Survey

found that on the day of the interview, in 1976, 14.3 percent of the
population had some limitation of usual activity due to a chronic
condition. It is true that this situation is particularly characteristic
of elderly persons (45.4 percent of those sixty-five and over), but it
is also true of 24.3 percent of those forty-five to sixty-four years old.

Moreover, if only one's major activity is considered, 10.8 percent of
the total population reported a limitation. This proportion rises to
19.1 percent of those aged forty-five to sixty-four.
In considering these figures, one should keep in mind that all
these data refer to the civilian, noninstitutionalized population. Ear-

lier, we some 0.8 percent of the popu-
referred to the likelihood that
lation are, at one time, in long-term At
institutions for sick people.
this point it is appropriate to add that the surveys also do not in-
clude people who, at the time of the survey, were in short-stay

hospitals. During 1976 there were an estimated 14,100 hospital dis-
charges per 100,000 persons. The average length of stay was 7.9
days. Or, to put it another way, about 10.6 percent of the popula-
tion was hospitalized at least once during the year preceding the
interview in the Health Interview Survey. This figure does not in-
clude persons who died in the hospital, but it does include hospital
deliveries (National Center for Health Statistics, 1977a).
Overall Morbidity. In closing my review of the data, I refer
to two studies of morbidity without detailed consideration of specific
pathologies. Chapter Two deals with this problem conceptually; for
the time being, I refer to the issue only to test the hypothesis posed
initially in this chapter. In 1965 a probability sample of 6,928 adult
residents of Alameda County, California, completed questionnaires

that used four central measures: ability to perform certain basic
daily activities; presence during the previous year of one or more
chronic conditions or impairments; presence during the previous year
of one or more symptoms; and subjective rating of general energy
level. On the basis of the detailed responses to this set of questions,
respondents were placed on a seven-category ordinal scale, ranging
from severe disability to high energy level (Belloc, Breslow, and
Hochstim, 1971).
Of the total sample, representing a community of some one
million adults, 29 percent were classified in the no-complaint cate-
gories (23 percent low to medium energy, 6 percent high energy).
At the other extreme, 7 percent reported a severe disability, an ad-
ditional 8 percent a lesser disability, 9 percent two or more chronic
conditions, 19 percent one chronic condition, and 28 percent at
least one symptom. There were, of course, sharp differences by age.
If a cutting point of at least one chronic condition is used (that is,
including the four poorer categories and excluding symptoms), we

find 43 percent of the population included, ranging from about one
fifth of the youngest age group, through close to half of those aged
forty-five to fifty-four, to over three fourths of those sixty-five and
over.What should be stressed is that the no-complaint-high-energy

group remained almost constant at 6 or 7 percent in all age groups
through ages sixty-five to seventy-four and then dropped to 4
percent.
In contrast to the first study, whose major goal was to ob-
tain an epidemiological picture of the health status of a community,
the study reported by Collen ( 1977) presents data assembled in the
context of multiphasic screening in a prepaid health care plan orga-
nization. Data are given on 6,285 adults who had not seen a physi-
cian within one year and who asked for a multiphasic checkup.
Thus, on the one hand, these persons evidently felt well enough to
go without seeing a doctor for a year; on the other hand, they had
some motivation for having the checkup. We have no way of know-
ing how representative the group was of a total community popula-
tion; certainly there is no reason to think that the error was on the
side of being unduly ill. (In a personal communication, Collen in-
formed me that Kaiser data show that persons who come for multi-
32 Health, Stress, and Oping
phasics have been found to be healthier than a control group of

members who do not come.) Recall, too, that they were members
of a prepaid health care plan. Those classified as well (no signifi-
cant medical complaint or problem and no clinically significant
finding or abnormality) constituted 56.8 percent; worried- well
(complaint but no finding), 11.6 percent; asymptomatic-sick (no
complaint but clinically significant finding), 3.9 percent; and sick
(both complaint and finding), 27.7 percent. Thus sick care, within
this organization, was needed for 32 percent.

Methodological Considerations. Before we consider the sig-
nificance of the data for the central concern of this book, it is
appropriate to make a number of methodological comments. First,

I would reiterate the issue of overlap, which has been mentioned in
passing a number of times. Individuals who contribute to the nu-

merator in one rate may also contribute to the numerator in another
rate. For example, our point of departure, the Baltimore study,
reported an average of 1.6 chronic conditions per person. Yet, as
noted with regard to the discussion of cardiovascular diseases, this
overlap is not always as great as one might think. Unfortunately, the

pathogenic orientation, which focuses on specific diseases, has pre-
vented any possibility of a firm conclusion on the question of over-
lap. It may well be that many older people have multiple conditions.
But, as the data on disability and activity limitation as well as the
Health Interview Survey reports indicate, those under sixty-five do
not present a picture of blooming health.
A second issue to be noted is conceptual confusion — in part
inevitable —of risk factors, pathology, and consequences. I have

tried to limit this confusion without forfeiting data. Thus, I men-
tioned alcoholism and obesity but not smoking. Only cultural defi-
nitions legitimate referring to alcoholism and obesity as morbid or
pathological. In all likelihood, smoking is more implicated in mor-
tality than is either alcoholism or obesity, yet somehow our medical
culture regards alcoholism and obesity as pathological but sees smok-
ing only as a risk factor. In the same vein, note should be taken that
I make no mention of accidents, suicide, or homicide (respectively,
the third, ninth, and tenth causes of death in the United States in
1973, considering all cardiovascular diseases jointly; see National
Health Education Committee, 1976, p. 289). Accidents and vio-
—
lence are causes of morbid conditions, however, and hence presum-

ably are reflected in the data on pathology.
The third issue that affects the reliability of the rates of
morbidity reported here is that of the "iceberg" phenomenon so well
known to epidemiologists —that is, the gap between the "true" prev-
alence^ — if onlywe were able to count every truly existent case,
whatever that may mean with regard to a given morbid condition
and the reported or counted prevalence. To what extent this factor
has led to the data presented being underestimates cannot be known.
It is appropriate that the issue be mentioned; but there is, I believe,
no need to consider it in evaluating the extent to which the hypoth-

esis of the chapter is supported by the data.
No doubt, however, vested interests have, to some extent,
overestimated the prevalence of pathology, a fourth consideration.
Whatever the safeguards built into a clinical prevalence study of
chronic illness, such as the Baltimore study, my hunch is that errors
tend to occur on the side of overestimation. Doctors, after all, have
been trained to find diseases. But I am here referring to a far more
serious phenomenon: estimates from organizations that have a

vested interest in reporting high prevalence of that disease with
which they are concerned while at thesame time showing that
things are getting better thanks to the work of that organization.
Two examples may be given of such partisan use of data,
both taken from The Killers and the Cripplers (National Health
Education Committee, 1976). In answering the question "How has
medical research paid off against tuberculosis?" the report notes that
"between 1945 and 1973, the tuberculosis death rate has declined
95 percent —due in major part to the medical-research discovery of
the effectiveness of streptomycin in 1945 and isoniazid in 1952"
(p. 255). Knowingly or not, the authors disregard the fact that
tuberculosis mortality has been declining for well over a century. In
England and Wales, for example, only 14 percent of the decline in
respiratory tuberculosis mortality between the mid nineteenth cen-
tury and 1971 took place after the introduction of specific effective
drugs (McKeown, 1976, p. 52). A dramatic-looking histogram on
p. 246 of The Killers and the Cripplers shows an 89 percent decline
in tuberculosis death rates since 1952. If, however, we superimpose
the histogram on a curve showing the decline since 1900 in the
United States, we have a good illustration of how to lie with statis-

tics while telling the truth. In 1900 the tuberculosis death rate was
200/100,000; by 1952 it had declined to 15.8/100,000, and by
1973 to 1.8/100,000. "Despite these epidemiological findings,"
Winkelstein (1972, p. 72) concludes, "vast and elaborate commu-
nity-wide case-finding programs were extensively promoted after
World War United States at great cost and probably with
II in the
very little on the established trend of the disease."
effect
The second example that suggests the skepticism that must

be exhibited toward prevalence data presented by those with vested
interests is found in the estimate (National Health Education Com-
mittee, 1976, p. 186) that about 250,000 persons are suffering from
multiple sclerosis; this figure gives a prevalence rate of about 120/
100,000 persons in the United States. The cited source is one of the
classic texts in the field. It so happens that I have engaged in some
epidemiological work on multiple and was surprised by this
sclerosis
rate. My suspicion led me (McAlpine, Lums-

to check the source
den, and Acheson, 1972). Neither the first two chapters, devoted
to epidemiology, nor the index revealed any basis for the claim. The
highest prevalence rates cited for sizable populations are 60-70/
100,000 (p. 13 ). A list of some forty surveys reveals only three pop-
ulations with rates of over 80 (p. 6 Rochester, Minn. Orkneys and
: ;
Shetland; and southeast Norway). I am not suggesting intentional

deception. I am pointing out that a vested interest shapes one's per-
ception of the facts.
The fifth methodological issue to be considered is that of a
seeming paradox. In a 1973 National Health Interview Survey,
48.7 percent of the population sample reported (or the spokes-
person of the family reported) that they were in excellent health,
38.4 percent in good health, 9.4 percent in fair health, and 2.8 per-
cent in poor health (see National Center for Health Statistics, 1976,
p. 243 ) Even
. if we take into account the fact that the sample repre-
sented only the noninstitutionalizcd part of the population, the sub-
jective evaluation of 87.1 percent that they were in at least good
health seems to stand in sharp contradiction to the mass of data
presented in this chapter. A similar report was issued over television
by the Department of Health, Education and Welfare only a few
days before these lines were written. The explanation is fairly simple.
Researchers generally agree that posing such broadside questions

about satisfaction (about one's family, work, health) almost always
elicits a widespread positive response. I am not at all saying that
subjective perception of one's state of health is unimportant. As a
social scientist, Iwould be the last to do so. My point is methodo-
logical. Research experience shows that the more an interviewer
establishes rapport and goes into detailed questioning, the more the
interviewer discovers elements of discontent, concern, and difficulty.
Thus at one level people report that they are quite healthy; at an-
other level they reveal substantial ailments. Both sets of data have
meaning, and there is no contradiction between them.
Finally, I must take note of a possible explanation of the
data that is not warranted. The claim might be made that aggregate
data, such as those presented here, obscure the fact that the highest
rates of morbidity are found among the elderly, the poor, blacks,
and other minorities. No doubt these groups indeed have much

higher rates, quite across the board, than does the rest of the popu-
lation. My own work has contributed to substantiating this gen-
eralization. Nonetheless, careful examination of the data shows that
higher rates for these groups do not come even close to explaining
the rates of morbidity for the entire population. The salutogenic
question is hardly less pertinent to white, middle-class, and non-
elderly people than it is to the more discriminated-against.
Significance of Salutogenesis
I trust that I have provided some measure of evidence in

support of my contention that even in the 1970s and even in the
most industrialized societies, with, relative to earlier eras, high levels
of standard of living, environmental control, and medical tech-
nology, a far greater proportion of the population is ill, by any con-
one might think to be the case. I antici-

servative definition, than
pated that would be true on the basis of the observation that the
this
bugs are ubiquitous and, one might add, most sophisticated. It

should, then, be clear that the mystery of health is indeed intriguing.
But why is it socially and humanly important?
Let me, at the outset of my attempt to support my position
that the question Why do people stay healthy? is today the crucial
question, make it crystal dear that I do not contend that it should

replace the older question (not that I think there is any such
danger). My point is to make a plea for the allocation of a small
fraction of the resources — of the funds, of the manpower, and, most
important of all, of the thinking of those concerned with health and
illness —now overwhelmingly, if not exclusively, devoted to the study
of pathogenesis.
If, then, we can begin to understand this mystery —the mys-
tery of survival, the mystery ofwhy some people's health is such that
they go through some of the time with relatively little
life for
—
pain and suffering we might begin to think about applying this
understanding to reduce pain and suffering among the rest of us.
There is, of course, no promise that it will be within our power to
do so. Conceivably, the only answer to how to stay relatively healthy
is not to live —
which is no answer. But if we do not attempt to even
—
pose the question if we remain forever in the sphere of chasing
and learning to kill this or that bug or of being sophisticated and
seeking general weapons that kill a number of bugs simultaneously
or of avoiding "unnatural" behavior —we limit ourselves to the con-
finesposed by the study of pathogenesis; and, as glorious as the
achievements of that study have been in the past, we thereby handi-
cap ourselves unnecessarily.
I would point to three specific reasons why the focus on
pathogenesis is likely to handicap us in grappling with both the
scientific problem of understanding why from deviant
illness is far
and the human problem of somewhat reducing pain and suffering.

In each case, the salutogenic approach offers a viable alternative.
First, the pathogenic approach pressures us to focus on the
disease, on the illness, on the alteration of body fluids or structures,

and to disregard the sickness, in Eric Cassell's terms (1976). That
is, it blinds us to the subjective interpretation of the state of affairs
of the person who is ill. Some medical scientists and clinicians man-
age to overcome this narrow vision. Thus Cassell points to general
characteristics of sickness — loss of connectedness and of feelings of
omnipotence, omniscience, and control — as consequences of illness
and entry into the patient role (pp. 40-44). His humanity and sen-
sitivity bring him to the realization that these feelings also cause
suffering, which the physician should try to alleviate. But even if
emphasis on the disease state does not blind one to the person who
has the disease, the pathogenic model binds one to an etiologic
focus on the disease — that is, one limits oneself to asking, What has
caused this specific disease? Even Cassell fails to ask whether the
characteristics he discusses may not be of etiologic significance for
disease in general. Salutogenesis, by contrast, opens up, or even com-
pels us to examine, everything of import about people who are ill,
including their subjective interpretations of their state of health.

Second, thinking in pathogenic terms is most comfortable
with the "magic-bullet" approach —one disease, one cure —which
explains the resistance of many to the concept of multiple causation.
But even those who do not view matters simplistically find that, by
concentrating on a single disease, they can seldom account for much
of the variance in its development in either populations or indi-
viduals. They are, in a way, worse off than those who search for the
necessary even if not sufficient cause, for eliminating necessary causes
does do away with a specific disease. But in either case the assump-
tion is that we are cleverer than the bugs and can eradicate them
one by one. Rid of this illusion, we can only become pessimists. But
using a salutogenic model, asking questions about what maintains
health, allows us to go beyond "not smoking" or "vaccinating
against polio" or "eliminating smog."
Third, pathogenesis by definition is a model that postulates
a state of disease that is qualitatively and dichotomously different

from a The individual is sick or well. The organ
state of nondisease.
is diseased or nondiseased. The condition is pathological or non-
pathological. If the data presented in this chapter mean anything,

they suggest that such dichotomization blinds us to a conceptualiza-
tion made by a salutogenic model, namely, a multidimen-
possible
sional health-illness continuum between two poles that are useful
only as heuristic devices and are never found in reality: absolute
health and absolute illness. Pathogenesis asks, Why does this person
enter this particular state of pathology? (or the epidemiological
equivalent of this question) . Salutogenesis asks, What are the factors
pushing this person toward this end or toward that end of the
continuum?
Chapter Two
Measuring Health
on a Continuum
Kuhn (1962) has taught us that all scientific research and the
social institutions in which it is conducted and applied are charac-
by a given pervasive paradigm. The dominant paradigm of
terized
Western medicine, the pathogenic orientation, has indeed produced
great triumphs. Western industrialized societies have reached the
most advanced stage in history in their understanding of and capac-
ity to cope successfully with at least the physical suffering to which
1
people are heirs. Chapter One, framed within the constraints of
1 This statement is not sarcastic I am fully aware of the ills of mod-
ern societies, but the evidence points clearly to the conclusion that, by any
reasonable criteria for health, Western industrialized societies have achieved
the highest levels of any sizable population in history. Of course, if one
wishes to prove the contrary and drags in criteria such as alienation, ugliness,
and mass deceit to show how sick modern societies are, no discussion is
possible.
I am also aware of the relatively recent "discovery" of the state of
physical and emotional grace, almost, of the true natural person, as it were,
38
Measuring Health on a Continuum 39
the pathogenic paradigm, nonetheless demonstrated that even in

such a substantial proportion of the population manifests
societies
disease. This paradox, I suggest, cannot be explained unless we re-
phrase the central question and ask, Given the ubiquity of bugs,
why does anyone ever stay alive and reasonably healthy? Or, to put
it more formally, we cannot explain the paradox unless we adopt a
salutogenic paradigm.
At the core of the pathogenic paradigm, in theory and in
action, is a dichotomous classification of persons as being diseased
or healthy. Our linguistic apparatus, our common sense thinking,
and our daily behavior reflect this dichotomy. It is also the con-
ceptual basis for the work of health care and disease care profes-
sionals and institutions in Western societies. Consideration of the
problem of the origins of health, however, leads us to face the ques-
tion of whether the dichotomous approach is adequate or whether
it may not be imperative to formulate a different conceptualization
of health.
This chapter, then, is devoted to an analysis of the com-
pelling link between the pathogenic paradigm and the health-disease
dichotomy and to the formulation of a continuum model compati-
ble with and appropriate to a salutogenic paradigm. Let us look
first at what goes on in the appropriately named disease care system.
Clinical Model
At any one time, people are placed, as a first approximation,

in one of two categories by use of a simple, observable, and objective
in hunter-gatherer societies, living in harmony with the surroundings and

biologically well adapted to them and Eyer and Sterling,
(see Powles, 1973,
1977, who present this thesis, with a good bibliography). At present this claim
derives from an ideological commitment and is hardly a hypothesis that is
testable. There is certainly general agreement that my text statement is
accurate with respect to the comparison of industrial and agricultural societies
and that the first agricultural revolution some ten thousand years ago initiated
a chain of increased food supply, brought about the aggregation of popula-
tions of substantial numbers,and saw the beginning of the predominance of
infectious diseases (see McKeown, 1976, pp. 66-67). But whether Homo
sapiens, in the hundreds of thousands of years before this, was healthy in a
near-idyllic manner seems most doubtful. Until we have convincing evidence
to the contrary, I suggest that malnutrition, violence, and accident made life
quite unpleasant and short, even if not as much so as during the agricul-
tural era.
criterion : If they have not appeared in the formal, institutionalized
disease care system — if they have not become patients —they are
well. If they have so appeared, they are But having someone ill
ill.
is distressful for personnel in the disease care system since they have
never been trained to deal with ill people but rather with illnesses.
The well-trained physician looks for signs that suggest a disturbance

of the organs or body fluids characterized by structural alteration or
biochemical change. Where data are adequate, the patient can be
classified inone or multiple categories by use of the conventional
rubrics of pathology, whose ultimate development is found in the
International Classification of Diseases. The physician can then, one
hopes, engage in treatment.
There are two residual categories. First, there is the person
who shows no "objective" signs of pathology. Such a person used to
be called a malingerer and was given short shrift; this is still in
good part the case in those social situations, such as an army, where
being classified as ill often pays off. Today, more frequendy, people
in this category are presumably treated more kindly; they are told
that the problem is emotional or psychological. The second residual
category contains those people who, having entered the disease care
system and undergone a diagnostic scrutiny, have not provided ade-
quate data for subclassification, though they clearly have something
wrong. It should also be noted that the diagnosed-patient category
is subdivided into those for whom nothing therapeutic can be done
and those who can in some way, it is to be hoped, be helped.
Subclassification also goes on in the world outside the disease
care system. Given the social costs of illness, there is great pressure
in industrialized societies against entering the disease care system.
(There are, of course, counter-pressures, and the pressures against
entry are differentially applied. Analysis of this issue, however,
would take us too far afield. ) But there is some legitimate scope for
"sick" people who are not patients. It is acceptable to feel bad gen-
erally or to enter a diagnostic category defined by oneself or another
layperson without entering the disease care system. In such a case,
the categories may bear little relationship to those in the Interna-
tional Classification of Diseases, and the therapies considered appro-
priate may be shaped by the fads of the Madison Avenue repre-
sentatives of drug manufacturers or by the sociocultural group of the
:
sick person. One finds, though, that laypersons more than profes-
sionals are tolerant of and comfortable with feeling bad in general,
without lay or professional diagnostic categorization, if feeling bad
is limited to relatively short, nonrepetitive time periods.
In sum, the dominant conceptual model of laypersons and
practitioners alike is one that classifies people at any given point in
time in the following way
A. A nonpatient
1. healthy
2. sick
a. feeling bad in general

b. sick with a particular diagnosis of a particular "disease"
as defined by oneself or another layperson
B. A patient
1. diseased
a. with signs sufficiently clear to allow subclassification as
(1) a case of disease X
(2 ) a case of disease Y and/ or
(n) a case of disease n
b. with insufficient signs to allow specific disease subclassifi-
cation
2. not diseased : malingerer, crock, hypochondriac, or emotion-
ally disturbed person.
The pathogenic paradigm leads to and is reinforced by this system

of classification. It constantly presses toward the dichotomous loca-
tion of people in Category Al—healthy nonpatients—or in Category
Bla —a case of disease X, Y, and/or n in treatment.
For the purpose of facilitating the work of the ambulatory-
care physician, this conceptualization of the health-illness phenom-
enon seems appropriate and useful. It is also an effective tool in the
hands of the practitioner in the acute-care hospital, whose interest,
concern, and responsibility are self- and socially defined as facilitat-
ing the removal of the patient from the hospital as rapidly as possi-
ble, consistent with the perceived best interests of the patient.(Un-
less, of course, the reimbursement system or a syndrome fascinating
to the medical staff tempts one to keep the patient in treatment.)
But, more important, this conceptualization is useful for persons
suffering from a constellation of painful and anxiety-arousing sensa-

tions, running the gamut from very mild to extreme, who have
come to believe that the disease care institution can be of help to

them.
To put the matter another way, on any given day, some-
where between 1 and 2 percent of the population of Western in-
dustrialized society (or any segment of it that has at least some
services available) will be in contact with the disease care system.
(The calculation is based on three components: an average of four
to five ambulatory visits to the doctor per person per year, or a
monthly consulting rate of 250/1,000; about 10 percent of a popu-
lation being admitted to an acute-care hospital for an average stay
of eight days in the course of a year; and about 0.8 percent of a
population being in a chronic-disease institution at any one time.
See Thacker, Green, and Salber, 1977.) Given the present level of
knowledge, understanding, and technology, we have good reason to
believe that the magic-bullet, diagnose-subclassify-treat approach, a
concomitant of the dichotomous model, is of considerable value to
those who enter the disease care system. It works, in many cases, in
decreasing suffering, in avoiding complications, and, occasionally,
in solving problems. And lest someone sneer at the "considerable
... in many cases . . . occasionally" of 1 to 2 percent, let me
remind the reader that we are talking about 1.5 million people a
day who hurt (in the United States). The biologically oriented re-
search establishment, which employs the same dichotomous model,
and the clinical establishment, then, to my mind, are not to be dis-
missed. Nor, I might add, do I see the slightest evidence that any
sizable segment of the population of any Western society wishes to
do so. And I do not share the arrogance of some who see backward-
ness or false consciousness or brainwashing in their behavior.

I think that a great many serious criticisms can be leveled at
the disease care systems as they exist in the Western world. But to
go into this issue would require a different kind of book —which
may more important than this book, but it is not the one I
well be
am writing. Importantand fundamental deficiencies might well be
remedied, changes and modifications might well be introduced that
are not in conflict with the dichotomous model of health and illness
we have been discussing. This model is compatible in many ways

with solo practice, with Kaiser-Permanente, Group Health Insur-
ance, the Health Insurance Plan, and other forms of group practice.
compatible with the quasi-socialized medicine found in Israel
It is
and with the socialized medicine of Great Britain and the Soviet
Union.
I am here concerned with the unfortunate consequences of
the dichotomous model —consequences that will obtain as long as
this model continues to hold sole sway in the training and practice
of physicians and other personnel. I touch on this issue only briefly
in this chapter, as a way of introducing the necessity of at least an
additional, if not an alternative, paradigm. In Chapter Eight, de-
voted to an analysis of the health care or disease care institution in
the light of the fundamental thesis of this book, I contrast the im-
plications for this institution of alternative paradigms. There, both
ideological and social-structural issues are considered in full. For the
present, I would but take note briefly of the inherent implications of
the dichotomous model.
Eric Cassell (1976), among others, has movingly called at-
tention to the general psychological needs that patients bring into
every encounter with the doctor. He does not preach, urging physi-
cians to be moral and humane, but cogently argues that the "heal-
er's art" of considering such needs is an essential part of the diagnos-
tic and therapeutic procedure. Patients simply get better more
thoroughly and more quickly or suffer somewhat less when the curer
also seeks to be a healer. (For a more profound and systematic
analysis of the distinction between illness and disease, see Kleinman,
Eisenberg, and Good, 1978.) My colleague Shuval has raised the
same issue from a sociological point of view, inquiring into the
latent, "nonmedical" functions of the medical institution (Shuval,
Antonovsky, and Davies, 1970).
Cassell traces the failure to meet psychological needs to "those
factors in the history of medicine that, in artificially separating the
person from the disease, have directed our awareness away from the
nexus of the problem" and to "the failure of both physicians and
society to realize that medicine is inherently a moral profession"
(p. 119). But Cassell does not see that the trained incapacity to
focus on anything but the specific disease at hand is intimately
linked with what he himself has thoroughly internalized the : dichot-

omous model. Throughout, Cassell urges physicians to see patients
as persons, by which he essentially means that physicians should be
concerned about the emotional as well as the organic aspects of dis-
ease. But if one conceptualizes all sickness problems as disease prob-
lems (as qualitatively different from a state of health), no matter
how broadly one views the disease problem, I am skeptical about
how one can come to see the person. This is all the more true when
the technology at one's disposal provides a fascinating set of toys.
The medical on inter-
student, trained only in hospitals with a focus
esting diseases and armed with the dichotomous model, will in-
evitably continue to be concerned with the particular, organic
disease.
Another consequence of the dichotomous model is the failure
to ask why the patient came into the disease care system. The task
of ascertaining the appropriate diagnostic category and providing
appropriate therapy is either quite simple and routine and uninter-
esting or complex, mysterious, troubling. In either case, the patient's
history as a person is hardly seen as relevant. When diagnosis is
simple, the patient's history is superfluous; when diagnosis is com-

plex, there is too much else to do; the medical history in such a case
is hardly adequate, to put it mildly.
A regular feature of the New England Journal of Medicine,
"Case Records of the Massachusetts General Hospital," demon-
problem most depressingly. In the issue of October 13,
strates this
1977, Case 41-1977 is about a sixty-eight-year-old man. Poor?

Rich? White? Black? Married? Widowed? We do learn the details
of his rich disease history. After his most recent hospital spell of
forty-three days, he was home for four weeks before his present ad-
mission. "At home he was lethargic and anorexic." But as to what

might have gone on in whatever is called home that might help to
explain his present condition and be relevant to prognosis, not a
word from any of the learned, outstanding clinicians.
If the model we have been discussing, which divides people
into patients/ diseased and nonpatients/ healthy categories, is related

to failure to see the patient as a person even in the social situation
of present interaction with the physician and if it is related to failure
to see the past of the person as relevant to that person's current and
future health, then we should certainly not be surprised that it leads

to the disease care institution's totally excluding nondiseased persons
from consideration and denying responsibility for them. Fortunately
or not, as we have seen in Chapter One, such exclusion, over a
relatively short period of time, affects only a minority of the popula-
tion (when there are no barriers to access). That this is the case
may well be linked precisely to the dichotomous model — to the fail-
ure to allow, even in one's thinking much less institutionally, any
perceived relationship between the person as a healthy nonpatient
and the same person as a diseased patient. To put my thesis here
within the terms of reference of this book, unless the problem of
salutogenesis is confronted, confronting the problems of pathogenesis
is likely to be a Sisyphean task.
Public Health Model
Thus, the disease care institution, organized around the prob-

lem of pathology, is most comfortable with the individual patient
who has come down with a diagnosable disease for which effective
therapy is available. What may with some legitimacy be called, by
contrast, the health care institution of Western societies has a
radically different cast. First and probably foremost, it is concerned
with groups —whether geographically, politically, or socially de-
fined —rather than with individuals. Second, it is concerned with
control of the environment, its aim being to prevent the outbreak of
diseases.
Epidemiology is one of the major scientific disciplines that
have developed in the service of the health care institution. There

isno doubt in my mind that the epidemiological conceptualization
of the health-illness phenomenon, the model or paradigm used by
epidemiologists, is powerful and, for some purposes, far more power-
ful than the clinical model we have been discussing.
First, epidemiologists are aware of the iceberg phenomenon.
They assume, with adequate evidence, that for every case of a dis-
ease that has been brought to clinical attention — that has come
within the scope of the disease care system —there are additional
cases below the surface — cases not known to the "authorities."
But, second, if there is thus an inherent temptation for epi-
demiologists to find more cases, they are kept in check by methodo-

logical sophistication and compulsiveness, which are probably the
central norms of the profession. The epidemiologist is tough minded,
unable to accept a case unless clear criteria are rationally specified
and measurable. Clinicians are kept in line by conscience and by

whatever system of peer review they may be subject to,which pre-
sumably prevent mistaken diagnoses; but intuition, art, and clinical
skills are necessarily acceptable in arriving at a conclusion. After all,
one must act The epidemiologist has the luxury of rejecting such
subjectivism.
Third, the sine qua non of the epidemiologist's professional
activity is to go beyond description and enter the field of analysis, to
deal with causation. As such, it rounds out, complements, the field
of laboratory and clinical research. But its core and strength are its
understanding of causation as based on the study of group rather

than individual differences.
The epidemiological, or public health, model, then, is quite
different from the clinical model. It is concerned with "real" num-
bers; it is concerned with the distributions of cases in groups; and it
is concerned with causation.
Much as I acknowledged considerable value in the concep-
tual model used by clinicians, I would certainly grant value (of both
a pragmatic and a scientific nature) to the public health model.
Rational planning of care-delivery systems -is dependent on the in-
formation that flows from work based on this model. Further, the
model opens up the possibility, hardly available to the clinical
model, of allocating resources to prevention and early detection of
disease as supplements or alternatives to diagnosis and therapy.
(For one of the most sophisticated attempts to cope with the prob-
lems that arise for a self-maintaining health care plan (Kaiser-
Permanente) —a plan that has an commitment
explicit conceptual
to a modified clinical model that allows for prevention and early

—
detection see Garfield, 1970, and Garfield and others, 1976. The
fourfold classification is described in Chapter One.) Scientifically,
epidemiological study, reinforced by the concept of multiple causa-
tion, opens the way for gaining understanding of disease causation.
Having, then, given what may sound like even higher marks
to the public health model than those I assigned to the clinical
—
model, must take note of the common element of the two models
I
the element that I maintain is a crucial limitation Both models are :
based on a dichotomous classification of all people as healthy or

diseased. The clinician wants to diagnose and cure the patient. The
epidemiologist wants to prevent people from coming down with a
disease or, at the least, wants to identify the disease early so that
action might be taken to prevent degeneration. But the shared
focus is always on the disease. The epidemiologist, perhaps precisely
because of the compulsion for scientific classification, insists on the
dichotomy as much as does the clinician.
Why the strength and persistence of this dichotomous ap-
proach? The source, I suggest, for both clinician and epidemiolo-
gist, is the fundamental commitment to the question of patho-
genesis. As long as one asks why people get sick, one must inevitably
begin to talk not about disease but about diseases or a disease. Once
again, let me make it crystal clear that in putting matters this way
I am not in the least belittling the work of clinicians and epidemiol-
ogists. The question of pathogenesis, the dichotomous model of dis-
ease and health that it requires, and the research and action that are
shaped by the question and the model are, as I have tried to spell
out, important. But they — probably other phenomena charac-

as all
terizing human existence —have inherent

their theycontradictions;
not only limit inquiry but militate against even seeing that there are
boundaries. What lies beyond these boundaries I now explore.
Continuum Model
do not claim that one can arrive at a continuum model

I
only by posing the problem of salutogenesis. But, unless one is

brought to the "right" answers by the "right" questions and unless
one finds that these answers give one power, for theoretical or ap-
plied purposes, such answers may soon be relinquished as fads. Let
us briefly consider a number of examples of formulations of the
continuum model of health and disease. I trust that it is clear that
I have chosen these only as examples and certainly do not mean to
imply that Cochrane (1972b), Susser ( 1974), and Fanshel (1972)
are the only ones who have developed continuum models.
Cochrane (p. 89) reviewed "the evolution of the ideas of my
colleagues and myself in relation to the measurement of ill health."

As a medical student he had been "taught that there was a simple
dichotomy between ill and healthy people." But after collecting a
great many biological measurements from the community, he was
struck by what seemed to be the extensive distributions of "abnor-
malities." These could not be clearly separated into the clinician's
two groups of sick and well; in fact, they showed no bimodal dis-
tribution whatsoever. Nor did he find the notion of normality and
standard deviations of any use whatsoever in view of the funda-
mental question he posed What should a clinician do? Unlike most
:
who have asked this question, Cochrane, who did not allow himself
to remain oblivious to the iceberg aspects of the data he had col-
lected, did not reject the continuum model. Instead, he hit upon the
solution of concentrating on "finding the point (s) on the distribu-
tion curve where treatment begins to do more good than harm."
This, then, is one step forward. Cochrane and his colleagues
have rejected the ill-healthy dichotomy and have said Let us look :
at a wide variety of radiological, hematological, electrographical,

and biochemical indices; on the basis of current knowledge, which is
always changing, let us act where we have good evidence to show
that our actions do more good than harm. They seem to have sub-
stituted these indices for the more traditional disease categories.
Thus, instead of saying that the population is divided into those
who have diabetes and those who do not,- they would say that the
population has a given distribution curve of blood-sugar levels and
that, given present knowledge, X seems to be the cutoff point where
treatment will do more good than harm. This for blood-sugar levels,
hemoglobin, intraocular pressure, and so forth. Cochrane grants

that the great weakness of this approach is "its limitation to uni-
variate analysis at present" (p. 92) — that is, to a total failure to
confront the problem of overall health and illness. The rejection of
the dichotomous model inculcated by medical schools and of the
disease-categorization bind marks what may be a useful guide to
clinical practice; it does not, however, open up the way to new
theoretical questions.
Quite a different conceptualization of health and illness,
legitimately identified as acontinuum model, is advanced by Sus-
ser (1974) in his analysis of the place of value systems and social
. —
structure in the "definition of health." The health care or disease

care system, Susser argues (p. 541 always deals in some way with
three levels of health: the organic
—"organic and
),
physiologic dis-
order best described as disease in process) impairment
static and
(if
persistent)"; the functional

—"a or as
subjective state of psy-
(if
chologic awareness of dysfunction best described as illness (if in

process) or as disability (if static and persisting)";and the social
"a state of social dysfunction, a social role assumed by the individ-
ual, best described as sickness (if in process) or as handicap (if
."
static and persisting)
Susser points out, as I read him, that all three levels of health
involve conceptions of normality. Normality, however, can be con-
ceived of in three senses. "In a first sense of pathology, normality is
generally perceived as dichotomous; the disease is either present or

absent." In the statistical sense, "normality for a specified condition
is defined from its modal distribution in a population" ; such condi-
tions include mental deficiency (see Mercer, 1972, for an excellent

discussion of this issue), hypertension, and obesity. In the social
sense, normality is defined by values, by notions of how things ought
to be. In actuality, Susser seems to relate his discussion of nor-
mality to the organic level, but I see no reason to limit it in this way.
The focus of Susser's paper is on how values and social structure
shape the attention paid to each of the three levels of health and ill-
ness, the choice of the appropriate concepts of normality (and, sub-

sequently, the breadth of social phenomena that are subsumed in the
workings of the health care or disease care institution)
I have introduced this summary of Susser's paper because I
think it presents one significant way of arriving at a continuum
model of health and illness, even though Susser does not explicitly
commit himself to this model. Both the statistical and social senses of
normality, it seems to me, compel adoption of a continuum model.
But Susser's paper makes a further contribution in that he takes us
beyond the sole concern with the organic. "The functions of the
health professions relate to each of these levels of organization of
health states" (p. 542), he points out, and it behooves us to consider
consciously how this relationship works.
Susser, then, has provided an expanded model in two senses.
First, he views it as essential that health and illness be conceived of
in broader terms than those of the unidimensional organic-patho-

logical model. Second, at least implicitly and at least with respect to
the functionaland social levels, he suggests that people are ranked
on a continuum. Also, he says that "values and social structure ac-
count for much of the lack of correspondence between the existence
of organic disease, of illness, and of the sick role" (p. 541 ), though
he does not attempt to develop a systematic account of the relation-
ship among the three levels. These contributions notwithstanding,
we are left with a model that is still shaped by the pathogenic orien-
tation, with its constant pull toward dichotomy and its concern with
the disease, the illness, and the sick role, or the impairment, the dis-
ability,and the handicap.

A good example of a third approach to the definition of
health using a continuum is found in a paper by Fanshel (1972).*
Explicidy committed to an operational definition useful "in making
decisions affecting the allocation of resources and the kind of re-
search required in the health services" (p. 319), Fanshel built his
model around one variable: functioning. "A person is well if he is
able to carry on his usual daily activities. To the extent that he can-
not, he is in a state of dysfunction, or deviation from well-being"
(p. 319). Fanshel has, then, committed himself to the third of

Susser's levels. Fanshel recognizes that functioning is a matter of the
social definition of appropriate activities for specific individuals.
Application of the index would require subdivision of the target
population into cohort subsets, with specification of the socially ex-
pected role behaviors for each cohort.
The index itself consists of a series of states and describes
"the extent to which anyone in a specified state is able to carry on
his usual daily activities" (p. 319). Eleven ranked states are briefly
2 This is only one of the continuum models developed

in recent years
that, because their concern is operational, employ mathematical took in
constructing health indices. The Clearinghouse on Health Indices regularly
reports both publications and research projects that use continuum models.
Since my purpose here is not to review the field, I have limited myself to
considering Fanshel's paper. It is of some interest, incidentally, that Fanshel,
a professor of electrical engineering, relies on the approach developed by
one of the leading theoretical sociologists (Parsons, 1951, chap. 10).
described, ranging from well-being through dissatisfaction to isola-

tion, coma, and death. It would not be in place here to discuss the
details of this scale; the subsequent issues raised by Fanshel relating
to prognosis, population, and duration for given states; or the at-
tempted mathematical application to the specific problem of invest-
ment of health resources in a program for control of venereal
diseases.
There is little doubt that this paper is a contribution toward

the purpose for which it is intended : to develop an operational tool
making in the allocation of health
to facilitate rational decision
resources. One may be somewhat skeptical about the extent to which
the instrument is indeed valid and reliable. One may reject the
proposal that the social value judgments admittedly necessary to
apply the instrument be made by "those responsible for the . . .
[health-delivery] services, the secretaries of state and their admin-

istrators" (p. 324). But for our purpose, that of understanding
the factors that are salutogenic, the school that this paper repre-
sents cannot be of much help. Fanshel writes, "It is important to
note that no statement has been made as to the cause for lack of
well-being" (p. 320). When he develops this point a bit further, as
well as in his later application to venereal diseases, the fundamental
distinction between Fanshel's approach and my own becomes clear.
When he thinks of causes, as well as when he thinks of action, his
reference is still to concrete diseases.
Again, do not wish to be misunderstood and thought to be
I
disparaging. I see the development of this continuum scale and of

others of a similar nature, which focus on the functional conse-
quences of diseases, whatever they may be, as valuable and of major
practical significance. (See, for example, Rutstein and others, 1976,
who write: "There are no easily measured quantitative definitions
of *bad health,' 'average health,' or 'good health.' Our proposed
system overcomes this difficulty by establishing quantitative negative
indexes of health. Cases of unnecessary disease and unnecessary dis-
and unnecessary untimely deaths can be counted" (p. 586).)
ability
Such scales are major attempts to answer the clinical and epi-
demiological questions of pathogenesis. But posing the question this
way cannot help us explain the mystery of how people manage
somehow to keep close to or move close to the healthy end of the

continuum, however defined.
WHO Definition
Before turning to a discussion of the breakdown concept,
which is my tentative proposal for the appropriate conceptualization
of health, I wish to consider what is probably the most famous defi-
nition of health in the second half of the twentieth century, that of

the World Health Organization (WHO), as stated in the preamble
to its charter: "Health is a state of complete physical, mental, and
social well-being and not merely the absence of disease or infirmity."
admit to having passed through three phases of reaction to
I
At first, it seemed to me a most laudable formulation
this definition.
of one of humanity's most ardent wishes. What could be more ap-

propriate than the statement of such a Utopian goal in a world led
jointly by two cultures, differing in many ways but united in a
philosophy of permanent, inevitable progress in all spheres? But not
only the Utopian goal and the accentuation of the positive appealed
to me; even more significant, the sociologist in me thought, was the
explicit realization that one could not isolate the physical from the
mental and social.
This reaction, however, was not long lasting. As a researcher,

as favorably disposed as I was to theoretical conceptualization as a
guide to empirical studies, I soon came to see that so global a defi-
nition wasremoved from an operational definition of

far health.
This second phase of benevolent neutrality toward the WHO defini-
tion as a useless but harmless statement, however, has given way to
my present position, which is one of sharp opposition. As
hope I I
make clear in the following pages, the resemblance between the

focus on positive health and the problem of salutogenesis is quite
superficial. There are two fundamental elements in my opposition.
I am not particularly troubled by specification of Utopian

goals. But when their social function is to keep our minds on, as the
old song has it, "pie in the sky," one must be skeptical. Such goals
then are not just different from, more abstract than, and more gen-
eral than other goals; they are contradictory to and diversionary
from such alternatives. The reader of Chapter One will not be sur-
prised that at this level of generality I much prefer Dubos' definition

of health as "a modus vivendi enabling imperfect men to achieve a
rewarding and not too painful existence while they cope with an
imperfect world" ( 1968, p. 67) Such an approach, while grounded
.
in philosophical pessimism in contrast to the Panglossian optimism

of the WHO
definition, provides the dynamic orientation that is
needed in the perpetual struggle that makes some improvement
possible.
But it is not only on the philosophical level that I find the
WHO definition sorely wanting. I also find it to be the quintessen-
tial expression of medical imperialism, of the assumption that every-
thing in life falls within the jurisdiction of the health care or disease
care system and of those who control that system. To avoid misun-
derstanding, I should stress that I do not question the motivations
of the framers of the definition. In formulating a Utopian goal, they
no doubt had the finest intentions. Moreover, the "mental" and "so-
cial" references no doubt are based on a crucial and most welcome
awareness of the impossibility of isolating biochemical and physio-
logical elements of human existence from other elements. But moti-
vations, intentions, and insights are not adequate. The WHO defini-
tion does not speak of physical (and perhaps even emotional)
well-being as being shaped by or as interacting with social well-
being. It declares flatly that everything people feel about their state
of well-being is part of health and hence within the province of the
health institution. From here, it is but a minuscule jump to saying
that all aspects of a person's well-being are appropriately under the

control of those who control the institution —a formulation that, as
will become quite clear in Chapters Five and Eight, is absolutely
contrary to the thesis of this book. The sense of coherence, to antici-
pate, is hardly strengthened by having the health institution respon-
sible for all aspects of people's lives.
To put the point in other words that, unfortunately, have

not been left as mere words since the WHO
definition became a
sacrosanct phrase: Whatever the powers that be do not like enters
the proper sphere of medicine: political dissent, whatever the social
system, has led to locking people up "for their own good"; and sex
education, family planning and and homosexual-
abortion, divorce
ity, along with underachievers and overachievers, dropouts and
—
jocks and grinds — all these and many more fall within the province
of health with the blessings of WHO. There is no inherent leftist
or rightist, liberal or conservative bias to the WHO approach. The
appropriate distinction is between the powerful, who can use so
benevolent a definition with a clear conscience, and the powerless.
I am thus convinced that the WHO definition of health de-
serves severe criticism. Yet it would be quite unfair to end on this
note without considering one important venture that is based ex-

plicitly on the WHO definition. Breslow (1972, pp. 347-348)
presents his argument as follows: "Our concept and measurement
of health has generally focused .... This focus on
on ill health.
pathology in the measurement of health probably arose from the
fact that for most of human existence the health problem facing
society,and medicine in particular, has been overcoming disease.
... By the mid twentieth century [the health problem] no
. . .
longer consisted solely, or even largely, of being threatened by early

death or specific diseases. . . . The control of previously epidemic
diseases and the fact that the chronic diseases developed insidiously
. . . created an essentially new kind of health problem."
It seemed to Breslow and his colleagues (Belloc, Breslow,
and Hochstim, 1971) in the Human Population Laboratory of
Alameda County, California, that this analysis led to the WHO
definition of health. They set about examining this concept of health
and formulating, in quantifiable terms, a measurement of "health
in the generic sense." Being committed to the "triumvirate nature
of the [WHO] concept —
physical, mental, and social well-being
they thought it useful to seek a single method that would permit
simultaneous assessment of all three components" (Breslow, 1972,
p. 350).
At this point, however, conceptual clarification seemed to
give way which finally led to a seven-point
to methodological effort,
spectrum as a measure of physical health. The table detailing the
scale is headed "based on disability, impairments, chronic condi-
tions, symptoms, and energy level" and consists of: severe disability,
less disability, two or more chronic conditions, one chronic condi-
tion, symptomatic, low to medium energy level without complaints,
and high energy level. Two further scales were devised. The first is
an eight-item index of mental health — for example, feelings of being

depressed or on top of the world. The second, a social-health index,

measures the extent to which the individual is a functioning member
of the community. It incorporates employability, marital satisfac-
tion, sociability, and community involvement.
Clearly, Breslow and his colleagues have gone a long way
toward meeting the objections that the WHO definition of health is
impossibly abstract, philosophically Utopian and misleading, and

static. They have made a most important contribution in develop-
ing and testing a usable, sensible, and scientific tool. Their theoreti-
cal bias —an orientation toward the healthy end of the spectrum — is
one I share. They have, however, failed to overcome two major

weaknesses. First, their index of physical health disregards its multi-
dimensionality and reflects a professional medical bias. Second, they
have blithely fallen into what I believe to be the dangerous trap of
the WHO definition in devising indexes of mental and social health.
I shall, toward the end of this chapter, return to this issue.
Breakdown, or Health Ease/Dis-ease, Continuum
The central thesis implicit throughout this chapter is the con-

tention that as long as we remain in bondage to the question of
pathogenesis, we will fail to see that there are two distinguishable,
fundamental questions. The first seeks to explain why a given person
comes down with a specific disease and another person does not
come down with that disease. On the epidemiological level, the
question is simply expanded to higher and lower group rates of that
disease. An important offshoot of this approach is, for those con-
cerned with action, to find an appropriate immunological or thera-
peutic solution for the specific disease even before it is understood
or explained. For the nth time, I should stress that I do not disparage
the question of pathogenesis or the health care or disease care insti-
tution that functions on its basis.
But only when we turn to the mystery of salutogenesis, when

we become aware that, with all the action and all the research in the
world, so many of us so much of the time have been, are, and will
continue to be sick, can we begin to see that there is a radically dif-
ferent question. One can put this question in a number of ways. My

own preference is to ask, How can it be explained that a given
individual, in this miserable world of ours, has not broken down?

Or, in a group version, how come this group has such a relatively
low proportion of people who have broken down? Obviously, this is
a first approximation of the question. But it is adequate to point up
that the focus of concern is the ease/dis-ease continuum rather
than the health-disease dichotomy.
To ask about a specific disease is to narrow one's search to
specific, disease-relevant factors. To ask about ease and dis-ease is to
ask about generalized factors that are relevant to all diseases. And
to ask about health ease, that what facilitates
is, to seek to explain
our movement toward the most salutary end of the breakdown con-
tinuum, is to search for weapons that may be far more potent in
decreasing human suffering than is any specific disease-preventing or
disease-curing factor. (Theoretically, it is quite possible to derive
hypotheses about generalized factors through the study of individual
diseases. This can happen when one is struck, time and again, by the
fact that the same factor seems to show up in disease after disease.
In practice, such generalization from specifics does not seem to
have happened.) I fully grant that the search for generalized factors
may turn out to be totally chimerical or that if some answer or an-
swers are discovered, there will be no way to apply them. Not to ask
the question, however, would be, I profoundly believe, a sad mistake.
Sources of ideas are mysterious. Possibly my interest in gen-
eralized factors relevant to dis-ease stems from my own childhood
experience that chicken soup was the appropriate preventive, cura-
tive, and rehabilitative solution to all problems. Perhaps this intui-
tive knowledge was reinforced much later by my becoming per-

suaded that tender loving care is a more sophisticated kind of
chicken soup. Whatever the case may be, my serious proposal at this
point is that the search is worthwhile. I do have some faith, as will
become evident, that the tentative answer I propose —the sense of
coherence — is in the right direction. But even if my answer turns out
to be of little help, my commitment to the proposal remains.
I have now set the stage for the explicit discussion of my
formulation of the dependent variable. In my first public presenta-
tion of the concept (Antonovsky, 1972), I used the term break-
down. I then indicated that I would have preferred to use dis-ease.
But we have all been so conditioned that it is well-nigh impossible
word with a soft s or to assign any meaning to it

for us to read the
other than some medical category. The term breakdown seems to
have caught on, and I shall continue to use it, asking the reader to
bear with me and to keep in mind that the fully appropriate term
is the ease/dis-ease continuum.
Having worked closely with Louis Guttman for years, I was
influenced in defining breakdown by his facet theory (Guttman,
1974). This theory enjoins one, in formulating at the conceptual
level a definition that will be subsequently operationalized, to specify
the facets, or dimensions, of the phenomenon one is trying to define.
Facet theory based on the assumption that social phenomena are,
is
by and large, most adequately understood when they are seen as

multidimensional. Obviously, no ironclad rule can specify which
facets of a phenomenon are sufficiently significant to be chosen
for inclusion in a definition. One designs a tool with care and
thought and then subjects it to public test.
With all due sensitivity to dynamics and change, the defini-
tion of breakdown starts out by viewing it as a state or condition.
Further, its referent is the ecosystem of the individual organism.
Once a person is located, at a given point in time, in a given cate-
gory on the breakdown continuum, one can consider going further
to extrapolate in time and in social space. That is, one can ask
about that person's location on the continuum in the past and in the
future; one can get data about other persons in the family, com-
munity, social class, and so forth. But the starting point is to ask,
Where is the person now?
The sources of a faceted definition —the literature (scientific
and otherwise), theoretical considerations derived from other
areas, such as biology and criminology, intuitive hunches and
representations —are varied. It seemed to me that one could most
usefully conceptualize breakdown by selecting four crucial facets:
pain, functional limitation, prognostic implication, and action
implication.
Pain. Pain an extremely complex, litde-understood phe-
is
nomenon, and yet universally known as a "personal, private

it is
sensation of hurt." (See Weisenberg, 1977, for a comprehensive re-

view. The quote is from R. A. Sternbach, cited therein, p. 1009.)
Laypeople and scientists alike may be baffled in trying to define the
concept precisely. None of us, however, from the time we acquire

language, whatever language it may be, have any difficulty in know-
ing what we mean when we say "it hurts." Nor does anyone have
difficulty in answering a question along the "How much are
lines of
you in pain?" Cultures may vary in the extent to which they have
developed fine distinctions in vocabulary, but none have ignored the
phenomenon. It also seems clear that the phenomenon is closely as-
sociated with some health disturbance, with a pathological process.

It is crucial, moreover, to differentiate between pain in its
health-related context and a variety of negative sensations and emo-

tions for which the same words arc often used. We often feel pain,
we say, when a when we lose a tennis match, when

loved one dies,
our candidate fails to get elected, when we are rejected. Yet sad and
sorrowful as such blows may be, it would be falling prey to what I
earlier called medical imperialism to include such suffering within
the sphere of breakdown. My tendency, then, is to limit the use of
the pain concept in this context in the following way. I am con-
vinced that pain an essentially subjective phenomenon and must
is
be measured as such. Whatever the ultimate achievements in neuro-

physiological research, the emotional-motivational aspects of the
pain phenomenon are crucial at least for behavioral, response out-
come. This conviction also leads to the conclusion that one must
allow the person whose breakdown status one is measuring to de-
cide whether the pain is related to health or not. I would, then, ask:
"Is there any state or condition of your health, general or specific,
that you feel is painful? Do you have no painful condition, mild
pain, moderate pain, severe pain?"
I am fully aware of the wide cultural and individual (and
even mtraindividual-situational) variation that characterizes re-
sponses to such a question when objective observation (that is,
clinical examination) would lead one to expect relative homoge-

neity. This variation poses no difficulty; rather, it points up the
importance of defining breakdown in multifaceted terms because
such variation leads one to ask questions that might not be asked
when one is limited to a single-faceted definition. Concretely, en-
countering a person who reports severe pain but no functional limi-
tation or "medical" condition (see below), someone with a multi-
faceted definition is stimulated to ask questions about the "deviant"

such as : What explains the "inconsistency"? Will the person's health
behavior differ from that of others in the same categories of the
other facets who report no pain? What is the likely breakdown
prognosis (that is, can we find typical paths of movement up and
down the breakdown profiles)? I believe these are extremely im-
portant clinical and research questions.
Functional Limitation. It is quite flattering to a sociologist
to see how nonsociologists have become enthusiasts of role theory.
The extent to which functional limitation, which more often than
not is used primarily in the sense of role performance, has been seen
as a core component of sickness is considerable. There is no doubt
in my mind that this is not only a valid insight but that it is uni-
versally the case. That is, when people in any given culture talk
about illness, they invariably imply a limited capacity to engage in
appropriate role performance. This limitation may be temporary or
enduring, partial or total, but when one is socially defined as sick,
one is excused from fulfilling the normative functional expectations.
Thus Fanshel (1972) explicitly built a health index on the basis
of the extent to which one is able to carry on one's usual activities.
Breslow (1972) also stresses functional limitation as a major basis
for a physical-health index, while Susser (1974) uses functioning
as one of his levels of health.
There are, I suggest, three major difficulties with this formu-
and probably foremost, insofar
lation of functional limitation. First
as the implications for many, many persons go, there is an inherent
tendency to see the functional limitation and the person as coexten-
sive. This connection has been most perceptively analyzed by Scott
in his study of people who have visual limitations of varying extents
(1969, particularly pp. 22-23). He points out that we have in our

society gone far beyond the fact that "the absence of vision prevents
a person from relating directly to his distant physical environment"
and have created an all-encompassing social category, the blind. Oc-
cupants of this category are then socialized to adhere to the special,
all-encompassing norms appropriate to a member of the blind. We
have similarly categorized the "functional limitation" that ranges
from perfect pitch to total tone deafness. "Mental retardates" too
are related to in our society, by and large, as are the blind rather
than as people with varying limitations on their performance of

specific activities.
The second weakness often apparent in the use of functional
limitation with respect to health status is, in a sense, the obverse of
the first. If the person with little vision or with low intelligence (or
perhaps, in pre-Rooseveltian days, with little lower-limb mobility)

is seen as coextensive with his particular limitation and cut off from
functional activity that he may be quite capable of performing, one
also finds the person who, though perfectly capable, technically, of
performing the social roles deemed appropriate by society chooses
to opt out. Thus the objective observation as to whether a person is
performing the social roles one (society? Big Brother?) would expect
him or her to perform places in the diseased category those who do
not accept the dominant normative expectations of their society. In
less benevolent eras, such people were punished or ostracized; today,
they are seen as sick, and the attempt is made to heal them.
The third weakness I would note is related to the second but
differs in that it focuses on the relationship between normative ex-
pectationsand given objective facts. In a hypothetical society in
which adults are supposed to become parents can sterility be classi-
a functional limitation? As long as one accepts the dominant
fied as
normative expectation as the sole criterion, the answer is yes. If,

however, the criterion used is the internalized expectations of the
person whose health status is being examined, then the answer can
be yes or no, depending on the person's expectations.
Consideration, then, of the importance of this facet, on the
one hand, and of the caveats discussed, on the other, has led me to
the following operational question concerning functional limitation:
"To what extent, if at all, does the state or condition of your health
prevent you from carrying out the activities of living that you feel
it is appropriate for you to engage in —
no limitation, mild limita-
tion, moderate limitation, severe limitation?"
This approach, no
doubt, poses the specific problem of self-deceptive and "unrealistic"
expectations. I doubt that too many people are likely to feel severely
limited because their health prevents them from climbing Mt
Everest, running a marathon race, or becoming star performers in
)
the Metropolitan Opera. But this approach has the virtue of meet-
ing the three objections raised above without discarding the concept
of functional limitations. It might be that, in most cases, phrasing
the question in the usual way (that is, not adding the phrase about
appropriateness) and phrasing it as I have proposed would lead to
the same response. Until this possibility is tested, we have no way of
knowing, and it seems to me to be preferable to be guided, in this
facet no less than in the pain facet, by an explicit subjective crite-
rion. (I regard the issue of subjectivity as of crucial significance
for the intervention implications of the breakdown approach. It
will, I hope, become clear in Chapter Eight, where I discuss the
relationship between people in various categories of breakdown, the
sense of coherence, and health professionals, that I think that it is
at least reasonable to hypothesize that movement along the break-

down continuum is closely related to who defines the situation.
Prognostic Implication. up to now,
If, I have gone beyond
the bounds of conventional modern medicine, it is my explicit con-
cern here to include within the definition of the health ease/dis-ease
continuum professional, institutionalized knowledge. In doing so,
however, I do not think that I am departing from my search for

universally applicable facets of a definition of breakdown. In every
culture, I suggest, a given set of signs and symptoms is classified as
having a particular prognostic implication. The same set might well
be differently A seizure has been
classified in different cultures.
viewed variously as a sign of beatification and blessedness, as a

passing and not particularly noteworthy event, as a promise of im-
provement (when following administration of electroshock), and
as an expression of imminent death. But any given culture tends to
have a predominant health-practitioner view of the set of signs and
symptoms. To ignore this facet in defining breakdown would, I
believe, be erring just as badly as to continue to limit oneself to the
model presented by the International Classification of Diseases.
I would suggest three axes along which, in our culture or in
others, a syndrome (or, if one wills, a disease) is always classified
for prognostic purposes; such classification follows the lines of the
natural history of that disease as perceived by the formal health
care or disease care institution of the society. To dichotomize or
dichotomize each of these three axes is, of course, an oversimplifi-
cation and raises problems for the reliability of the classification. In

order, however, to make the tool manageable, I nonetheless think
it possible to dichotomize and trichotomize without doing too
much violence to data. The three axes are the severity of the
condition (mild or serious) ; whether it is acute or chronic; and
whether its natural history is viewed as essentially self-limiting,
stable, or degenerative.
In order to treat such a set of data as a single facet, with a

manageable number of categories that can be ordered, I think it
reasonable to disregard those combinations of the data that on face

value or empirically are not likely to occur. The criterion for in-
clusion is the extent to which a category indicates a clear and

present danger to life.
In contrast to the pain and functional-limitation facets, this

facet must be based on the consensual fund of knowledge available
to the formal authorities in the disease care institution of the so-
ciety — in Western societies, to doctors. Once the physician has
arrived at a diagnostic label, using the tools available, he or she is
asked: "In your professional opinion, considering the medically

most serious state or condition you have observed, would you
say that the person has no acute or chronic condition; a mild,
acute, and self-limiting condition; a mild, chronic, and stable
condition; a serious, chronic, and stable- condition; a serious,
chronic, and degenerative condition; or a serious, acute, and life-
threatening condition?"
In the one which this classification was put to
field trial in
the test in a limited fashion, there was some indication that, at the
very least, as a first approximation, it is useful (Antonovsky, 1973).
Four physicians, following a relatively limited period of training
and clarification, were asked to classify 697 middle-aged women
whom they examined. The classification seemed to make sense to
them, and they reported difficulty with categorizing fewer than
17 of the women. Neither reliability nor validity was examined,
however, and hence the proposal must be considered as most
tentative.
Action Implication. A given condition — a lump
obesity, in
the breast, a lower back pain, caries —or a given behavior—a tic,
visions, smoking, contraception —may or may not be regarded as

health-related behavior in different cultures or subcultures. Once
a condition is defined as related to health, however, all cultures have
norms as to the appropriate action response, including the response
of inaction, even though the norms may differ widely. This obser-
vation leads me to propose that an action facet be included in the
definition of the breakdown, or health ease/dis-ease, continuum.
The question to be asked of the professional is as follows: "Con-
sidering all aspects of the person's health, would you say that he
or she requires no particular health-related action; efforts at reduc-
tion of known risk factors; observation, supervision, or investigation
by the health care system; active therapeutic intervention?"
Several comments are warranted. First, it may be claimed
—
that a wide variety of actions or, indeed, everything one does is —
related to health. In this sense, no person would ever be classified
in the first category. I have, however, intentionally phrased the
other alternatives to refer explicitly to more or less agreed-on (in a
given society) risk factors or conditions related to disease. Thus,
the first category is residual.
Second, I have explicitly committed myself to placing the
onus for categorization on health professionals rather than on the
persons to be classified. I have done so on the grounds that knowl-
edge and expertise not available to the layperson provide the only
rational basis for such classification. I trust it is clear that my ref-
erence to health professionals applies to professionals within a given

society. What happens when professionals from different societies
observe the same phenomenon but arrive at different conclusions

about action is a fascinating question. Study of such situations
would teach us a great deal, but the issue would take us too far
afield here. Further, I hope it is clear that my discussion here is
limited to classification, which should not be confused with decision
making as to action. Epidemiological studies using the breakdown
classification do not require that the health professional take action.
The situation of the clinician who is using the breakdown approach

is somewhat different. The clinician, having classified, is greatly
tempted to make the action decision. This decision making does not,
however, necessarily follow. My own preference is for the clinician
to make available to the person an action recommendation and its
rationale and for the person involved to decide. This preference not
only derives from my own value prejudices but is linked to my
theory of therapeutic efficacy, as will become clear in Chapter
Eight
Finally, a word about the last category, active therapeutic
intervention. In some situations in all societies knowledge is un-
available, though one wishes that something could be done. These
an activist society. The most
situations are particularly frustrating in
dramatic example is what is defined as terminal illness. In
cases of
such cases, nothing can be done except to make the patient as com-
fortable as possible, provide dignity, and let matters take then-
course. Since such situations are patendy more serious than those
that call for no more than supervision, I would classify them in
the last category.
Breakdown Profile
In sum, I have proposed a conceptual definition of the

breakdown, or health ease/dis-ease, continuum as a multifaceted
human organism. Operationally, I have
state or condition of the
suggested that at any one time a person can be described as having
a particular profile —that is, a score on each of the four facets. (I
trust it is clear that under the term breakdown, which must serve
me until a better term comes along, I include the entire gamut of
typeson the multidimensional continuum I have presented. "Low
breakdown" refers to the healthy end of the continuum.)
The mapping-sentence technique involved in facet theory
makes a succinct summing up of the approach possible. The sen-
tence is presented in Table 5. Theoretically, this multifaceted def-
inition of breakdown provides a total of 384 possible profiles
(4X4X6X4). In practice, in the study of middle-aged women
referred to above (Antonovsky, 1973), almost two thirds of the
women were classified in fourteen profiles; almost half, in six
profiles. The study sample, it should be noted, included women
from five different Israeli ethnic groups, ranging from Arab village
women to middle-class, urban women of European origin.
It would take us too far afield to review the data and then-
implications, which I discuss in the published paper. (I cannot,
Table 5. Mapping-Sentence Definition of the Health

Ease/Dis-ease Continuum
A. Pain
Breakdown is any state or 1. not at all
condition of the human 2. mildly
painful;
organism that is felt by the 3. moderately
individual to be { 4. severely
B. Functional
Limitation
1. not at all limiting for the

that is felt by him/her 2. mildly performance of life
to be 3. moderately activities self-defined
I 4. severely as appropriate;
C. Prognostic Implication
1. not acute or chronic

2. mild, acute, and self-limiting
3. mild, chronic, and stable
that would be denned by
4. serious, chronic, and stable
the professional health condition;
5. serious, chronic, and
authorities as a
degenerative
6. serious, acute, and life-
threatening
D. Action Implication
1. no particular health-related
action
2. efforts at reduction of
and that would be seen by known risk factors
such authorities as 3. observation, supervision,
requiring or investigation by the
health care system
4. active therapeutic
intervention
however, refrain from noting the finding that only 9 percent of

the women in the study sample were classified in breakdown pro-
file 1-1-1-1. This finding, in a study conducted in 1970, was sig-
nificant in shaping my interest in salutogenesis. ) My concern here
has been to propose both a conceptual and an operational approach
to the definition of health and illness —an approach that overcomes
the variety of inadequacies I have noted in reviewing alternative
approaches. I would stress that my concern derives from posing
the problem in salutogenic terms.
. —
Why do I insist on putting the problem in terms of saluto-

genesis?Does it make any difference, or is it not even more con-
genial (and hence presumably more acceptable) to those engaged
in this area to use a concept such as general vulnerability or gen-
eral susceptibility, as my colleague Syme has proposed? He writes
of "compromised disease defenses and increased general suscepti-
bility"; "a better understanding of factors that compromise host
resistance and increase vulnerability to disease, factors that . . .
influence generalized susceptibility [and] those that influence the

development of specific diseases" (Syme and Berkman, 1976, p. 6;
Syme and Torfs, 1978, p. 47 )
In principle, I have no objection to Syme's proposal to study
generalized factors that make for moving down the breakdown
continuum. It is, in fact, identical with my own proposal in the
original presentation of the breakdown concept, as suggested by
the very use of the term. I do, however, hesitate to use generalized
factors because the orientation remains clearly pathogenic. In the
conclusion to Chapter One, I note three reasons why the question
of pathogenesis is inadequate: it blinds us to the subjective in-
terpretation of the state of affairs of the person; it pushes in the
direction of the single disease and single bullet; and it postulates
a dichotomous, qualitative distinction between a state of disease
and a state of nondisease. In the present chapter I have developed
the thesis that locating a person on the health ease/dis-ease con-
tinuum (or characterizing a population according to its distribution
on the continuum) poses a different problem than does reference
to particular diseases and opens the way for different kinds of
solutions.
My hunch is that posing the problem of generalized vulner-
ability rather than salutogenesis tends to be subject to the above
objections. Further, it tends to focus on the "more vulnerable,"
continuing to assume that they are the deviants, rather than on
the real deviants —those close to the ease pole of the continuum
and rather than on the factors that facilitate moving toward this
end of the continuum.
If these tendencies are indeed not found, then I certainly
do not have the slightest objection to the use of vulnerability, sus-
ceptibility, or any similar term. Similarly, I do not insist on the
term breakdown, or health ease/dis-ease, continuum. I should like

to think that my concern is substantive and not semantic.
One final important remark is in order. There would seem
to be reasonable ground for charging me with disregarding the
central imperative of my own salutogenic orientation. The health
ease/dis-ease, or breakdown, continuum as presented here essen-
tially seems to formulate the most desirable health category in
negative terms: an absence of pain, no functional limitation, and
so forth. There may be some truth to this charge.
Nonetheless, if the reader is troubled by the absence of a
superhealthy, positive health category, my point has not been fully
clarified. The salutogenic orientation is not concerned primarily
with explaining how people reach perfect health — at best, a heuristic
notion —but rather with understanding the factors involved in re-
maining at a given point or moving up the breakdown continuum,
wherever one is located on it at a given point in time. This orienta-
tion is what makes it no less a tool for understanding the fate of
patients recovering from myocardial infarction than for understand-
ing people located toward the salutary end.
Yet it may be valuable, if we are to study really healthy
people, few as they are, to have some way them be-
of identifying
yond the 1—1-1-1 category. To this end, I would propose an ad-
ditional question, to be asked after the first four questions have been
answered with the first alternative in each case: "You have said
that your state of health is not painful and imposes no limitations.
The doctor's report gives you a clean bill of health. But these are
negative things. Would you say that your state of health goes be-
yond this, that you feel an abundance of energy, that you are what
people call a picture of perfect health?"
Well-Being
have been careful so far in specifying that my concern is

I
with the health ease/dis-ease continuum. As a behavioral science

teacher of medical students, I have often found myself in a bind.
On the one hand, it is crucial that they learn to see health in a
broad context going far beyond the physiological level. On the
other hand — as indicated in my comments above on the WHO
definition of health and Breslow's approach I have considered it —

of the utmost importance that a clear delineation be made between
health ease/dis-ease and other realms of well-being.
In Dynamics of Wellness, a book devoted to health well-
being by and large consonant with the salutogenic orientation, the
opening paper (Sorochan, 1970, pp. 9-10) reflects what I regard
as unnecessary and damaging confusion. "Health is made up of
many kinds of personal well-being. . . . High-level wellness is
characterized by . . . adaptability to cope with and to overcome

all types of stresses in everyday living, . . . ability to give way to
creative imagination, . . . feelings of being a worthy member of
society, . . . feelings of responsibility for others." In other words,
health includes everything that can possibly be regarded by some-
one, or in some culture, as desirable.
In a study conducted in Israel some years ago (Antonovsky
and Arian, 1972), which was part of an international study, I in-
vestigated what may be called the concept of well-being. Health,
in the more limited sense, certainly assumed an important role in
this study, but it was only one of literally scores of significant con-
cerns of people. Some people were terribly troubled about the

possibility of war with the Arabs; others, about the "immoral be-
havior" of their children; still others, about their inability to get out
of debt. But there were other people who found tranquility in
religion, who very much enjoyed their work, or who enjoyed sex.
The study of such social indicators has been possibly the fastest-
growing field of social research in the world since the late 1960s.
One report developed a conceptual model that, after considerable
empirical work, reduced the inquiry to about 100 concerns of per-
ceived life (Andrews and Withey, 1974; Andrews and
quality
Withey, 1976; compare Taeuber, 1978).
My point is that by defining health as coextensive with the
many other dimensions of well-being, one makes the concept of
health meaningless and impossible to study. It is, of course, folly to
deny the interaction between health well-being and other dimen-
sions. I deal with this issue in Chapter Seven. But the nature of this
relationship is one that must be subjected to theoretical clarification

and empirical investigation. Health well-being must be measured
separately, whether in the way I have proposed or in some other
way. Only then can we clarify the forces that shape the individual's
or group's location on the health ease/dis-ease continuum.
Having defined the problem as that of salutogenesis and

having clarified the meaning of health by using the breakdown
concept, I can now turn to the development of the conceptual

model that I believe offers the greatest promise in leading to a
solution of the problem. What set of factors, we shall be asking,
can help us understand location on and movement up the break-
down continuum?
Inevitably, both because I have been conditioned as well
as everyone else by the question of pathogenesis and because the
overwhelming part of the data available asks this question, I too
shall slip into asking, Why are people located on —or why do they
move down toward —the dis-ease end of the continuum? I shall
seek to avoid doing so and ask the reader to join me in this effort.
But for those readers who have not been persuaded that saluto-
genesis is a different question, I would at least insist that they un-
derstand that it is dis-ease, and not disease, that is of concern in
this book. As I put it when I first advanced the breakdown concept
(Antonovsky, 1972, p. 540): "Given the 'right' constellation of
factors, one will 'look around' for a way to break down. There
are, as it were, always additional factors in one's internal or ex-
ternal environment which one 'chooses' and which facilitate the
expression of the breakdown in one specific disease or another."
It is to the study of the "right" constellation of factors that we
now turn.
Chapter Three
Stressors, Tension,
and Stress
I would no more deny the pathogenic role of stressors than I
would the role of the tubercle bacillus as a causative factor in

clinical tuberculosis. The evidence is quite strong that, in a large
number of diseases, stressors appear as statistically significant risk
factors. Posing the problem of salutogencsis, however, brings one up
short. When the tubercle bacillus is not present, one does not get
tuberculosis. But is it the case that when stressors are absent, one
stays at or moves toward a low level of breakdown? The absurdity
of the suggestion is not immediately self-evident in reading the stress
literature. There, discussion always focuses on the highly stressed
group, suggesting that the controls have not experienced stressors.
In this light, let me state the central thesis of this chapter in

ele-
mentary form: Stressors are omnipresent in human existence. In

response to a stressor, the organism responds with a state of tension.
70
Stressors, Tension, and Stress 71
This state can have pathological, neutral, or salutary consequences.

Which outcome results depends on the adequacy and efficiency of
tension management. Poor tension management leads to the stress

syndrome and movement toward dis-ease on the continuum. Good
tension management pushes one toward health ease.
In this chapter I seek to clarify the nature of stressors and
place their role in salutogenesis and pathogenesis in what seems to
me to be a proper perspective. The concept of tension management
is the focus of attention of Chapter Four. There a variety of re-
sistance resources come under systematic scrutiny. This attempt at
systematization led me to complete the salutogenesis model with the
development of the sense of coherence, the subject of Chapter Five.
What Is a Stressor?
The human organism has evolved homeostasis-maintaining

and homeostasis-restoring mechanisms. These regulate body tem-
perature, blood pressure, blood calcium, and so forth, so that we
may stay alive. As Cannon (1929, 1939) demonstrated about half
a century ago, these exquisite mechanisms are in constant, auto-
matic operation, responding to the constant disturbances of homeo-
stasis that living involves. Parallel learned mechanisms in the
development of personality, the cultural, subcultural, and idiosyn-

cratic-individual responses to the minutiae of an ever-changing en-
vironment, maintain social and psychological homeostasis. In other
words, in the course of living, we constantly engage in the minor,
automatic expenditures of energy required to keep ourselves on
even keels. Surely this is a fascinating field of study. But it would
be foolish to define the relatively slight changes in our environ-
mental field, thousands of which occur daily, as stressors.
The view of life as a constant, dynamic cybernetic system,

however, is a valuable point of departure for a discussion of the
stressor concept. Without qualifying as an acceptable definition, the
notion that information about change requires energy expenditure

for the restoration of homeostasis suggests the crucial facets of a
useful definition of a stressor. First, it implies that the initial source
of the disturbance can be the external or the internal environment
of the organism or both. Second, it suggests that homeostasis has
.
been upset —that a demand has been made on the organism that,
given its mode of functioning at that moment, it is not capable of

meeting. An additional source of energy must be called on. Note
that no assumption about pain or unpleasantness is made.
The difference between a stressor and other types of stimuli,
I would suggest, is, at first sight, a matter of degree. A routine
stimulus is one to which the organism can respond more or less
automatically, one that poses no problem in adjustment. The feed-
back mechanism is part of one's routine repertoire, with energy
readily available to allow its functioning. A stressor, however, can
be defined as a demand made by the internal or external environ-
ment an organism that upsets its homeostasis, restoration of
of
which depends on a nonautomatic and not readily available energy-
expending action. I do not know whether it will ever be possible
to identify empirically at what point a routine stimulus becomes
a same way that we can identify zero degrees Celsius
stressor in the
as the point at which water becomes ice. More likely than not,
there is a transition zone. This does not mean, however, that there
is not a qualitative difference between stressors and other stimuli.
Given this position, the implication would seem to be clear

that whether a given phenomenon, a given experience, a given
stimulus is a stressor or not depends both on the meaning of the
stimulus to the person and on the repertoire of readily available,
automatic homeostasis-restoring mechanisms available. This ap-
proach is not at all original. It is well-nigh identical with that taken
by Lazarus and Cohen (1977), who view the problem in "trans-
actional" terms and focus on the "mediating processes" of cognitive
appraisal and coping. They state that stressors are "demands that
tax or exceed the resources of the system or, to put it in a slightly
different way, demands to which there are no readily available or
automatic adaptive responses' ' (p. 109, emphasis in original). Sim-
ilarly, my approach to the definition of a stressor is concordant with
the "problem-solving model" of Scott and Howard (1970). And
I do not see any significant discrepancy between my approach and
that ofMoss ( 1973 ) even though, in referring to Scott and Howard,
,
he writes, "Their model, we think, should not be presented as a

general integrative model" (p. 51 )
There is, then, an emerging consensus with regard to the
stressor concept that, in Lazarus' term, sees it as a transactional

phenomenon based on the meaning of the stimulus to the perceiver.
Whatever the many disagreements and the lack of clarity in the
theory of stress, on this one issue there seems to be consensus. Yet
I cannot go along with the implication of this consensus, para-
doxical as this position may seem. Of course events mean different
things to different people and are differentially threatening because
of past experience, because man is a symbolic animal, and because
of differences in available repertoires. But if we return to our
starting point —the

problem of salutogenesis I suggest that this —
understanding, true as it might be, does not matter much.
Let me be explicit about what I am saying does not matter
much. I agree that there are individual or group differences in the
differential perception of phenomena as stressors or as routine
stimuli. But ifwe seek to understand different levels of breakdown,
then what is common to us all (or almost all) is far more important
than such differences. My argument is based on two interrelated
by the very nature of the human organism at the
points. First,
biochemical, physical, and psychological levels and by the very
nature of all human cultures, there is a wide sphere of consensus
about what would be perceived as a stressor. Second, and even
more crucial to my argument, is the observation that even if we

do differ in labeling phenomena as stressors, the overwhelming
number of human beings are, most of the time, in the throes of
confronting what they define as stressors. The phenomena defined
as such may differ; the extent of confrontation differs relatively
little.
Are Stressors Objective or Subjective?
Let me start this discussion by granting one qualification at

the outset. Some individuals unfortunately are extremely deviant.
They define experiences that the rest of us tend to see as minor
stressors or as routine stimuli as overpowering stressors. Taking an
examination, speaking in a group, entering an elevator, being
touched by another are phobias in the realm of psychopathology.
My case does not at all rest on the idea of phobic stressors. (If,
however, George Orwell in 1984 is correct in suggesting that each

of us may well have his particular phobic interpretation of a given

stimulus that Big Brother can identify, then such phobias are part
of the picture.)
As a point of departure, we may note that none of us would
have any hesitation in identifying certain human exigencies as
stressors. I here refer to physical trauma of sufficient suddenness
and massivity to destroy or decisively injure a vital organ and

hence result in sudden death. Perhaps falling just short of such
events are what Lazarus and Cohen identify as cataclysmic phe-
nomena (1977, p. 91), which they define as "life events . . .
affecting large numbers of people, usually outside of the control

of individuals or groups, and assumed to be more or less univer-
sally stressful." They refer to natural disasters, wars, bombing,
and relocation.
In part of my own research, ranging from my 1963 study
of multiple sclerosis (Antonovsky and Kats, 1967) to my 1972
study of overall health (Antonovsky, 1974), my commitment was
to the question "Did you encounter this experience?" rather than
to the question "Did you encounter experiences that you would
define as stressors?" It seems to be a reasonable assumption that
almost everyone would agree in defining as stressors experiences
like being in a situation where the people around one are being
killed, having the head of the family unemployed for months,
having one's child die, or migrating from one country to another.
In taking this was following the lead largely of
approach, I
Selye ( 1956) Since he was concerned mosdy with stress, defined as

.
a general adaptation syndrome, he seemed to define a stressor as

"that which produces stress" (p. 64). Working in the laboratory,
he administered stressors that were noxious agents by definition:
"[One can produce the general adaptation syndrome] by injecting
foreign substances (tissue extracts, Formalin). Subsequent experi-
ments showed that one can produce essentially the same syndrome
with purified hormones. . . . One can also produce it with physical
agents, such as cold, heat, x rays, or mechanical trauma; one can
produce it with hemorrhage, pain, or forced muscular exercise; in-
deed, I could find no noxious agent that did not elicit the syndrome"
(pp. 29-30). Note that Selye is not defining stressor in circular
terms. He assumes broad consensus among laboratory workers that
—
certain things are noxious agents. In precisely the same way, I con-
tended that there is indeed broad cultural, if not universal, con-
sensus that certain experiences are noxious or are stressors.
This objective, consensual approach to stressors at first seemed
to be accepted by Harold Wolff and his colleagues at the Human
Ecology Study Program at Cornell. Study of thousands of patients
in the late 1940s and early 1950s led them to the conclusion that
certain types of life events were empirically observed to cluster at
the time of disease onset (Wolf and Goodell, 1968). Holmes, one of
Wolff's colleagues, working with Rahe, later systematized these
events in a rating scale (Holmes and Rahe, 1967). The events in-
cluded in the scale, whether ordinary or extraordinary, socially
desirable or undesirable, had one common theme: Each was one
"whose advent either is indicative of, or requires, a significant change
in the ongoing life pattern of the individual" and as such "evoked,
or was associated with, some adaptive or coping behavior" ( Holmes
and Masuda, 1974, p. 46). In subsequent studies conducted in a
variety of countries, Holmes and his colleagues found that the forty-
three items in their Social Readjustment Rating Scale were given
much the same rank order in weighting by many different popula-
tions. Using this scale, they assign people Life Change Unit (LCU)
scores, which have been used in many studies and have been shown
to be correlated with vulnerability to illness. Holmes, then, seems to
be arguing that at least these forty-three life events are, objectively
and universally, stressors, though of different magnitudes.
Holmes' colleague Rahe, however, suggests a departure from
this view and opts for what I earlier called an emerging consensus.
In his view, "In dealing with large samples one can use these mean
LCU values. ... In dealing with small groups of subjects, how-
ever, individual variation in LCU scaling may assume some im-
portance" (1974, pp. 76-77). Rahe then proposes a Subjective
Life Change Unit scaling system, in which each individual assigns a
score to those of the forty-three life change events he experienced
the score representing "the amount of adjustment you needed to
handle the event." Conceptually, Rahe describes this technique as
"the past experience filter" (Rahe, 1974). (For a succinct review
of the life-events school, see Rahe, 1978.) Wolf and Goodell (1968),
colleagues of Wolff, explicitly commit themselves to this subjective
approach: "Illness often occurs when an individual perceives his

life situation as peculiarly threatening to him, even though his life
situation may not appear to be threatening to an outside observer"

(p. 206).
Hinkle, Wolff's co-organizer of the Human Ecology Study
Program from which so many of the important stress
at Cornell,
studies have come, is likewise committed to the meaning of an ex-
perience as a potential stressor: "From a physiological point of view
a man may be expected to react to the leaning' of information he
obtains from his social environment and not necessarily to the 'ob-
jective' features of it that are discernible by others" (Hinkle, 1974,
p. 10).
At the same time, one can detect some ambivalence among
these scientists. Wolf and Goodell, in a lengthy and detailed discus-
sion, write,"Important as are the upheavals of rapid social change
and the cruel circumstances imposed on man by his fellows, the ordi-
nary vicissitudes of daily life offer their share of challenge as has
been pointed out. It may indeed be stated that man is always under
stress of one sort or another." They then go on, in a lengthy discus-
sion, to detail what American life is like in the twentieth century
from this perspective, making the assumption that all the things they
point to are stressors and are interpreted as such by all those who
experience them (pp. 212-221). Hinkle too comfortably slips into
phrases like "major social deprivations and demands and . . .
major changes in . interpersonal relations" (p. 24).

. .
In other words, though many investigators seem to pay lip
service to the importance of conceptualizing stressors as subjectively

experienced and interpreted, "meaningful" phenomena, in the last
analysis they are aware, implicitly or explicitly, that most people

viewed a fairly wide range of experiences as stressors.
Ubiquity of Stressors
It is, however, most important to take this discussion one

step forward. Conceivably, there could be a large measure of agree-
ment among people that certain experiences or phenomena are
stressors. Yet at the same time they might well differ considerably in
the extent to which they personally experience such stressors. Thus,

in my own studies, I found considerable variation, with respondents

having very low to very high stressor scores. Similarly, all the studies
using the life-events scale have reported great variation. And from
common observation and daily living, all of us can point to people
to whom everything seems to happen and to others who seem to
walk between the raindrops. In fact, two of my publications would
suggest that I am at least as aware as other investigators of the sub-
stantial differences among individuals and groups with regard to
their exposure to stressors. I cannot pretend that the concentration
camp survivors I studied lived through no more stressors than did
the control group (Antonovsky and others, 1971). Nor would I
claim that poor people suffer fewer stressors than do the nonpoor
(Kosa, Antonovsky, and Zola, 1969).
My thesis, rather, is that all of us throughout life, in even
the most benign and sheltered of environments, are fairly contin-
uously exposed to what we define as stressors. The range of human
experience in exposure to stressors is not from very low to very high.
It is, rather, from fairly serious and lifelong —
in, shall we say, the
typical experience of many, though far from all, comfortable middle-

class readers of this book —
to the unbelievable hell on earth of so
large a part of the world's population. We are able to get low scorers
on stressor experience because we do not ask the right questions or
do not ask patiently enough and not because there really are any low
scorers. But then why on earth should we expect otherwise? If in
the last 150 years well over sixty million human beings have died as
a result of human violence alone, how have the rest of us lived? Let
us thus turn to a somewhat systematic analysis of the stressors that
confront human beings.
It is, perhaps, most appropriate to start with cautious and
tentative reference to two exciting fields that are largely alien to
social scientists, genetics and physics. While I would not even begin
to pretend to anything but the most superficial acquaintance with
the ideas I shall touch on, their importance is so considerable in the
present context that I have gone out on a limb. The concepts of
entropy and of genetic mutation are, at the physical and biochemi-
cal level, most pertinent to the thesis that stressors are ubiquitous
in human existence.
The Second Law of Thermodynamics states that the en-
tropy —roughly, —
a measure of disorder of a closed system will
always increase. Maximum entropy is reached when a permanent
state of no observable pattern of events in the system occurs. The
total entropy of a system must increase or at least remain constant.
But the law applies only to a closed or isolated system. Clearly, the
human organism is not a closed system. As Schrodinger puts it
(1968, p. 145) "Thus a living organism continually increases its

:
—
entropy or, as you may say, produces positive entropy and thus —
tends to approach the dangerous state of maximum entropy, which
is death. It can only keep aloof from it, i.e., alive, by continually
drawing from its environment negative entropy. . . . Thus the de-
vice by which an organism maintains itself stationary at a fairly
high level of orderliness (
— fairly low level of entropy) really con-
sists in continually sucking orderliness from its environment."
The concept of negative entropy is referred to again in
Chapter Four. For the time being, our concern is with a considera-
open and closed systems differ only in that
tion of the possibility that
closed systems cannot be saved by negative entropy, by sustenance
or information. (See Buckley, 1968, pp. 143-169, for discussions of
the links between the concept of entropy and information theory.)
Thus there must be inexorable, unavoidable, immanent factors that
produce entropy; in our terms, these factors are called stressors. We
may find, I believe, some hint of what these factors are if we turn
to genetics.
In brief, there is a genetically programmed, built-in, and
ultimately victorious pressure toward senescence and death of the
organism. Burnet (1974 and 1971, especially pp. 154ff.) posits two
evolutionary requirements for the individuals of any species: sur-
vival to a reproductive age and death when survival no longer
offers any advantage for reproduction. He particularly focuses on
"abnormal mutant cells which by developing toward malignancy
threaten survival" (p. 131). (In Chapter Four, Burnet's concept
of immunological surveillance is considered in the context of our
consideration of resistance resources. ) The suggestion is that, whether
through mutation of cells or other senescent pressures, the organism
is constantly assailed by the stressors, the challenges, that push
toward entropy.
While Burnet's interest is largely in genetically determined
stressors, he by no means ignores the microbiological and other

stressors of the external environment. These certainly are not to be
disregarded in an age that has become conscious of environmental
pollution, radiation hazards, and the like. But Dubos, too, in his text
on infectious diseases (1965, p. 35), argues that endogenic micro-
organisms are, at least in modern societies, the major category of
biological stressors: "In the classic infections of exogenous origin,
the detennining etiologic event of the disease is exposure to the in-
fective microorganism. In endogenous microbial disease, the imme-
diate cause is the environmental factor which upsets the biologic
equilibrium that normally exists between the host and the microbial
agents. . . . The methods ... of the past will not prove to be
effective in the control of the disease states caused by microbial
agents which are ubiquitous in our communities in the form of
dormant infections."
Technically, Dubos' focus on the innumerable, omnipresent
endogenic microorganisms, which maintain a symbiotic relationship
with the host, is best referred to as a concern with potential stressors.
The dormant infections of which he writes become patho-
latent or
genic only under certain conditions. We now understand that these
conditions are not only physical, chemical, and biological but also
psychosocial. Dubos' analysis, then, is most consistent with the
present thesis of the ubiquity of stressors; it presses us to view the
environment as made up of inextricably interwoven strands. And it
also poses, to anticipate, what will be the key question of Chapter

Four: How are potential stressors prevented from becoming
pathogenic?
My concern here has not been to make any pretense of
expertise in the physical or biological sciences but only to indicate
that they cannot be ignored in a systematic consideration of the
ubiquity of stressors in human existence. I simply wish to avoid the
mistake of disregarding literal bugs in a discussion focusing on fig-
urative bugs. Let us, however, now turn to the latter, to what gen-
erally has come to be called psychosocial stressors. Today the litera-
ture is overwhelming. At the time Hinkle and Wolff were developing
their laboratory at Cornell, however, the link between psychosocial
stressors and health was hardly the concern of many. Thus, I find it
of interest to note that the central theme of this chapter the —
ubiquity of stressors —
was caught in an image that appears in what
I regard as a landmark book of the early 1950s, Beyond the Germ
<r
Theory. In it, the editor writes (Galdston, 1954, p. 13) : Viewed
thus, dynamic homeostasis can be likened to a tight-man walking a
rope from one end to the other, balancing himself even while he
changes clothes and takes on and discards a variety of other ob-
jects." Let us consider the tightrope image systematically.
Reviewing the sources of stress, Lazarus and Cohen ( 1977)

consider three major types. The first two differ in the number of
people involved. Cataclysmic phenomena, such as natural disasters,
bombings, internment, and relocation, are experiences shared by
whole population groups. Events of the second type, including
bereavement, terminal illness, and being laid off from work, also
have the same powerful and sudden impact on the persons involved,
but these typically happen to fewer persons or to a single individual.
Both these types of stressors occur, Lazarus and Cohen imply, with
relative infrequency — or, if frequent, only to special segments of
the population. My hunch is, as I suggest below, that the data would
indicate that such events are far more frequent than one might
think. But it is when we turn to their third type of stressor daily —
hassles —that we come close to an examination of the ubiquity of
stressors. But the Lazarus-Cohen account, while extensive, is hardly
systematic.
Let me attempt a systematic account of psychosocial sources
of stress here. In the 1930s an international "progressive" writers'
congress was held in Paris. The tone for days had been apocalyptic;
speaker after speaker portrayed the imminent solution of all human
problems in the paradise abuilding in the Soviet Union. Andre
Malraux, leftist but iconoclastic, rose and, in a few words, damp-
ened the euphoria: "And what of the child," he asked, "who, even
in the perfect socialist society, is crushed under the wheels of a
tram?" My point is simple. In all human societies, whatever the
precautions taken, harsh accident is immanent. My concern here is
not with the victim of the accident but with the survivor who has
loved the victim and, in many cases, with the bearer of responsibility
for the accident. The victim has suffered a direct physical trauma;
the survivor is confronted with a psychosocial stressor. Given the
rates of accidental injury and death and the rate of homicide and
the fact that for each victim there are most often at least three or
four persons intimately involved with the tragedy, such accidents
can hardly be considered an unusual source of stressors. They are
not omnipresent, except as potential stressors. But their frequency,
particularly among certain groups (spouses of people who drive
daily, parents of children who play in the streets, workmates of
those exposed to industrial accidents), leads me to see them as a
constant threat in our lives.
Consideration of accidents leads us to take a further step.

By and on "What has happened to you?"
large, stress studies focus
Of the on the Holmes-Rahe scale of life events,
forty-three events
which is typical of or even better than most scales, five refer ex-
plicitly to other persons; several refer to interpersonal events. But
one other, the global "major change in the health or behavior of a

family member," is presumed to cover the many things that happen
to those close to us. Because we are embedded in networks of social
relationships, our own experiences are compounded by those of the
persons with whose lives our own are intertwined. ( For those many
persons not so embedded, this source of stressor is not significant;
the price paid for not having such headaches, however, may well be
farmore severe.)
I would continue this list of sources of stress by discussing
an area that has become oddly neglected. As a sociologist, I am
pleased by the increasing attention paid in the stress literature to
social disorganization, poverty, the workplace, and so forth. Yet it
seems to me that this rebellion against psychoanalysis has gone too

far. It behooves us to recall that serious demands and conflicts can
be rooted in the internal no less than in the external environment.
The very phrase life events, when it monopolizes discussion, signifies
a displacement from any concern with internal events (as stressors
and not as reactions). The things thathappen to us have become

the concern of investigators. Freud himself characterized the horrors
of the human condition (quoted in Schur, 1972, p. 398) : "The
terrors of nature, . . . the cruelty of Fate, particularly as it is
shown in death, . . . the sufferings and privations which a civilized

life . . . has imposed, . . . the perplexity and helplessness of the
human race, . . . the suffering which men inflict on one another."
Thus Freud in 1927, before Hitler, before the Gulag Archipelago,
before Leningrad and Coventry and Lidice and Rotterdam, before

Auschwitz, before Hiroshima. But one must go much further with
Freud, into the soul of the human individual, beyond the exposure
to the horrors of history. Whether drives are viewed as instinctual
and biologically determined or as socially learned; whether one ac-
cepts a narrow or broad interpretation of libido; whether one is
committed to viewing personality formation as an intrapsychic or

interpersonal process, a book that is closed in early years or one
that is rewritten throughout life —whatever the disputes about
Freudian theory — it is the utmost of naivete to disregard the per-
manent state of underlying conflict that characterizes our inner
emotional environment. The impermissible impulses are there; the
ambivalence is there; the fears and anxieties, if not of mutilation
then of abandonment and death, are there; the guilt and shame are
there. Individuals differ (as do cultures), I suggest, not so much
in the extent to which underlying emotional conflict is found. These
stressors are immanent. We differ, rather, in the extent to which
we can contain, cope with, and perhaps even exploit (or, as Freud
would have it, sublimate) these conflicts. But the stressors are there.
Again, one need not be committed to all the ideas of the
recent trend in zoology, ethology, and sociobiology to take these
ideas seriously in the present context. Thus L. Tiger and R. Fox
(as quoted in Becker, 1972, p. 40) see human behavior as shaped
by a biogram shared by all hunting primates, involving a basic ap-
petitive predisposition that orients them "in the direction of the
search for power and self-perpetuation within a hierarchy of dom-
inance and subordination characterized by competitiveness, the real
hunger for triumph, and the celebration of triumph." Whatever the
adequacy or inadequacy of this approach as an explanation of
violence, evil, aggression in human existence, I would suggest that
there is in us all at the least an elemental fear of being the victim

of At some level, we all fear being the victim of aggression,
evil.
mutilation, and destruction by others. Biograms may indicate no

more than potential for, predisposition to, or orientation toward
certain actions; ontogeny no less than phylogeny provides us with
more than enough experience to make the fear an integral, if sub-
dued, part of our everyday psyche.
Clearly, then, there are grounds for maintaining that the
human organism is subject to constant psychosocial stressors by the

very fact of having been born and survived. But we also live in
ongoing social environments. I would consider these at three differ-
ent levels of distance from the individual: the historical-universal,

the contextual, and the immediate. It may well be that before the
twentieth century what happened on a historical scale could not be
considered as a significant stressor unless one were directly swept up
in its wake. But today the leaps and bounds of the imagination are
facilitated by the mass media. More and more of us know that a
small war in one part of the world may engulf us all. We see the
fighting on our screens. We can visit Verdun.
Sometimes it would seem that our own lives are fairly stable
but that the immediate world around us is in the process of trans-
formation. We cannot avoid relating to it but it is not the world to
which we have learned to relate. A good case in point is a study
of rural residents in North Carolina; the comparison of interest was
between those who lived in counties undergoing differential rates
of urbanization (Tyroler and Cassel, 1964). Their concern is with
people whose own activities did not change much, but the demo-
graphic, economic, and technological contexts in which these ac-
tivities were carried out changed radically. ( For a broader, popular
discussion of the world around us in upheaval, see Toffler, 1970.)

Our assumption tends to be that such transformation is a peculiar
characteristic of our modern world. That this is hardly the case is
suggested by the following quote. In 1882, preparing the intro-
duction to a new edition of a work that had first appeared in 1858,
Oliver Wendell Holmes ( 1884, p. iv) wrote:
This book was written for a generation which

knew nothing or next to nothing of war and hardly
dreamed of it; which felt as if invention must have ex-
hausted itself in the miracles it had already wrought.
Today, in a small seaside village of a few hundred in-
habitants, I see the graveyard fluttering with litde flags
that mark the soldiers' graves; we read, by the light the
rocks of Pennsylvania have furnished for us, all that is
most important in the morning papers of the civilized

world; the lightning, so swift to run our errands, stands
shining over us, white and steady as the moonbeams,
burning, but unconsumed; we talk with people in the

neighboring cities as if they were at our elbow; and as
our equipages flash along the highway, the silent bicycle
glides by us and disappears in the distance. All these

since 1857, and how much more than these changes in
our everyday conditions! I can say without offense today
that which called out the most angry feelings and the
hardest language twenty-five years ago. I may doubt
everything today if I will only do it civilly.
But if history and the more immediate world around us

supply us with a considerable panorama of stressors, I suggest,
without in the least ignoring this panorama, that the daily social
structures in which we are all embedded are inevitably and per-
petually stressful. I would, moreover, remind the reader that my
concern is with stressors as defined earlier, and not with stimuli
that may upset homeostasis but for which we have readily available
restorative mechanisms. This is true throughout the life cycle.
For many years, the field of child psychology preempted the

attention of scientists concerned with development Infancy and
childhood, in the wake of the Freudian revolution, were viewed
as replete with inherent developmental crises. Adolescence, with
less was viewed as an age of turmoil by definition, with only
study,
occasional checks and correctives introduced by the anthropological
perspective. Reaching adulthood, however, meant achieving ma-
turity, stability, and more or less smooth sailing. With the change
in demographic structure and the increasing proportion of the el-

derly, gerontology was launched as an important field of inquiry.
The stressors of retirement, dependence, poor health, and imminent
death came to be investigated. However, only in recent years has
the field of developmental psychology emerged as an area that is
based on the paradigm of a total normal life cycle at the core of
which lies an endless series of challenges. True, sociologists had long
studied particular stressors of work, family life, migration. But there
had not emerged from this work a developmental view of conflict
or, in our terms, of stressors over the entire life span.
To take an example. When our team
began a study of first
ethnic differences in adaptation to menopause, we found two over-

whelming interests reflected in the literature. First, the particular
concern was with involutional psychosis, a pattern that at most

affected a small fragment of the population (truly, the medical-
model concern). Second, the assumption was that the cessation of
menstruation was invariably and almost solely the key stressor on
both the hormonal and psychosocial levels. With the exception of
Bernice Neugarten's team at the University of Chicago, which did
a series of studies on aging, no one seemed to be aware that many
other stressors were endogenic at this stage of the life cycle: for
example, the empty nest, the change in social status, illness, and the
possible death of one's spouse (Maoz and others, 1970) . Or to take
another example. One of the better collections of readings on hu-
man development, edited by one of the leading psychologists in the
United States, is concerned with "ecological validity —the study of
the developing person in the contexts in which he lives." Yet its
678 pages contain not a single word about "the developing person"
beyond adolescents and college students (Bronfenbrenner, 1972).
Only with the appearance of Erikson's familiar eight-stage
diagram of phase-specific psychosocial crises (Erikson, 1950) did
we gain a theoretical model that is useful for studying the proto-
typical stressors characteristic of all stages of the normal life cycle.
For present purposes, it matters little that Erikson himself has made
his most brilliant and important contributions to the study of the
adolescent identity crisis; it matters little whether there is adequate
evidence that the prototypical adulthood "generativity vs. self-
absorption" crisis is cross-cultural and universal. ( Certainly Erikson

is supremely aware of the historical-cultural context of psychosocial
crises; see Erikson, 1968, pp. 61-65.) What matters is that he has
called our attention to and given us a possible framework for an-
alysis of the psychosocial stressors that are inherent in human ex-
istence over the whole life span.
It is manifestly beyond my ability and would serve little
purpose to attempt to consider in detail these inherent stressors.
Readers can fill in the details for themselves whose and for those
biographies they know well. Such details, it should be remembered,
are limited largely to those stressors of which we are aware; others
have been banished from the forefront of our consciousness and
memory. Many a mother or father can only smile in recollection of
the first time they bathed their first child; many a teacher smiles
recalling his first lecture; many a spouse, sexually fulfilling and ful-
filled, forgets the period of early intercourse. Both those who fail
to overcome and those who do overcome stressors have good reason

to forget the terrors of such encounters. But they were there. My
intention, rather, is to suggest a number of theoretical issues in
thinking of the inherent stressors that accompany us throughout the
life cycle. These issues are, I believe, relevant to all our social roles
and to all our personal encounters. (A detailed consideration of

stressors in the life cycle is given in Academic Press' Life-Span
Developmental Psychology series and particularly in Datan and
Ginsberg, 1975.)
The first point to be made is implicit in the above discussion.
In the most benign and stable of worlds —a world certainly no
reader of this book will ever encounter —we constandy enter new
roles, new stages, new contexts. All of these pose challenges, make
demands —that is, they are precisely what we have called stressors.
Such new entries are numerous. I know of no one who has counted
such entries (which always, it should be remembered, also involve
exits and detachments) any population. The Holmes and Rahe
for
Social Readjustment Rating Scale is but a bare start in this direc-
tion, given its closed-end character and time-limited focus.

Second, we should note that anticipatory socialization is
inevitably inadequate. In training, we always know that we are, as

it were, playing a game, serious as it may be. The responsibility
and the authority are still not ours. One can always say "fingers"
(in the language of my New York City childhood) to stop the
game. Simulation has been brilliandy developed as a didactic tool,
but it is not the real thing. When we do enter a new situation, the
stressor is always considerable. We do differ on how adequately we
have been prepared to cope with it, but the most adequate social-
ization does not guarantee that the stressor will not be perceived as
we may have been well prepared to fight the last war;
real. Further,
but thenew war is inevitably somewhat different.

The third issue is raised tentatively because it requires a
fundamental assumption about human nature. I discuss it here
briefly and shall return to its consideration subsequently. Let us call
it the problem of underload. We tend to think of the stressor con-

cept, by its very definition, as referring to overload. Yet this is not
necessarily the case. There is good reason to think that an environ-
ment makes no demands on us, that does not confront us with

that
stressors, by that very fact, a stressor. The concept of underload
is,
postulates an inherent need for sensory stimulation of moderate

magnitude and complexity. This postulate can call on evidence
from a number of sources. It can be linked to the considerable work
on infant development, which relates comfort and sustenance to
stimulation. A wide variety of studies on sensory deprivation in
animals and in adults, as well as in children, point to the patholog-
ical consequences of underload. Psychophysiological research points
to the requirement of the central nervous system for arousal in
order to function adequately. Whatever the basis for the argument,
it seems reasonable to suggest that environments that do not pose
challenges for one are inherently stressful. (For reference to some
of the literature, see Bronfenbrenner, 1972, pp. 256-301; Frank-
enhaueser and Gardell, 1976; and Wohlwill, 1971.)
Up to this point I have taken pains to emphasize the in-
herent stressors that all humans confront by virtue of being human
and social beings. I have argued that whatever our social, cultural,
and historical locations, life is, to reiterate Galdston's phrase, a
walk on a tightrope with constant changes of clothes and taking on
and discarding of objects. The most elementary acquaintance with
history, with anthropology, and, above all, with literature —be it the
Bible, the Greeks, Shakespeare, Dante, or Dostoevsky — reveals the
rarity of tranquility in human existence. The content of the dis-
turbances varies in time and place. The magnitude, at most, varies
from that of Watergate to that of Vietnam.
But I would here focus not on what I earlier called cat-
aclysmic events. As important as they are, they are different from
the internal conflicts of everyday existence that are anchored in the
social and cultural organization of every society. Lazarus and Cohen
are profoundly though unintentionally misleading when they speak
of these as daily hassles. True, they do refer, under this heading,
to the stable, chronic, and repetitive hassles of poverty. But their
necessarily brief summation and citation of the literature are hap-
hazard. Different environments, they are saying, can be stressful.
I would suggest two fundamental reasons why all social environ-

ments are inherently stressful.
One need not be a Marxist to see, permitting myself some

license, the history of all social institutions as one of struggle; hus-
band and wife, parent and child, supervisor and subordinate, priest
and parishioner, doctor and patient, teacher and student, officer and
soldier, representative and voter, leader and rank and file all —
without exception relate to each other in a context of scarcity of
resources, of power, of different perspectives and interests and mo-
tivations. To deny love, mutual aid and support, cooperation,
complementarity, altruism — all that Nisbet ( 1973), in his survey of
Western thought and civilization, calls community^ — is to pro-

foundly distort But to blind oneself to the constant, inherent con-
flict in all social relations, group and individual, is no less a dis-
tortion. (The sociologist reader will recognize my indebtedness to

Simmel, 1955, and to Dahrendorf, 1966. For a popularized account
of the intrinsic stressors in American middle-class life, see Sheeny,
1974.)
The second reason for anticipating that all social environ-
ments will be stressful derives from Merton's classic analysis (1957)
of the gap between culturally inculcated goals and socially struc-
tured and legitimated means for goal achievement Merton was

particularly concerned to understand the structural origins of de-
viance, which he defines as rejection of means or goals or both.
Because he wrote his analysis originally in 1936, it is not surprising
that Merton takes as his prime example the American culturally
prescribed goal of the acquisition of material goods and, certainly
at that time, the socially structured impossibility, for most Ameri-
cans, of achieving the goal that they had presumably internalized.
Not all human societies base role placement primarily on
achievement rather than on ascription. But all societies have at least
some room for mobility, if only for a few individuals. Further, all
societies prize particular characteristics: beauty, skill, strength, char-
acter, virtue, whatever the particular interpretation given. Most

important, all societies have scarce resources and rewards, material
and nonmaterial. Even in the most rigid caste society, with the
most perfect indoctrination so that everyone knows one's place, the
lowliest of the low have goals that are, in theory, accepted by all as
legitimate. Yet no society in history has managed to avoid struc-
tural limitations or, to put it positively, has managed to provide
structured access to the goals that it has propagated.
In this context, of particular interest is a paper that seeks
to bring Merton's paper up to date (Simon and Gagnon, 1976).
Merton focused on the "fetishism of commodities" unavailable to

most Depression-era Americans. Merton's paradigm of different
types of anomic response to the means-goal gap (innovation, con-
formity, retreatism, and rebellion) seemed to be adequate to ac-
count for the major behavior patterns of that time. Simon and
Gagnon, however, raise the question of "the anomie potentially
generated by unanticipated affluence." In essence, they ask, what
happens when people do succeed in the goal achievement posited
by Merton? Many, though far from all, Americans are affluent, and
they are not particularly happy. The authors' expanded paradigm
of responses to this situation, though fascinating, does not concern
us here. What does concern us is the thesis that all societies are
characterized by the dilemma of the goal-means gap in two senses.
First, socially structured means for goal achievement are available
to only a few in any society. And, second, would it be going too

far to suggest that ultimate wisdom lies in Solomon's "Vanity of
vanities, all is vanity" and that such wisdom is not reached only
in old age?
Throughout this chapter I have emphasized the immanent
character of the stressors in human existence. Both as an Israeli
and as a sociologist, I am fully aware of the differential distribu-
tion of stressors, from historic period to historic period, from culture

to culture, from group to group within a society, and from the
unique life of one individual to that of another. Much of my own
research career has been devoted to precisely this issue, ranging
from early studies of ethnic relations, immigrant adjustment, and
discrimination to work on ethnic differences in adaptation to meno-
pause, concentration camp survivors, and social-class differences in
morbidity and mortality. Vietnam is not Israel is not Sweden, poor
is not rich, black is not white, and male is not female. However,
precisely because the overwhelming attention in stress research has
been on group or individual differences in exposure to stressors,
I have insisted that even the most fortunate of people and groups
know life as stressful to a considerable degree.
I have in this section pursued the thesis that stressors are
ubiquitous in human existence, and I have noted eleven sources

of psychosocial stressors. Since I have not used subheadings, it may
be useful to list these: accidents and the survivors; the untoward
experiences of others in our social networks; the horrors of history
in which we are involved; intrapsychic, unconscious conflicts and

anxieties; the fear of aggression, mutilation, and destruction; the
events of history brought into our living rooms; the changes of the
narrower world in which we live; phase-specific psychosocial crises;
other normative life crises — role entries and exits, inadequate so-
cialization, underload and overload; the inherent conflicts in all
social relations; and the gap between culturally inculcated goals

and socially structured means.
I have set this list of stressors inherent in the human con-
dition against the background of the potential stressors posed by
genetic and microbiological factors. My concern here has not been
to engage in a stringently systematic analysis of psychosocial stres-
sors. Some readers may find inadequate differentiation between one
source and another (accidents are part of the untoward experiences
of others). Others may be troubled by unproven assumptions (un-
derload as a stressor) or by inadequate consideration of certain
types of stressors.Most important, I have by and large disregarded
the historical and sociocultural contexts of stressors. I have done
so intentionally, not because I am unaware of how such contexts
shape and determine the particular expressions of stressors but be-
cause my concern has been to suggest that in all societies, under all
historical circumstances, psychosocial stressors are inherent in hu-

man existence.
I have presented this thesis orally before various audiences
(true, all with university affiliations, but such audiences probably
only accept the thesis with more difficulty). I have asked them, as
I would ask the reader, to test the thesis in the light of their own
lives. I have yet to find anyone who has not known tragedy, sorrow,
challenge, disruption, and conflict as steadfast accompaniments of
living.
Let me make my am fully aware

position completely clear. I
of the joys and and simple satisfactions, and
pleasures, the ecstasies
the sense of achievement, of happiness, and of creation that men
and women and children find in life. I could have, though possibly
with not quite equal cogency, written a paean to history and the
life of the species. One part of the picture need not blind us to the
other. But, it will be recalled, the problem I set out to investigate
was that of salutogenesis. One possible explanation of how people
move toward the ease pole of the breakdown continuum was that
they are subject to a low level of stressors. I trust that the inad-
equacy of this explanation is now obvious. I believe there is ample

evidence for rejecting the significance of the transactional concep-
tualization of stressors, which emphasizes the importance of per-
ception and of appraisal. We all walk a tightrope and are fully
aware of it.
The data of Chapter One perhaps now begin to become
comprehensible. But then the mystery thickens: How do any of us
manage and even manage well? Before we begin to deal with this
question, one further issue must be discussed, if only briefly.
Stressors on the Group Level
Thus far I have discussed stressors as experiences that in-
dividuals confront. The individual has remained the focus of at-

tention even when the same phenomenon is perceived by many
individuals as a stressor and has an impact on the lives of many
members of a group; this is the traditional epidemiological ap-
proach, which deals with the proportion of individuals in a given
population who experience X, in this case, a stressor. But it is no
less important to examine the possibility of applying the same
approach at the group level. To what extent, we now ask, can we
describe social groups —communities, nations, social classes, racial
—
and ethnic groups as characterized by stressors? Or, to put it a
bit more elegandy, as living in stressful situations? It is important
that the meaning be clear, for the same issue will emerge again
when we deal with the resistance resources available to, and the
sense of coherence of, a social group. I am referring to the objec-
tive, sociological situation of the group. How individual members of
the group perceive and react to the situation is a different question
(see Durkheim [1897], 1951).
The most extreme proponent of what is known in sociology
as consensus theory would grant that no social group is ever so
integrated, cohesive, and consensual that it knows no conflicts or
stressors. Nor would that person deny that human history has known
few groups so isolated and living in so benign an environment that

stressors from the external environment are unknown. The most
extreme conflict theorist would agree with the consensus theorist

that no group can last long when it is perpetually bombarded by

the stressors of internal conflict and attack from without.
What differentiates the two schools of thought is not so
much the extent to which a social group is stressor ridden. This is
a matter of empirical investigation (though it is true that, armed

with one model, one tends to find consensus; armed with another
model, one finds conflict). The crucial difference between the two
in viewing the stressors that any social group is in seeing
do exist in
them as primarily exogenic or endogenic. Consensus theorists tend

to see stressors that come from outside the group as accidental
and not inherent in the nature of the relationship of the group to
other groups. If the stressors are internal, they see as the source
inadequate socialization, failures in communication, or, by chance,
as it were, insufficient resources —a condition that leads to scarcity's
becoming a stressor. The stressors, in this view, are not really stres-
sors, or rather they are stressors because we, the members of the
group, see them as such. If we can learn to view things differently,
to talk things through, to give up unrealistic demands, we will not
have stressors.
My own commitment is to a conflict model of social pro-
cesses (meaning not that I like it better but rather that I think
it explains reality more adequately). Over and above the exogenic,

accidental stressors, I suggest, are stressors that are inherent in
the nature of intragroup and intergroup processes. An individual
all
family, a socialmovement, a social class are, by their very natures,

confronted with both internal and external conflict. This is not to
say that the stressors found in a family that is stable are of the
same order of magnitude or of the same content as those found in
a family with multiple problems. Black and white, poor and rich
differ. But the difference is, as I have put it earlier, between a
seriously moderate level of stressors and the level of almost constant

emergency. It is, I contend, therefore absurd to search for the ex-
planation of high group levels of health ease in the direction of
the absence of stressors.
Character of Tension
this chapter, I pointed to what seemed

In the early part of
to be an emerging consensus about the transactional character of
stressors. In 1974 I wrote, "But if anything has been learned in the

study of stressful life events, it is that what is important for their
consequences is the subjective perception of the meaning of the
event rather than its objective character" (Antonovsky, 1974, p.
246). In this chapter, I have argued that this position is accurate
but not important, in large measure because all of us confront what
we define as stressors much of our lives. What, then, is the response
of the organism to a confrontation with stressors? Overwhelmingly,
the literature posits a state of stress. I suggest instead that it is of
the utmost importance to distinguish between two states that I have
proposed (Antonovsky, 1971 ) be called tension and stress.
Selye originally defined the stress syndrome as consisting of

all the nonspecifically induced changes in response to stressors. He
more precisely defined it as the well-known, three-stage general
adaptation syndrome: alarm reaction, resistance, and exhaustion.
Selye was careful to underscore that these were stages in adaptation.
Evocation of the alarm and resistance stages, he suggested, and
even of the stage of exhaustion, provided it is reversible, need not
be damaging to the organism. "In the course of a normal human
life, everybody goes through these first two stages many, many times.
Otherwise we could never become adapted to perform all the

activities and resist all the injuries which are man's lot" (Selye,
1956, p. 64). But, in his most recent clarification, he goes one step
beyond suggesting the functional, positive consequences of the re-
sponse to stressors. "With more recent sophisticated methods it is
show that essentially the same syndrome is also elicited
possible to
by demands for adaptation, experienced as agreeable or beneficial;
these are designated as 'distress' in opposition to 'distress' " (Selye,
1975, p. 39).
Unfortunately, it is not quite clear what is "experienced as
agreeable"; is it the demand —the stressor? Is it the emotion ac-
companying the stages of the adaptation syndrome or the con-

sequences of the stress situation? Nor do I find that the literature
is of much help. Stressors, by and large, are presumed to be bad:

we are unhappy when we encounter them; the physiological and
biochemical changes in response to stressors are accompanied by
unpleasant affect; and the ultimate consequences are sad. Here
and there one finds qualifications. Holmes and Masuda (1974)
insist that the important thing about stressors is the adaptive de-
mand made on the organism, irrespective of whether the life event

is happy or sad. They disregard affect and are concerned solely with
negative consequences. Most often those who are committed to the
transactional approach point out that with adequate preparation,
adequate coping skills, and resources, there is no stress: "Most life
situations are not experienced as threatening because they do not
tax people's capacities." But if we are not prepared to cope, then
we will experience "anxiety, fear, depression, general discomfort"
as a result of having been "exposed to some stress situation" — that
is, "physical, social, and cultural conditions likely to be discomfort-
ing." (Quotes are from Mechanic, 1968, pp. 304, 299, and 297.)
Quite in contradistinction to the overwhelming emphasis in
the field (though not to hints here and there) I suggest that: (1)
The definition of stressor given in the early part of this chapter
carries no implication of any necessary value judgment by the per-
son or group confronting the stressor. (2) The response of the or-
ganism to a stressor —a response I propose calling tension —can be
accompanied by either or both negative and positive affect. (3)
The consequences to the organism of having entered a state of
tension can be negative, neutral, or salutary.
In order to clarify this position, let us take as a possibly
homey example moving one's family residence, particularly when
the move is to a new community. Millions of Americans move each
year. Anyone who has ever moved surely .will agree that, by and
large, it is quite a stressor. Weissman and Paykel (1972) conducted
a most interesting study of the relationship between moving and

depression in women. They note that moving often involves finan-
cial burdens and risks, loss of familiar people and habits and sights
and involvement with new ones, technical problems including
breakage and loss, and energy drain. The stressors are clear, and
there is no doubt that we enter a state of tension that is, the first —
two stages of Selye's general adaptation syndrome and possibly
even the third stage, with all the biochemical, physiological, and
emotional accompaniments. But what are these emotional ac-
companiments? They may well be, as was the case for the women
studied, negative. But they need not be. One may feel excitement,
exhilaration, challenge, relief —not because everything about the
move was smooth sailing, not because one felt no tension, not be-
cause the demands were not defined by one as stressors. Moving

unquestionably taxes people's capacities, in Mechanic's terms.
"The American geographical mobility," Weiss-
stressful effects of
man and Paykel (p. 28) conclude, "have been underestimated.

Moving often places inordinate demands on the individual to adapt
and raises continued challenges to his identity. There are a . . .
substantial number of persons who do experience incapacitating

suffering." But this only after they have noted that "many people
move each year and experience no problems or only transient
ones."
The point, I trust, is clear. Tension is aroused by stressors.
But we can be delighted or pained or, — for that matter, both

simultaneously —by the tension. This point begins to be under-
standable if we consider the frequency of the voluntary search for
stressors. I am not referring to masochistic tendencies, nor am I
concerned with the acceptance of tension as a price to be paid in

order to achieve a goal. Possibly the root is in what many posit as
a fundamental exploratory urge. Others have spoken of the wish for

new experience. The pattern may be related to the earlier-discussed
unpleasantness of underload. Whatever the source, human beings
constantly choose to enter stressful situations — in bed, in football
matches, in risk-taking ventures —not only as a gamble for ultimate
drive reduction but for the sheer pleasure of the tension. (See
Radloff and Helmreich, 1968, for a detailed study of people who
voluntarily enter a situation with a high stressor load, exploration
of the continental shelf.)
The term drive reduction can take us to the next step. I
have argued that tension per se can be pleasurable. But what are
its consequences? It is not only inadequate but misleading to say,
Well, if drive reduction is reasonably quick, if homeostasis is rapidly
regained, or, as I prefer to put it, if tension management is efficient
—no harm done. The consequences of a state of tension can be

not only unharmful but indeed salutary. Perhaps the most forceful
example that has come to my attention is found in what my col-
league Shuval (1957-1958) calls the hardening hypothesis. In her

study of the adjustment of concentration camp survivors to the
rigors of Israeli life in the early 1950s, she found that survivors,
under certain conditions, were better adjusted than those in control
groups. Their point of reference for current difficulties was the

horror they had somehow managed to overcome. The exercise
model is another case in point. We place a special load on our-

selves by demanding more than routine, automatic energy re-
sponse, and we foster the emergence of possibly hitherto unknown
capacities. We learn, and we can apply such learning to future
stressors. The point need not be belabored because it is so familiar
in everyone's experience.
I would like to borrow from another field the term that
seems to me to apply perfectly to the possible salutary consequences
of tension. In a report of some rat experiments, we read, "In sum-
mary, we have shown that environmental stressors not only can de-
press immune responsiveness but can also enhance it. Both sup-
pression and potentiation [are possible]" (Monjan and Collector,
1977, p. 308). Potentiation —the calling up of hitherto potential
resources and thereby enriching one's repertoire —
is precisely a
possible consequence of tension. One might go even a step further

and posit protection as a positive result of tension. The conse-
quences of initial exposure to microorganisms are too well known
even to be cited. As another example, "the development and
toleration of . . . anxiety is not only inevitable but also desirable
both as a stimulus to early infantile development and as an essential
prerequisite for the construction of adequate defenses in all danger

situations" (E. R. Zetzel, quoted in Roazen, 1968, p. 281).
Tension, then, must be distinguished from stress. Stress is a
contributing factor in pathogenesis. Tension can be salutogenic, but
it also can lead to stress. Unless a distinction is made between the
two concepts, this connection cannot be seen. When the distinction
is made, the next crucial question can be asked: What determines
whether a state of tension will be transformed into a state of stress
or will have neutral or salutary consequences? Or, in the frame of

reference of this book: What determines whether a person in a state
of tension will be pushed in one direction or the other on the health
ease/dis-ease continuum? The phrase I have used (Antonovsky,
1971, p. 1580) in dealing with this question is tension management,
"defined as the rapidity and completeness with which problems
are resolved and tension dissipated." Consideration of the possible
salutary consequences of tension, however, suggests that the word
rapidity may be misleading. Sometimes tension may be resolved so

quickly that it prevents full beneficial consequences. Thus, for ex-
ample, premature ejaculation can be most unsatisfying; compulsory
arbitration to avoid a strike may prevent full clarification of the
underlying issues. With this caveat, the crucial question becomes:
What are the determinants of successful tension management? The
answer I have suggested is found in the construct of resistance
resources.
Let me conclude this chapter by quoting from Vickers' dis-
cussion of the mismatch between "symbolic representations of what

ishappening and of what 'ought' to be happening" (Vickers, 1968,
p. 358) "A model of conflict does not necessarily tell us anything
:
at all about pathological stress. Conflict is endemic; breakdown

is still, happily, relatively exceptional. We need not—and therefore
must not —assume that conflict in itself is noxious. . . . We need
to understand both the noxious nature of a given stress in relation
to the organization of a given organism and the vulnerability of a
given organism to a particular form of stress. That, of course, in-
volves an elaboration of the conceptual model far ahead of any-
thing on which I have yet touched." We now turn to consideration
of what I propose as a second major element in such a model.
Chapter Four
Tension Management
and Resources
for Resistance
If bugs are ubiquitous, endogenic in human existence, and ex-

tremely resourceful; if we are all subjected to a constant barrage of
what we ourselves would define as stressors or to those bugs that,
though not in the forefront of our consciousness (such as microor-
ganisms, unconscious conflicts, and strivings or social pressures), we
would quite were we aware of them;
readily agree to call stressors
if levels of tension and imbalance range from moderate but real to
unimaginably high for different individuals and subcultures; if

tension is not at all necessarily pathogenic; and if the data indicate
that a surprising number of us indeed manifest pathology at any
one time —then the crucial question becomes, How do we manage
98
Tension Management and Resources for Resistance 99
tension and prevent it from leading to stress? What are the re-
sources at our disposal that enable some of us or, rather, all of us,
as long as we are alive, to resolve tension at least some of the time?
Meaning of Resistance Resources
In my paper for the 1973 Stressful Life Events Conference

(Antonovsky, 1974), I referred to resistance resources as "the most
exciting things to be studied." I was far from being alone in taking
this position. A pathogenic orientation, however, particularly with-
in the framework of the traditional medical model of the single
bullet, immunological or therapeutic, pressures one to focus on
specific resistance resources relevant to a particular disease. Even
when, under the influence of Selye's concept of the general adapta-
tion syndrome, I formulated the breakdown concept of dis-ease,
I was still tempted to think in terms of particular diseases. Not

until I had developed a salutogenic orientation — at first about
"deviants" such as those concentration camp survivors, poor people,
or members of minorities who do stay at a fairly high level of
health ease and then, in view of the ubiquity of stressors, about
—
everybody did I become aware of the full significance of gen-
eralized resistance resources (GRRs). Only then did I appreciate
what I myself had written (Antonovsky, 1972, p. 541) "Because :
the demands which are made on people are so variegated and in

good part so unpredictable, it seems imperative to focus on de-
veloping a fuller understanding of those generalized resistance re-
sources that can be applied to meet all demands." At the most
a GRR as any characteristic

general, preliminary level, I defined
of the person, the group, or the environment that can facilitate
effective tension management.
This is not to deny the importance of specific resistance

resources. They are many and are often useful in particular situa-
tions of tension. A certain drug, telephone lifelines of suicide-pre-
vention agencies, or an understanding look in the eyes of an
audience to whom one is lecturing can be of great help in coping
with particular stressors. But these are all too often matters of
chance or luck, as well as being helpful only in particular situations.
One can go even further. It is the GRR that determines the extent
to which specific resistance resources are available to us. Thus, if
I may anticipate, being literate or being rich —which I see as a

GRR—opens the way to exploitation of many specific resistance
resources — for example, knowing one's way around a hospital
bureaucracy.
One further preliminary point should be made. Rejection of
the health/ disease dichotomy opens the way for application of the
GRR concept to all people at all times. Whatever one's location
at a given point in time on the health ease/dis-ease continuum,
the extent to which GRRs are available to one plays a decisive role
in determining movement toward the healthy end of the continuum
or, at least, holding one's own.
In my initial exploration of the GRR concept, writing
largely intuitively, I identified three kinds of general resistance re-
sources: adaptability on the physiological, biochemical, psycholog-
and social levels; profound ties to concrete, immediate
ical, cultural,
others; and commitment of and institutionalized ties between the

individual and the total community. The approach made sense to
me. It seemed to fit some data, particularly epidemiological. The
job of research, then, was to move from the intuitive to the syste-
matic consideration of the GRR concept. The remainder of this
chapter is devoted to this enterprise.
Avoidance of Stressors
As a point of departure, we would do well to consider that

type of generalized resistance resource that is essentially "negative"
in character: preventive knowledge, attitudes, and behavior. This
resource is employed in order to avoid — or, more accurately, has
the consequence of avoiding —exposure to stressors. The emphasis
of Chapter Three on the ubiquity of stressors was not at all meant
to suggest that individuals and groups do not differ in exposure to
stressors. One of the reasons for such differences is the differential
distribution of the GRRs that enable one to avoid exposure.
The most directly pertinent GRR of this type is what has
been called a preventive health orientation. Briefly put, Rosenstock
and his colleagues have hypothesized that persons with this orienta-
tion are likely to engage in behavior that, in our terms, helps them
avoid stressors: they are vaccinated against polio; they have pre-
ventive dental and general medical checkups, breast examinations,
Pap smears, and so on (Rosenstock, 1960; for a critical develop-
ment of this approach, see Mechanic, 1968, pp. 130ff., and
Antonovsky and Kats, 1970).
But this GRR is of interest in even more general terms.
Essentially, it is the underlying concept of preventive medicine, of
the health care orientation, of the philosophy so cogently expounded
by McKeown (1976). In part, the approach is oriented to specific
diseases, but this is not necessarily the case. Not smoking, eating
a balanced diet, engaging in physical exercise are essentially to
manage tension, in our terms, by not getting into a state of ten-
sion — thatby avoiding stressors.

is,
Let me be clear that I in no way denigrate this approach

(much as I do not at all denigrate the traditional medical model).
The society or community that adopts the Medizin-Polizei ap-
proach, in the more traditional public health terms in democratic
societies or in the more extreme terms in totalitarian societies, will
undoubtedly chalk up significant gains in the health of its popula-

tion. The evidence of past and present is strong. On the individual
level, there is also no doubt that prevention and early detection pay.
But we are often prey to enthusiasm unwarranted by the data, as
expressed in some material of the cancer societies (see Antonovsky
and Hartman, 1974, for an evaluation of the efficacy of early de-
tection). The cholesterol controversy points to a second problem,

that of arriving at action proposals without adequate knowledge
(Cochrane, 1972a). It often strikes me that converts are a danger
to a good cause. They tend to become true believers, thinking that
X is the total, instant solution to the only problem that matters.

Not only does this single-mindedness alienate others from consider-
ing X; it sometimes inhibits any serious evaluation of the efficacy
of X. A good current example is physical exercise and jogging in
particular. I would predict that the religious enthusiasm now found
will have to confront the double-edged consequences that this X
might have.
A third problem of the preventive approach consists in
ignoring the price paid for adopting a health care, risk-avoidance
action as if health care were the only human value. Consider the
financial cost of many screening tests or health education programs

that utilize scarce resources without adequate evaluation and
weighing of alternative resource allocations. Or consider the price
in anxiety of people whose constant question becomes, "Is it bad
for my health?" Or consider the more serious price of the medicali-
zation of human
existence, professional control, and government
intervention.These caveats notwithstanding, there is no denying
the importance of the preventive health orientation based on the
concept of the avoidance of stressors.
Having said this, one must add that limitation of the GRR
concept to this one area would be most unfortunate. Even in
Samuel Butler's Erewhon the bugs were smarter. People became 31.
Moreover, much of the advice often given to potential victims is
gratuitous and insulting: "Don't moonlight. Take the fanuly on

a relaxing vacation. Don't eat so many starchy foods. Don't walk
in darkened areas where you may be raped." Over and above the
misplaced ideological implications of blaming the victim, the
changes in life style are often forced on us by the social organization
within which we live and in which individual choice plays only a
minor role.
GRRs, then, undoubtedly have the one function of avoiding

stressors. This discussion of avoidance of risk factors, as it is usually
called, is brief only because it has already been considered so much
more adequately than I could possibly do, particularly by Mc-
Keown (1976). What has not, in my view, been adequately dis-
cussed what happens when, as is
is inevitable, stressors are not
avoided and tension is created? What GRRs are available to man-
age tension?
Systematic Consideration of GRRs

Two factors have shaped the following discussion of GRRs.
has been guided, both in organization and in content, by a
First, it
simple mapping-sentence definition of a GRR. The second factor

is even more compelling, namely, my own limited knowledge of
the
extant literature I believe to be relevant. But, as will be seen at
the close of the chapter, it is not particularly important that the
mapping sentence is inadequate or that not every profile has been
considered or that an important reference has been omitted. The

crucial question, as will be seen, will What do these GRRs
become,
have in common? What is it that makes them GRRs? I do hope,
however, that the mapping sentence in Table 6 is of some help at
least for organizing the material.
Table 6. Mapping-Sentence Definition of a Generalized

Resistance Resource
1. physical
2. biochemical
3. artifactual-material
4. cognitive characteristic
A GRRisa
5. emotional / of an
6. valuative-attitudinal
7. interpersonal-relational
8. macrosociocultural
1. individual \
2. primary group / that is ( 1. avoiding | a wide variety
3. subculture I effective in (2. combating
J
of stressors
4. society I
and thus preventing tension from being transformed into stress.
Physical and Biochemical GRRs. I have, perhaps unwisely,

not refrained in earlier chapters from including physical and bio-
chemical agents under the rubric of bugs, or stressors. Having thus
been foolhardy, I believe no further harm can be done by at least
briefly alluding to GRRs at this level.
The concept of immunological surveillance is closely as-
sociated with the work of Burnet (1971, pp. 131, 151ff.). His
particular concern has been with cancer. Thus he writes of "a
surveillance function perpetually patrolling the body, as it were,
he speaks of the "complex mechanisms
for evildoers." Elsewhere
. . . body for any abnormal mutant cells which
[that] police the
by developing toward malignancy threaten survival." But he
seems to go somewhat beyond this in discussing aging, where he
implies an even more general resistance resource: "a fading-out of
maintenance. . . . Nature loses interest after age x." Unfortun-
ately, he does not discuss just what he means by "maintenance" or
"nature." In a more recent paper with the same focus, we read

(Oettgen, 1977, p. 488) : "Agents such as BCG and C. parvum
are known to increase immune responsiveness to a variety of an-
tigens, and they can be counteracted by immunosuppressive meas-
ures. It seems likely that they affect the growth of antigenic tumors
by immune responsiveness and so are
raising the general level of
apdy named immunopotentiators. The determination of how
. . .
immunopotentiators work is a major challenge." Indeed it is. I

mean to imply no magic bullet, but it does seem possible that it
might be useful not only to immunologists to conceptualize im-
munopotentiators as a GRR. It may well be that one of the links
between the GRRs discussed in this chapter and tension manage-
ment is precisely immunopotentiating mechanisms.
Dobzhansky quite some time ago suggested a genetic basis
for human plasticity: "Genotypes are evolved which permit their
possessors to adjust themselves successfully to a certain spectrum of
environments by homeostatic modification of the phenotype." If in-
deed this proved to be the case, it would certainly be legitimate to
regard plasticity as a significant GRR. The existence of a genetically
determined capacity for rolling with the punches or, more formally,
"plasticity of behavior under a wide variety of conditions" cannot,
however, be proved "at this time by observation or experiment"
( 1955, p. 345; see also Lerner, 1968, p. 231 ).
Dubos has, possibly more than anyone else, focused on adapt-

ability as humankind's most distinctive characteristic. "The plas-
ticity of man's nervous system enables him to make a wide range of
behavioral adjustments without having to depend on the slow pro-
cess of biological evolution" (1973, p. 95). This neurological de-
parture from tropismatically shaped, inflexible responses to stimuli
is one of the distinguishing features of the human species. Some
other species may have found far more comfortable ecological
niches or, in present terms, environments that pose far fewer stressors
than those confronting human beings. It may well be that our choice
of environment, our "commitment," was unwise on the part of our
species. Having made it, however, we are dependent on plasticity
as a vital GRR. This is not to say that plasticity of the nervous sys-
tem is a completely clear, objectively definable, and measurable
characteristic. The significance of the concept, much as of other

GRRs discussed here, is that it opens up avenues of research.
This neurological departure leads us directly into the area of
more particular concern and prominence in this book : the psycho-
logical, social, and cultural GRRs and groups
available to people
in coping with stressors. If we
endowed innately with such
are
plasticity, our lives would be most chaotic, Dubos points out, were
our behavior (in the broadest sense) not ordered, constrained, and
guided by culture and society. Much as the bugs that confront us
are no less psychological and sociocultural than physical and bio-
chemical, so the GRRs available to us are in all domains.
But before turning to a discussion of these GRRs, let me ex-
plain what may seem to be a remarkable omission. Selye, one of the
leading figures in stress research, is the originator of the construct of
the three-stage general adaptation syndrome: alarm reaction, resis-
tance, and exhaustion. The term generalized resistance resource

seems to suggest that the second stage is most germane to our discus-
sion. Yet a close examination of Selye's description of this stage
(1956, especially pp. 87fT.) reveals that one must make an im-
portant distinction. The GRR
concept refers primarily to charac-
teristics that facilitate dealing with and overcoming the stressor.
Selye's concern, in his discussion of the resistance stage, is with

actions of the organism that are directed at containing and offset-
ting the expressions of the alarm reaction, in order that the organism
not enter the third stage of exhaustion. Thus resistance, in Selye's
terms, does not relate to coping with the stressor. It is, as it were, a
holding action, designed to prevent damage in the hope that the
somehow go away. Selye's 1975 paper confirms the dis-
stressor will
tinction made here.

The approach of WolfT and his colleagues, who have made
such a major contribution to the stress field, is similarly largely not
pertinent here (Wolf and Goodell, 1968). Again, their concern is
with how the body reacts to a "circumstance of threatening signifi-
cance to the organism," in the sense of the etiology of bodily dis-

eases and disorders. This focus is likewise characteristic of psy-
chosomatic medicine in the tradition of Alexander (1950). To
reiterate Our concern is not with the important question of how the
:
body copes with the biochemical and physiological expressions of

tension but with those GRRs that are applicable to the stressor
direcdy and that function to resolve the problem set by the stressor.
Artifactual-Material GRRs. At the outset of our considera-

tion of psychological, social, and cultural GRRs, I would pay ex-
plicit attention to one GRR. With all our sophistication, we often
overlook the obvious. In extended discussions of coping with stressors,
one seldom finds reference to wealth. A search not only in my own

memory and work but in the indexes and texts of some of the major
publications on stress reveals no reference to material resources as a
GRR. The overwhelming emphasis in the literature is on the stressors
confronted by poor people. Bound by the pathogenic orientation,
we think in these terms.
Access to money, the symbolic equivalent of resources, is, I
suggest, an important GRR in all societies. Money buys a safe abor-

tion to deal with the stressor of an unwanted pregnancy. Its pos-
session may assuage ego conflicts. Not only does money directly
facilitate coping with stressors; but, linked to the acquisition of other
GRRs, it also is indirectly powerful. wealth mat-

In some societies
ters less than in other societies; in some historical and social situations
the possession of money may be a stressor. Whether we like it or not,

however, these are quite unusual exceptions.
Reference to money allows us to consider an issue that,
though seemingly methodological, is a central theoretical issue ap-
plicable to many GRRs. Is there, we might ask, a simple linear

correlation between a GRR
and success in coping with stressors or
is there a qualitative leap beyond a given level? In my work on
social class and overall mortality (Antonovsky, 1967b), I sug-
gested — again, pathogenically oriented — that the data indicated that
the lower-lower class was from other classes.
significandy different
Is it the case that those who are wealthy enough, in J. P. Morgan's
classic phrase, not to have to ask how much a yacht costs have a
GRR that is qualitatively different from the money GRR of the rest
of us?
It is superfluous to extend this discussion of material re-
sources as a GRR. Two further points of clarification, however, are
necessary. First, it is wise to distinguish analytically between mate-
rial resources and interpersonal relations. Material resources are
)
money, physical strength, shelter, clothing, adequate food, and the

like. Interpersonal relations include power, status, availability of
services. Second, in order to assess the significance of a given type of
material GRR, one must know the particular culture. In one cul-
ture or subculture, physical strength may be a potent weapon in
coping with stressors ; in another, money may be more effective.
Cognitive and Emotional GRRs. It would be both unneces-
sary and impossible to canvass the entire psychological literature to
identify important intrapersonal cognitive and emotional GRRs. I
therefore limit the discussion here to two characteristics knowledge- :
intelligence and ego identity. This limitation has the virtue of allow-
ing the discussion to remain on the appropriate level of generality
rather than entering the domain of specific traits and skills. Further,
it covers both the cognitive and emotional domains, implicidy indi-
cating the direction of further research in identifying intrapersonal

GRRs. (I have intentionally excluded from the present discussion
certain conceptualizations that focus on personality, in particular
those of George Engel and his colleagues, of Eleanor Wertheim, and
of Martin Seligman. I have excluded them here because of their
crucial importance, on a higher order of abstraction, to the central
thesis of this book. Thus discussion of these approaches is reserved for
Chapter Five. This point is also applicable to the discussion below

of sociocultural GRRs and the work of Melvin Kohn.
Knowledge-intelligence as a GRR is used here in the broad-
est cognitive sense to encompass both a storehouse of information
about the real world and skills that facilitate acquiring such knowl-
edge. In modern societies, knowledge-intelligence is to a consider-
able extent contingent on literacy and formal education; in more
traditional societies, the skill that best called wisdom is the central
is
component of knowledge. The specific reference here is to a fund

of information — to stick to the above examples, about smoking-
cancer data, techniques of contraception —and the skill that facili-
tates acquiring such information. One might equally well speak of

knowing about Tay-Sachs disease and amniocentesis or, for that
matter, about medicinal herbs. Or, to go one step further, of know-

ing about (and knowing one's way about) community services, op-
portunities, and rights. Of no less importance is the cognitive skill
that enables one to distinguish between quacks or fakers and heal-

era. Elsewhere, I have suggested that one of the reasons one finds a
generally inverse relationship between social class and mortality is
the differential availability of knowledge and cognitive skills (An-
tonovsky, 1968; Antonovsky and Bernstein, 1977). The examples
given are from the health field, but many others might well be given
from other areas. By and large, we find that those individuals and
groups who possess knowledge in one realm are likely to possess
knowledge in other realms of life (Caplovitz, 1963, especially chap.

12; Knupfer, 1947). This overlap of knowledge allows us to speak
of knowledge-intelligence as a GRR.
I have no illusions that knowledge-intelligence is the decisive
GRR in coping with stressors. The inordinate difficulties and, by
and large, the failures of health education based on a cognitive ap-
proach are painfully familiar. Even though I know the epidemio-
logical data well and am of reasonable intelligence, I still smoke. I
have pointed out to many physicians enthusiastic about sex educa-
tion the futility of teaching about reproductive physiology in the
attempt to help adolescents cope with sexual stressors. Knowledge-
intelligence, nonetheless, is seldom a handicap in coping; it often
goes hand in hand with more potent weapons, and occasionally it
may be decisive.
On the emotional level, the central concept is ego identity,
which I regard as a crucial GRR. Again, the literature is vast, and I
have selected a small part for discussion.

Like many other commentators with an essentially patho-
genic orientation, Schachtel ( 1962) starts his discussion by focusing
on the extreme of nonidentity: the number of the concentration
camp inmate or the paper identity of the person in a bureaucratic
social system. Schachtel goes on to speak of someone who lacks
identity as having a sense of being an impostor, of wanting to hide
something important about oneself, consciously or not, or of not
having something one ought to have that would make one a real
person. One travels, as he puts it well, with a false passport. One
yearns for an identity that is real, inner, acceptable, and stable.
Pain leads one to seek things to build up one's ego, to make one feel
good. Schachtel wisely points out that this public image version of
the ego is the "most important model of an alienated concept of
identity." One grabs hold, as it were, of some fixed characteristic,
role, or relationship, insisting that this is the fixed I. The result of
adopting this reified concept of the ego is that one cuts oneself off
from the dynamic, real world. This is true no less for those with a
positive than for those with a negative, alienated identity concept.
Assuming a fixed role, in the present context, makes it impossible
to cope with stressors. Dorian Gray, in Schachtel's final image, is
hopelessly lost; a picture is inevitably impotent.

I have opened the discussion of identity as a GRR with
Schachtel's nonidentity because it so well offsets the essence of
meaningful ego identity. To further explicate, we turn to Erik
Erikson (1960, pp. 47, 49) : "Identity formation neither begins nor
ends with adolescence; it is a lifelong development largely uncon-
scious to the individual and to his society, ... an evolving con-
figuration —a configuration that is gradually established by succes-
sive ego syntheses and resyntheses — . . . integrating constitutional
givens, idiosyncratic libidinal needs, favored capacities, significant
identifications, effective defenses, successful sublimations, and con-
sistent roles" (emphasis in original). It would take us too far afield
to follow Erikson in his profound discussion of the relationship
between ego identity and social roles, group membership, and ide-
ology, and its function as a basis for intimacy, generativity, and in-
tegrity. For our purposes, the central elements of ego identity as a
GRR have been indicated: a sense of the inner person, integrated

and stable, yet dynamic and flexible; related to social and cultural
reality, yet with independence, so that neither narcissism nor being
a template of external reality is needed.
Two further points are indicated. First, in much of the
literature on mental health, a strong ego identity is taken to be the
equivalent of mental health. This, it seems to me, is an expression of
the confusion and difficulty in defining the term. Mental health is a

more complex issue than is ego identity. But even if the two were
accepted as identical rather than as dependent and independent
variables, this identity would be of little help in explaining the loca-
tion of a person on the health ease/dis-ease continuum.
The second issue is central to the thesis of this book. Chapter
Five develops the concept of a sense of coherence. Clear distinction
must be made between this concept and the concept of ego identity.
Ego identity refers to a picture of oneself; the sense of coherence, to
1 10 Health, Stress, and Coping
a picture of one's world, which, of course, includes the self. It may

well be that a strong ego identity will turn out to be a decisive, or
even a necessary, precondition for a strong sense of coherence. I
trust that the distinction between the two concepts will become
clear in Chapter Five. For the time being, I simply take note of my
view that a strong ego identity is a GRR and not the equivalent of
a strong sense of coherence.
Valuative-Attitudinal GRRs. Heretofore, we have focused
on essentially intraindividual factors that characterize individuals
and that they may bring to bear in coping with a variety of stressors.
We now move to consider what have usually been called coping
styles. These too are individual characteristics, but they are best
understood when they are seen, to use Lazarus' term, in a transac-

tional context.
A note of caution is in place. Our cultural preconditioning
tends to prejudice us to assume that what may be called a mastery
coping style or orientation is ever and always the only effective and
respectworthy style of coping with stressors. One question raised
about the Holocaust is, Why did Jews behave like lambs led to the
slaughter? Only those who understand Jewish history are capable of
appreciating that a central component of Jewish culture, developed
over many centuries of harsh experience, was a mode of coping
quite different from that of a mastery orientation. This other mode
was That it may not have
extraordinarily effective for survival.
been appropriate in the face of the unique and utterly incompre-
—
hensible Nazi program of genocide much as no other mode of cop-
ing could have been appropriate— beside the is point. In discussing
cultural GRRs below, I shall return to this issue. For the present, it
is appropriate simply to call the reader's attention to our cul-

tural bias.
Unfortunately, there is little clarity in the concept of coping.
Sometimes confused with the outcome of an interaction with a
it is
stressor. Thus, for example, Shalit (1977) studied the character-

istics of tasks that facilitated or inhibited coping, in the sense of
resolution ofa problem. Rahe (1974, p. 74) takes a similar but
narrower position in defining coping "as one's abilities to reduce
his physiological activation, . . . such as a subject's ability to relax,
and thereby alter his psychophysiologic activation"; he refers to the
successfuloutcome of dealing with a physiologic problem, what-

ever its However, many writers do view coping as a behavior
cause.
pattern of dealing with a problem. Whatever one does, then, be-
comes a coping pattern. Thus Roeske (1978), in discussing how
women deal with hysterectomy, writes of "normally controlled and
usually assumes the leadership" (p. 15) as exemplifying one coping
pattern and using "illness to attract attention and extract sympathy"
(p. 18) as another such pattern. Anything, then, can be said to be
—
coping a view that hardly clarifies the concept.
Lazarus, one of the leading workers in the field of coping
with stressors, has made major contributions to conceptual clarifica-
tion. Let examine the Lazarus-Cohen transactional model
us, then,
of stress (1977), focusing on the second and particularly the third

of its five key features: cognitive appraisal and coping. In Chapter
Three, I noted that cognitive appraisal referred primarily to the
meaning of the stressor for the person. But one aspect of cognitive
appraisal is germane here: "the individual's judgment of the coping
resources available or of vulnerabilities in the face of given dangers"
(p. 111). In describing the coping processes, Lazarus and Cohen
write: "Most persons utilize a variety of coping strategies,
. . .
anticipatingand evaluating what might happen and what has to be

done, planning and preparing, changing the environment, retreat-
ing when necessary, postponing action for maximum effect, toler-
ating frustrationand pain, and even deceiving themselves in order

to feel better and to maintain hope and a sense of self-worth"
(p. 112). Let us overlook the lack of clarity in including "judgment
... of vulnerabilities" in cognitive appraisal and "evaluating what
might happen" in coping strategies. What is of greater concern and
relevance is that Lazarus and Cohen, in their eagerness to reject
seeing coping as a fixed "structural property of a person," throw the
baby out with the bath water by implying that there is a necessary
contradiction between a transactional model of stress and a sys-
tematic analysis of coping modes. Such systematic analysis would be
much more helpful than an endless list of coping modes or actions.
We should, however, heed Lazarus and Cohen's warning
that people use different coping strategies contingently, that is, in
relation to the particular stress situation within which they are
acting. If, however, the thesis of Chapter Three is accurate—that
we all live at a moderate to high level of confrontation with

stressors —then there is some basis for thinking that, given our in-
dividual cultural, historical, structural, and personal-historical back-
grounds, we each tend to work out a typical coping strategy. This

proposal is, I submit, consistent with the transactional approach.
The task before us, then, is to specify the appropriate pat-
terns of alternative coping strategies that do not violate the recogni-
tion that coping strategies are always conducted in a historical-
cultural as well as in a situational context. But the task is even more

difficult. Our entire discussion is based on the premise that we can
identify coping strategies that serve as GRRs. This identification
requires ranking according to effectiveness. Unfortunately, as far as
I know, we can obtain little help from the empirical literature. The
attempt, then, must be viewed as a proposal subject to testing and
not one based on evidence.
Let us define a coping strategy as an overall plan of action
for overcoming stressors. One traditional way of categorizing coping
strategies is the fight-flight-freeze triad. Scott and Howard (1970)
reformulate this category with more sophistication, using the terms
assertive, divergent, and inert responses. For descriptive purposes,
this approach may be useful. But if we wish to identify the coping
strategies that are effective GRRs, such classification is misleading.
Despite our cultural prejudice in favor of fighting (at least for
males), clearly an assertive response is not always the most effective
strategy for coping with stressors.
I would point major variables that enter into every
to three
coping strategy: rationality, and farsightedness. These, I
flexibility,
suggest, are not situation-contingent characteristics. The more a

coping strategy is high on these variables, the more effective a GRR
it will be.
Rationality here is the accurate, objective assessment of the
extent to which a stressor is indeed a threat to one, given who one
is, in the broadest sense. To adopt a transactional approach — to
stress the meaning of a stimulus to the perceiver — is not to deny
objective reality. At one extreme, one's rational definition of a situa-
tion is decisive in determining the outcome. The objective observer
would indeed say that a student who freezes in examinations is be-
ing rational in defining the examination as threatening (assuming
it is important to the student to pass the examination) But in . many

situations a coping strategy is irrational, in that there is a totally
inaccurate assessment of the stimulus and of who one is. The Jew
in Germany in the 1930s who saw himself as immune; the teenage
girl who considers the chances of becoming pregnant as a result of
intercourse as nil; the person who pooh-poohs the data on smoking
and lung cancer are not thereby guaranteed immunity. To put it
simply: The importance of W. I. Thomas' dictum "If men define
situations as real, they are real in their consequences" has not been
fully understood. One's definition of the situation does not neces-
sarily transform the reality, though it may. (Compare Petrie's 1967
discussion of a genetically determined tendency to be an "aug-
menter" or a "reducer.")
Flexibility refers to the availability of contingency plans and
tactics and of a Given the dynamic
willingness to consider them.
character of coping with stressors, the strategy that is open to con-
stant, built-in evaluation and subsequent revision that is open to —
new information — is bound to be more successful, other things be-
ing equal, than other strategies. Note that receipt of new informa-
tion and readiness to change one's course does not mean that one
necessarily will do Nor does it mean that action
so. is perpetually
postponed, pending new information.
The third element of a coping strategy that contributes to
making it a GRR is farsightedness, or, if one wills, being a good
chess player. Farsightedness is linked to rationality and flexibility
but goes beyond them in that it seeks to anticipate the response of

the environment, inner and outer, to the actions envisaged by the
strategy. The most rational and flexible coping strategy for our
student may be to steal a copy of the examination and at the same
time pay a fellow student to cooperate in cheating. But the coping
strategy will be more effective if he anticipates the pangs of con-
science or what happen if he is caught, weighing these as addi-
will
tional data in deciding on behavior (see Mechanic, 1962).
To avoid misunderstanding, I should add that a coping
strategy is a plan for behavior, not the behavior that eventually
results to cope with the stressor. Such behavior isshaped by many
variables, including coping strategies. Further, I would remind the
reader that our concern is not with the substance of a strategy for
coping with a given stressor at a given time. I have posited the pos-
sibility of an enduring, general coping strategy that is characterized
by a high level of rationality, flexibility, and farsightedness. Such a
coping strategy, I suggest, is an important GRR.
It may is a fourth important
well be that emotional affect
characteristic a coping strategy. Lowenthal and Chiriboga
of
( 1973) suggest that some people prototypically
respond to stressors
by feeling overwhelmed, while others are challenged. I doubt that
this is the case for more than a few people, though many might
show a tendency one direction or the other. More important for

in
present purposes, however, I see no basis for arguing that such an
overall tendency is consistently functional. Unlike rationality, flexi-
bility, and one cannot say that the more one tends
farsightedness,
to perceive a stressor as a challenge, the more effective is one likely
to be in coping with the stressor. It puts Don Quixote in mind.
Interpersonal-Relational GRRs. We now come to what may
be the most substantiated and promising field of relevant research.
It has more often than not been called social supports. For a long
time, particularly in the area of mental health research, the patho-
genic orientation led to a focus on social isolates, on those who
might be called social destructs. Only recently has attention been
paid at least to the entire continuum and, here and there, to what,
in the present context, I would call the GRR of deep, immediate
interpersonal roots.
Let me start by a number of
selectively calling attention to
studies that shed light on Gove ( 1973 ) has analyzed

social supports.
cause-specific mortality data and has shown the protective function

that simply being married plays, particularly in relation to those
causes of death where one's psychological state plays a direct or in-
direct role. He further reports, however, that being married is much
more advantageous, in this sense, for men than for women. A paper
by Phillips and Feldman (1973) analyzes the dates of death of
famous people in relation to their dates of birth and of Jews in
pre-World War I Budapest in relation to Yom Kippur. In each case
a significant dip in deaths was found before the ceremonial occa-
sion. The authors, in my view correcdy, hypothesize that the occa-
sion marked a reaffirmation of the social ties of the individual to a
group. Hochschild (1975) takes this issue even further in her the-
)
oretical reanalysis of disengagement theory, suggesting that social

and psychological disengagement is generally much more proxi-
mate to death than we generally think.
Probably the most recent and thorough analysis of the rela-
tionship between social supports and health outcomes is that of
Berkman (1977). In 1965, data had been collected on a random
sample of adults in Alameda County, California. The 682 persons
who died in the following nine years, out of almost 7,000 persons in
the sample, were identified. Included in the original data were four
measures of social ties : marriage, close friends and relatives, church
membership, and informal and formal group associations. Each of
the four factors predicted mortality independently. A social-network
index, combining all four measures, showed that people with many
social contacts had the lowest mortality rates. This association was
found to be controlling for a variety of sociodemographic and risk-
factor measures. Though Berkman's concluding formulation focuses
on "the hypothesis that isolation and lack of social and community
ties may influence host resistance and increase vulnerability to dis-
ease in general" (p. 3), her findings are completely consistent with
a salutogenic formulation; such ties can be viewed as GRRs. One
might go further and postulate that if isolation and lack of ties are
crucial, they can best be understood as a stressor. We would then
anticipate a qualitative leap between the mortality rates of isolates
and the rates of all others. Instead, we find a linear association, which
suggests that social ties are indeed a GRR; it is not only their ab-
sence that matters. (See also Cobb, 1976, for a review of studies of
social supports.
Finally, mention should be made of the Israeli menopause
studies (Datan, Antonovsky, and Maoz, forthcoming), reviewed
in the Introduction to this book. It will be recalled that in designing
the study of ethnic differences in the adaptation to climacterium,
we had anticipated a linear relationship between the modernity-
traditionalism continuum and successful adaptation. We differed on
women (Israeli Arab village women)
whether the most traditional
or the most modern women (Israeli Jewish, middle-class urban
women of central European origin) would be best or worst adapted,
using a wide variety of measures of adaptation. Instead, we found
(to summarize the findings in overall and necessarily oversimplified
terms) a curvilinear relationship. The European women, closely

followed by the Arab women, were the best adapted; the North
African, Turkish, and Persian women adapted much more poorly
in the climacteric period. The most powerful explanation we could
present germane in the present context. The content, styles, roles,
is
statuses, norms, and contexts of the lives of the European and of the
Arab women differed considerably. But what both groups of women

shared, by and large, was being embedded in networks of relatively
stable social relationships and in coherent, relatively integrated
subcultures.
The most adequate concept, I would suggest, that subsumes
the findings of much of the work that deals with interpersonal-rela-
tional GRRs is that of commitment. The pathogenic orientation has
led to a fairly extended consideration of the complex concept of
alienation. A salutogenic orientation leads us to focus on commit-
ment. (The idea that commitment is the quite unexplored opposite
end of an alienation-commitment continuum is proposed in An-
tonovsky and Antonovsky, 1974.)
In a totally different context, Kanter (1968) has given us
what seems to me to be the most adequate theoretical analysis of
the concept of commitment. She defines commitment as "the
process through which individual interests become attached to the
carrying out of socially organized patterns of behavior which are
seen as . . expressing the nature and needs of the person"
.
(p. 500) She identifies three analytically and empirically separable,

.
though interdependent, types of commitment. Continuance commit-

ment is essentially a cognitive judgment that it is worthwhile for one
to remain within a group. Consciously or not, we balance the profits
and losses and arrive at a tentative conclusion. Cohesion commit-
ment refers to a cathectic orientation, the extent to which one feels
affective ties to one's group and to its members. Identification, soli-
darity, and gratification from interpersonal interaction are part of
this dimension. Control commitment is an evaluation along the lines
of good-bad of the legitimacy of the group, the moral lightness of

the group's norms, its ways of doing things, its authority structure,
its goals, and its means for their achievement.
The extent to which one is embedded in social networks to

which one is committed, I suggest, is a crucial GRR. I am fully
aware that it is a complex, multidimensional concept, not at all
easily made must be aware of a crucial

operational. Moreover, one
hidden assumption: If I am deeply committed to a given network
and to its members, does this necessarily mean that the group or
people in it are necessarily reciprocally oriented? In order that
commitment serve as a GRR, such reciprocation is crucial.
I would add one further point here on an issue that has be-
fuddled much of the literature on alienation. In our study of the
kibbutz (Antonovsky and Antonovsky, 1974), we pointed out that
one must distinguish between alienation-commitment with reference
to primary and to secondary groups. There is no necessary contra-
diction between having extremely strong commitments to one's
family, friends, work group, or other primary groups and extreme
alienation from larger social structures such as community, union,
social class, and nation. The two must be considered separately,
which most of the measures of alienation certainly do not do. For
present purposes, however, our concern is, as indicated by the sub-
heading of this section, with those immediate social settings in
which daily life is lived.
Macrosociocultural GRRs. This distinction can serve as a
transition to our consideration of the last type of GRR. My tempta-
tion is and refer the reader to Malinowski's classic
to stop here
article on culture, which appeared in 1931. The article discusses the
fundamental points I would make here more adequately than I

could hope to do. Let me, then, summarize these points as they
relate to salutogenesis and generalized resistance resources on the
macrosociocultural level.
In essence, Malinowski says that culture gives each of us our
place in the world. We are given (or learn to acquire) a language
in which communicate, a role set and a norm set, and a larger
to
world in which to fit (or not fit). In Chapter Three, I defined a
stressor as a demand made on one for which one does not have an
automatic and readily available response capacity. From this point

what culture does, in giving us our place in the world, is to
of view,
an extraordinarily wide range of answers to demands. The
give us
demands and the answers are routinized: from the psychological
point of view, they are internalized; from the sociological point of
view, they are institutionalized.
1 18 Health, Stress, and Coping
But what happens when, as is inevitably the case, life re-
mains full of stressors? Or, to put it another way, what happens

when cultures fail to provide routinized answers, as all cultures do,
inherendy and inevitably, to some extent or other? Does a culture
make generalized resistance resources available to individuals and
groups? This, then, is the present crucial question. As Malinowski
answers "But however much knowledge and science help man in
it:
allowing him to obtain what he wants, they are unable completely

to control chance, to eliminate accidents, to foresee the unexpected
turn of natural events, or to make human handiwork reliable and
adequate to meet all practical requirements. ... In this field . . .
there develops a special type of ritual activities which anthropology

labels collectively as magic. . . . The domain of magic,
richest
however, is, in civilization as in savagery, that of health" (pp. 634-

636).
Magic, however, "is distinguished from religion in that the
latter creates values and attains ends direcdy, whereas magic con-
sists of acts which . . . are effective only as a means to an end.'*
But religion too is, in our terms, a GRR: "Religious belief and
ritual, by making the critical acts and the social contracts of hu-
man life public, traditionally standardized, and subject to super-
natural sanctions, strengthen the bonds of human cohesion. . . .
Religion in its ethics sanctifies human life. ... It grows out of

every culture, because knowledge which gives foresight fails to over-
come fate; because lifelong bonds of cooperation and mutual inter-
est create sentiments, and sentiments rebel against death and disso-
lution" (p. 642).

One might go on in this vein, as Malinowski does, to discuss
play, secular rituals, mythology, and artistic expressions of society as
well as ideologies and philosophies that Malinowski, oriented to
preliterate societies, tends to disregard. But there is no need. The
point, I take it, is clear: At one ideal pole of the continuum, not to
be remotely approximated in human history, a culture provides its
members, group and individual, with ready answers, clear, stable, in-
tegrated; with keening for a death, an explanation for pain, a cere-
mony and a form for disposition and accession of
for crop failure,
leaders. At the other extreme, which at times becomes a reality for
individuals and groups, there is only utter chaos; there are no an-
swers. Ready answers provided by one's culture and its social struc-
ture are probably the most powerful GRR of all.
Absence of a GRR as a Stressor

Before we move on to integrate the concepts discussed in this
chapter, it is important to point out a fundamental characteristic of
GRRs. Throughout, I have considered them as resources; that is,
I have viewed a GRR as something which, in the possession of a

group or individual, makes possible either the avoidance of stressors
or the resolution of tension generated by stressors that have not been
avoided or both. I have seen a GRR, then, as something that one
has, as something that characterizes one.
But there is another way of looking at the matter with
respect to some GRRs. What happens when one is low on a given
GRR? One possibility is that of substitutability. Merton (1968,
pp. 86-91) has warned us of the danger of that variant of func-
tionalism that assumes that if consequence X is a function of phe-
nomenon A, then A must be maintained if one wishes to be X
maintained. Thus, for example, if social cohesion is strengthened by
religious homogeneity, religious homogeneity is viewed as essential
to social cohesion. It would take us too far afield to explore the

question of substitutability of GRRs, but we should keep it in mind.
The implication of the question in the present context will, I trust,
become clear in the pages that follow, in which the relation of
GRRs to the sense of coherence is considered.
But even if we assume that a GRR can be substituted for
another in managing tension, we must note that the absence of some
GRRs can become a stressor. Perhaps the simplest example is that
of money. Although having money obviously does not solve all
problems, it helps with many. But not having money is not simply a
matter of not having a given resource at one's disposal. Being in
such a circumstance often direcdy and immediately is a stressor. Not
only is access to need satisfaction blocked. But also the knowledge
that one is penniless is a source of anguish in and of itself —a situa-
tion that can hardly be appreciated by those who have never been
in such straits.
In Chapter Seven we shall consider the particular problem
.
of circularity in the central model of this book. It will there become

clear that a given location on the breakdown continuum not only is
a consequence of other variables but also serves as a GRR (if one is

high on the scale) or as a stressor (if one is low)
Keeping these examples in mind, I would ask the reader to
reconsider the absence of GRRs discussed in this chapter as possible
stressors. In keeping with my salutogenic orientation, I have not
raised this issue until the present. It would, however, have been
misleading to have totally neglected its consideration.
GRRs as Providers of Negative Entropy
Consideration of the absence of a GRR as a stressor opens

the way for further development of the phenomenon under review
in this chapter. In Chapter Three, I suggested that the ubiquity of
stressors in human existence is analogous to the Second Law of
Thermodynamics; it points to an inevitable increase of positive
entropy, or disorder. But unlike the closed systems with which the
law deals, the human organism is an open system. The input of
negative entropy may well be adequate, at any given point in time,
or even over a longer period of time, to offset the inherent ten-
dency to disorder. It strikes me that it may be extremely useful to
view the period of early childhood as one in which the input of
negative entropy from the external environment or, in Schrod-
inger's (1968) vivid image, particularly apt here, the "sucking of
orderliness from [the] environment" is (or, rather, can be) suffi-
cient to lower the level of positive entropy, of disorder, in the child.
Is this not the essence of growth — the increasing orderliness of
language, of muscle control, of interpersonal relations, and so on?
In these terms, is it not appropriate to see stressors, originat-
ing in either the internal or external environment, as entropic, in-

creasing the level of disorder of the system? And to see GRRs, again
whether originating in the internal or external environment, as neg-
atively entropic, decreasing the level of disorder of the system?
If GRRs can be defined as characteristics that introduce
negative entropy into the system, the mapping sentence (Table 6)
needs to be modified. It is indeed quite useful in providing a sys-
tematic basis for considering a wide range of phenomena as possible

GRRs. from thinking of ego identity

It prevents the psychologist
and ignoring all else; it reminds the sociologist and anthropologist
that they tend to forget genetics and immunology, and the biologi-
cal scientists that the social scientists have a contribution to make.
But, in the last analysis, inclusion of a generalized resistance resource
for discussion in this chapter was intuitive, or if data suggested a
correlation between it and tension management, inclusion was em-
pirical. The mapping sentence is not yet a guide for determining
whether a given phenomenon or characteristic is a GRR.
To put the matter another way, the mapping sentence does
not provide a culling rule. It is essentially tautological, in that it
states the criterion for inclusion as a GRR to be an empirical corre-

lation with tension management. This makes it impossible
criterion
to test a hypothesis that a GRR facilitates tension management. It

does not tell us why a phenomenon is effective in combating a wide
variety of stressors.
We can resolve this problem by modifying the mapping sen-
tence within the framework of the concept of negative entropy.
Thus, a generalized resistance resource is a physical, biochemical,
and so forth (see Table 6) characteristic, phenomenon, or relation-
ship of an individual, primary group, subculture, society that pro-
vides extended and continued experience in making sense of the
countless stimuli with which one is constandy bombarded and facili-
tates the perception that the stimuli one transmits are being received
by the intended recipients without distortion. Making sense, in in-
formation theory, refers to the provision of information rather than
noise. But, following Thomas, I would also include music. As
Thomas ( 1974, pp. 22-25) writes: "It is one of our problems that
as we become crowded together, the sounds we make to each other
. .become more random-sounding, accidental, or incidental, and
.
we have trouble selecting meaningful signals out of the noise. . . .
We are only saved by music from being overwhelmed by nonsense.

. . . The need to make music, and to listen to it, is universally ex-
pressed by human beings."
This definition of a GRR, moreover, goes beyond providing
a criterion for identifying GRRs. By providing a unifying theme, an
overarching concept, it makes possible a resolution of the incredible
complexity of stress research. In a review of this field from an inter-

actionist perspective, Chan (1977)a wide variety of intraper-

cites
sonal, interpersonal, and situational variables that, in our terms,

would seem to be related to tension management. No degree of
methodological and statistical sophistication, I suggest, will resolve
the problem of which Chan (p. 100) writes: "disentangling the

many intriguing complexities arising from the person-situation inter-
action." Only a more adequate theoretical formulation can do so.
Significance of GRRs
We can now, I propose, take the decisive step in formulating

the significance of GRRs. The extent to which our lives provide us
with GRRs is a major determinant of the extent to which we come
to have a generalized, pervasive orientation that I call a strong
sense of coherence.
In a paper that speculates about the possible exploitation of
information theory in analyzing biological situations, Rapoport
(1968) discusses Maxwell's Demon, whom he familiarly calls
Maxie. Maxie is "a being whose faculties are so sharpened that he
can follow every molecule in his course." He "works with 100 per-
cent efficiency pumping information [negative entropy] into a
system." A person with a sense of coherence, I suggest, is like Maxie,
though it is inconceivable that any human being or social group can
ever remotely approximate the theoretical efficiency of this im-
probable imp. The extent to which this approximation is realized,
however, determines the efficiency with which tension is managed.
Chapter Five
Perceiving the World

as Coherent
The time has now come for a formal definition of the concept that
and explicit, and that is
integrates all the foregoing details, implicit
a crucial variable in explaining movement on the health ease/ dis-
ease continuum. The sense of coherence
is a global orientation that
expresses the extent towhich one has a pervasive, enduring though

dynamic feeling of confidence that one's internal and external en-
vironments are predictable and that there is a high probability that
things will work out as well as can reasonably be expected. Almost
invariably when I have presented this concept, people have had an
immediate, intuitive response of "I know very well what you're
talking about." In the particular cultural context in which the con-
cept has been presented, it is often taken to mean "I am in control"
and associated with the concept of an internal locus of control. This
123
cultural bias is discussed in detail below. Once this misinterpretation

is clarified, I have found that the concept makes important sense to
people.
One of the reasons it does is that people can identify in their
personal experience —both in their own lives and in their reading of
literature — individuals with a strong sense of coherence as well as
individuals with a markedly weak sense of coherence. Thomas
Mann's vivid portrayal of Mynheer Peeperkorn in The Magic
Mountain illustrates a strong sense of coherence. In The Brothers
Karamazov, Fedor Dostoevsky has given us a masterly contrast of
the two extremes Ivan and Dmitri Karamazov, each of whose sense
:
of coherence is, in its own way, minimal compared with that of

Alyosha. Perhaps the supreme example in literature, particularly
appropriate because one individual moves from one extreme to the
other, is that of Job.
Intuitive approval of a concept that is intended as an ex-
planatory and hypothesis-generating tool, however, is only a start-
ing point. It is crucial to analyze all the facets of the definition and
to clarify possible misunderstanding and the precise implications of
each of the terms.
Analysis of the Definition
As defined, the sense of coherence' explicitly

and unequivo-
cally isa generalized, long-lasting way of seeing the world and one's
life in it. It is perceptual, with both cognitive and affective com-
ponents. Its referent is not this or that area of life, this or that
problem or situation, this or that time, or, in our terms, this or that
stressor. It is, I suggest, a crucial element in the basic personality
structure of an individual and in the ambiance of a subculture,
culture, or historical period.
This does not mean that there are no ups and downs. A
particular experience, a specific situation, a detailed success or fail-
ure can effect a temporary and minor shift in one's sense of coher-
ence. ( In fact, being impervious to particular changes in one's en-

vironment is one indication of what I subsequently call a fake sense
of coherence. ) But such changes occur around a stable location on
the continuum.
Perceiving the World as Coherent 125
The term dynamic in the definition takes us much beyond

such minor fluctuations. meaning will become fully clear, I
Its
hope, in the discussion below of the sources of the sense of co-

herence. I certainly am not committed to understanding the sense of
coherence as being determined forever and anon by genes or early
childhood experience. It is shaped and tested, reinforced and modi-
fied not only in childhood but throughout one's life.
To take an example. The pathogenic orientation, particu-

larly in the hands of clinicians, clarifies the dynamic pattern of a
deteriorating sense of coherence. Thus the psychotherapist seeks to
point out to the neurotic patient how he or she continually gets in-
volved in endeavors that by definition are doomed to failure; the
psychotherapist then focuses on the sources of this neurotic pattern.
He or she may even note how such continued experiences weaken
the sense of coherence. A salutogenic orientation, however, can lead
to working with the patient to engage in goal-oriented behavior that
promises success, thereby strengthening the sense of coherence.
Similarly, a radical change in one's structural situation — in
marital status, occupation, place of residence —can lead to a signifi-
cant modification in one's sense of coherence. But in addition one's

sense of coherence, strong or weak, plays a significant role in deter-
mining one's choice of remaining in or changing one's structural
situation. Thus, to stick to the above example, the neurotic, with a
weak sense of coherence, may choose to avoid the "danger" of
entering psychotherapy. In this sense, one tends to choose situations
that reinforce the level of one's sense of coherence.
The same dynamic approach can be applied to different
areas of life. If experience in one area tends to weaken while ex-
perience in another area tends to strengthen one's sense of coher-
ence, the person with a stronger sense will seek to change the area
that weakens it, while the person with a weaker sense will gravitate
toward that area. In this way, there is a constant albeit dynamic
tendency toward consistency and generalization, stability and
continuity.
Stability and continuity bring us to the crux of the matter.
A strong sense of coherence involves a perception of one's environ-
ments, inner and outer, as predictable and comprehensible. In the
image that has been and will be used often, it means that the
stimuli that impinge on one are meaningful, as information or as

music. Even more important than the immediate response, however,
is the overall expectation that stimuli will continue to be meaning-
ful. This is what one's world is seen to be like. Receiving stimuli is
only one side of one's transactional relationship with one's world,

however. No less important is the confidence that the stimuli one sends
will be received without undue distortion. Thus, one's world has
form and structure, is choate and comprehensible. We should take
note of one frequent situation. We receive stimuli that are perceived
as noise. The person with a strong sense of coherence locates the
trouble outside himself; the stimuli are nonsensical. But in the
case of the stranger or the migrant, people all around speak a lan-
guage that makes no sense, but one knows very well that the im-
pediment is in oneself.
If one understands what is going on, however, and if the
world is seen as predictable, outcomes may still not fulfill needs.
And a person with a weaker sense of coherence will indeed tend to
anticipate that things are likely to go wrong. When things make no
sense and are not predictable, it is difficult to expect that needs
will be fulfilled, except by sheer luck or blind chance. One can
clutch at straws; one can engage in privatized (not culturally rout-
inized and ritualized) magic. But one remains without much hope.
The person with a stronger sense of coherence is quite able to see
reality, to judge the likelihood of desirable outcomes in view of the
countervailing forces operative in all of life. One is not blinded by
confidence. It is in this context that the phrase as well as can rea-
sonably be expected has been included in the definition of the con-
cept. Malinowski's (1931) Trobriander fishermen, even after
having engaged in all the proper rituals prior to fishing in the open
sea, know full well that the rituals are no guarantee against drown-
ing or a poor catch.
To put the matter another way: A strong sense of coher-
ence is not at all equivalent to feeling that everything in life is
handed to one on a silver platter or that one has the Midas touch.
Quite the contrary may even be true. Life may well be seen as full
of complexities, conflicts, and complications —which one under-
stands. Goal achievement may be seen as contingent on immense
investment of effort. Moreover, one may be fully aware that life
involves failure and frustration. The important thing is that one

has a sense of confidence, of faith, that, by and large, things will
work out well. Not that things will have a Hollywood happy end-
ing. This is why the proviso "as can reasonably be expected" is
added. A strong sense of coherence includes a solid capacity to

judge reality.
What makes frustration, failure, and pain tolerable without

vitiating a strong sense of coherence is, to introduce another crucial
term, the perception of lawfulness. Job was shattered not because
of his terrible suffering, not because almost all that he had was
taken from him. When fate is capricious, when events are arbitrary,
when there is no lawfulness —and not at all when there is no
omnipotence —the sense of coherence is shattered. In this lies the
terrible brilliance of modern totalitarianism, foreshadowed in Franz
Kafka's Castle. One never knows when the doorbell will ring or for
what reason.
The ideas of predictability and lawfulness may suggest too
great an emphasis on the cognitive aspect of the sense of coherence.
A belief in lawfulness does not necessitate intellectual understanding
of the logic of the laws. An orthodox Jew even regards such an
attempt as apostasy —a view that led to the excommunication of
Spinoza. The intellectual task, as well as the emotional aspiration,
is know God's laws; the behavioral task is to obey them. The
to
rank-and-file party member need not understand the laws of his-
tory; this is the realm of the leadership. It is enough to maintain
one's faith in God or in the party in order to maintain a strong
sense of coherence. Such faith makes everything comprehensible, at
least affectively. It is when faith collapses that the sense of coher-
ence dissipates. The Holocaust, violating all previously known law-
fulness, thus posed a most torturing problem for believing Jews.
An approximation of their problem is found in Crossman (1950),
in which six well-known ex-Communists analyze their break with
the party.
We can now take the final step in clarifying the definition
of sense of coherence. It is of the utmost importance that I did not
choose the more familiar phrase sense of control, which clearly
implies and is overwhelmingly used as meaning "I am in control."
This conceptualization reflects a superfluous cultural bias, an
issue discussed below. A sense of coherence, as I trust has become

clear, does not at all imply that one is in control. It does involve
one as a participant in the processes shaping one's destiny as well
as one's daily experience. The orthodox Jew, striving with all his
might to obey the 613 commandments, is not at all passive. The
Calvinist on whom signs of grace have been bestowed in response
to his utmost effort has been extremely active. But this does not
mean that it is they who are decisive in the outcome.
The crucial issue is not whether power to determine such
outcomes lies in our own hands or elsewhere. What is important is
that the location of power is where it is legitimately supposed to

be. This may be within oneself; it may be in the hands of the head
of the family, patriarchs, leaders, formal authorities, the party,
history, or a deity. The element of legitimacy assures one that
issues will, in the long run, be resolved by such authority in one's
own interests. Thus a strong sense of coherence is not at all en-
dangered by not being in control oneself.
Case Histories
In order to convey more adequately what is meant by the

sense of coherence, I have chosen three case histories. The fact
that two of them relate direcdy to health is not accidental. But
since this issue is explored in detail in Chapters Six and Seven, it is
not the primary concern here. We start with Norman Cousins

1
(1976).
In August, 1964, I flew home from a trip abroad with a

slight fever. I was hospitalized
. . . [and had] tests, some of . . .
which seemed to me to be more an assertion of the clinical capa-

bility of the hospital. . . .I turned them [lab technicians] away
and had a sign posted on my door saying that I would give just one
[blood] specimen every three days.
I had a fast-growing conviction that a hospital was no place
for a person who was seriously ill [because of bad] . . . sanitation,
. . sometimes promiscuous use of x-ray equipment, the seemingly
.
indiscriminate administration of tranquilizers and powerful pain
1 Reprinted, with
Norman Cousins' kind consent, by permission from
The New England Journal of Medicine, Vol. 295, pp. 145&-1463, 1976.
killers, more for the convenience of hospital staff, . . . the reg-

ularity withwhich hospital routine takes precedence over the rest
requirements of the patient. . Perhaps the hospitaTs most . .
serious failure was in the area of nutrition.

My doctor . . .was able to put himself in the position of
the patient. . . . We had been close friends for more than twenty
years. . . . We had often discussed articles in the medical press.
. . . He felt comfortable about being candid with me. . . . Dr.
Hitzig called in experts. . . . He leveled with me, admitting that
one of the specialists had told him I had one chance in 500.
All this gave me a great deal to think about. Up to that
time, Ihad been more or less disposed to let the doctors worry
about my condition. But now I felt a compulsion to get into the
act. I had better be something more than a passive char-
. . .
acter. . . . I thought as hard as I could about the sequence of

events immediately preceding the illness. I wondered whether . . .
the exposure to the hydrocarbons, . . . individual allergy, . . . a

condition of adrenal exhaustion [when] . . . I was less able to
tolerate a toxic experience [could be the cause]. . . . Our last
evening in Moscow had been . . . an exercise in almost total

frustration.
Assuming this hypothesis was true [exposure to diesel and
a time of adrenal exhaustion], I had
jet pollutants at . to re- . .
store what Walter Cannon, in his famous book The Wisdom of the
Body, called homeostasis. . . . I remembered having read . . .
Hans Selye's classic book, The Stress of Life. . . . If negative emo-

tions produce negative chemical changes in the body, wouldn't the
positive emotions produce positive chemical changes? . . . Even
a reasonable degree of control over my emotions might have a
salutary physiologic effect.
. A plan began to form in my mind . .
for systematic pursuit of the salutary emotions. . . .Two precondi-

tions for the experiment: . . . medication [could not be] toxic to
any degree . . . [and] I would have to find a place somewhat
more conducive to a positive outlook on life.
With Dr. Hitzig's support, we took allergy tests and discov-
ered that I was hypersensitive to virtually all the medication I was
receiving. It was unreasonable to expect positive chemical
. . .
changes to take place so long as my body was being saturated with,

and toxified by, pain-killing medications. I had one of my research

assistants at the Saturday Review look up the pertinent references
in the medical journals. . . . The history of medicine is replete
with instances involving drugs and modes of treatment that were

in use for many years before it was recognized that they did more
harm than good. Living in the second half of the twentieth
. . .
century, I realized, confers no automatic protection against unwise

or even dangerous drugs and methods.
Pain is part of the body's magic. . . . I could stand pain
so long as I knew that progress was being made in meeting the
basic need . . . [of] the body's capacity to halt the continuing
breakdown of connective tissue. . . . I recalled having read in the
medical journals about the usefulness of ascorbic acid. . . .
Couldn't it also combat inflammation?

I wanted to discuss some of these ruminations with Dr. Hit-
zig. He listened carefully. . . . He said that what was most im-
portant was that I continue to believe in everything I had said. He
shared my sense of excitement . . . and liked the idea of a partner-
ship.
Even before we had completed arrangements for moving out
of the hospital, we began the part of the program calling for the full
exercise of the affirmative emotions as a factor in enhancing the
body chemistry. . . . A good place to begin, I thought, was with
amusing movies. . . . The nurse was instructed in its [the movie
projector's] use. . . . So we took sedimentation-rate readings just
before as well as several hours after the laughter episodes.
The arrangements were now complete for me to move my
act to a hotel room. The sense of serenity was delicious. . . . . . .
/ found him [Dr. Hitzig] completely open minded on the subject [of
ascorbic acid]. . . . It seemed to me that, on balance, the risk
was worth taking, know whether we were on the right track.

. . . to
There was no doubt in my mind that I was going to make it
back all the way. Two weeks later, my wife took me to Puerto Rico
for some sustained sunshine. . . . I must not make it appear that
all my infirmities disappeared overnight. . . . But I was back at
my job at Saturday Review full time again. . . . Is the recovery a
total one? Year by year the mobility has improved. . . . [I hit a]
tennis ball or golf ball, . . . ride a horse, . . . play the Toccata

and Fugue in D Minor. . . . My neck has a full turning radius

again.
What conclusions: . . . The will to live is not a theoretical
abstraction, but a physiologic reality. . . . I was incredibly fortu-
nate to have as my doctor a man who knew that his biggest job was
to encourage to the fullest the patient's will to live and to mobilize
all the natural resources of bodyand mind to combat disease. Dr.
Hitzig was willing to set aside the large and often hazardous arma-
mentarium .when he became convinced that his patient might
. .
have something better to offer. . . . The principal contribution

made by my doctor to the taming, and possibly the conquest, of my
illness was that he encouraged me to believe I was a respected part-
ner with him in the total undertaking.
Even a quick reading reveals five striking characteristics of

the case. First, Cousins was quite ready to be a sharp critic of the
hospital: the efficiency of its organization; its level of adequacy in
sanitation, nutrition, and use of equipment; its concern for the
patient. Second, a remarkable patient-doctor relationship is re-
vealed. They were long-time friends. The doctor served as liaison to

was honest, supported the patient's active orientation,
specialists,
and "shared my sense of excitement and liked the idea of a . . .
partnership." Third, and perhaps most central, we note the defini-

tion of Cousins' own role, with an unusual set of norms for a pa-
tient : get into the act, assume responsibility for therapeutic decision
making, take initiative, participate in assembling data and review-
ing literature, formulate hypotheses. "We took sedimentation-rate
readings." Fourth, we see that Cousins' nonpatient roles are rele-
vant. He was familiar with medical literature. He had a staff to do
research for him. He could afford to be "hospitalized" in a hotel.
And, finally, we read of his attitude toward pain and the idea that
"the risk was worth taking."
His experience, I submit, is a marvelous example not so
much, as Cousins puts it, of a will to five, but of a total behavior
pattern shaped by an extremely strong sense of coherence, rooted
in a concrete social structure and cultural setting. More important,
this general orientation quite evidently characterized Cousins before
his unfortunate trip to the Soviet Union. Note that I do not claim
that his sense of coherence caused the happy outcome, just as it did
not prevent the onset of his illness.
Let us look next at the case of Sigmund Freud. Max Schur

(1972), Freud's personal physician from 1928 until Freud's death
in 1939, wrote:"One of the aims of this study is to trace the grad-
ual development of different responses used by Freud to prevent
situations of great stress and danger from becoming traumatic and
to discover the ways Freud utilized to achieve mastery without
resorting to denial" (p. 40). Schur's work is indeed of the utmost
relevance for us, particularly his discussion of the most critical and
difficult—and least healthy—years ( 1892-1899) of the first half of
Freud's life and 3). On the one hand, relative poverty
(chaps. 2
and major responsibilities, a near total break at about age thirty-six
with the professional world around him and his own past, a publicly
known plunge into the most chaotic (and, to Vienna, most repel-
lent) of worlds; on the other hand, his wife Martha's love and the
friendship of Fliess, and an extremely strong sense of coherence.
Unlike the article by Cousins, Schur's work contains no brief
selection that is appropriate here. I have, therefore, chosen an alter-
native, which comes close to the mark. In his Introduction to the
abridged edition of Jones' Freud ( 1961 ), Lionel Trilling writes :*
There is yet a third reason for the interest that Freud's life
has for us — . . . the style and form of the life itself, . . . the
consonance that we perceive between Freud's life and his work. The
work is and ordered and courageous and magnanimous in

large
intention; and of the life we can say nothing less.
Overtly and without apology, Freud hoped to be a genius.
. . The commitment to achievement of both his family and his
.
culture was reinforced by the ethical style which a traditional edu-

cation proposed. . . . A heroic English Puritanism joined with the
ancient ideal of public virtue to confirm the necessarily more private
but no less rigorous notion of how a life must be lived: with stern-
ness, fortitude, and honor.
Excerpted from the Introduction, by Lionel Trilling, to The Life
*
and Work of Sigmund Freud, by Emest Jones, edited and abridged by

Lionel Trilling and Steven Marcus, ©
1961 by Basic Books, Inc., Publishers,
New York. Abridged from The Life and Work of Sigmund Freud, 3 volumes,
© 1953, 1955, 1957 by Ernest Jones. My appreciation is extended to Diana
Trilling and to Basic Books for their consent.
— —
Freud's intellectual achievement must be thought of as a

moral achievement. I have two things in mind saying this. One has
reference to the courage of a man in middle life, with family respon-
sibilities and a thoroughly conventional notion of how these must be
met, who risked his career for the sake of a theory that was ana-
thema to the leaders of his profession. . . .
The other thing I would imply by speaking of the moral

nature of Freud's achievement is suggested by Freud's own sense of
his intellectual endowment. With this he was never satisfied. . . .
*/ am not really a man of science, not an observer, not an experi-

menter, and not a thinker. I am nothing but by temperament a
conquistador an adventurer . . . with the curiosity, the boldness,
and the tenacity." . . .
Freud's last years were his darkest. Despite the high demand
he made upon life, despite his notable powers of enjoyment, he had
long regarded the human condition with a wry irony; and now by
a series of events the cruel and irrational nature of human existence
was borne in upon him with a new and terrible force: the defec-
tions of two of his most valued collaborators [and] the shadow . . .
of death —
[friend and patron], his beautiful daughter Sophie, . . .
. . Sophie's son Heinz [for whom Freud had a special love].

. . . .
In 1923 he learned that he had cancer of the jaw. [He had] thirty-
three operations. For sixteen years he was to live in pain.
. . . . . .
The prosthesis he wore was awkward and painful, distorting his face
and speech. He had, of course, no religious faith to help him
. . .
confront the gratuitousness of his suffering. Nor did he have any

tincture of "philosophy." He is as stubborn as Job in refusing to
take comfort from words —even more stubborn, for he will not
permit himself the gratification of accusing. The fact is as it is.
Human life is a grim, irrational, humiliating business nothing —
softens this judgment. . . . Yet nothing breaks him and nothing
really diminishes him. He often says that he is diminished, but he is
not. He frequently speaks of his indifference, but the work goes on.
Civilization and Its Discontents . . . appears when he is seventy-
three. At his death at eighty-three he is writing his Outline of Psycho-
analysis. He sees patients up to a month before he dies. . . .
This heroic egoism is surely . . . the secret of his moral

being. "Mit welchem Recht?" "By what right?" he cries, his eyes
blazing, when he is told in his last days that when the diagnosis of
cancer was made, there had been some thought among his
first
friends of concealing the truth from him. Through all his . . .
years of very great pain —

near the end he spoke of his world as be-
ing "a little island of pain floating in a sea of indifference" —he took
no analgesic drug and only at the last did he consent to take aspirin.
He said he preferred to think in torment to not being able to think
clearly. Only when he felt sure that he had outlived himself did he
ask for the sedative by the help of which he passed from sleep into
death.
All comment would be superfluous.
The sense of coherence, as I have suggested, is a concept

that is applicable to groups as well as to individuals. One of the most
striking illustrations is found in Brinton's
1965 ) analysis of similari- (
ties that characterized the prerevolutionary periods in England,

the United States, France, and Russia. Brinton's portrayal of the
ruling class on the eve of revolution (pp. 53-59) in contrast with
the ruling class that rules and believes in its rule catches the essence
3
of a weak and of a strong sense of coherence, respectively.
What may be called the ruling class seems in all four of our
societies to be divided and inept. . . . [Earlier, it was they] who
seemed to lead dramatic lives, about whom the more exciting scan-
dals arose, who set the fashion, who had wealth, position, or at least
reputation, who, in short, ruled. . . . It seems likely that the great
masses of poor and middling folk . . . really accept the leadership
of those at the top . . . and dream rather of joining them than of

dislodging them.
[Just before the revolutions] the ruling classes in our societies
seem . . . to have been unsuccessful in fulfilling their functions.
. . . A
mixture of the military virtues, of respect for established
ways of thinking and behaving, and of willingness to compromise,
and, if necessary, to innovate is probably an adequate rough ap-
proximation of the qualities of a successful ruling class.
When numerous and influential members of such a class

begin to believe that they hold power unjustly or that all men are
• Reprinted with permission from the book The Anatomy of Revolu-
tion by Crane Brinton, ©

1965 by Prentice-Hall, Inc. Published by Prentice-
Hall, Inc., Englewood Cliffs, N.J. 07632.
brothers, equal in the eyes of eternal justice, or that the beliefs they
were brought up on are silly or that "after us the deluge/' they are
not likely to resist successfully any serious attacks. . . . [Their]
decadence is not necessarily a "moral" decadence. . . . The vir-
tuous Lafayette was a much clearer sign of the unfitness of the

French aristocracy to rule than were Pompadour or even DuBarry.
The Russians here provide us with a locus classicus. . . .
Russian aristocrats for decades before 1917 had been in the habit of
bemoaning the futility of life, the backwardness of Russia, the Slavic
sorrows of their conditions. . . . [They had] an uneasy feeling that
their privileges would not last. Many of them, like Tolstoy, went
over to the other side. Others turned liberal and began that process
of granting concessions here and withdrawing them there. . . .
When who had positions of politi-

those of the ruling classes
cal power did use force, they used it sporadically and inefficiently.
. . . [They were] more than half ashamed to use force and there-
fore used it badly.
Perhaps nowhere better than in France is to be seen one of
the concomitants of the kind of disintegration of the ruling class we
have been discussing. This is the deliberate espousal by members of
the ruling class of the cause of discontented or repressed classes —
upperdogs voluntarily sitting with underdogs. . . . It is necessary
to point out that the existence of rebellious radicals in the upper
classes is only one symptom. . . . [They] must be relatively nu-
merous as well as conspicuous in a society in disequilibrium. They,
and the wasters and the cynics, must set the tone for the class —
. .the same mixture of weariness, doubt, humanitarian hopes,
.
and irresponsibility.
Except perhaps in America, we find the ruling classes in the
old regimes markedly divided, markedly unsuited to fulfill the func-
tions of a ruling class. Some have joined the intellectuals and de-
serted the established order; . . . others have turned rebels, less
because of hope for the future than because of boredom with the
present; others have gone soft or indifferent or cynical. Many, pos-
sibly even most, . . . retained the simple faith in themselves and
their position which is apparently necessary to a ruling class. But the
tone of life in the upper classes was not set by such as these. The
sober virtues, the whole complex series of value judgments which
guards a privileged class from itself and others, all these were out of
fashion.
From Brinton's analysis of the disintegration of the ruling

classes in prerevolutionary periods, we can reasonably draw four
major inferences that can be applied in our context. First, though
the point is most clearly seen in considering a ruling class, a strong
sense of coherence can characterize any social unit, from the Jones
family to a neighborhood, a city,a region, or a country; from a
local voluntary association to an apocalyptic religious movement;
from underdogs to overdogs. Second, it is not the number of in-
dividuals in the group with a strong sense of coherence that matters
but, as Brinton puts it, who sets the
tone. Concomitantly, in a group
with a strong sense of coherence, defection of individual members is
viewed as random, accidental, and a sign of the weakness of the
defectors. When defection is viewed as inevitable, continuous, and
with sadness, it is a sign of a disintegrating sense of coherence.
Third, whatever the particular culture of the group with a strong
sense of coherence, there is a firm commitment to a set of goals
(hazy as that commitment might be) and a set of means to achieve
these goals —means that are held to be not only legitimate but the
embodiment of morality. Finally, the future is perceived as inevita-
ble: stable continuity into infinity for the overdog; unavoidable
subservience for the underdog.
Sources of the Sense of Coherence
In Chapter Four, I proposed that "the extent to which our

lives provide us with GRRs is a major determinant of the extent to
which we come to have a generalized, pervasive orientation that I
call a strong sense of coherence." The GRRs discussed in that chap-
ter were considered in essentially an intuitive albeit ordered manner.
Not until the end of the chapter was the inductive question asked,
What is it that all these GRRs have in common? Only then could
the linkage be proposed between GRRs and the sense of coherence.
At this point, having defined and clarified the concept of
the sense of coherence, we can take a further and crucial step. The
:
discussion of GRRs in Chapter Four in good part ignored the ques-

tion of the sources of GRRs. One acquired these resources mate- —
rial, personal, cognitive, social, and so forth — to a greater or
lesser extent, it could have been inferred, almost by happenstance.
This, of course, is not the case. It is not at all accidental that certain
individuals and social groups are likely to have a much stronger
sense of coherence than others. Particular social-structural and cul-
tural-historical situations are quite likely to provide the develop-

mental and reinforcing experiences that result in a strong sense of
coherence.
In this light we turn to a consideration of the work of Engel
and his colleagues, Seligman, Wertheim, Kohn, Rose Coser, and
Kardiner. Their work, encountered at various stages in the develop-
ment of my own model, is important in clarifying the concepts of
generalized resistance resources and the sense of coherence. By con-
sidering similarities, parallelisms, and differences among the various
approaches, we can go to the heart of that most important question
What conditions are central in providing adequate GRRs to de-
velop, reinforce, and maintain a strong sense of coherence?
Psychological Sources. The work of George Engel and his
colleagues at the University of Rochester has been of great impor-

tance in advancing the unfortunately named field of psychosomatic
medicine. Guided by a pathogenic orientation, they have developed
the central concept of the giving-up syndrome, whose meaning is
best summarized in their own words (Sweeney, Tinling, and

Schmale, 1970, p. 378).
The affective response of "giving-up," when it
follows real, threatened, or symbolic loss of a highly

valued form of gratification, tends to precede the onset
or exacerbation of somatic as well as psychic disease . . .
[in that it] allows or facilitates whatever disease potential

or predisposition exists in the individual or environment
to become manifest. . . . [It] includes a loss of self-
esteem, a disruption in object relationships, a decrease in

motivation, and an expectation that such a state may be
enduring. . . . The "giving-up" reaction can be divided
into two phenomenologically distinct subtypes, which are
best described by the terms helplessness and hopelessness.

The qualitative differences between these affects are
postulated to have developmental bases. Helplessness is
defined as a feeling of being left out or abandoned where

loss of gratification is perceived as caused by external
events or objects. . . . With hopelessness . . . the indi-
vidual feels that he alone is responsible for the loss and
that there is nothing he or anyone else can do to over-
come it.
Subsequently, Schmale (1972) reviewed the extensive re-

search that had been conducted in clarifying the giving-up syn-
drome and its origins. In the same year, Engel and Schmale
reformulated the basic concept in a profound endeavor to link it to
the process of biological adaptation of organisms. They proposed the
concept of conservation-withdrawal, "a basic biological anlage serv-
ing survival" that comes to be "reflected in the behavior and
psychological experience of man." Giving-up is viewed as the "inner
experience of the person in whom the conservation-withdrawal
mechanism has been activated" (1972, p. 72). In brief, the argu-
ment is as follows. All organisms are confronted, periodically or by
chance, with unfavorable environmental conditions, in which a
withdrawal rather than an active response is highly adaptive and
serves "an elemental survival function" (p. 64). Appraisal of the
situation indicates "either a too intense input which cannot be as-
similated" (overload) or "a deficient input which indicates unavail-

ability of supplies" (underload) (pp. 68-69). In such cases, activity
is not only fruitless but is a waste of scarce resources. We have,
then, become biologically equipped with both response modes, ac-
tivityand withdrawal. The hibernation mode, however is risky: The
supplies taken in previously for repair, renewal, and growth are,
after all, limited.
who has adopted habitual withdrawal as a
In essence, one
dominant mode of behavior is unable to take sustenance from the
environment while existent supplies are rapidly exhausted. The
affinity with the concept of the sense of coherence is quite clear
(except that the focus is pathogenic). In our terms, withdrawal as
a general orientation is a marked inability to be accessible to the
information and music that one's environment might well be send-
ing, or, putting it otherwise, an inability to utilize the GRRs that

do exist as potentials or to develop GRRs.
The conservation-withdrawal response mode becomes a
classic, habitual, dominant syndrome through the repeated ex-
perience, particularly in the first six years of life, of failure to master
feelings of helplessness and hopelessness, to learn to give up when
appropriate, and to internalize a characterological balance between
activity and withdrawal (Schmale, 1972). The resulting lifelong
orientation toward conservation-withdrawal a is, in our terms,
weak (The relationship of this approach to the
sense of coherence.
work of Bowlby, 1977, on attachment theory is quite clear.)
There is considerable compatibility between the work of
Engel and his colleagues and that of Seligman (1975). Seligman,
however, is an experimental psychologist and developed his model
in the framework of psychological learning theory. "Helplessness,"
he writes, "is the psychological state that frequently results when
events are uncontrollable" (p. 9). He points to two crucial ele-
ments that determine helplessness. First, the individual responds
to stimuli in a locked manner rather than in a voluntary manner
(contingently), a response that can be affected by reward and
punishment. Second, and more crucially, the outcome of behavior
is seen as noncontingent on one's behavior. Learned helplessness
"produces a cognitive set in which people believe that success and
failure is independent of their own skilled actions" (p. 38). On
the motivational level, it "undermines response initiation quite gen-
erally" (p. 36). One observes "the striking giving-up sequence."

Emotionally, learned helplessness involves first a fear-protest stage
and then a helpless-depressed feeling.
Seligman explicitly stresses that what is crucial is not failure
per se but "the loss of control over reinforcements"; no matter
what one does, the outcome —which may even be positive — is not
perceived as contingent. The similarity found in Seligman's work
to the sense of coherence becomes even clearer in his discussion of
safety signals and relevant feedback (p. 121 ) . Predictability is made
possible by safety signals, even if the signal connotes an undesirable
outcome. At least in our terms, there is information, not noise.
Seligman contends that learned helplessness has its origins
in early childhood and in the continued experience of response-
outcome independence. Referring to Rene" Spitz' studies of maternal

deprivation, he argues that what is crucial is not the deprivation
of stimulation per se but the deprivation of synchrony between
action and feedback. This does not, he points out, mean that such
deprivation has an inevitable result. On the contrary, he notes that
the experience of the child in coping with anxiety and frustration,
if the child masters these stimuli by his actions, is essential in build-
ing immunity against learned helplessness. Children who for what-

ever reasons constandy experience an absence of controllability
learn to expect that nothing they do matters.
Two crucial points in Seligman's work should be stressed.
First, though he is never quite explicit, he seems to regard it as
essential, if learned helplessness be avoided, that the environ-
is to
ment be comprehensible, ordered, and consistent, whatever the con-
tent of the signals. If one never knows what is coming, one never
knows how to organize one's behavior. Seligman does not quite
commit himself to this view, however. On the one hand, his key
notion is that of contingency; on the other hand, he stresses that
experience in mastery is vital: "Objective control, however, is a
necessary condition for the development of the perception of con-
trol" (p. 138). The matter of control brings us to the second point.
Seligman's total commitment is to a body of work rooted in the
internal locus-of-control concept, a totally unnecessary and con-
fusing bias towhich I return later.
would make reference here to the work of Wertheim
I
(1975, 1978), whose detailed analysis and reconceptualization of

early child development make a considerable contribution to un-
derstanding the development of what she calls autonomy and com-
petence. As she defines them, there is good reason to posit an affinity
between them and the sense of coherence. 4 Her starting point is
the assumption that, from the time of intrauterine experience, a
"doing-undergoing" dichotomy always characterizes the "person-
4 In a personal communication, Wertheim writes: "The concepts of
autonomy and competence represent an attempt in this direction, akin to

your 'sense of coherence.' My ultimate aim is to arrive at a conceptual under-
standing of 'anthropogenesis,' or the capacity to be 'human.' ... In this
context it may be necessary to ask whether 'anthropogenesis* is a necessary
but not a sufficient condition of 'salutogenesis,' or both are merely different

aspects of the same process."
significant other" system and the "person-physical environment"

(including one's own body) system. "An adaptively optimal person-
environment system should be characterized throughout life by an
appropriate dynamic balance between the 'power' of the person
and the 'power' of the environment" (p. 2). In the case of both
autonomy and competence, which are parallel, interdependent, and
dynamic processes, Wertheim writes of "age-, sex-, and culture-
—
adequate control" that is, judgment as to the adequacy of the
autonomy and competence balances can be made only in a given
context. Such judgment, however, "can be viewed objectively or
subjectively. ... In the former case, autonomy and competence
are evaluated from the point of view of specific sociocultural and/or
clinical criteria. In the latter, perceptual processes, which are
treated as the critical intervening variable mediating the individ-

ual's subjective sense of autonomy and competence, are examined"
'
(p. 2).
I have referred to Wertheim's work here not because she
has an answer to the question "of the roots, optimal developmental
conditions, and subsequent developmental course of autonomy and
competence" (p. 6) but because she has, more than others, I think,
emphasized a number of important points in seeking an answer.
First, she insists on an interactional framework for understanding
development; living is a matter of constantly working out and

reworking out a modus vivendi with the relevant persons and en-
vironments. Second, she stresses that such persons and environments
form systems, and she does not focus on action-reaction patterns
between individuals. Third, she opens the way for avoidance of the
cultural bind of identifying coherence with "I am in control" both
by speaking of a "doing-undergoing" balance and by her insistence
on and culture-adequate control."
"age-, sex-,
Although Wertheim does refer to a lifelong process of re-
working autonomy and control, her detailed analysis focuses on
early (and often very early) child-environment system transactions.
Thus we come away from her work, and from the work of Engel
and Seligman, with the important understanding that childrearing
patterns that place a premium on autonomy and competence, an
appropriately balanced withdrawal-conservation response mode,
and the opposites of the giving-up syndrome and of helplessness
are highly relevant sources of a strong sense of coherence. But none

of them probe the social conditions that foster or impede such
childrearing patterns. Nor do they consider the conditions of adult
life weaken the developmental directions set in
that reinforce or
early life, beyond making passing reference that this can happen.
Childrearing patterns, one might think, are located in a sociocul-
tural vacuum.
Social-Structural Sources. For sociocultural connections, we
can turn to the work of Melvin Kohn. His professional career has
included three seemingly different major areas of study: social
class and schizophrenia, parental values and childrearing patterns,
and the structural conditions of occupational life. Though Kohn

himself has explicitly made the link among these three areas, not
all who know his work are aware of the close relationship. The
question of the source of the sense of coherence highlights the in-
tegral connection among them.
Let us start from Kohn's work on schizophrenia (1968,
1973, 1976a) . He notes that almost all of the more than fifty studies
on and schizophrenia have shown highest rates in the
social class
lowest social He proposes an explanatory etiological model
class.
that is based on the joint occurrence of genetic predisposition; a

socially determined high level of stressors; and a particular socially
conditioned, orientational cognitive-emotional system. This cogni-
tive-emotional system is of particular concern to us here. The
orientational system proposed by Kohn as being an etiological
factor in schizophrenia is almost precisely the essence of a weak sense
of coherence (1976a, pp. 179-180) : "[The schizophrenic is char-
acterized by] fearfulness and distrust and by a fatalistic belief that
he is at the mercy of forces beyond his control and often beyond
his understanding. ... An orientational system predicated on
conforming to the dictates of authority results in a perception of
social reality that is too simplistic and fearful to allow the in-
dividual to take advantage of options that might present them-
selves. It is too inflexible to permit effective coping with precisely
those problematic and stressful circumstances that most require
subtlety, flexibility, and a perceptive understanding of larger social
complexities."
Kohn, then, is describing a certain "perception of social
reality." His hypothesis, well supported by a considerable body

of evidence, is "that the constricted conditions of life experienced
by people of lower social class position" are most conducive to the
internalization of the described orientational system. In the same
way, he shows that the life conditions of persons in other social
class positions are conducive to alternative orientational systems.
What are these conditions of life that are relatively more
characteristic of lower-class persons? Here, I suggest, we must care-
fully distinguish between conditions of poverty and poverty in a
given historical setting. Being poor in a stable, generally legitimized
social structure is not as likely to produce the perception of social
reality that Kohn describes as much as is poverty that occurs "when
a stratified social and economic system is breaking down or is
being replaced by another as in the case of the transition from
feudalism to capitalism or during periods of rapid technological
change . . . [or] imperial conquest . . . [or] detribalization"
(Lewis, 1970, p. 69). Lewis puts his finger on the core of the
matter when he says that "the lack of effective participation and
integration of the poor in the major institutions of the larger society
is one of the crucial characteristics of the culture of poverty" (p.
70). It is not that they are isolated from such institutions. But in
all their contacts with such institutions they experience over and
over again powerlessness, arbitrariness, and bewilderment, and,
whichever way they turn, they are victimized, without any ap-
parent rules. This is the experience of reality, and this is the image
of reality that is built up. Moreover, such an image is hardly weak-
ened by the truly extensive participation of the poor in such social
institutions as the legal system, the military, or the public relief
system. (For an interesting and relevant study using the Holmes-
Rahe change instrument, see Justice and Duncan, 1976. They
life
describe the life of child-abusing parents as one of prolonged crises

that constantly knock them over and in which "the unpredictability
of all kinds of changes" is the key variable. They do not ask what
orientation system evolves out of this kind of life experience.)
Without in any way diminishing the importance of under-
standing the conditions that deprive lower-class people of GRRs
and shape a weak sense of coherence, I believe the pathogenic
focus prevents us from a full understanding of the phenomenon.
It also blinds us from asking about those who, despite such con-
ditions, do make it. We can move ahead by returning to Kohn's
work. For the majority of people in the Western world, after all,
daily life is not the unpredictable and nearly chaotic world of the
very poor. As a matter of fact, Kohn's studies on the conform-
ity/self-dircction orientations, to which we now turn, by and large
exclude the very poor. In his study on social class and parental
values, Kohn writes (1969, p. 189) : "The essence of higher social
class position is the expectation that one's decisions and actions can
be consequential. . . . Self-direction —acting on the basis of one's
own judgment, attending to internal dynamics as well as to external
consequences, being open minded, being trustful of others, holding
personally responsible moral standards — this is possible only if the
actual conditions of life allow some freedom of action, some reason
to feel in control of fate. . . . Self-direction, in short, requires op-
and experiences that are much more available to people
portunities
who are more favorably situated in the hierarchical order of
society."
Turning to the empirical examination of such opportunities
and experiences led Kohn to the study of people's jobs. Within the
established, institutionalized, and internalized daily routine, our
major role activity takes up the bulk of our time and energy and
shapes, more than anything else, our place in society, be we teach-
ers, executives, houseworkers, machinists, or sanitation workers.
How, Kohn has asked, does this routine shape our orientation?
Kohn's central variable is the substantive complexity of
work: "the degree to which the work, in its very substance, requires
thought and independent judgment" (p. 1). (This summary of
Kohn's work and the quotations are based largely on an unpub-
lished paper —
Kohn, forthcoming. For the published work on these
issues, see Kohn, 1976b, 1969; and Kohn and Schooler, 1978.)
Kohn's concern is with "the impact of work on [a] sense of self
and orientation to the rest of the world" (p. 3). It is crucial to
understand that Kohn is concerned with the objective conditions
of work and their effects on personality and not with one's aware-
ness of such conditions. "Substantively complex work, by its very
nature, requires making many decisions that must take into ac-
count ill-defined or apparently conflicting contingencies. . . .
[Complexity] is at the heart of the experience of work. More than

any other occupational condition, it gives meaning to this experi-
ence" (pp. 8-9).
In some of the most sophisticated empirical work of which
I am aware, Kohn has shown (for men and, more recently, for
employed women) that the substantive complexity of one's work
"has a strong, independent relationship to many facets of psycho-
a relationship stronger than that of any other

logical functioning,
dimension of occupation (subjective or objective) we have studied"
(p. 10) . Further, a follow-up study has shown that the "substantive
complexity of work . . . has a causal impact on psychological
functioning" (p. 10). Kohn does not deny that personality in-
fluences what occupation one goes into and how one molds one's
job. His contribution is in showing that at the least "cognitive
processes do not become impervious to environmental influence

after adolescence or early adulthood . . . [or] well into mid
career" (p. 15).
Although Kohn's investigation of causality has been limited
to substantive complexity and intellectual flexibility, he feels that
he has at least a strong prima facie case that the substantive com-
plexity of one's work also is causally related to "values, self-con-
ception, and (p. 17) and particularly to the
social orientation"
powerlessness, normlessness, and self-estrangement components of
alienation. These are even closer to our concerns. As Kohn puts
it (pp. 18-20) "Thus, men who do complex work come to
:
exercise their intellectual prowess not only on the job but also in
their nonoccupational lives. They become more open to new
experience. They come to value self-direction more highly. . . .
[Complex work strengthens] one's sense that the problems one

encounters in the world are solvable and manageable [in- . . .
creases] one's valuation of self-direction and one's tolerance of

deviant belief." In an article closely related to Kohn's concerns,
Coser (1975) analyzes the psychological-orientational implications
of embeddedness in "complex role sets and differentiated roles."
(Kohn, it should be noted, has put the emphasis on the conditions
of work, stressing substantive complexity and seeing complexity of
work with people as but one type of substantive complexity. Coser
sees complexity of work with people, conceptualized in terms of
role sets, as central and other types of substantive complexity as

secondary. For present purposes in seeking the sources of the sense
of coherence, both approaches should be taken into account.)
Though Coser's central focus is on the self-estrangement com-
ponent of alienation or, on the positive side, individuality, she also
deals with the socially structured preconditions for "innovation,
flexibility, reflection, and self-direction . . . and empathy." As she
puts it (p. 259) : "It is in a differentiated where
social structure,
individuals are segmentally involved, where they are encouraged

to take distance and to articulate their roles and their thoughts,
that people are able to develop that degree of individuation that
goes together with rationality and flexibility. . . . Complex role
sets and differentiated roles are not alienating restrictions on in-
dividuality; they are its basic structural precondition."

I have concentrated on Kohn's and Coser's work here be-
cause they, more than anyone else I know of, have been concerned
with the socially structured conditions —embedded in a system of
social organization defined more than anything else by the social
class dimension (though by far not the only dimension)

this is —
that foster the orientation with which we are concerned. Working
at a certain kind of job (Kohn) or occupying a certain kind of
role-setcomplex (Coser) leads one to see the world as complex as —
offering alternatives and choices, making sense in different ways,
allowing considered rationality, facilitating planning, having
room for one's own action. Certain people, they say, more than
others, because of where they are socially located, continually have
such experiences; over time, they come to see the world in this
complex way.
Finally, brief mention must be made again of Kohn's work
(1969) on parental values and childrearing patterns. Parents set

frameworks of expectations within which their children grow up.
Compared with almost all other areas of life, childrearing provides
parents with the opportunity to exercise an enormous amount of
power. They set the rules and have the wherewithal, to a consid-
erable extent, to enforce these rules. Kohn
is particularly con-
cerned with the dimension of self-direction/conformity, which can

be seen in a broad sense to be congruent with Kohn's conceptualiza-
tion of the general orientation of the lower social class as described
in his work on schizophrenia. Kohn is thoroughly aware of social

class structure. He knows that in all societies one's social class is
one of the key variables that determine one's daily existence. His
concern with social class brought him to ask, What is it about
social class membership that molds the way one looks at the
world? and to focus on one important element that he calls the
substantive complexity of the job. Kohn's overarching thesis, then,
can now be stated. Engaged in a job that offers extensive substan-
tive complexity (buttressed by Closer's complex role set), one comes

to have an orientation that is one of complexity and flexibility,
alternativesand self-direction, meaning, consistency, choice, and

a sense that problems are manageable and solvable. This orien-
tation comes to expression, above all else, in how one raises one's
children. Childrearing in turn develops the orientation of the

next generation —an orientation that guides them in shaping their
own lives and in selecting alternatives (such as jobs of substantive
complexity).
I have taken the liberty of extrapolating from Kohn's con-
cept of the self-direction/ conformity orientation to the concept of
sense of coherence. In doing so, I may have gone too far, for
there is a fundamental difference, though not a contradiction,
between the two. Self-direction is consonant with a strong sense
of coherence in a particular sociocultural context. The availability of
options and choice facilitates rationality in coping. The existence
of alternatives affords the luxury of failure; one can try a different
tack. But such self-direction is contingent on living in a society that
approves of a variety of ways of doing things. Further, if the world
is to be seen as comprehensible, self-direction must be accompanied

by a considerable degree of tolerance of ambiguity. In other societies,
where self-direction is hardly praiseworthy, substantive complexity
and tolerance of ambiguity may well be inimical to a strong sense
of coherence. Kohn's essential contribution to the exploration of the
sources of a strong sense of coherence, then, is twofold. On a gen-
eral theoretical level, he has shown how social structure shapes
orientation. In the particular Western societies he has studied, he
has carefully delineated one particular pathway.
Before turning to a consideration of alternative cultural
pathways, we should note a fundamental though not intrinsic short-
coming in Kohn and Coser's work —a shortcoming most germane

to our present concern. In their discussions of complexity, they fail
to note two conditions that are essential in allowing a high degree
of complexity to contribute to a general orientation in which the
world is and comprehensible. First, the com-
seen as predictable
plexity must not be overwhelming; the number of choices must
be somewhat limited, the number of impinging factors compre-
hensible. Second, and related to the first, the resources available
to one must be such as to make it possible, at least a considerable
part of the time, to solve the problems confronting one with a
reasonable degree of success.
In reviewing studies of sex differences in coping, Shalit
(1977) found articulation (the ease with which concepts can be
ranked and differentiated from each other) to be most potent in
successful coping. Thus he suggests that providing presurgical
patients with too much information hinders coping, whereas
selective attention, which makes articulation simpler, strengthens
cognitive control and helps coping.
The second precondition that is necessary if job complexity
is to facilitate development of a strong sense of coherence —the
availability of resources for problem solution — is considered clearly
in a Swedish study of sawmill workers. Kohn found a high cor-
relation between job complexity and control by the worker over the
work process. The authors of the Swedish study, however, describe
a situation in which this is not the case. As Frankenhaueser and
Gardell (1976, p. 44) describe the objective characteristics of the
job, it involves a "demand on the worker to make skilled and eco-
nomically important decisions" —a characteristic that at least seems
to be related to Kohn's job complexity. The sawmill workers,
however, have to make these decisions "at a pace set completely
by the machine system." Thus these workers were confronted both
by too much job complexity (quantitative overload) and "too
low a degree of personal control over the work process" (qualita-
tive underload). In such a double bind situation, one's work
experience cannot become a GRR, as the data from the study
5
suggest.
5 The reader may have noted that, with the exception of a passing
reference to employed women, the focus of this section has been on men. I
Cultural-Historical Sources. Consideration of Kohn and

Coser has taken us beyond the psychological focus on childrearing
patterns and led us to consider location in a social structure as no
less important in shaping one's GRRs (and, by implication, the
patterned, prototypical GRRs of social groups and categories of
persons). Kohn's work has also provided a link between the adult
and child, the sociological and the psychological. We can now go
further by introducing a cross-cultural and historical dimension.
The present thesis is that certain cultural and historical situations
fostera strong sense of coherence.

The work of Kardiner (1939, 1945) in the development
of the basic personality structure concept, though written over three
decades ago, is still extremely valuable. "The basic personality type
for any society is that personality configuration which is shared by
the bulk of the society's members as a result of the early experiences
which they have in common. It does not correspond to the total
personality of the individual but rather to the projective systems,
. . . the value-attitude systems which are basic to the individual's
personality configuration" (1945, p. viii). Working largely in small
preliterate societies, Kardiner and his colleagues developed a the-
oretical model that is useful in clarifying the issue of the sources
of a sense of coherence. Briefly, and with oversimplification, let me
present Kardiner's model by arbitrarily taking as a starting point
a given natural environment. People enter this environment and
work out a culture and social structure, a set of answers to the
fundamental problems of social existence. Among these answers
is a prototypical family structure and childrearing patterns. Given
the high order of coherence on the institutional and behavioral
level throughout the society, the child is socialized in a particular
direction that is most appropriate to, that fits best into, that society's
social organization and culture. There
a hierarchy of systems in
is
the basic personality structure, ranging from profound, unconscious
would point out that the theory of job complexity and control over job
process can be applied directly to the major role activity of anyone: worker,
housewife, pensioner, prisoner, patient, student. Unfortunately, the literature
on male workers predominates. I know of only one serious study of the situa-
tion of the housewife that lends itself to this analysis (Oakley, 1975). Goff-
man's (1961) study of total institutions is certainly also germane, though
again the focus is pathogenic.
projective systems through learned systems to pure empirical

reality systems. Cutting across all these are value systems and
ideologies (p. 34).
Having constructed this ideal-typical theoretical model,
Kardiner then applied it to a number of concrete societies. His
concerns were essentially two, and these are of direct significance to
us. First, he asked, what is the content of a given basic personality
structure in a given society? As a collaborator with anthropologists,
he was able to systematically dispose of the notion that there is a
fixed human nature. Each culture develops its own relatively stable
human nature. For our purposes, this formulation raises the crucial
—
question of whether different human natures in the sense of differ-
ent basic personality structures —are more or less consistent with
a strong sense of coherence. would go further and suggest that
I
the sense of coherence is a major component of the basic per-
sonality structure.
Kardiner's detailed analyses of the Comanche and Tanala
cultures (p. 99) point to two remarkably different culture patterns.
In Comanche culture, group solidarity is based on the ego strength
of the individual, the absence of segmental vested interests, the
prohibition of overt expression of tension, and full cooperation
without subordination. Among the Tanala, group solidarity is
built about obedience and passivity, with fixed statuses, vested

interests, and protection of the weak by the strong. These cultures
are radically different from each other, but in both cases a con-
sistent, lifelong experience that one's culture has clarity and con-
sistency and makes sense leads to what I infer is a strong sense of
coherence. These cultures are in extremely sharp contrast to the
culture that occupies the bulk of Kardiner's book, the Alorese,
whose essence can be summed up briefly (pp. 228ff.) : "the low
range of affectivity, the distrust, the low aspiration level, the
amorphous aggression patterns, the limited capacity for strong
attachments, the sensitivity with all varieties of defenses it mo-
bilizes, the repressed predatory trends, . . . complete absence of
anything that can be called government or even status, . . . com-
plete lack of systematization." Clearly, this culture is not likely to
place at the disposal of its members a considerable store of GRRs.
Kardiner's second central concern may be even more ger-
mane to us. All human beings live in a changing world. When,

because of a change in climate and topographic conditions, the
dry method of rice production became obsolete, all of Tanala
culture underwent a radical change. "In Tanala we see a society
which under one set of conditions had strong stability and through
the change of subsistence economy became extremely unstable by
virtue of the enormous amount of mutual hostility unleashed by
the new economy" (p. 418). What is crucial for us here is not the
earlier-noted issue (that some cultures are "better" than others
for a strong sense of coherence) but that radical change and in-
stability are, by their nature, not conducive to a strong sense
of coherence.
In the last section of his study and in a subsequent study of
American Negroes (Kardiner and Ovesey, 1951), Kardiner sought
to apply his concepts to complex societies. In historical perspective,
dealing with the emergence of capitalist industrial societies, Kar-
diner considers the basic personality structure "required by" such
societies and best exemplified by Calvinism. Though it is tempting,
it would take us too far afield to explore the Calvinist basic per-
sonality structure and its relation to a strong sense of coherence.
Fromm (1941) has done this far more adequately than would be
possible here. His concept of social character is no less an appro-
priate conceptual tool in analyzing cultural-historical sources of the
sense of coherence than is Kardiner's basic personality structure.
In this can be made to Erikson
context, too, useful reference
(1958), where we read (p. 263): "The answer lies in man's
capacity to create order, which will give his children a disciplined
as well as a tolerant conscience and a world within which to act
affirmatively."
The importance of Kardiner's work (as well as that of
Fromm and Erikson) for the present study is that he gives us the
historical and cross-cultural perspective needed in order to gain
an understanding of the sense of coherence. He leads us to ask
two fundamental questions: To what extent does the society or
subculture under consideration require a basic personality structure
that involves a strong sense of coherence? And, no less important,
to what extent does it provide a clear cultural image of the world,
whatever the content of the image may be?
I do not see how it is possible to exaggerate the importance

of the last point. and comprehensibility are the
Predictability
central themes of all the approaches we have considered. They are
necessary albeit not sufficient prerequisites for the development of
GRRs. Beyond this, we can specify additional conditions. What I
trust has become clear in this lengthy exposition is that specifiable

social conditions — certain specifiable childrcaring patterns and
subcultural and cultural patterns of social organization —provide a
continued series of experiences that build up the GRRs that are
crucial to a strong sense of coherence: a strong constitution, money,
a clear ego identity, a flexible coping style, social supports, and
so on. Over and over again, these GRRs allow us to see our internal
and external environments as meaningful, predictable, and ordered.
With this perception, we can reasonably hope that we can emerge
victorious much of the time, though not necessarily in every en-
counter. It allows us to develop an orientation at whose core is
"I (or we) can overcome."

I would be the last to claim that our knowledge is sufficient
to specify the precise conditionsunder which a strong sense of

coherence emerges. There is a great need for detailed research,
both on the individual and on the group level. I should like to
think, however, that the framework provided here allows us to
formulate research questions —
the key to any meaningful research.
Let me give but one example. It strikes me that one common sub-
stantive theme that is shared by all the studies we have considered,
from Engel's giving-up complex to Kardiner's Calvinist man, is the
continuous experience of participation in shaping one's fate. How,
we might ask, is such participation related to GRRs and to the
sense of coherence?
Danger of Bias
This last point brings us to a most important issue that has

been touched on here and there but has not been fully explored.
monograph ( 1966), a significant amount

Since Rotter published his
of research has been conducted on his locus-of-control concept.
Much of the work has been summarized in Lefcourt's book ( 1976)
and most recently in an important paper by Perlmuter and Monty
(1977). Because of its similarity to the concept of the sense of

coherence and because many of the readers of this book are
familiar with this body of work, I find it most important to point up
the fundamental distinction between the two concepts. (See the
exchange of letters between myself and Perlmuter and Monty in
Antonovsky, 1978.)
Rotter's fundamental distinction is between an internal
and an external locus of control. Someone with an external locus of
control has "the pervasive expectancy . . . that rewards and pun-
ishments do not occur because of one's own actions but are due
instead to forces outside of one's control such as fate, luck, chance,
or powerful others" (Naditch, 1974, p. 111). An internal locus of
control, by contrast, locates one's fate in one's own hands. The
tendency is most ethnocentrically powerful to equate sense of co-
herence, sense of control, and internal locus of control, using the
model of the autonomous individual extolled in the litany of
Western societies since the Industrial Revolution or, perhaps more —
appropriately, the Protestant Reformation. This ideological par-
adigm dominates our own lives and shapes our science.
My point can be seen by reference to the first example given
by Seligman (1975, chap. 1) to illustrate his definition of control-
lability. He first describes the behaviors of a competent, almost
certainly suburban American housewife who smoothly and ap-

propriately handles the quasi-emergency of her five-year-old's
suffering a gash in the leg. That night the child's fever rises
and her leg is swollen. In the interminable hours of helpless and
ignored bystandership in the hospital emergency room, the mother
is in a totally different psychological state. Seligman uses this
illustration to introduce the two crucial, intimately related con-
cepts of voluntary response and response-outcome independence.
There is no question in Seligman's mind that the sense of control
is totally related to the freedom of the individual to choose among
available alternatives and to perceive the outcome of the dynamic

situation as completely contingent on the choice he or she makes.
Thus, in analyzing the housewife's experience at home,
Seligman would predict a strong sense of control (or, in Rotter's
terms, an internal locus of control). The mother is "a competent
parent with a smattering of first aid"who knows the importance
of hugs and reassuring words, cleaning the cut, stemming the

flow of blood, meanwhile telling the child of her own experience
as a and applying antiseptic. Her sense of control is con-

little girl
tingent on the outcome of her action, and it is not until her "little
girl is happy again and the bleeding has stopped" that she can
feel in control. Seligman's point is that in this situation the mother
could, at a glance, diagnose and apply appropriate therapy all by

herself. By contrast, the hospital situation was one in which "most
of [her] actions were to no avail. The hospital staff paid no at-

tention to [her] plight, lost [her] forms, and ignored [her] request
for an explanation; [her] child recovered without [her] having

brought it about. The course of events was uncontrollable —the
outcome was independent of each of [her] voluntary responses."
The crucial issue here is posed by the failure to distinguish

between a desirable outcome that is contingent on one's voluntary
responses and one that, although not as contingent, nonetheless
does not confuse and bewilder one. The latter, too, is consistent
with a strong sense of coherence.

Suppose the mother had a positive experience with the
emergency room. Suppose the forms had been filled out only after
care for the child had been initiated and had not been lost. Suppose
the nurse or doctor had come out several times and turned to
the mother, keeping her informed, consulting with her, dealing
with her as a patient who needed reassurance, and responding to
her requests. In other words, suppose everything had happened that
was conducive to the mother's feeling that things were under con-
trol, that she was part of the picture, though not in control, and
that a positive outcome was most likely without there being any
need whatsoever for her intervention. Is it not likely that her sense
of controlwould be high?
We can now place control in a historical-cultural context.
The writer and most of the readers of this book, as well as Selig-
man and his suburban mother, have been socialized to mistrust
any situation in which we are not personally in control, in which
it is not our actions that shape the outcome. Our version of the
Protestant Ethic has taught us that we cannot rely on others. In
the last analysis, there is no family, no friend, no priest; there is
even no God who shows mercy and helps.

This, of course, is the model. In practice, none of us could

take the awesome anxiety that would inevitably accompany an
existence based on such beliefs. And so we compromise our prin-
ciples, as it were. Let me make most explicit the two implications of

the previous paragraphs. First, I have suggested that even in our
most extraordinary culture in its insistence on "my" being in con-
trol, in many areas and situations, we are perfectly willing to
allocate control to the expert without feeling that our sense of
coherence is being violated. The second point is, as it were, a
contradiction of the first, yet I believe that both are true. By
the very nature of our culture, which leads us to insist so con-
sistently on "my" being in control, we are never comfortable, we
remain ambivalent, when we do allocate or when we are compelled
to allocate control to others. I shall later consider the implications
of this tension and ambivalence in discussing the problem of a fake
sense of coherence. For the present, I have been concerned only
to point to the distinction, even within our extreme middle-class
culture, between being in control over things and things being
under control.
This distinction is even more apparent, and the cultural
bias more striking, when we turn to a consideration of other cul-
tures. Contrasting these with the dominant value orientation of
Western industrial society highlights two issues. First, as suggested
above, there is the question of "I am in control" versus "things
are under control." Second, there is the issue of the significance of
the existence of alternatives and choice. In many societies, the
sense of coherence not only is not impaired but isenhanced by the
fact that control is located in a deity or in the hands of powerful
others. The crucial question is whether, as I have indicated earlier,
there is a serene belief that those in control have legitimate power
and act in one's own interest. This formulation allows, but does not
insist on, the co-identity of the person whose fate is being deter-
mined and the decision maker. Think of a Calvinist with a pro-
found belief in predestination. What would someone with a belief
that God has already determined his destiny score on Rotter's
scale? Familiarity with Calvinist life (or with Christian Science
or with orthodox Judaism) suggests the answer to the seeming
paradox; I find it in the word participation. If life offers one the
chance of confirming one's predestined salvation by doing the

prescribed right things, one can have a strong sense of coherence.
Only when there is no deity, no writ, but only meaningless chaos
does one's only hope lie in an internal locus of control. There are,
then, many cultural roads to a strong sense of coherence.

The issue of available alternatives as essential to a strong
a second aspect of the danger of cultural
sense of coherence reflects
bias. I have referred to danger of overload
this issue earlier as the
and hence only make note of it here. As Bensman and Rosenberg

have put it (1963, p. 187), "Actual freedom and individual in-
tegrity, something often called autonomy or spontaneity, much
praised by philosophers and psychologists, have frequently pro-
duced a feeling of helplessness." It may be that Americans have
been dulled to what for me, having been out of the United States
for many years, is still a fresh response to watching television com-
mercials or entering a supermarket ; the choices can be so bewilder-
ing that they result in a sense of paralysis rather than control. One
must add, of course, that in a culture where no deity is potent or
relevant, some choice is unquestionably an essential ingredient of
Having raised the question of cultural or, as one might put —
it more sharply, ideological —
bias with respect to the locus-of-
control approach, I think it only appropriate that the same ques-
tion be directed at the concept of sense of coherence. Two major
inferences might be drawn that could point to such bias. First, a
strong sense of coherence could be held to be a good in and of
itself. Second, the emphasis on conditions of stability and predict-
ability, whether in childrearing patterns, social structure, or cul-
ture, would seem to be inherently conservative. Neither inference
is warranted.
I have hypothesized that a strong sense of coherence is
salutogenic. Let us assume that the data will support this hypothesis
and, further, that health is universally held to be good. Health,
however, is only one value. Many in the health professions would
have it that it should be the supreme value. This bias leads them
to disregard other values or blithely to assume that there is never
any between health and other values, that no price is ever
conflict
paid for doing what is good for one's health. (See Lewis and Lewis,
)
1977, for an analysis of the implications for health of the growing

participation of women in the labor force. ) I would make it quite
clear. Even if my thesis is supported and a strong sense of co-
herence is valued because it is good
a sense of coherence
for health,
is still not necessarily praiseworthy as a good in itself. A person with
a strong sense of coherence is quite capable of being what many
would consider an insensitive, unpleasant, inconsiderate, exploita-
tive bastard. For a good example of what I mean, think of George
Bernard Shaw's portrayal of Undershaft in Major Barbara. This
is precisely one of the reasons why I included the quote from
Brinton. In fact, probably the best historical examples of social
groups with a strong sense of coherence are ruling classes in their
heydays.
The second charge of ideological bias may be seen as even
more serious. I have taken the position that social upheaval, rapid
social and personal change, and severe conflict may well disrupt
a strong sense of coherence. That this position necessarily carries
conservative connotations can be refuted on three grounds. First,

social conflict is not necessarily disruptive in this sense. It may well
provide a firm basis for enhanced rootedness in a subgroup, for
strengthened clarity of the meaning of life (see Coser, 1956).
Second, so what? Conflict and mobility may be preferred on other
grounds even if one pays the price in an impaired sense of co-
herence. (As a voluntary migrant,
I can hardly take any other
would stress that a reasonable degree of repeated

position. ) Third, I
experience of challenge and successful resolution strengthens one's
sense of coherence. It need not make life incomprehensible, par-
ticularly in a society that is tolerant of change, has socialized its
members to prefer an ambiance of orderly change and to be toler-

ant of some ambiguity, and provides the structural conditions for
successful resolution of challenge. (See Wertheim, 1974, for con-
autonomy
sideration of the implied conservatism in the position that
and competence are tied to fitting in neatly into the age, sex, and
other slots assigned by one's culture.
To put the matter even more bluntly. A strong sense of
coherence is not compatible with a totalitarian or other type of
regime based on terror and naked power (or with analogous per-
sonal situations). Nor is it compatible with normless libertarianism,
—
"doing one's own thing." It is, however, quite compatible with a

variety of other types of social and personal order: legitimated con-
servative systems no less than dynamic, democratic consensual sys-
tems. There are, to repeat, many sociocultural roads to a strong
sense of coherence.
Opera tionalizing the Sense of Coherence
If the sense of a concept that seems to have

coherence is
internal and face validity and theoretically offers promise as a pow-

erful research tool, then the next task to be confronted is that of
operationalization. Beyond full recognition of the importance of the
problem, it would not be appropriate here to explore this issue.
Three important points, however, are in order.
First, there is no need —indeed, it would be unfortunate
to commit oneself to a particular methodology. The concept re-
quires much exploration before it can become a systematic tool.
Survey questions, a focused interview schedule, observation, use

of informants are all in order at this stage of the game. Ultimately,
of course, if the concept is to become useful, commitment must be
made to a specific operational definition. This can, however, be
done prematurely.
Second, I have throughout referred to a strong and a weak
sense of coherence. This implied dichotomy, used for the sake of
convenience, is misleading. To have written "located toward the
strong (or weak) end of the sense-of-coherence continuum" would
have been awkward. I hope that the reader has understood that,
much as with references to health, I view the concept as referring

to a continuum. The researcher who intends using the concept
should certainly be aware that it refers to a distribution that is
assumed to be normal (bell shaped) rather than bimodal. In this
regard, it should be noted that, though Rotter suggested that locus
of control refers to a continuum, most studies have dichotomized
the scale.
My on a continuum leads us, finally, to consider
insistence
what a fake sense of coherence. A salutogenic orientation,
I call
committed to studying those who are indeed toward the strong

end of the continuum, is particularly prone to this danger. Lefcourt
(1976, chap. 11) indicates that those who are extremely internal
on the locus-of-control scale are prone to paranoia and delusions
of grandeur. I would suggest that the core of such fakeness is
When there is a contention that all problems
hysterical rigidity.
have an answer, when challenge or doubt is intolerable, when
there is no flexibility to adapt to changing circumstances, when one
claims to be in control of all things or to understand everything,
when there is a denial of sadness, and when there is an incapacity
to admit to the uncontrollable without being overwhelmed — there is
a clear indication that we are confronted by a fake sense of

coherence.
Chapter Six
Relation of the Sense

of Coherence to Health
Up to now we have unavoidably dealt with the issues in piecemeal

fashion. Having initially set the problem as that of salutogenesis, we
put aside health as the dependent variable so that the concepts of
stressor, generalized resistance resources, and the sense of coherence
could be clarified. But before the overall salutogenic model can be
presented in Chapter Seven, we must consider one crucial question :
What and evidence for the presumed relationship

are the grounds
between location on the sense of coherence continuum and location
on the health ease/dis-ease continuum? This continuum was, after
all, the point of origin in this endeavor. Asking the salutogenic ques-
tion has led me to the sense of coherence as a proposed answer. It
would take a major research program to test the power of this an-
swer. One can, however, begin to anticipate the results of such a
160
Relation of the Sense of Coherence to Health 161
program — or, at least, judge its possible fruitfulness —by considering

data now available.
We face two fundamental difficulties in examining the evi-
dence for a relationship between the sense of coherence and health.
First, the concept of the sense of coherence has not been opera-
tionalized and data have not been assembled that could put the
hypothesis to a real test. The data that will be considered were as-
sembled for other purposes. Only by reinterpreting these findings
can we consider the possibility that the sense of coherence offers a
more adequate or a more parsimonious explanation (or both) than
the one offered by previous researchers. Such interpretation is al-
ways legitimately suspect. I may be sensitized only to such data as
seem to fit my hypothesis and may ignore contradictory data. I can
say only that I have not done so consciously, though I make no
claim to a mastery of all the data that might be relevant.
The second from the fact that the over-
difficulty arises
whelming part of the data was assembled with a pathogenic orien-

tation. More particularly and more seriously, almost all but the
overall mortality data relate to specific diseases. Seldom do we learn
anything about the proportion of a population in a study that is
healthy, whatever the definition. For all we know, a higher rate of

coronary artery disease in group A than in group B may also be
accompanied by a lower rate of cancer in group A.
It seems, then, most sensible to consider the discussion of
"evidence" below as no more than an attempt to establish a prima

facie case for the plausibility of the hypothesis. I do not wish to
sound too diffident. I am persuaded that the sense of coherence is a
powerful construct that can be most helpful in understanding a
wide variety of data. But it has yet to be tested rigorously.
Somewhat Direct Evidence
In Chapter Five, I referred to the work of Kohn (1968,

1973, 1976a) in which he arrived at an overarching synthesis of the
many studies linking social class to schizophrenia. The conditions
of life of the lowest social class, he argued, foster limited and rigid
conceptions of social reality. This orientation, interacting with
powerful stressors and a genetic predisposition, explains the much

higher rates of schizophrenia consistendy found in the lowest social
class. I suggested that a very weak sense of coherence is the precise
equivalent of the orientation Kohn described. In a way, however,
there is such a great linguistic similarity between the way I have
described a weak sense and the way schizophrenics look
of coherence
at the world that one might well suspect tautology. At best, it might
be said that a weak sense of coherence is implicated in schizophrenia
but not in other diseases.
The work of Engel and his colleagues, however, is not sub-
ject to this charge. Their work on a concept analogous to the sense
of coherence, over them to conclude (Schmale,
two decades, led
1972, p. 23) : 'The giving-up was found as an antecedent
process
to diseases of all categories, from infectious and metabolic to those
of degenerative and neoplastic origin in the medical group and
from acute organic brain syndromes and schizophrenic reactions to
psychoneurotic disorders and the clinical syndromes of depression in
the psychiatric group."
But perhaps the most direct support of a link between the
sense of coherence and health status is from the study whose data
led me to formulate the concept of the sense of coherence (An-
tonovsky and Shoham, 1978). A sample of 389 Israeli men and
women aged forty to forty-nine were selected from the rosters of
seven neighborhood health centers in all areas of a middle-sized
Israeli city and were interviewed. Some 85 percent of the area
population belongs to the health-insurance fund that operates the
centers. The dependent variable, health status, is based on an earlier
version of, but is essentially similar to, the breakdown continuum

presented in Chapter Two. A considerable variety of measures of
generalized resistance resources were covered in the largely closed
questionnaire: flexibility, self-esteem, a variety of social supports,
material resources, and so forth.
One such GRR is of interest here. Deriving it from Engel
and Schmale's ideas of helplessness and hopelessness, we called it
coping ability. Operationally, four single items and one seven-part
item were used. 1 Only when we examined the total structure of the
*The items, translated from the Hebrew, are: (1) How often do you
run into problems that you think you can't solve? (very often . . . often . . .
—
data, using a smallest-space analysis, did we seemed we realize that
to be dealing with a somewhat different order of phenomenon. Not

only were the correlations with the health measure substantially
higher than were those between health and other variables, but the
data structure suggested to us that coping ability played an inter-
mediary between a variety of other resources and health. The
role
index based on the seven-part item correlated 0.49 with the measure
of health; only the children's needs item showed a correlation of
less than 0.40. No other measure in the study had such a high cor-
relation with health except for a life-satisfaction index that is, —
what was called, in Chapter Two, other dimensions of well-being.
Indirect Evidence
The studies and data that provide indirect evidence for a

relationship between the sense of coherence and health are grouped
below in five areas: social-structural, cultural, psychological, situa-
tional, and animal studies. In each case, the fundamental argument

is the same, namely, that the independent variable provides life ex-
periences related to the level of the sense of coherence and hence to
health and illness.
Social-Structured Variables. The literature relating social
class to illness is vast. My own work relates to overall mortality and

life expectancy, cardiovascular morbidity and mortality, and infant
mortality (Antonovsky, 1967a, 1967b, 1968; Antonovsky and
Bernstein, 1977). The work on infectious diseases is too well known
to be cited. Syme and Berkman (1976),
in a succinct review of a
wide variety of data, say: "In summary, persons in lower-class
occasionally rarely or never). (2) To what extent do you feel that you
. . .
succeed in solving the problems that you run into? (always or almost always
succeed . often
. . occasionally
. . never or almost never succeed).
. . . .
(3) Does it happen that you feel that you can't give your children what you
would like to give them? (very often often
. . . . . . occasionally . . . rarely
or never). (4) During the past few
(Same as item 3 for husband/wife. ) (5)
months, have you been in a situation where you felt you were in a trap
that something extremely unpleasant could happen to you or to someone close
to you and you were helpless to do anything about it? Did this happen to
you (yes or no) with respect to: (a) friends or acquaintances, (b) husband/
wife, (c) children, (d) financial matters, (e) work (or major role activity),
(f) your health or that of someone close to you, (g) anything else?
.
groups have higher morbidity and mortality rates of almost every

disease and illness" (p. 2). One needs, then, an overall explanation
for such vulnerability to illness in general. They suggest that re-
search be focused on "the more generalized ways in which people
deal with problems in their everyday life" (p. 5), an approach
clearly consonant with the concept of the sense of coherence. If,
then, one premises that a central characteristic of lower-class life is
a weak sense of coherence, it seems most reasonable to accept this

body of work as suggestive of a relationship between sense of coher-
ence and health status.
Bcrkman's doctoral dissertation ( 1977; see also Berkman, in
press) is one of the most valuable germane reviews of the literature.
Concluding her documented review, she writes (p. 18): "Thus,
people who are widows, mobile, or migrants as well as those who are
poor or come from certain parts of the country or who are members
of certain ethnic groups have higher than expected morbidity or
mortality rates from many diseases. . . . One is directed towards
the search for some characteristic or set of characteristics which
these groups have in common and which in itself has been thought
to have deleterious health effects." Berkman proposes that what is
common to the poor, blacks, widowers, people who are highly mo-
bile socially or geographically, relocated elderly people, and people
with important characteristics different from those of the commu-
nity in which they live is that they are all subject to powerful
socially structured constraints "on an individual's ability to main-
tain enduring and effective social ties" (p. 19)
Using questionnaire data from a sample of almost 7,000
people aged thirty to sixty-nine, collected in 1965, and the re-
corded deaths of 682 of the respondents in the following nine-year
period, Berkman found strong support for her hypothesis. Her so-
cial-network index was found to have a clear association with mor-
tality independent of a wide variety of other risk factors. It would
take us too far afield to consider the study in detail, but a number of
findings merit note in the present context. Men aged fifty to fifty-
nine of lowest socioeconomic status with a high social-network
score have a lower mortality rate than do those of highest status
with a low network score; women aged thirty to forty-nine in these
two polar status/social-network categories have the same lowest
.
mortality rate (pp. 147-148). Though Berkman's theoretical and

empirical focus was on those with the lowest social-network score,
her data show a linear association — the higher the score, the lower
the mortality (pp. 291-292).
In her discussion of what it is about social contacts that may
be critical to health, she suggests that social contacts provide tangi-
ble support, appraisal support, and emotional support (p. 229). It
seems to me reasonable to propose that the concept of a sense of
coherence provides a more powerful and parsimonious explanation
of differential mortality rates. A
strong sense of coherence, on the
one hand, is fostered by social supports (and debilitated by migra-
tion, being poor, being black in a given society) and, on the other
hand, enables one to mobilize tangible, appraisal, and emotional

support in coping with stressors.
The same reasoning can be applied to Gove's study (1973)

of mortality rates and marital status in the United States from 1959
to 1961. He clearly shows that the "mortality rates of the unmarried
. . . are, controlling for age, higher than those of the married, and
the differences between the married and unmarried . . . are much
greater for men than for women" (p. 61 ). The variations are par-
ticularly large "where one's psychological state ( 1 ) appears to play
a direct role in death, as with suicide, homicide, and accidents, (2)
is directly related to acts such as alcoholism that frequently lead to
death, and (3) would appear to affect one's willingness and ability
to undergo the drawn-out and careful treatment required" (p. 61 )
If being married, then, provides protection, particularly for men, is
it not possible that this structural variable facilitates a strong sense

of coherence?
Perhaps the most intriguing indirect evidence linking the
sense of coherence to health status via social-structural variables is
found in the phenomenon of voodoo death. Seligman's section on

death from helplessness in humans (1975, pp. 175-188) reviews
reports ranging from Cannon's examples of hex death to the death
rate of American prisoners of war in the Philippine campaign of
World War II. Common to all cases is a social situation in which
the person is traumatically stripped of social identity and in which
his or her expectations of stability are dramatically destroyed. A
nonperson, Seligman suggests, cannot remain alive, particularly
when the social structure that gives validity to existence predicts his
or her demise.
A high degree of status integration —occupancy of social
roles that set compatible norms —may also be conducive to a
strong sense of coherence. In a study subject to methodological criti-
cism but pertinent to our present topic, Dodge and Martin ( 1970)
review a considerable body of evidence relating status integration to
what they call stress diseases. Their focus is ecological rather than
individual. Of the 136 correlation coefficients on rates of chronic-
disease mortality and status integration, 89 percent were negative
as predicted.
A review of the literature on the particular social-structural
changes involved in bereavement led Jacobs and Ostfeld (1977,
p. 344) to conclude that "the attributable risk of mortality in per-
sons suffering a conjugal bereavement . . . may be as high as 50
percent." They apply Parkes' concept of psychosocial transitions
(1971) in analyzing this phenomenon and link it to Engel and
Schmale's (1972) conservation-withdrawal reaction. In the eight
studies covered, "a basic pattern of excess mortality in the widowed,
especially in males, is discernible. . . . The duration of the ele-
vated risk ... is no more than two years" (p. 349). The two
qualifications of the overall findings are particularly germane. The
sense of coherence of widows in Western cultures, I suggest, is less
vulnerable to the death of a spouse than is that of widowers. Fur-

ther, if one does indeed survive the traumatic impact on one's sense
of coherence, it is likely that recovery to the previous level can take
place in a relatively short time.
We now turn, in this review of social-structural factors, to
the work of the late John Cassel and his associates (Cassel, 1974,
1976; Kaplan, Cassel, and Gore, 1977; Nuckolls, Cassel, and
Kaplan, 1972). "A remarkably similar set of social circumstances,"
Cassel writes, "characterizes people who develop tuberculosis and
schizophrenia, alcoholics, victims of multiple accidents, and suicides.
Common to all these people is a marginal status in society. They
are individuals who for a variety of reasons
have been de- . . .
prived of meaningful social contact" (1974, p. 474). He later

speaks of "disordered relationships, this failure of various
. . .
forms of behavior to elicit predictable responses. . . . These be-

havioral acts are in a sense inappropriate. . . . The actor is not
receiving adequate evidence [feedback] that his actions are leading
to anticipated consequences" (pp. 476-477). These notions, de-
rived from animal studies (see below), are then applied by Gassel
to suggest an explanation of elevated blood-pressure levels among
Zulu urban migrants, of Syme's work on coronary heart disease
among migrants, of his own studies on heart disease in counties
varying according to degree of urbanization and the health status
of first- and second-generation factory workers, as well as of work
on mental disorders and stroke.

In a study summarized by Cassel and detailed in Nuckolls,
Cassel, and Kaplan (1972, p. 438), "the adaptive potential for
pregnancy score" was used to measure psychosocial assets. The
fascinating result of this study, for present purposes, is that "in
the presence of mounting life change . . . women with high psy-
chosocial assets had only one third the complication rate of women
whose psychosocial assets were low." To transpose into my terms:
Given a high level of stressors, which leads to tension, a strong
sense of coherence (psychosocial assets, social supports) acts to pre-
vent complications in pregnancy. In my own recalculation of the
data, there is even a hint that being high on stressors, given high
social supports, is salutary. This interpretation would fit the saluto-
genic hypothesis even more neatly.
The Kaplan, Gassel, and Gore (1977) paper is an attempt
of the Cassel group to clarify the social-support hypothesis and re-
view the literature, showing its consequences for biomedical events.
They conceive of a stressor situation as involving "the inability of
the individual to obtain meaningful information that his actions are
leading to desired consequences." Protective factors — for example,
social supports —are seen as interacting with stressors to "determine
to a considerable extent the susceptibility of the organism to physi-
cochemical disease agents" (p. 48). A creative and extremely use-
ful analysis of social supports follows.But the authors unfortunately
stop short of putting the question that begs to be asked: Why are
social supports related to health status? They have almost given the
answer in characterizing stressors: Social supports enhance the
ability of the individual to obtain meaningful information, or,

in my terms, enhance the sense of coherence. (For an additional
review of the social-support literature, see Cobb, 1976.)
A final study to be noted here is a review by Kiritz and
Moos (1974) of "dimensions of the social environment." Though
most of their citations refer to short-term physiological changes,
they do include a number of studies of morbidity — for example,
growth retardation and susceptibility to disease in infants, hyper-
tension, peptic ulcer, coronary heart disease, and rheumatoid
arthritis, as well as deaths among older persons. For present pur-
poses, what is most relevant is their systematic conceptualization of
three basic types of dimensions that characterize social climates:
relationship dimensions —involvement, affiliation, peer cohesion,
expressiveness; personal-development dimensions —autonomy, a
practical orientation, responsibility; and system-maintenance and
system-change dimensions — order, clarity, control, work pressure,
innovation. Under these headings, the authors review studies that

show how a given social environment, characterized by being high
or low on these dimensions, relates to certain health outcomes.
Again and again, one is struck by the ease with which one can
translate these dimensions into the overarching framework provided
by the concept of the sense of coherence. Thus, for example, they
cite a study of air-traffic controllers who, working under conditions
of heavy responsibility and extreme time -pressure, show higher risk

and earlier onset of hypertension and peptic ulcer than do those in
a control group (Cobb and Rose, 1973). This finding, in my view,
relates almost directly to the sense of coherence: the constant
bombardment by stimuli in a situation fraught with trepidation
that they cannot be handled coherendy and fraught with knowl-
edge of the possibility of disastrous consequences.
Cultural Variables. There is considerable overlap between
and cultural variables. A number of the studies
social-structural
mentioned above (Berkman, Dodge and Martin, Cassel) do not
find it necessary to distinguish between locations in a social structure
that impair the sense of coherence of their occupants and locations
in a cultural milieu that act in the same direction. Indeed, the
distinction is largely analytical. But it may be useful to briefly take
.
note of a number of additional studies that focus directly on

cultural variables.
We turn first to those studies that consider the urban en-
vironment, which is most aptly characterized as a cultural pattern
of information-input overload. While most of the work that is not
largely speculative is based on laboratory animal studies or does
not indicate clear pathology as a dependent variable, some results
do point to the relationship between noise, crowding, and so
forth and essential hypertension. The crucial variable to be studied
is, as Ostfeld and D'Atri (1975) put it, "an inability to structure
one's social and physical environment" in an urban culture. That

it is indeed the sociopsychological rather than the physical factors
of the urban environment that are crucial to health status is
convincingly demonstrated in a major collaborative effort (Hinkle

and Loring, 1977, particularly the papers by Hinkle and by Kasl).
Reviewing several hundred studies, Kasl does not accept the
oversimplified notion that "there is no evidence of any direct effects
of the residential environment on health." Despite some evidence of

such direct effects, he suggests that a more powerful model of the
complex relationship between environment and health should use
"the notion of a person-environment fit" as "the integrating con-
cept" (1977, p. 108).
More direct evidence on the relationship between morbidity
and cultural incongruity, or cultural transition or change, or how-
ever it is variously termed by different authors — all clearly related
to the sense of coherence —
is found in Wolff's edited work (Wolf
and Goodell, 1968). The editors write of American Indian tribes

"taken from their home land and put into reservations within a
few miles' distance, in essentially the same physical environment,
but in a setting of social disorganization, with a resultant appalling
increase in mortality from tuberculosis" (p. 8). A similar result
is reported for Bantu natives who had been moved from the coun-
tryside outside of Johannesburg into the environs of the city (p.
193). As a third example of morbidity and mortality associated

with a radical change in physical and social environment, they cite
the study of epidemics of meningococcus meningitis that occurred

with the onset of barrack life in the U.S. Army (p. 193 )
a
Most pertinent of all is their summary of the widely known

study of 100 Chinese expatriates in New York. With detailed in-
formation on health histories, Hinkle and Wolff (1957) were

able to distinguish between those who were consistendy healthy
and those who were, with a substantial degree of consistency, ill.
Wolf and Goodell summarize the study (pp. 204-205): "They
had shared a life in which they had all been exposed to a rapidly
changing culture, repeated disruptions of old social patterns, and
many physical dislocations. The healthiest members of this
. . .
group are people who are able to tolerate with some ease such
recurrent disruptions of their life patterns, partly because they re-
gard such changes and disruptions as a normal and expected part
of a life pattern. . . . Hinkle and Wolff infer that ill health . . .
appears to occur when an individual exists in a life situation which

places demands upon him that are excessive in terms of his ability
to meet them."
It seems reasonable to view these findings as corroborating
the hypothesis that when people, for whatever reason, are somehow
able to translate a difficult, complex bombardment of stimuli into
a whole that is meaningful, high health levels are likely to be
maintained. Thus, while there is little doubt that the evidence
points to cultural disruption and transition, which involve a weak-
ening of the sense of coherence, as dysfunctional for health, some
people are capable in these circumstances of maintaining a sense
of coherence. The crucial GRR in such cases, I would suggest, is
what I have called a rational, flexible, and farsighted coping

strategy.
In the major summary of the pertinent studies done in
Hinkle and Wolffs Cornell Human Ecology Study Program —
summary that reviews some of the above studies as well as those
of American telephone workers, managers, Hungarian refugees,
—
and others Hinkle (1974) writes in conclusion: "If a culture
change, social change, or change in interpersonal relations is not
associated with a significant change in the activities, habits, in-
gestants, exposure to disease-causing agents, or in the physical
characteristics of the environment of a person, then its effect upon
his health cannot be defined solely by its nature, its magnitude,
its acuteness or chronicity, or its apparent importance in the eyes
of others" (p. 42). Yet Hinkle has studied many people who have
undergone such changes without becoming ill. His key explanation
is that certain "psychological characteristics . . . help to 'insulate'
them from the effects of some of their life experiences" (p. 40).
The core of such psychological characteristics, I would suggest, is

In the early 1960s, Cassel's team conducted a number of
studies of first- and second-generation factory workers and of residents
in areas undergoing rapid urbanization and industrialization (Cas-
sel, Patrick, and Jenkins, 1960; Cassel and Tyroler, 1961; Tyroler
and Cassel, 1964). On the basis of their findings, they provided one
of the earliest formulations of the cultural-disorganization hypoth-
esis, totally consonant with the construct of the sense of coherence.
They write, "Rapid culture change is likely to have deleterious
health consequences when it leads to the development of incon-
gruities between the culture of the population at risk and the de-
mands and expectations of the new social situation" (Tyroler and
Cassel, 1964, p. 167). Unfortunately, this proposal seems to have
been too daring to them, for in closing their paper they retreat to
an explanation somewhat more palatable to the medical model and
focus on "deleterious changes accompanying urbanization . . .
[such as] diet, level of physical activity, cigarette smoking, and

meeting of deadlines" (p. 175). Not until a decade later (Cassel,
1974) was Cassel to develop his earlier, most important insight.
As a final note in this section, I would mention one of the
first works in the field. In a study of enlisted American naval per-
sonnel and civilians, Ruesch (1948, p. 91) found that "among
ulcer bearers there is an unusual frequency of individuals in the
process of culture change. . . . The results indicate that accultura-
tion and social mobility are one of the most important sources of
stress among ulcer bearers."
Psychological Variables. Since Rotter (1966) published his
work on the locus-of-control concept, a considerable amount of
research done on this psychological construct (admirably sum-
marized by Lefcourt, 1976) has been closely related to the sense of
coherence. For the most part, however, the studies have focused
on the consequences of the locus of control for problem solving
and other performance measures and for physiological parameters,
.
and they have been conducted in experimental laboratory situa-

tions. The focus on more direct measures of health has been
limited to animal studies (see below)

Not book (1975) was
until the publication of Seligman's
the attempt made to establish a clear link between external locus

of control —
or, in Seligman's terms, learned helplessness and the —
clinical entity of depression. I consider Seligman's work in Chapter
Five, where I try to show its close relationship to the sense of
coherence. I mention it again here in order to refer to Seligman's

chap. 5, which is devoted to analyzing the relationship between
learned helplessness and depression. Though providing no hard
evidence, he does provide a careful, systematic analysis of six
characteristics common to the two variables; this commonality pro-
vides reasonable evidence that the two variables are indeed related.
In this regard, and as a way of linking social-structural variables
to depression, note should be taken of Pearlin and Johnson's work.
Their review of relevant studies ( 1977; see also Pearlin, 1975) pro-
poses a link between "durable, structured conditions of life" (of
unmarried people) and depression; Seligman's hypothesis might
well be understood within this context.
Frank (1975) stresses the centrality of the "patients' own
healing powers" in "the restoration of healthy equilibrium." Despite
the fact that he clearly operates within the healthy-sick dichotomous
model and both as clinician and theoretician is concerned primarily
with restoration, his exploration of the issue of "expectant faith"
takes us one step beyond the amorphous will-to-live concept. In
addition to some of the studies mentioned above, Frank reports
others that have a direct bearing on the sense of coherence and
its relationship to health. Thus, for example, he writes of the
failure of a group of soldiers with schistosomiasis to recover as
expected. Interviewed, they were characterized as having a "de-
structive mental state"; ambiguity, a feeling that nobody was con-
cerned about their welfare, a lack of authoritative information,
alarmist rumors, and a perceived serious threat to survival were
the sources of their state of mind.
Other studies cited by Frank refer to information provided
before tonsillectomies, recovery from brucellosis and influenza, and
speed of healing of a detached retina. In all, Frank sees the psy-
chological state of the patient as intimately and causally related
to recovery. A patient with a high level of expectant faith is a

promising patient. His analysis goes on to consider the miracle
cures of faith healers and placebos, but for our purposes his in-
terpretation of the data is crucial. The physician, he writes, is able
to facilitate the desired state of expectant faith in the patient

because (my emphasis) "he mobilizes in the patient the attitudes
of trust and dependency that a child feels toward a good parent.
. . . His treatment is validated by a theory which both expresses
and confirms the world view which both he and
of the society in
the patient function. ... A shared world view both makes sense
out of life and reinforces the sense of group belongingness" (p. 52).
I have extremely serious reservations both about Frank's assump-
tions, indicated in this quote, about the reality of the patient-
doctor relationship and about the extremely manipulative approach

he suggests throughout his paper. These issues I explore in Chapter
Eight. For present purposes, it seems reasonable to infer that the
work reviewed by Frank has a clear affinity to the concept of the
sense of coherence.
Frank also goes on to cite the controversial work of Bahn-
son and Leshan with regard to cancer patients. Leshan's work
(1978; for a scathing review, see Medawar, 1977) can hardly
be cited as scientific evidence of the causal role psychological factors
may play in cancer, but germane because of his theoretical
it is
approach. A on working with some

psychotherapist, he reports
seventy cancer patients. He was repeatedly struck by a loss in these
patients of a reason for existing, which followed a loss of a relation-
ship of deep meaning (compare with Engel's concept of giving-up),
and by an absence of direction or goal. "They felt," he writes, "a
lack of any stable reference points for themselves in the universe,
... a and unrelatedness." This world view, he
sense of loneliness
argues, antedates the Both Leshan and Bahnson (1966,
illness.
1974) propose that cancer patients, more than controls, manifest

a personality style of repressing or denying unpleasant affects such
as depression, anxiety, and hostility.
Probably the most exciting and certainly the best known
work that bears a close relationship to the concept of the sense of
coherence is Ray Rosenman and Meyer Friedman's development
of the relationship between coronary heart disease and the Type A

behavior pattern. (For present purposes, I have thought it adequate
to cite, from the extensive literature, only the paper by Mathews
and others, 1977.) Serious evidence shows that individuals who are
"competitive, achievement oriented, involved in multiple activities
with deadlines, . . . impatient with slowness in others, [and who]
like to set a rapid work pace and tend to be hostile and aggressive"
have higher prevalence and incidence rates of coronary heart
disease (p. 489). In an attempt to identify the crucial elements in
the rather global Type A pattern, a detailed study was conducted
of 62 coronary cases and 124 coronary disease-free controls. All
were selected from a large-scale prospective study conducted in
California in 1960-1961. The data were collected when all re-
spondents were coronary disease-free. On the basis of the original

interview, each respondent was classified as having given a Type
A or Type B response on thirty-seven content items, three speech-
stylistic items, and four clinical-judgment items. Factor analysis
revealed these forty-four items to cluster in five scores: competitive
drive, past achievements, impatience, nonjob achievement, and
speed. Only competitive drive and impatience scores were found
to be significantly higher among patients than among controls.
I have gone into detail in reporting this study because it
is, for the time being, the culmination of a large body of serious
work on the major chronic disease facing Western societies. The
crucial question I would pose is: What is it about the Type A be-
havior pattern that explains its association with coronary heart
disease over and above the more conventional risk factors (smoking,
hypertension, cholesterol)? We would do best to consider the
discussion by Mathews and others (my emphasis) "A high drive :
level, coupled with impatience and hostility, is readily apparent in
the characteristic tendency of Type As to seek ever-expanding goals

and achievements. They have tried to change but have
. . .
reverted to their hard-driving activities as they found themselves

becoming increasingly anxious about work which still needed to be
finished and goals that had not yet been attained. . . . Impending
lack of control is experienced as anxiety arousing. . . . Pattern A
may indeed be interpreted as a response style for coping with
threats to a sense of environmental mastery and control" (p. 496).
Unfortunately, the authors do not explicitly take the final

step that would make their description of the Type pattern almost A
a perfect equivalent of a weak sense of coherence. The step just
begs to be taken by their own words. It is not just the competitive

drive and impatience that matter. The two together, and considera-
tion of the specific items that make up these factor scores, point
to an underlying global orientation that leads one to expect that no
matter how hard one tries life is so organized that one will never
succeed in its being under control. One of the authors, writing
elsewhere (Glass, 1977), indeed concurs with this interpretation:
"Type As are engaged in a struggle for control. . . . Pattern A be-
havior is a strategy for coping with uncontrollable stress; enhanced
performance reflects an attempt to assert and maintain control
after its loss has been threatened" (pp. 181-182).
In this paper, Glass also touches on an issue that is most

salient here. He makes the necessary distinction between Type A
as a behavior pattern and as a global orientation of learned help-
lessness (a weak sense of coherence), a concept he takes from Selig-
man. He suggests that Type A behavior, following "extended ex-
perience with salient uncontrollable stress, results in enhanced
vulnerability to helplessness among Type As" (p. 184), and hence
it is a precursor to coronary disease. Wedo not have evidence that
Type As are less vulnerable to other diseases. But even if this should
prove to be the case, it would not contradict the hypothesis that a
weak sense of coherence is crucial to dis-ease. Quite the contrary
may be true. In other words, a weak sense of coherence combined
with Type A behavior and failure to maintain environmental con-
trol (combined with other variables, genetic, constitutional, en-
vironmental, and behavioral) may lead to coronary heart disease.
A weak sense of coherence combined with other variables may lead
to cancer, and so on. The salutogenic orientation, as I have re-
peatedly stressed, compels us to search for themes common to health
ease, rather than factors in specific diseases.
Situational Variables. We now turn to the fourth major
category of studies that may be seen to provide indirect evidence
for the existence ofa relationship between a strong sense of co-
herence and health. Because the literature is vast, I limit myself to
the one volume that is the most valuable concentration of work in
.
the area, the proceedings of a conference on stressful life events
organized by Barbara and Bruce Dohrenwend in 1973 (1974; see

also Gunderson and Rahe, 1974).
The concern with life events as possible factors in the
etiology of various somatic and psychological disorders in this
century can be traced to W. B. Cannon and Adolph Meyer. For
present purposes, I confine my discussion to what has become the
major school in this area rather than consider life events as
equivalent withall stressors. The Holmes-Rahe Social Readjust-
ment Rating Scale and the Schedule of Recent Experiences have
been the most widely used instruments in this school. As Holmes
and Masuda (1974) describe the development of the forty-three-
item rating scale, it originated in the Cornell Human Ecology
Study Program when investigators studying the protocols of thou-
sands of patients were struck by repeated reporting of life events
"whose advent is either indicative of, or requires a significant
change in, the ongoing life pattern of the individual. The emphasis
is on change from the existing steady state and not on psychological
meaning, emotion, or social desirability" (p. 46). Holmes and his

colleagues were persuaded "that a cluster of social events that re-
quire change in ongoing life adjustment is significantly associated
with the time of illness onset" (p. 47). They turned to devisinga
weighting scale for each of the forty-three items, based on a broad
degree of consensus in a considerable number of populations, both
American and non- American. Thus death of a spouse is scored 100,
marriage, 50, and minor violations of the law, 1 1
Before the final development of this instrument, retrospec-
tive studies using the same approach had shown that "life events
cluster significandy in the two-year period preceding onset of tuber-
culosis, heart disease, skin disease, hernia, and pregnancy" (p. 57).
Later studies have shown relationships between life change scores
and sudden cardiac death, time of onset of myocardial infarction,

and fractures. Subsequendy, work began on prospective studies, the
best known of which is Rahe's study of U.S. Navy personnel aboard
three cruisers at sea for six months; he found that life change
Holmes and Masuda conclude
scores could predict reported illness.
their review of the data by saying (p. 67), "The greater the life
change or adaptive requirement, the greater the vulnerability or
lowering of resistance to disease, and the more serious the disease
that does develop. . . . Thus, the concept of life change appears

to have relevance to the causation of disease, time of onset of
disease, and severity of disease. It does not seem to contribute much
to an understanding of specificity of disease type."
Holmes has insisted that this approach was derived empiri-
cally and has worked consistently. Rahe (1974) has sought to de-
velop a more sophisticated theoretical understanding of why one
should expect to find a relationship between life change scores and
illness. He proposes a chain of events involving life changes, psy-
chological defenses that filter out some changes as nonsignificant,
physiological reactions to those events not filtered out, coping — at-
tempting to reduce physiological reactions, illness behavior, and

identified illness.
There are clearly many serious methodological problems,
and much more research is required before any firm conclusion is
possible about the life-events approach. The Dohrenwend volume
raises most of these problems and includes findings that stress the
negative life events, or events involving exits from social relation-

ships, as being important in facilitating illness. If indeed it should
prove to be the case that negative life events are critical to break-
down, I suggest such a connection would be consistent with the
sense-of-coherence approach. Rahe's theoretical position suggests
this relationship. The adult with a strong sense of coherence is,
if my analysis has been correct, certainly capable of mobilizing

resources to cope with the adaptive demands of life changes,
whether positive or negative. The person with a weak sense of
coherence meets the adaptive requirement with a sense of helpless-
ness, which becomes a self-fulfilling prophecy; he or she sees the
life change as not making sense and therefore is incapable of
successful adaptation.
It may be that Holmes' original position will prove to be
correct —that is, that life change scores per se are decisive for
health and illness, irrespective of whether people have a strong or
weak sense of coherence. My own position is unquestionably differ-
ent. There is, of course, a relationship between life events and the
sense of coherence, which I spell out in Chapter Seven. But, I
would hypothesize, given the same life-events score, people with
different strengths of the sense of coherence manifest different
health outcomes. And, conversely, I anticipate that people with
the same level of the sense of coherence but different life-events

scores are equally healthy. As a matter of fact, as suggested earlier,
I would not even be surprised if a moderately high life-events score
that does not shake up the sense of coherence turns out to be
salutary.
The life-events data, then, constitute one set of evidence
that does not support the sense of coherence thesis. As Rahe's recent
work suggests, however, there may not be as much of a contradic-
tion between the two approaches as first seems to be apparent. In
any case, I believe there is good reason to support considerable re-
search to test the alternative approaches.
A series of studies have dealt with one life situation that,
if anything, involves stability rather than change and has long

been presumed to be related to health. Crowding in the sense of —
both persons per room and density in a geographical area has in —
itself, over and above other factors associated with it, like poverty,
been implicated as an etiological factor. The common sense hy-

pothesis, based on considerable evidence that failed to separate out
crowding per se from other variables as a cause of infectious
diseases, merits closer analysis.
In one review of the evidence, Cassel (1977) casts con-

siderable doubt on the presumed relationship between crowding
and even infectious diseases. As he reads the evidence, the crucial
variable is not at all any measure of density but rather "the pat-
tern of relationship that exists" among the people being studied.
Disordered relationships, he suggests, are crucial to morbid con-
sequences. They "often have in common a failure to elicit antic-
ipated responses to what were previously appropriate cues and an
increasing disregard of traditional obligations and rights" (p. 133).
Cassel goes on to hypothesize: "In human populations, increased
susceptibility to disease should occur when, for a variety of reasons,
individuals do not receive any evidence [feedback] that their actions
are leading to desirable and anticipated consequences" (p. 133).
In other words, as the most recent evidence on crowding, including
studies in Hong Kong, Holland, and Canada (Booth and Cowell,
1976), suggests, when a high degree of crowding is accompanied by
social and interpersonal disorganization, there will be deleterious
consequences for health. The compatibility between Cassel's words
.
and the concept of the sense of coherence should be noted. In all

justice, I should make it explicit that Cassel's work and his words
have been a direct and influential factor in my development of the
concept.
Animal Studies. I do not hold much of a brief for using
laboratory studies of animals as evidence for a relationship between
the sense of coherence and health. Handled carefully, however, a
consistent pattern of results can give rise to fruitful hypotheses
about people. This is indeed the approach Cassel has used (1974,
1977). An early study with a unique twist —the idea for the study
is attributed to Spitz' work on maternal deprivation is reported —
by Liddell (1954). His experimental procedures were quite suc-
cessful in showing how "psychological stress" can lead to neurotic
behavior and death in sheep and goats. The same stress introduced
in the presence of the animal's mother, however, brought no ad-
verse consequences. Cassel (1974) and Seligman (1975) both re-
view the experimental evidence, all of which seems to indicate that
it is the social conditions under which stressors electric shock, —
insoluble problems, crowding, deprivation —are imposed that are
decisive in health outcomes. Unfortunately, these experiments have
all focused on pathogenic outcomes, with much less careful manip-
ulation of the variables that might be decisive in salutogenic
outcomes.
Engel ( 1974), it seems to me, has come closest to formulat-
ing an interpretation of many animal studies. Cleverly, he "asked"

rats, as a good clinician would do, when they develop ulcers. The
answer he was "given" is interpreted in accord with the giving-up
hypothesis. Engel's words, however, are almost closer to the con-
cept of the sense of coherence than to his own formulation: "the
immobilized rat [was unable] to do anything about his situation.
. . . The most effective protection against ulcerogenesis was
afforded by devising situations in which the animal received re-
liable feedback information that what it was doing was the right
thing to avoid shock. . . . The organism [lost its ability] to predict
and maintain control over its environment" (p. 1090)
I have reviewed a considerable number of empirical studies

relating psychosocial factors to health outcomes. With few excep-
tions, the results of these studies are at least compatible with hypoth-
eses that would be derived from the concept of the sense of coherence.
This compatibility suggests that there are serious grounds for a large-
scale empirical program of research. Yet I would go much further.
I propose that this construct is not just another variable to be stud-

ied. It is, a most parsimonious way of subsuming a great
rather,
variety of discrete variables, from mother's presence through work
overload-underload balance to information provided to patients.
—
But two further clarifications or, rather, reiterations are —
required. First, I have explicitly committed myself to the thesis that
the concept of a sense of coherence is a global orientation that is
pervasive and enduring. My contention has been that fluctuations
over relatively limited periods of time or in different situations are
relatively minor (barring traumatic experiences; see Chapter
Seven). This contention must, of course, be put to the test. The
methodological implications are considerable as far as populations,
situations, and tools of measurement are concerned.
Second, I have intentionally made no mention of the mech-
anisms and channels through which the sense of coherence is related
to health. Not only is this question outside the scope of my profes-
sional competence, but a number of workers in the field have writ-
ten quite plausible postulates concerning the all-important link. The
reader concerned with this question might do well to look at the
work (going beyond Selye, of course) of Cassel (1974), Glass
(1977), Kagan and Levi (1974), and Levi (1974). In considering
this work, I would point up one distinction that, as I indicated in
Chapter Three, seems to me to be crucial — the fundamental distinc-

tion between a state of tension and a state of stress. Not all those
who analyze the stress responses of organisms to psychosocial stressors
— or, in my terms, the response of people to living in general, what-
ever the strength of the sense of coherence —are sensitive to the dif-
ference. Kagan and Levi come fairly close (p. 227) : "It has been
claimed that if a sympatho-adrenomedullary stimulation lasts too
long or is be unpleasant func-
repeated too often, the result will first
tional disturbances in various organs and organ systems. . . .
Such a dysfunction, if long standing and/or intense, may result in

permanent structural changes of pathogenic significance at least
in predisposed individuals."
The work, however, that seems to me to be most congenial

to the salutogenic approach is Solomon and Amkraut
that of
( 1974). True, they and emotional distress
initially state that "stress
may influence the function of the immunologic system via central

nervous system disease and endocrine mediation." Their concept of
immunological balance, however, opens up the way for the study of
how a strong sense of coherence may be linked to maintaining one's
location on or moving toward the ease end of the health ease/ dis-
ease continuum. It may be that when confronted with stressors, the
person with a strong sense of coherence can activate a variety of
immunological factors, thereby preventing tension from being trans-
formed into stress.
I have now completed the inevitably partial discussion of
each segment of the salutogenic model. The time has come to put all
the pieces together. For this, we turn to Chapter Seven.

Chapter Seven
The Salutogenic
Model of Health
In Chapter One, I posed the problem of salutogenesis. Chapter Two

proposed a solution to the problem of the measurement of health
status consonant with the salutogenic orientation. At that point, the
core of the question was put as the need to explain the location of a
person near the ease end of the health ease/dis-ease continuum.
— —
Chapter Three considered and rejected the hypothesis that the
answer could be stressor avoidance. In Chapter Four, an initial
alternative answer was presented: the availability of generalized
resistance resources. The initial question was also broadened to con-
sider maintenance or improvement of one's position on the break-
down continuum, irrespective of location at any given time. Analysis
of the nature of generalized resistance resources, of why they are
hypothesized to facilitate tension management and avoid stress, led
to the formulation of the central construct of the book, the sense

of coherence, considered at length in Chapter Five. The final build-
182
The Salutogenic Model of Health 183
ing block in what I call the salutogenic model appears in Chapter

Six, which presents the evidence for linking the sense of coherence
and health status.
Inevitably, detailed consideration of each building block ob-
scures the integral character of the model as a whole. Paradoxically,
further, it oversimplifies a complex of interrelationships. Finally,
it tends to obscure the links between the variables in the model.
The function of this chapter, then, is to overcome these difficulties.

The full picture, as I see it, is presented in Figure 1. A model frozen
in a diagram unfortunately has a static character. It takes a leap
of the imagination to transform both the elements in the model and
the arrows and lines indicating their interrelationships into a dy-
namic whole in space and particularly in time. But the diagram is
the best I can do, and I would ask the reader to refer to it as each
element and link is discussed.
Sense of Coherence
Studying a diagram or discussing it in words requires an ele-
ment of arbitrariness in selecting a point of departure. It is nonethe-

less not accidental that I start the discussion from the sense of
coherence. This is, after all, the core of my answer to the problem
of salutogenesis. The sense of coherence is measurable; each of us is
located at some point on the sense-of-coherence continuum, which

can be seen as an ordinal scale. Sense of coherence is an orientation
that is not situation- or role-specific. Although there may be situa-
tions or issues with regard to which a person with a strong sense of
coherence can be utterly perplexed, these are essentially peripheral
to one's life or mark minor fluctuations around a fairly stable loca-
tion on the continuum. Given the nature of human existence, it is

difficult to conceive of anyone being extremely high on the con-
tinuum. This would require an unimaginably stable world, an in-

conceivably unchanging internal and external environment. Only
someone who is totally out of touch with reality could claim to have
an absolute sense of coherence. In fact, in Chapter Five I suggest
that a good clue to a fake sense of coherence would be an extremely
high score. Most of us, then, would score from extremely low to
moderately high.
Figure 1
The Salutogenic Model
TIN
Socle
Cetera!
and
Historical
Context
Potential Endogenic and Exogenic Stressors
A. Psychosocial Stressors
Public
1 . accidents and survivors 8. other normative
and
2. others' experiences crises
Private
3. horrors of history, 9. conflicts in
(avoid or neutralize) Health
direct and vicarious social relations
Measures
4. intrapsychic conflicts 10. goals-means gap
5. fear of aggression
6. Immediate world change b. Physical and Biochemical
7. phase-specific crises Stressors
Other Ease/Ois-ease
Continua
Line K: A strong sense of coherence, mobilizing GRRs and SRRs, defines stimuli
as no n stressors.
Arrow L: Ubiquitous stressors create a state of tension.
Arrow M: The mobilized GRRs (and SRRs) interact with the state of tension and
manage a holding action and the overcoming of stressors.
Arrow N: Successful tension management strengthens the sense of coherence.
Arrow 0: Successful tension management maintains one's place on the health

ease/dis-ease continuum.
Arrow P: Interaction between the state of stress and pathogens and "weak
links" negatively affects health status.
Arrow Q: Stress is a general precursor that interacts with the existing potential
endogenic and exogenic pathogens and "weak links."

Arrow R: Good health status facilitates the acquisition of other GRRs.
Note: The statements in bold type represent the core of the salutogenic model.
Key to Figure 1
Arrow A: Life experiences shape the sense of coherence.

Arrow B: Stressors affect the generalized resistance resources at one's
disposal.
Line C : By definition, a GRR provides one with sets of meaningful, co-

herent life experiences.
Arrow D: A strong sense of coherence mobilizes the GRRs and SRRs at
one's disposal.
Arrows E : Childrearing patterns, social role complexes, idiosyncratic factors,

and chance build up GRRs.
Arrow F: The sources of GRRs also create stressors.
Arrow G: Traumatic physical and biochemical stressors affect health status

directly; health status affects extent of exposure to psychosocial
stressors.
Arrow H : Physical and biochemical stressors interact with endogenic patho-

gens and "weak links" and with stress to affect health status.
Arrow I: Public and private health measures avoid or neutralize stressors.
Line J : A strong sense of coherence, mobilizing GRRs and SRRs, avoids
stressors.
Line K: A strong sense of coherence, mobilizing GRRs and SRRs, defines

stimuli as nonstressors.
Arrow L: Ubiquitous stressors create a state of tension.
Arrow M: The mobilized GRRs (and SRRs) interact with the state of ten-
sion and manage a holding action and the overcoming of stressors.
Arrow N: Successful tensionmanagement strengthens the sense of coherence.
Arrow O: Successful tension management maintains one's place on the
health ease/ dis-ease continuum.
Arrow P: Interaction between the state of stress and pathogens and "weak
links" negatively affects health status.
Arrow Q: Stress is a general precursor that interacts with the existing poten-
tial endogenic and exogenic pathogens and "weak links."

Arrow R : Good health status facilitates the acquisition of other GRRs.
Note: The statements in italic type are the core of the salutogenic
model
To say that the sense of coherence is stable, enduring, and

pervasive does not, however, compel us to say that it is immutable.
In what sense, then, is it dynamic? We would do well to divide the
answer to this question into two parts. The first focuses on the de-
velopment of the orientation in childhood, adolescence, and early
adulthood. The second considers modifications throughout subse-
quent life. The two parts overlap: there can be sharp changes of
direction in childhood; there is development throughout life until
death. But it is analytically useful to deal separately with the emer-

gence of a sense of coherence and then with its modification.
Life Experiences
As indicated in the diagram, life experiences (Arrow A) are

crucial in shaping a sense of coherence. From the time of birth, or
even earlier, we constantly go through situations of challenge and
response, stress, tension, and resolution. The more these experiences
are characterized by consistency, participation in shaping outcome,
and an underload-overload balance of stimuli, the more we begin
to see the world as being coherent and predictable. When, how-
ever, one's experiences all tend to be predictable, one is inevitably
due for unpleasant surprises that cannot be handled, and one's sense
of coherence weakened accordingly. Paradoxically, then, a meas-
is
—
ure of unpredictable experiences which call forth hitherto un-
—
known resources is essential for a strong sense of coherence. One
then learns to expect some measure of the unexpected. When there
is little or no predictability, there is not much one can do except
seek to hide until the storm (of life) is over, hoping not to be no-
ticed. Or else one strikes out blindly and at random until exhaustion
sets in. No defense mechanisms can be adequate.
We must note an implicit assumption here. If a strong sense
of coherence is must be not only by
to develop, one's experiences
and large predictable but also by and large rewarding, yet with some
measure of frustration and punishment. The outcome depends on
the underload-overload balance. But what if one's life experiences
are largely consistent and predictable but frustrating and punishing?
Again, the answer is a matter of degree. Frustration and punish-
ment can be so devastating that survival is put into question. If
they are not so extreme, then defense mechanisms become possible
and a reasonably strong sense of coherence begins to form.
One emerges from childhood, then, with some formed albeit
tentative sense of coherence. In adolescence, the crucial stage for ego
identity, tentativeness begins to be transformed into definitiveness.
If one's experiences continue to be by and large cut of the same
cloth as earlier experiences, one's sense of coherence is reinforced.
—
Yet considerable change is possible. The important point is that

there is increasing room for choice. The child maintains a small
number of salient relationships. He gets feedback from relatively
few people. The stimuli are not too variable. The adolescent has
greater options in choosing or encountering experiences that en-
hance or weaken his or her sense of coherence.
Entering young adulthood, one has acquired, as it were, a
tentative level of the sense of coherence, a picture of the way the
world is. One now makes major commitments: marriage and a new
nuclear family; the work at which one will spend most of one's wak-
ing hours; a style of life; a set of social relationships. These provide
one with a relatively stable set of life experiences, day after day
and year after year. By the time a decade or so has passed, if not
sooner, the tentativeness has been transformed into a considerable
degree of permanence. One selects and interprets experiences to
conform to the established level of the sense of coherence. It is un-
likely, then, that one's sense of coherence, once formed and set, will
change in any radical way. Fluctuations will be minor.

But unlikelihood is not certainty, which brings us to the
second part of our answer to the question about the dynamic nature
of the sense of coherence: modification of the sense of coherence.
We can point to two major ways in which an adult's sense of co-
herence can undergo fairly significant transformations. First, there
is the cataclysmic stressor, in either a broad or a personal sphere,

which transforms a great variety of life experiences, often in a brief
period of time, through a considerable change in one's GRRs
(Arrow B) One has had no hand, no choice, in this experience and
.
often no preparation for it. Perhaps the classic example is sudden

widowerhood. War, forced migration, the death of one's child, los-
ing one's job because the plant closed down, a natural disaster
central to all these events is not primarily that they are largely un-
anticipated, in a personal sense, but that they bring in their wake a
variety of unpredictable experiences. Inevitably, then, they result in
a significant weakening of one's sense of coherence.
Is such weakening necessarily permanent? To ask this ques-
tion is to point to the second major way one's sense of coherence
can undergo a significant modification. By contrast to the first, it is
never sudden, almost always has an element of choice (conscious or
The Salutogcnic Model of Health 189
unconscious), and can result in movement in either direction on the

sense-of-coherence continuum. Let us take widowerhood as an ex-
ample. Whatever one's previous level of sense of coherence, this is
inevitably a major disruption of one's life, particularly when there

has been no anticipatory socialization. Slowly and painfully, one
can choose experiences that, offering meaningful stimuli, rebuild
one's sense of coherence. Or, if one is lucky, such experiences are
thrust on one. No less can the opposite pattern characterize one's
life (see Parkes, 1972; Parkes, Benjamin, and Fitzgerald, 1969). In
parallel fashion, and not necessarily as a result of a cataclysmic
stressor, a woman can go out to paid work after decades of beinga
housewife; an illiterate person can learn to read and write; one can
undergo psychotherapy; one can embark on a substantially different
kind of work; one can marry or divorce. Change, then, can take
place. But change of this type is always within the context of one's
previous level of the sense of coherence, is always slow, and is always
part of a web of life experiences that transmit stimuli that are more
or less coherent. Movement toward the strong end of the continuum
always requires hard work.
Generalized Resistance Resources
If one's life experiences, then, shape one's sense of coherence,

what shapes one's life experiences? What determines whether they
consist of coherent or incoherent stimuli and are characterized by
consistency, participation in shaping outcome, and neither under-
load nor overload? Part of the answer the effect of stressors was — —
given above (Arrow B) But the greatest part of my answer is one's
.
generalized resistance resources (Line C). GRRs are discussed in

great detail in Chapter Four. By definition, a GRR provides one
with sets life experiences. Thus Line C is a
of meaningful, coherent
symbol more for tautology than for causality. If material resources
or a flexible coping strategy or social supports by definition provide
coherent life experiences, if that is their hallmark, one cannot
quite say that the relationship is causal, for there is no way of testing
the truth of the statement. The value, however, of separating GRRs
from life experiences in the diagram, and the meaning of Line C, is
that we are thereby provided with a theoretical criterion, a culling

rule, for identifying GRRs. The empirical prediction, which can be

tested, is the relationship between GRRs and the sense of coher-
ence, which can be defined and measured independently.
At the present stage in the development of the salutogenic
model and without considerable empirical research, we have no
basis for predicting the structure of the relationship between GRRs
and the sense of coherence. Is a given GRR—for example, a clear,
stable ego identity or social supports —a necessary or even a neces-

sary and sufficient condition for a strong sense of coherence? Or, if
such generalizations are impossible, are some GRRs more useful

than others in coping with certain (Arrow D, Figure 1 )?
stressors
One of the advantages of the salutogenic model is that it allows

us —indeed, even stimulates us — to ask such questions, whatever the
answers turn out to be.
Sources of GRRs
We can now move further "back" in the diagram. Arrows E

point to the sources of the GRRs. These have been discussed at
length in Chapter Five. I would only briefly point up a number of
issues that may have been slighted earlier. First, whatever the some-
what cavalier approach I may seem to have taken with regard to
the role of stressors in influencing health status, I did so only to
offset the strong current concern with stressors. As will be evident
shortly, this position must be qualified. I take it here in order to
point out that there are also direct links between the sources of
GRRs and stressors(Arrow F). Someone growing up and living in
a society withan annual per capita income of $250 confronts differ-
ent stressors and has different GRRs at his or her disposal than does
someone growing up and living in a society with a per capita income
of $2,500. Living in a world with limited means of transportation
and communication and weapons of destruction is quite different
from living in a "civilized" world of satellites, 747s, and nuclear
weapons. Whether one's society has pacific or hostile relations with
its neighbors matters a great deal both for the GRRs and the
stressors in one's life. It is perhaps even more important, in these
terms, to distinguish between people who are members of different

social classes, sexes, or ethnic groups within one society.
Second, no grounds for assigning priority to one or the

I see
other of the two major sources of GRRs (childrearing patterns and

social-role complexes) in shaping the GRRs at one's disposal. At
first sight, it might seem clear that there is a greater affinity between
childrearing patterns and ego identity and between present social
class position and material resources. But ego identity and adult
major role activityand role set are deeply intertwined, as are pa-
rental social class and present material resources. The relationships,
then, are complex. Then again, they may well be not as complex as
seems to be the case. There is at least some reason and empirical
evidence to think that there is a wholeness in the direction in which
one is constrained to go by the complex of one's childrearing pat-
terns and social-role complexes, in a given sociocultural and histori-
cal setting. The strain toward consistency among these acts to push
one in the direction of a greater or lesser degree of GRRs. This
consistency enables us to speak of the prototypical life chances of
an individual in a given subculture — for example, of a poor, white
housewife in Appalachia.
The relationship between cultural and individual factors
brings us to a third point. Whatever the very major power of socio-
cultural and historical factors in shaping the GRRs at one's dis-
posal, we are witness to substantial individual differences. Intel-
ligence, however it may be measured, is distributed on a normal
curve. Beauty, charm, strength, and a myriad of other personal
characteristics, however these are measured, vary from person to
person. Of course these are measured and evaluated differently in
different cultural settings. Of course they are influenced and per-
ceived differently depending on the social context. But these idio-
syncratic characteristics and tendencies are not therefore irrelevant
in shaping the GRRs at one's disposal. One can make reasonable
probabilistic predictions knowing a person's sociocultural world as
to where he or she will rank on GRRs. But the prediction will never
be close to perfect for a given individual.
Finally, we must note the role of chance as a source of
GRRs. The luck that confronts us is often far from a matter of luck.
Further, one person may take advantage of a lucky opportunity
while another may not.One may even be so strong a determinist as
to claim that everything that happens had to happen. There are,
—
nonetheless, chance events that may often be of considerable sig-

nificance in shaping one's GRRs. Rarely does one buy a sweepstakes
ticket, chat with a stranger while waiting for a train, register for a
class because it is held at a convenient hour, leaf through the per-
sonal ads in a magazine, or get invited to a party and thereby em-
bark on a chain of events that substantially alters one's life. Rare.
Far less important than childrearing patterns and social-role com-
plexes. But it does happen.
Stressors
Let us now assume * we have accounted adequately for

the emergence of a given ">f the sense of coherence that char-
acterizes a person at a g b le and return to the center of the

diagram. Before we can i the relationship between the sense
of coherence and health, \< t focus on the place of stressors in
the salutogenic model. I on this issue above in considering
the impact stressors can 1 i GRRs (Arrow B). Now we look
elsewhere.
Chapter Three opent jl with a statement pointing to the evi-
dence of the relationship between stressors and movement toward
the dis-ease end of the health ease/dis-ease continuum. This state-
ment requires some clarification. Biochemical and physical stressors
droughts, bombings, invasion, pests —much as psychosocial stressors,
can have an impact on GRRs. But unlike psychosocial stressors,
whose impact is always mediated through GRRs and the sense of

coherence, biochemical and physical stressors can be of such direct
traumatic magnitude as to bypass interaction with the sense of co-
herence. A noxious gas,
a pc isonous substance, a bullet, or a car can
act direcdyon the health status of an individual (Arrow G; double
arrow shows two-way causation). Alternatively, a cumulative harsh
overload of such stressors (smoking or exposure to asbestos or to
high noise levels) can act mdirecdy on health by exploiting the
endogenic potential pathoge is and "weak links" in interaction with
a state of stress (Arrow H; double arrow shows interaction).

There is, indeed, good reason for the pathogenic model to
have dominated thinking about disease for most of human history.
The three-pronged power of stressors (Arrows B, G, and H), which
included perhaps above all nutritional deprivation and the most

primitive level of sanitation, was sufficient to overcome even sub-
1
stantial resistance resources. When, however, the standard of living
in a society (or in some segments of a society) reaches a rough level
of adequacy, differences in health level no longer are overwhelm-
ingly determined by biochemical and physical stressors. At this
point psychosocial stressors and, above all, the sense of coherence

become crucial variables. And at this point salutogenesis becomes at
least as intriguing and important a question as pathogenesis.
Management of Tension
Having analyzed the sense of coherence as a dependent

variable,we now turn to consider it as an independent variable.
The achievement of a roughly adequate standard of living does not
do away with physical and biochemical stressors. They remain ubiq-
uitous. For the first time in history on a large scale, however, it has
now become possible to cope with them. Success has been remarkable.
The triumphs of public health and of the microbiological sciences
have been great (Arrow I). But there is a built-in limitation, which
Dubos ( 1960) has profoundly analyzed. The bugs, as I have put it,
are smarter. Not always, not every bug, and they have retreated. In
this era of chronic diseases (and not much less applicable to infec-
tious diseases in such an era) the single-bullet approach can no
longer be seen as viable in and of itself or even as the dominant
weapon. In this context the sense of coherence becomes important.
As shown in the model, the role of the sense of coherence is
three-directional. First, by mobilizing the GRRs at one's disposal

(Arrow D), (SRRs), a strong
as well as specific resistance resources
sense of coherence can avoid one's being subjected to some stressors
(Line J). Second, it allows us to define some stimuli, which others
might perceive as stressors, as innocuous or even as welcome (Line
K) But whether we like it or not, none of us can in such ways keep
.
1
Yet even during the worst plagues, some remained healthy and some
recovered. Had our focus been salutogenic, we might have learned much more
about GRRs than we know today. For a brilliant if tongue-in-cheek paper
analyzing the remarkable plague immunity of an ethnic minority in ancient
Egypt thanks to a powerful GRR technically called Bohbymycetin, see
Caroline and Schwartz (1975) on chicken soup.
stressors out of our lives. Day in, day out, throughout our lives, we
find that stressors put us repeatedly in a state of tension (Arrow L).
Periods of calm and stability, of homeostasis, are rare in human
existence. At this point the third direction in which a strong sense of
coherence operates is decisive. It would hardly be important were
stressors reducible to an occasional experience (which is almost
never the case), much as it is not very important when stressors
are overwhelming. We respond to a state of tension, if we have a
strong sense of coherence, by mobilizing those GRRs that we have

at our disposal and that we judge to be appropriate in seeking to
resolve the tension by overcoming the stressor (Arrow M; double
arrow shows interaction). (See Pearlin and Schooler's 1978, pp.
6-7, discussion of the three functions of coping responses: to modify
situations, to control the meaning of situations, and to control the
stress.)
We can now clarify the dual function of GRRs. Earlier we

noted their function in creating life experiences that produce a
strong sense of coherence. In this sense, they are constandy active.
But they also function as a potential. Someone with a strong sense
of coherence, whatever its sources, confronted with tension, can call
on the GRRs to manage the tension successfully. One brings to bear

one's wealth, one's knowledge, one's strong ego identity, one's social
supports, and so forth. Note that this approach does not obviate the
need for GRRs of the internal environment to perform a holding
action. Physical and biochemical resources are required to prevent
too rapid a transformation of tension into stress (Selye's stage of
exhaustion). But the crucial role of GRRs is in overcoming the
stressor and thereby resolving the tension. One must add that the
person with a strong sense of coherence can also directly mobilize
SRRs appropriate to the particular stressor. Finally, it bears repeat-
ing in the present context that overcoming a stressor and resolving
tension is a life experience that in turn reinforces the sense of co-
herence (Arrow N). By overcoming a stressor we learn that exis-
tence is neither shattering nor meaningless.
From a dynamic, historical point of view, the dual function
of GRRs can now be seen as one. Since conflict and stressors are
ubiquitous throughout life and hence tension is at least as character-
istic of human beings as is homeostasis, one is from earliest infancy
calling on whatever GRRs are at one's disposal. When they suffice
to provide a life experience that makes sense to us —that is consistent

with our expectations, allows us some participation in determining
outcome, and has neither too few nor too many stimuli for us to
handle —and thus allow us to resolve tension, another building block
is added to our sense of coherence. This theoretical approach, it
should be noted, underlies my analysis of the relations between peo-

ple and communities and the health care institution in Chapter
Eight.
Stress
Given the initial statement of the problem as that of saluto-

genesis, there remains but one issue to consider at this point, the
feedback impact of health status. That is, in the never-to-exist soci-
ety in which people are never harmed by traumatic or cumulative
physical and biochemical insults, protected as they are by public
and private health measures; in which all persons have a very strong
sense of coherence of mobilizing GRRs and
and hence are capable
SRRs — in such a never-never land, one knows much tension but
never stress. And so all would be on the extreme ease end of the
health ease/dis-ease continuum, at least until one emulates Oliver
Wendell Holmes' wonderful one-horse shay in a dramatic ending.
Over and over again, we would manage tension successfully, thereby
reinvigorating our sense of coherence (Arrow N) and at least
maintaining our easeful health status (Arrow O). But given the
brilliance of the bugs and the inevitable inadequacies of the sense
of coherence and of exogenous GRRs, even the most fortunate are
bound, on occasion, to fail to manage tension well. Cousins, it will
be recalled, did come down with ankylosing spondylitis. Freud did
suffer from cancer of the jaw.
There is no need here — particularly since I do not intend to
deal with theories of diseases — to discuss the relationship between
a state of stress and movement toward the dis-ease end of the health
continuum (Arrow P). This is the focus of attention of almost the
entire stress literature. I can make no contribution. Of far greater
significance, the thrust of this entirebook is to propose a shift in
concern to the study of successful tension management. Two points
are germane here.
First, it seems clear to me that stress is a general precursor.
.
Only when stress interacts with the existing potential endogenic and
exogenic pathogens do pathological consequences occur (Arrow Q)
As Selye puts it (1975, p. 41) (notwithstanding his unclear use of
the word tension) "Although stress itself is defined as the 'non-
:
specific response of the body to any demand,' the weakest link in a

chain will be the one that selectively breaks under tension. Simi-
larly, the weakest part of any animate or even inanimate machine
will be the one that fails when a nonspecific, general demand is
made upon the performance of the whole." A most important corol-

lary of this approach is the rejection of the concept of psychoso-
matic disease. Other than the massive traumata that leave none
unscathed (Arrow G), all diseases are usefully understood as psy-
chosomatic. In other words, almost all breakdown involves stress.
Stress, however, does not determine the particular expression of the

breakdown.
My second point is crucial to the salutogenic approach. The
pathogenic orientation asks: What become ill
causes a person to
with a particular disease? The salutogenic orientation, by contrast,
asks: Whatever the person's particular location at any given time
on the health ease/dis-ease continuum, what are the factors that
facilitate his or her remaining at that level or moving toward the
more salutary end of the continuum? Thus no assumption is made

that one is well and becomes sick. On the contrary, the commitment
is to seeing people at some point on the health continuum at any
given time and continually confronted with stressors and hence with
the problem of preventing tension from becoming stress. In this way,
the sense of coherence is always hypothesized to be a relevant factor.
Health
Which
brings us to the final issue. Heretofore, we have
viewed one's location on the health ease/dis-ease continuum as a
dependent variable. We have seen it as the final outcome of a long
chain of phenomena. Such analytic albeit complex neatness is dis-
torting. One's health status can be usefully viewed as an independent

variable in three ways. First, it can affect the extent to which one is
exposed to stressors (hence Arrow G points in both directions). At
a high health level, conflicts in social relations may be attenuated or
.
phase-specific crises borne with equanimity. Second, good health is
in itself a significant generalized resistance resource by the defini-

tion of a GRR as a factor that fosters meaningful and sensible life
experiences. Third, in the same way that the other GRRs are inter-
related, being on the healthy end of the health continuum can
facilitate the acquisition of other GRRs (Arrow R)
Throughout, and particularly in the final section of Chapter

Two, I have insisted that the health ease/dis-ease continuum is not
to be regarded as coextensive with the entire realm of well-being.
Other ease/dis-ease continua exist. To have entered into a sys-
tematic discussion of what these continua are, how our locations on
them are determined, and how they relate to our concern with
health would have been an impossible and unnecessary task within
the scope of this book. Suffice it to say, then, that a nod has been
made in their direction; they are highly relevant to and intertwined
with health, but they are distinct. If our interest is in understanding
health, then location on the family-relations or social-relations or
material-resources ease/dis-ease continua can usefully be viewed as
a GRR.
Reality, for better or for worse, is complex. The attempt to

understand reality of necessity oversimplifies in that it must select
and abstract. In this chapter, I have sought to minimize such over-
simplification. Hence the perhaps bewildering array of arrows and
lines and boxes in Figure 1. Doubdess the attempt has fallen short
in one sense or another. Thus, for example, I have certainly dealt
inadequately with genetic and constitutionalGRRs and with public
and private health measures. Genetic and constitutional GRRs are
too complex and are beyond my capacities to explore here; public
and private health measures are considered in Chapter Eight and
in the Epilogue. I would be troubled, however, only if the central
paradigm of the book was not clear. This paradigm finds its graphic
expression in the diagram. Stripped of all qualifications and com-
plexities, the thick lines in the diagram and the italicized words in
its key are what I have to say.
Chapter Eight
Implications
for an Improved
Health Care System
Let me make my position clear at the outset of this chapter, which

discusses the implications of the salutogenicmodel for relationships
of individuals and groups with health workers and the health care
system. I do not in this chapter pretend to have any instant therapy
or guide to the perplexed. In my discussion of the sense of coherence
I have viewed it as a deeply rooted, pervasive orientation. I have
located its sources primarily in childrearing patterns and social-role
complexes and have described it as developing and being rein-
forced over the course of many years. To blame a weak sense of
coherence, as many do, on doctors or the health care system is sheer
demonology. A culture tends to manifest a strain toward consistency.
If, indeed, many of the pressures impinging on a given population
group are in the direction of forming a weak sense of coherence,
198
Implications for an Improved Health Care System 199
then it is members' experiences with health care work-

likely that its
ers and the health care institution will point in the same direction.
It is easy but fruitless to play the game of blaming the doctors. The
precise opposite is no less true. To imagine that the individual health

worker or health care institutions in general can go far in fostering
a strong sense of coherence when other social institutions pressure
1
in the opposite direction is illusory.
The Patient and the Health Care System
Having said this, however, is not at all tantamount to saying

that the nature of the life experiences of persons acting within the
health care system has no relationship to the sense of coherence. This
relationship can be analyzed on three levels: traumatic situations;
incremental and relatively slight reinforcement or weakening of the
sense of coherence; and the more significant impact of the health
care system in the community. But before these are analyzed, a
number of preliminary considerations are in order.
First, we are concerned with the prototypical patterns of en-
counters and not with the idiosyncratic relationship that may de-
velop between a patient and a doctor. That relationship is shaped
largely by personality factors. Personality factors do, of course, enter
any particular relationship, whether barefoot doctors or Harley

Street consultants, confused adolescents or strong longshoremen are
involved. But the arena for play of personality, by and large, is not
that great. Close observation of large series of encounters or rela-
tionships would reveal systematic patterns that are shaped by struc-
1 In paragraph, I have referred interchangeably to doctors and

this
to other health care workers. It will be awkward to continue to do so. I will,
therefore, refer only to doctors. In principle, my analysis of the relationship
between the individual (or family) and the doctor can include relationships
with nurses, dentists, medical social workers, and so forth or with health care
teams. In practice, for good sociological reasons, the physician is most often
central in the relationship even when authority is delegated to others. Another
terminological difficulty is the use of the word patient. Given the fact that our
health care system is predominantly a disease care system, the word patient
implies that the person is in some way sick. In the framework of the health
ease/dis-ease continuum, what I mean by patient is anyone who is in a
reciprocal role relationship with a physician, whatever the location on the
continuum. Below I use particular terms for patients and for doctors as
indicative of the particular type of role relationship.
tural and cultural factors. Hence our central question is not the
moral, What should be the relationship? nor is it the abstract, What
can it be? Rather, we shall
ask, Given a certain normative relation-
ship between person and health worker, what are its consequences?
Second, if consequences, then one must specify consequences
for what. We are not concerned with an overall analysis of the
functions of the patient-doctor relationship. The dependent variable
in this chapter is the sense of coherence. The case has been argued
that a person's sense of coherence is a major variable in the deter-
mination of his or her health status. Let us assume that this has
been demonstrated. To the extent, then, that what the doctor does
or does not do influences a patient's sense of coherence, it thereby
affects his or her state of health.
Third, and of most importance, my analysis is best under-
stood in the context of the theoretical model of salutogenesis. As a
teacher of medical students, I have often put it that I view my role
as training one type of generalized resistance resource, the doctor.

But I can now see, in retrospect, that though it sounded good —the
doctor would help his patients fight all the ubiquitous bugs with the
particular weapons at his disposal — it was a fuzzy notion. I did not
understand why resistance resources are important and how they
deal with stressors. Once one adopts the concept of the sense of
coherence, the link becomes clear. The doctor, as a potential GRR,
has the possibility of structuring life experiences for people that
reinforce their sense of coherence. If encounters with the health care
system indeed systematically are characterized by consistency, par-
ticipation in shaping outcome, and an underload-overload balance
of stimuli, the sense of coherence is maintained or reinforced.
But note that the word potential has been used. As I pointed
out in Chapter Four, some GRRs, stood on their heads, can weaken
the sense of coherence. This is true for doctors. There is no need to
exaggerate. A patient with a strong sense of coherence will not
emerge from encounters with doctors, whatever their nature, with a
significant change. But this is not to say that there can be no impact.
The matter, rather, is one for empirical study. What must be re-
membered is that inherent in all such encounters, to a greater
or lesser extent, are anxiety, uncertainty, unpredictability, and
dependence.
Traumatic Situations. I have tried to place the assessment of
the potential impact of the patient-doctor encounter on the sense of

coherence in proper proportions. Consideration of the anxiety and
dependence inherent in the encounter, however, brings us to situa-
tions in which the nature of the encounter can have a fairly decisive
influence on the sense of coherence. Such situations may constitute
a small fraction of the over one billion (in the United States) en-
counters that take place annually between patients and doctors. But
absolutely they do involve a substantial number of people and
are the most dramatic and memorable encounters.
Let us take two examples. The first almost invariably in-
volves interaction with the physician; the second may or may not.
In Chapter Seven, I noted that one's location on the health ease/
dis-easecontinuum may have a feedback impact on one's GRRs. A
woman who has undergone a mastectomy, someone who has had a
leg amputated or has suffered a stroke or severe burns often con-
fronts the danger of rapid radical erosion of the sense of coherence.
True, one's pretraumatic level a decisive variable in determining
is
what will happen. Some people may have so strong a sense of co-
herence that even such a trauma has a blunted impact. But, for
many, the word shattered is appropriate. The physician, I suggest,
can in such situations — since he or she is in any case closely in-
volved —play a significant role in the slow, painful, and difficult
task of reestablishing the sense of coherence.
The second example is a bit more complex and bears a direct
relationship to the nature of the health care system. Traumatic
psychosocial stressors can have a direct impact on GRRs (Arrow
B in Figure have already alluded to the studies showing the
1 ) . I
sharply increased mortality rates of widowers in the six months
following death of the wife. Whatever the strength of one's sense of
coherence, a fairly sudden, radical change in one's GRRs ranging —
from a chemical engineer aged fifty-two losing his job to a loyal
member of the Communist Party confronted with Khrushchev's
speech at the Twentieth Party Congress —has the potential of rock-
ing the foundations of a sense of coherence. A personal or structured
relationship between doctor and patient or a health care system that
institutionalizes channels for providing information of such traumata
to personnel opens the possibility for the contribution of the physi-
cian in reestablishing the sense of coherence.
The literature on what has come to be called crisis interven-
—
tion is vast. Much thinking (though precious few randomized con-

trol trials) has been invested in how the health and other well-being
services can contribute to the welfare of people in such situations.
What has been lacking, I believe, is a theoretical understanding of
what happens to people who go through such situations. Concep-
tualizing the process as the disintegration and reintegration of the
sense of coherence via concern with the life experiences provided by
GRRs, I submit, may open the way for possible intervention. (For
an important analysis of the possible role of the physician in such

situations —
an analysis that is most congruent with the concept of
—
the sense of coherence see Parkes, 1972.) The conditions under
which reintegration is more or less likely to occur are best analyzed
by looking at the routine encounter.
Routine Encounters. It bears repeating to say that, whatever
the individual variation from patient to patient and situation to
situation, predominant modes characterize the interaction of any
one doctor with his patients and indeed of patient-doctor interac-
tions in any given sociocultural setting. Three elements present in all
settings influence this regularity.
First, the patient invariably perceives the situation as involv-
ing anxiety, uncertainty, and ambiguity. In our terms, the life situa-
tion of patient-doctor interaction is inherently antithetical to a sense

of coherence. This perception may vary in intensity from that
experienced during a routine checkup to that experienced in an
interaction with perceived critical consequences for health outcome.
But even in the most innocuous situation, few patients show total
equanimity.
Second, the fact that a person has entered the interaction
means that he or she has acknowledged that the doctor has a particu-
lar technical competence in dealing with certain kinds of problems
a competence that the patient does not have. Even in situations
where entry is involuntary (compulsory examinations in industry,
the military, prison), one finds such acknowledgement, though the
doctor may be perceived in different terms from the way a doctor is
perceived in a voluntary situation. The doctor always assumes a

relatively superior professional competence.
Third, there is an inevitable built-in difference in the per-
spectives of the patient and of the doctor. The patient always
.
normatively is concerned with his or her self-interest, whatever the

cultural limits of propriety may be in a given setting. The doctor
can never, Parsons notwithstanding, be solely concerned, norma-
tively, with the patient's self-interest. (Parsons' stress on the collec-
tivity orientationwas first formulated in Parsons, 1951, chap. 10.

In his latest comments on the subject, in 1978, there is no indica-
tion that he has revised his position.) This is not a matter of a pos-
sibly different perception of the patient's self-interest. The extent
to which other concerns impinge on the doctor's orientation can
vary greatly. But even in the most extreme situation the physician, —
privately employed by a monarch or tycoon, who has thoroughly
internalized the norm of "for the good of the patient" the doctor —
must keep one eye out for his or her own welfare. Other than in
these rare situations, the doctor always has commitments to other
patients, to his or her other roles, to norms relating to not being
taken in or exploited. Such norms are based on the unwritten obli-
gation the physician has assumed toward society, in return for some
degree of professional autonomy, to minimize the social costs of
illness. This obligation is most often taken to mean to cure the
patient; but it also involves a stance against malingering and the
like (see Merton, 1976, p. 68)
Though Illich has not explicitly analyzed the doctor-patient
relationship in these terms, it would be consistent with his position
to argue that handing over responsibility for coping with anxiety to
the doctor, acknowledging the doctor's technical competence, and
the doctor's being concerned with interests other than those of the
patient are key factors in iatrogenesis. "The main source of pain, of
disability, and of death is now engineered, albeit nonintentional,
harassment. Our prevailing ailments, helplessness and injustice, are
largely the side effects of strategies for more and better education,
better housing, a better diet, and better health" (Illich, 1976,
p. 262; my emphasis). But his blanket, prophetlike indictment ob-
scures rather than enlightens the analysis of the structural modalities
of the doctor-patient relationship that influence the sense of co-
herence of the patient.
Given the above three elements in the relationship, which
modalities foster restoration or maintenance of the sense of coher-
ence and which weaken it? We now turn to this question.
Perhaps the most eloquent confrontation with the problem of

anxiety is Eric Cassell's 1976 book. He distinguishes between ill-
ness and disease. Illness refers to what the patient feels is wrong;
disease refers to what medicine says is wrong. The physician's
responsibility, Cassell argues, is to deal with both problems, to heal
and to cure. The thrust of Cassell's argument is that it is much more
humane and decent for the physician to accept this dual burden
rather than limit himself or herself to the responsibility for curing.
Among the core elements of illness he describes are a loss of the feel-
ings of omnipotence, of connectedness, and of personal indestruc-
tibility and also a loss of the power of rational reasoning and of the
sense of control (pp. 30-44). The relationship between these ele-
ments and the sense of coherence is self-evident. If indeed the sense

of coherence is related to curing, then it follows that, for the physi-
cian to cure, he or she must cope with the problem of a lessened
sense of coherence. Cassell does not make this connection explicidy.
His argument rests essentially on moral rather than on theoretical,
instrumental grounds.
But his very humaneness makes him oblivious to a substan-
tial danger. Cassell's choice of extreme vocabulary —omnipotence,
personal indestructibility — is misleading. He seems to suggest that
these feelings characterize us when we are well and that the physi-
cian's responsibility is to restore us to such a grand state. The physi-
cian is obviously much closer, both because he is well and because
he is a physician, to such a state. This belief can lead, though un-
intentionally, only to urging the patient to see the physician as
omnipotent and to place faith and fate in his or her hands. And
indeed, Cassell is at his most comfortable and most humanely mov-
ing in his writing on the terminal patient, for whom there is no hope
at all of returning to omnipotence, connectedness, and the other
extreme characteristics.
Cassell's analysis, then, is double edged. The physician-
paticnt relationship is structured, as he would have it, by the doctor's
concern for the total patient as a person, total commitment to all
the needs of the patient, and assumption of ultimate responsibility
for meeting these needs as best as is humanly possible. There is litde
doubt that such a relationship is at least consonant with the char-
acteristics of life experiences that I have argued foster a strong sense
of coherence. Cassell would be extremely sensitive to assuring that
the interaction with the patient is consistent and contains an ap-

propriate load balance (of information, for example). He is explic-
itly careful to assign maximum participation to the patient in

doing things for himself or herself, much as God requires a great
deal of the believer.
There are, however, three pitfalls in this approach that in-
evitably prevent the reestablishment of the sense of coherence. First,

while God may be omnipresent, the
or the parent of the small child
doctor is not, even for the terminal patient. The patient, socialized
into decision-making dependence, is helpless when the physician is
not present. From the point of view of the sense of coherence, there
are undoubtedly situations in which this mode of relationship is com-
pletely appropriate. Szaszand Hollender (1956) refer to the activ-
model in clinical situations of anesthesia, acute trauma,
ity-passivity
coma, delirium, and so forth. The only possible road to a return of
the patient's sense of coherence in such situations is physician deci-
sion making. But most physician-patient interactions, particularly in
an era when both prevention and chronic illness are increasingly
important, are not of this type. In such cases, a relationship that
leaves decision making in the hands of the physician can only im-
pair the sense of coherence.
The second pitfall refers to ignoring the cultural context of
the relationship. We are dealing with a situation of faith in the
omnipotence of the physician as the basis of restoration of the sense
of coherence. Such faith is inherent in therapy administered by God
or by His sacred representatives. It is quite efficacious in a cultural
context where God is omnipresent or when the patient is, for some
temporary period, in a total patient role being cared for by sacred
representatives, as well as in the clinical situations referred to above.
But when the patient is expected to be an adult in his or her other
role relationships, when the culture puts a premium on decision
making by oneself, such a structured role relationship with the physi-
cian is dysfunctional because it violates the need of patients, in cer-
tain cultural contexts such as those of Western societies, to equate
participation in determining outcome with decisional and not only
behavioral participation.
The issue, moreover, goes even further than participation in
decision making. Patients arrive at the encounter, having been

shaken by their illness, with an attempt at restoring their sense of
coherence. They do so, implicitly or explicitly, by casting their ill-
ness in a subjectively sensible etiological, diagnostic, prognostic, and

therapeutic framework. When doctor and patient share a common
language, asis the case in much non-Western medicine, the patient
can find only confirmation. But often, in situations where doctors

and patients are from different social class and cultural back-
grounds, the medical belief systems of the two differ. The doctor
most often does not even inquire into the patient's belief system.
When the doctor do so and acts solely within the system of
fails to
scientific medicine, which is often not even congruent with the
patient's system, the awesome omniscience of the physician can only
confuse the patient
Engel ( 1977 ) has put Cassell's case for healing in a far more
sophisticated and ethically neutral way. The tendency to reduce
medical action to the biomedical sphere not only disregards the
psychic needs of the patient, which are, in the patient's view,
legitimate medical problems, but makes it impossible to adequately
understand even supposedly straight somatic problems. Only when
the doctor sees and relates to the patient as a person in a sociocul-
tural context can the results of interaction be effective from the
point of view of the patient as well as of the doctor.
The approach of Cassell and Engel, then, in rejecting a
narrow, somatic, biomedical conceptualization of the appropriate
scope of the patient-doctor encounter, fosters the sense of coherence.
Engel does not fall into the trap of positing the omnipotent doctor,
inimical to the sense of coherence in a Western cultural context.
But he, like Cassell, also ignores a third pitfall that follows from
broadening the scope of the interaction. They both disregard the
problem of power in the relationship. The patient not only is anx-
ious but, by definition, is dependent. By entering into the relation-
ship, the patient hasacknowledged the legitimacy of the depen-
dence, though perhaps with ambivalence. Broadening the scope of
the interaction to include the doctor's viewing the patient as a total
person increases the dependence. The power relationship, then, is
inherently asymmetrical. By ignoring this issue, Cassell and Engel

are saying: Place your trust in the doctor (and in the professional
group that monitors its members' behavior) that he or she will not
exploit the asymmetry.
At this point, it is appropriate to recall the third inherent
element in the doctor-patient relationship. The doctor can at the
most only temporarily suspend his or her other commitments and
devote himself or herself to the needs of the patient. In other words,
there are always the temptation and the pressure to take advantage
of the asymmetry. This can be done in many ways. At one extreme,

the patient can be exploited financially or for the ego gratification
of the doctor or sexually. In academic centers, the patient can be
used for teaching and research purposes not in the immediate inter-
est of the patient. At the gentlest extreme, the doctor can act in
consideration of other patients.
I am not saying that such advantage is always taken. Nor
am I raising a moral issue. I am saying that the inherent nature of
the relationship not only makes taking advantage possible but pres-
sures in this direction. What is important in the present context is
that the extent to which the doctor's behavior is perceived by the

patient as anything other than in the total interest of the patient is
bound to weaken the sense of coherence. The crucial test is not that
of diagnosis or therapy proposal but rather that of decision making.
The patient who has the opportunity of saying "Go ahead" or
"Stop" or of saying "I trust you enough to accept your proposal" or
"You haven't yet gained my trust" is in a life experience character-
ized by participation in detemiining outcome.
Let me put my point in another way. Patients are always in-
strumentally dependent on the doctor, for they need the doctor to
give them what they believe they cannot, for whatever reason, get
for themselves. The power dependence is thereby
possibility of
created, for patients can never fully judge what is being done for
their own good. But the doctor can never act solely for a patient's
good. Unless counteracted, the consequences for the patient's sense
of coherence are detrimental. (For a most helpful set of theoretical
analyses based on data and germane to this section, see Ben-Sira,
1972, 1976.)
Modalities. We have, in part implicitly, in part explicitly,
considered the question of how the life experience of the patient-
doctor encounter can be characterized by a high or low degree of
—
consistency, participation in shaping outcome, and an undcrload-

overload balance. Three crucial factors have been identified as
shaping this life experience: scope of legitimate expertise, ranging
from a narrow, specific focus on the technical, somatic presenting
problem to perception of the patient as a total person in a sociocul-
tural context; goals of interaction (assuming that the patient always
focuses on his or her needs), ranging from the doctor's nearly total
focus on needs of other than the patient to total focus on the pa-
tient's needs as perceived by the patient; power allocation, ranging
from total decision-making power in the doctor's hands to total
decision-making power in the patient's hands.
The consequences for the sense of coherence are not simple.
The more the patient is perceived as a total person, the better. The
more the focus on the needs of the patient, the better. The more
is
decision-making power rests in the hands of the patient, the better.

However, in the real world fundamental structural properties of the
doctor-patient relationship—built-in characteristics of the relation-
ship independent of the particular patient and particular doctor
militate against such possibilities.
It becomes, then, helpful to depart from the general concept
of the doctor-patient relationship and to confront different possible
modalities or types of relationships. Having clarified the issues in-
volved, we can now ask: Which modality offers the best oppor-
tunity for the most efficacious interaction .to take place from the
point of view of the sense of coherence? Parsons (1978) proposes a
set of four models of social organization. His argument that the col-
legial modality is the most appropriate way of conceptualizing the
doctor-patient relationship need not concern us here. I will spell out
the details of the models, as I understand them, in the terms of the
above discussion. The labels used are mine, in the hope that they
clarify. I have also added what seems to me to be a crucial fifth
type. The details appear in Table 7. It is, I trust, clear that cate-
gorization is only for the sake of simplicity in presentation. No con-
crete interaction is ever a pure type but only tends in one direction
or the other.
A concrete doctor and a concrete patient can establish a re-
lationship that is not governed by the pressures of this or that
modality, thereby obtaining the advantages of this one and avoiding
Table 7. Alternative Conceptualizations of the

Doctor-Patient Relationship
of health care services. In this sense, the current trend in some

quarters, particularly in the United States, to speak of the patient
as a consumer is salutary. But the price to be paid is highly dys-
functional. The doctor is pressed to care for his own interests, only
qualified partially by the norm of "in the interest of the patient."
The consumer can rarely have a consumers' rights organization
sufficiendy present in the concrete situation to allay the anxiety and
is pressed to mistrust the doctor. Further, the doctor inevitably
focuseson his or her area of technical expertise; it takes too much
uncompensated time and energy to consider the total person.
The bureaucratic modality is, if anything, more detrimental
to the sense of coherence of the patient. This negative effect has
nothing to do with the pejorative implications of the term. Trained
incapacities, Parkinson's Law, red tape, and the like in a bureau-
cratic context only exacerbate the dysfunctional consequences in-
herent in the doctor-patient relationship. The very nature of the
context compels the physician to adhere to norms and rules estab-
lished in the interest of patienthood but not necessarily of the in-
dividual patient or of the organization. The amorphous idea of
the patient as a whole person is anathema to the classification re-
quired by bureaucratic organization. The client's rights and duties
are spelled out but by the organization. The only possible counter-
vailing force to such pressures is the establishment of institution-
alized procedures and structures explicidy designed to limit and
power of the organization
direct the — for example, when a trade
union owns or sets up a health care service for its members. This
solution, however, creates its own problems, such as the "pro-
(McKinlay, 1978). The iron law
letarianization" of the physician
of oligarchy should make us skeptical of whether such a counter-
vailing force can solve the problem of decision-making power.
The democratic, or friendship, modality, visualizing a phy-
sician-patient partnership, on the face of it seems most likely to pro-
mote a sense of coherence. In this modality, the physician must
be altruistic, subordinating all other interests to the patient's needs.
The relationship, being personal, is by definition all embracing
and not limited to the technical problem. Power is shared and de-
cisions are made jointly, with all the cards on the table. But, to
continue the metaphor, there are two crucial jokers. First, so per-
sonal a relationship inevitably militates against the affective neu-

trality that is essential if the physician is to meet the central technical
requirements. The physician, in diagnosis and formulation of pro-
posals for therapy, cannot do well if he or she is too emotionally
bound up with the patient. The issue is even more clearly seen
when one considers the inevitable reality of patients toward whom
the physician has negative emotions — bias, dislike, disrespect. To
ask the physician to be affectively neutral only with regard to his
negative emotions is a contradiction in terms. Second, the demo-
cratic modality disregards an essential property of all doctor-patient
relationships. Though the degree may vary, the patient is always

dependent. To ask full partnership in exercising decision-making
power is to place an intolerable overload on the patient. Even a
physician, when he or she enters the patient role, is subjectively in-
capable of assuming the burden of decision making. Anxiety in-
evitably prevails over technical knowledge. And, unlike the case
in the market modality, the patient cannot easily terminate the
relationship.
Parsons proposes that the collegia! modality "is the most
appropriate and the most likely to have a strong prospect of sur-
vival in the field of purveying of professional services" (1978, p.
452). He does not say appropriate for what. We may infer that he
means that this modality is the most adequate conceptualization
American medicine. This may well
of the reality of the practice of
be true. It may even be the modality most likely to survive in the
United States. But it is not necessarily the most conducive to a
strong sense of coherence.
Cassell's analysis, considered above, is one of the most force-
ful expressions of the collegial relationship at its most humane. He
has transcended, as noted, two of the limitations that I believe are
inherent in the collegial modality, namely, a focus on the specific

technical problem and only partially according primacy to the
patient's need satisfaction. Such transcendence is evidendy possible
but only because of unusual individual characteristics and motiva-
tions. The collegial relationship by its nature promotes dealing with
the technical problem, for, after all, this is the particular realm of
special expertise that is the basis of the claim of physicians for pro-
fessional autonomy and that brings the patient to the physician in
2
the first place. That collegial physicians have been trained in

teaching hospitals with a strong technical orientation may be a
historical accident. I return to this issue later. But even Cassell has
not transcended a third limitation, namely, that decision-making

power in the collegial modality remains in the physician's hands.
This is in the terms of the contract. The doctor is the expert and
knows what is best for the patient. In a sense, the collegial physician
who transcends the other limitations, thereby gaining the full trust
of the patient, paradoxically impinges on the possibility of sharing
decision making with the patient. One does not share power with
God. Damage is inevitably inflicted on the patient's sense of par-
ticipation in determining outcome. But this analysis of the situation
must be qualified; in subcultures or cultures fully accepting God or

His representatives as the physician, such damage is not inevitable.
Which brings us to the final modality, the sacred relation-
ship. Limited to thinking in American middle-class terms, Parsons
does not realize that the layperson-divine mode (divine in the sense
of priest rather than godlike) not only must be included in a sys-
tematic typology of modalities but is empirically found in all parts
of the world, including the fortress of Western scientific medicine.
The tremendous advantage for the sense of coherence of this
modality derives from the complete cultural congruence between
patient and doctor. Whether Christian Scientists, Taiwanese, Ethi-
opians, or from Sri Lanka, the patient and doctor fully share con-
ceptualization of etiology, classification of illness and disease,
probable course, appropriate therapy, and so on. The doctor still
remains the expert butis a medium who has been blessed. (See
Kleinman, Eisenberg, and Good, 1978; Obeyesekere, 1973; and

Young, 1976b, 1977.)
The sacred relationship, then, would seem to be the mo-
dality that most powerfully fosters restoration of the patient's sense
of coherence. The patient's anxiety and emotional needs are fully
taken care of, and he or she is related to as a total person. The
patient, appropriately socialized, has complete faith in the doctor's
commitment to the primacy of his or her needs. Though decision-
making power is not in the patient's hands, this is not at all
troubling. Carefully prescribed behavior, following doctor's orders,
assures the sense of participation in determining outcome.
.
The sacred mode, when practiced in a cultural context to

which it is appropriate —where there is total faith in God or the gods
and in representatives of divine power —does not present the dangers
noted in the discussion of Cassell's work. But Gassell works with an
armamentarium that is far more efficacious in coping with somatic
diseases than are all other forms of medical practice, the romantic de-
bunkers of what we know as scientific medicine notwithstanding. In
other words, in the world of reality, those who practice the sacred
mode, no matter how much they are concerned with healing, are not
all that good at curing. Does this not damage the sense of coherence?
Not at all, for what must be understood is that patients at home
in such a modality have a ready-made and totally self-convincing
explanation for such failure. Often, the patient's faith has been
inadequate. Young points to a more profound understanding when
he notes that our mistake is that we attempt "to study all medical
beliefs as attempts to control events within the bodies of sick per-
sons instead of studying them in the contexts of phenomenological
reality and the division of labor in society" ( 1976a, p. 147)
The failure to cure, then, is not the Achilles' heel of the
sacred modality from the point of view of the sense of coherence.
It would be, in my view, the most powerful modality in restoring
the sense of coherence for those with complete faith were it not for
the inherent danger of abuse. When all decision-making power is
in the hands of the divine, the sense of coherence is not impaired

as long as the layperson maintains complete faith. But divines are
people. They have inevitable tendencies to perceive themselves as
the Divine. This is compounded by two further tendencies. First,
the inherent uncertainty of medical practice, in the face of the need
to act, gives rise to anxiety. Second, absolute power raises the
possibility of corruption — of abandoning the primacy of the patient's
interests. Unless these tendencies are controlled, the chances are
considerable that the patient will perceive the fallibility of the doc-
tor (see Freidson, 1970). When this happens, the negative con-
sequences for the sense of coherence are substantial.
We have now considered five modalities of the doctor-
patient relationship. There are, as we have seen, considerable ad-
vantages to all but the bureaucratic mode for the sense of coherence
of the patient. (This mode, though, has advantages from other
points of view.) One might, then, think that the individual phy-
sician, alert to the importance of the problem of the patient's sense
of coherence, would structure a modality that selects the best of all
possibilities and avoids its pitfalls. Within narrow limits, this is
indeed possible. But the limits are narrow. The physician, after
all, works within a health care system and a broader social struc-
ture that impose considerable constraint. It is to this system and

structure that we now turn.
Impact of the Health Care System and the Social Structure
It would be manifestly impossible here to attempt an in-
tricate analysis of how the health care system in any given society
shapes the dominant modalities of the physician-patient relation-
ship. (An important but not very integrated endeavor in this di-
rection is found in Gallagher, 1978, and particularly in McKinlay's
paper therein.) I shall, therefore, only raise a number of issues
that seem to me to be most pertinent to the subject of this book.
A good starting point is the dominant system of medical
education. Of the 900-odd medical schools in the world, almost
all have adopted the ideal of living in the post-Flexnerian world.
This is not that much less true of the training institutions for other
health care workers. I do not in the least denigrate the significance
and achievements of scientific medicine. But . two inherent char-
acteristics of training inan academic, scientific institution have
major implications for the modality of the doctor-patient relation-
ship. First, the enormous achievements of biomedical science are
most consonant with a focus on the specific technical, somatic

presenting problem of the patient and make adoption of the demo-
cratic or sacred modalities difficult. Engel (1977), who has put
the case against the narrow biomedical orientation with so much
sophistication, has not asked himself why this model is adhered
to with so much tenacity.
Second, training for scientific medicine is closely integrated

with teaching hospitals and academic research institutions. Clinical
experience, then, is centered overwhelmingly on the horizontal

patient, most often with a complex somatic disease. Not only do
such patients constitute an unrepresentative sample of the patients
in the overwhelming number of doctor-patient encounters; but
they are by definition in a bureaucratic setting. The patient is
learning material. On all three counts — scope of legitimate ex-

pertise, primacy of patient's need satisfaction, and decision-making
power —the student is best prepared for the bureaucratic modality.
Moving out young physician faces
into the real world, the
one of three dominant alternatives, depending on the society in
which he or she lives. In a world increasingly committed to societal
provision of health care to its citizens as a right, the physician more
and more becomes a salaried employee of a bureaucratic organiza-
tion, a niche for which training has been most appropriate. Al-
ternatively, in capitalist societies, the market model is culturally

most appropriate. But the contractual terms imposed, exchanging
professional autonomy for normative commitment to the best
interests of the patient, often make the physician uncomfortable
in the market modality. The rewards of being the supreme culture
hero — all studies of occupational prestige assign pride of place to
the physician — are contingent on behavior conforming to this
norm. There is, then, pressure to adopt the collegial modality and
play God. This is even more true in those societies, such as those of
most of Western Europe, where the market modality is attenuated
by sick funds (private, public, or governmental) and contractual
arrangements with individual physicians or an organized group of
physicians.
The cultural world of the Western physician is a secular
world. It has its magic, its rituals,and its superstitions, but the
sacred is reserved for special days and special buildings, which do
not include the hospital or doctor's office. Hard science reigns
supreme, as an ideal. In such a world, the sacred modality has
little place. It is identified with "quackery" — Christian Science,
homeopathy, faith healers, snake-oil salesmen, and the like. This is
not to say that practitioners of the sacred modality do not exist.
Quite the contrary is true. This very fact pressures the "legitimate"
physician to stress the scientific character of the medical profession
all the more. Whatever the temptation to become a divine, the
physician must set himself off from the "quacks." Inevitably this
requirement leads one to steer clear of such murky notions as the
whole person and to focus on the soma, with which one is most
comfortable.
In these and other ways, then, the particular social struc-
turc — capitalist, welfare capitalist, social democratic, authoritarian

—
communist and the particular value system in which the physician
has been educated and within which he or she works shape the
structural properties of the doctor-patient relationship. This is, of
course, only one side of the coin. The patient too contributes to
the shaping of the relationship. Here matters are far more complex,
for patients from different social classes, with different ethnic back-
grounds and social roles, have different expectations and value
orientations. Thus, for example, the physician with a largely upper-
middle- or upper-class patient population is likely to share decision-
making power with the patient. He or she often assumes that the
patient's capacity for information load is large. Lower-class patients
are likely to an assumption on the part of the physician that

elicit
because they have different vocabularies, facility in language,

capacity for abstraction, and the like, they are incapable of par-
ticipation in determining outcome.
We now turn briefly to what may well be, in the long run,
the most significant consequence of the practice of medicine (and of
other health care professions) for the sense of coherence. Our
entire discussion so far has focused on the doctor-patient encounter.
Given the fact that the average person in the industrialized world
has some four or five such encounters each year, the impact can-
not be dismissed, though it should not be exaggerated. But such
encounters of doctor and patient are overwhelmingly limited, in
most countries of the Western world, to what is so erroneously
called the health care system. The more appropriate term is the
disease care system, as I have noted in Chapter Two. The en-
counter is almost by definition between a patient and a doctor
oriented to treating a disease and to disregarding dis-ease, not to
talk of ease.
I have now come full circle to the salutogenic orientation.
Overwhelmingly, the medical profession is organized, intellectually
and structurally, to assume responsibility for curing a disease when
the patient enters the disease care system. It would be futile, un-
wise, and inhumane to argue that this should be otherwise. Those
who do enter this system are the people who hurt most, who are
most in danger of dying. The thrust of my argument, however, has
been that a salutogenic orientation and the sense of coherence are
not luxurious frills. To ignore them is to be self-defeating. Perhaps
health care is not simply a matter of frantically plugging holes

in dikes. I deal with this long-range issue in the Epilogue. But a
care system that is organized to deal only with breakdown is a
system that can no longer make serious advances.
I shall leave to others, or to myself elsewhere, the question
of what form of broader social organization is most conducive to
a strong sense of coherence, even though this question is the most
overriding one. Nor do I think it appropriate to urge doctors qua
doctors to become involved in revolution, social work, social en-
gineering, and so forth. But at both the group and individual
level and given the assumption that the sense of coherence is re-
lated to health, onemust ask: Can the medical profession and the
individual physician engage in activities beyond the patient-doctor
encounter that affect the sense of coherence? I would suggest four
major activities that can positively affect the sense of coherence:
making health care available to all, promoting a preventive health
orientation, buttressing trust in the physician, and reaching out to
persons at high risk of damage to the sense of coherence.
I would start with what seems to me to be the underlying
value premise at the broadest possible level. In my original "break-

down" paper, even before I had formulated the concept of the
sense of coherence, I wrote (Antonovsky, 1972, p. 542): "The
ties between an individual and his total community are decisive
resistance resources. One indicator of such ties would be the extent
to which a society holds as a central value that the society as such,
rather than the individual alone, is responsible for dealing with

threat or for assisting the individual to do so." A society, then, that
has institutionalized a health care system that expresses consensus
that health care an inalienable right of all its citizens and is to
is
be made equally available to all on the universalistic ground of
being a resident of that society is a society that has taken a step

forward in strengthening the sense of coherence of its members.
If one is blocked from access to health care because one is poor or
a member of a racial minority or lives in a rural area, one's image of

the world as coherent — given the importance of health in people's
lives — is necessarily impaired.
Second, it will have been noted that one of the generalized
resistance resources included in Figure 1 is a preventive health
orientation. Doctors are not the only source of this orientation, but
surely the profession as such can play a key role in facilitating its
adoption. Again, what is important is not so much the action of

individual physicians as the institutionalized mechanisms for health
education, highlighting of risk factors, facilities for early detection,
and so forth. It seems reasonable to posit that strengthening of this

GRR contributes to a strong sense of coherence.
The third issue to be raised at the supraindividual level
refers to the image of the physician in society. This issue has been
dealt with, in a broader context than just that of the medical pro-
fession, in The Limits of Benevolence, by W. Gaylin and others.
In a review of the book, we read (Starr, 1978, p. 9) "Progressive :
reformers assumed a coincidence of values and interests between

caretakers and the dependent and were willing consequently to
entrust professionals with enormous discretionary authority. Today,
reformers increasingly assume a potential conflict between benev-
olent institutions and their clients." The more the medical pro-
fession is trusted, in the simplest of terms, the more likely is a
population to have a strong sense of coherence. The role of the
mass media is not to be discounted; both heroic exploits of doctors
in saving lives on television programs and headlining malpractice
suits have But far more important, in my judgment,
their effects.
are such matters as the perception of unwarranted income levels,
costs of health care, opposition to or promotion of health-insurance
programs, concern with environmental hazards, readiness for self-
policing, and availability for public scrutiny. The medical profession

is always likely to be a bearer of trust, whatever the barbs of
Moliere and George Bernard Shaw, more than any other occupa-
tional group. There is, however, a considerable margin. The level
of trust, at any given time in history, can affect the sense of
coherence.
The fourth and final issue, like the issue of a preventive
health orientation, derives from the salutogenic framework. I refer
to the availability of institutionalized mechanisms to identify high-
risk groups but in a broader than usual sense. A high-risk group
is generally thought to consist of those who smoke, are poorly
nourished, are old, are exposed to noxious agents, and the like. In
the present context, my concern is with those people subject to
particular stressors that have adversely affected their generalized
resistance resources. Thus, the widower, the adult who loses a job,
the parents of a child who has died may not break down directly,
but their sense of coherence is affected. Note that I have not here
suggested that the medical profession as such can deal with poverty,
unemployment, racial discrimination, migration, or other socio-
historical sources of a weak sense of coherence. My point, rather,
refers to the structured lines of communication between the doctor
and the population for which he or she is responsible. With a
structured mechanism for acquiring such information, the doctor
can reach out at a time when more good might well be done than
when the consequences of the trauma become manifest.
There is no need to discuss the role of the individual
physician in making health care available to all in the community,
in promoting a preventive health orientation, in buttressing an
image of trust in the physician, and in reaching out to high-risk
persons. Whether he or she works in this direction in some measure
depends on the individual personality and value system. But it
depends far more, as does the direction of action or inaction of
the medical profession, on the society in which the profession
operates and on the social structure of health care delivery and
its ideological premises.
I explicitly stated at the outset of this chapter that I had

no intentions of writing a guide to the perplexed. I have discussed
a number of issues that seem to me to be central in relating the
sense of coherence of people to the policies and practices of med-
icine (and, by implication, of other health professions). I have
discussed these issues in what is a sociological frame of
essentially
reference, which is at least one appropriate way of understanding

what doctors do and can do in strengthening the sense of coherence
of patients and potential patients —
that is, of all of us. I have
avoided dealing with many complex issues of health and health
care not directly related to the central themes of the book. Above
all, I have raised questions and not insisted on answers. If those
who more directly than I bear the responsibility for health care
take up the challenge and begin to formulate answers, such was
my intention.
Epilogue: Outlook
for Human Health
The central problems with which this book is concerned— survival

and the quality of life in the realm of health —have been with us
for a long time. Survival was the central health concern throughout
most of human history and in much of the world it remains central
even today. But the direction for successful coping with this prob-
lem is reasonably clear, as difficult as it may be to move in this
direction for socioeconomic reasons. In the industrialized world,
however, we are approaching the limits of life expectancy in large
populations. As Burnet puts it (1971, pp. 154, 168; see also all of
his chap. 8), "The 'allotted span' for any species is something
genetically programmed as a result of evolutionary processes. . . .
Nature is not going to cooperate with us in keeping men or any

other animal alive for much beyond the span she has allotted."
If this is indeed the case, then neither the sense of coherence
220
nor any other factor is of much relevance for the survival of those
who are near the allotted span of years, at least on an overall
population basis. Atfirst sight, it would seem that the real health
problem in industrialized societies is that of "premature" death.

Not that any death is seen as "mature," at least in our culture.
But when early deaths in one group blacks, poor people are — —
more frequent than in another group, there is good reason to apply
the concept of premature deaths. With the advent of chronic dis-
eases as the major cause of early mortality, the real health problem
of industrialized societies becomes the quality of life.
The issue, then, I would discuss in this concluding chapter
is that of the implications for the future of the analysis presented in
this book. It is, I think, useful to put this discussion in the context
of the presentation of the four most outstanding thoughtful con-
siderations of the question that have emerged in recent years. In
doing so, I follow the lead of Thomas McKeown's Rock Carling
Fellowship monograph (1976, pp. 160-180). After summarizing
the major approaches of Lewis Thomas, Rene Dubos, and Ivan
Illich, and that of McKeown himself, I will discuss the expectations
for coming decades that flow from the salutogenic model.
Thomas (1974, pp. 36-42), representing most contem-
porary biomedical scientists, distinguishes among three levels of
technology in medicine. When diseases are not understood, all med-
icine can do is provide the supportive therapy that tides patients
over, presumably until they recover or die. As indispensable as the
therapy is from a human point of view, its cost is extremely high.
Once this investment was largely in the infectious diseases; now the
chronic diseases require it. But it is no more than a holding opera-
tion, used because medicine does not know what else to do. The
second level of technology is the hospital-based, enormously ex-
pensive, "highly sophisticated, and profoundly primitive" tech-
nology of transplants, coronary-care units, chemotherapy, and the
like. This technology is designed to make up for the ravages of
disease or to postpone death. Again, at the human level, Thomas

cannot reject such expenditures.
Neither of these activities, in Thomas' view, is of significant
promoting advances in the health level of the population.
utility in
Real hope is to be placed in the third level of technology in med-

icine, which is based on a genuine understanding of disease

mechanisms. We now can deal effectively with diphtheria and
bacterial infections, syphilis and tuberculosis, and various nutri-
tional disorders because we have gained this understanding. This

effective technology also has the virtue of being quite cheap. For
Thomas, the application of basic knowledge has been the basis of
major advances in health levels in the past and is the direction that
is quite promising for the future. "If I were a policy maker," he
concludes (p. 41 ), "interested in saving money for health care over
the long haul, I would regard it as an act of high prudence to give
high priority to a lot more basic research in biologic science."
Dubos, no less a distinguished biological scientist than
Thomas, only in part shares his analysis of the past and his prog-
nosis for the future. It is true, he writes ( 1968, pp. 100-101 ), that
"many most destructive diseases of his [Benjamin Franklin's]
of the
time have now been brought under control. The application of
these methods [of modern medical science] to disease control offers
such diversified potentialities that one can anticipate scientific
solutions for almost any medical problem, once it has been clearly
defined." There are, however, three central reasons, in Dubos'
view, for being critical of Thomas' position. First, the position rests
on the doctrine of specific etiology or, as Dubos has put it, the
mirage of the magic bullet (1960, pp. 9 Iff.). The etiology of
chronic diseases is complex and can be understood only in terms
of multiple causation. Whatever the utility of Thomas' approach in
the past, it is not likely to be of much help in the future. Second,
Dubos stresses the fact that "economic, social, and ethical difficul-
ties often complicate or prevent altogether the practical utilization of
existent knowledge for the prevention and treatment of disease"
(1968, p. 101). But perhaps the most important thrust of Dubos'
position is his third observation, that "each period and each type
of civilization will continue to have its burden of diseases created
by the unavoidable failure of biological and social adaptation to
counter new environmental threats" (quoted in McKeown, 1976,

p. 163).
Dubos prefers to see himself as a "moderate optimist" ( 1973,
p. 184) . Clearly, he rejects Thomas' faith in the future, which rests
on the advances of biological science. But his own faith is equally

.
strong. Humanity will surely, he argues, find the will to work out
the conscious social and economic arrangements required to replace
the previous "spontaneous adaptive processes" in meeting new chal-
lenges. Dubos relies on the "ground swell of dissatisfaction" ex-
pressed by businessmen and scientists and in the countercultures of
youth, as well as on the historical precedents of the rise of Islam
and Christianity, which he sees as examples of humanity's capacity
to adapt to new challenges ( 1973, chap. 12)
While Thomas explains the advances of the past and bases
his hopes for the future in the realm of the biological sciences, and
Dubos, although with a tempered appreciation of the past con-
tribution of scientific medicine, places his faith in humanity's
adaptive capacities, Illich lays the crucial disease problems of con-
temporary industrial society at the doors of the disease care institu-
tion. Illich is a priest by calling and very much of a catholic
(small c) scholar. He first sounded his thesis in a general discussion

of technological development in modern society ( 1973). In looking
at the past, he grants that for a brief period from about 1913
through World War II, modern medicine had some effective treat-
ment to offer, as did many shamans and herb doctors; and this
treatment made a modest contribution to man's well-being. But,
by the mid 1950s, it became clear that it is "the malignant ex-
pansion of institutional health care [that] is at the root of the rising
costsand demands and the decline in well-being. Society can . . .
have no quantitative standards by which to add up the negative

value of illusion, social control, prolonged suffering, loneliness,
genetic deterioration, and frustration produced by medical treat-
ment" (1973,pp.&-7).
Illich's devastating, well-documented, and detailed attack
on medicine (1976) quickly became a best seller. As McKeown
summarizes (1976, pp. 166ff.), modern medicine
Illich's position
is thrice iatrogenic in that "medicine does more harm than good;

it breeds demands for its services and supports features of society
which generate ill health; most seriously, it diminishes the capacity
of the individual to deal with his own health problems and to face
suffering and death."
can be understood only as a prophet. It is his mission
Illich
to denounce the sins of Sodom and Gomorrah, not to qualify and

measure. As for the past, he is not concerned with a balanced

appraisal and assignment of credit but only with showing how the
technological institutionalization of medicine and of modern life
bears negative consequences for our health. Others will praise

achievement; the prophet's mission is to strip bare our illusions
and sins.
For Illich, then, the medical establishment, with the brain-

washing it has wrought in modern society, is the Devil. We must
learn not only to be stoical in the face of the inevitable pain and
misery of life but also to see how they are rewarding. Thus implicit
in Illich's position is the view that even if we manage to free our-
selves from social and cultural iatrogenesis, our health status is not
likely to improve. The only real strand of hope in his outlook is
found in his attack on medical iatrogenesis. In the unlikely event

that the Western scientific medical institution is abolished, people
will at least not be subjected to the damage done to our health by
turning to doctors.
McKeown's own position contrasts sharply with the views
of the men he discusses. His view of medicine (1976) is not as
dream (Thomas), mirage (Dubos), or nemesis (Illich). In the
distant past, the major dangers to health were those determined by
the primitive conditions of life: food shortage, homicide, accidents,
and predation. Then, with the agricultural revolution some ten
thousand years ago, infectious diseases joined inadequate nutrition
as the predominant cause of sickness and death. With the significant
improvement in the last century in the standard of living and
particularly in nutritional standards, as well as in public health
control of the environment, major strides in health status have
been made.
Today we face a somewhat different set of problems, but
McKeown believes that his analysis of the past points the way to
the conditions that must be met to allow us to cope with new
problems. As he puts it (1976, p. 100; compare p. 174) : "Those
fortunate enough to be born free of significant congenital disease
or disability will remain wellif three basic needs are met: they
must be adequately fed; they must be protected from a wide range

of hazards in the environment; and they must not depart radically
from the pattern of personal behavior under which man evolved,

for example by smoking, overeating, or sedentary living."
If we are, then, to avoid premature death and improve the
health quality of our lives, we must follow the two major paths of
the past century: reasonable nutritional standards and control of en-
vironmental hazards. But the relatively new great challenges of
the chronic diseases are best met in the realm of personal behavior.
Thus, for example, he writes, "In the past century the improvement
in expectation of life of mature males from all causes has been
reduced by at least half by smoking alone" (p. 99). Thus Mc-
Keown does not see as the key to the future Thomas' commitment
to biological sciences, Dubos' faith in man's social adaptive ca-
pacities, or Illich's hostility toward the medical establishment. His
on concrete changes in behavior on the individual level, and
stress is
he would have us change our eating, smoking, and exercise habits.

"As these are now the main determinants of health, it is hard to
believe that society will not wish to create conditions under which
such practices are encouraged" (p. 164).
I do not deny the importance of the views presented above.
Basic biological research will provide us with immunological and
therapeutic weapons against specific diseases. Dubos' stress on
multiple causationand the need to study the conditions under
which endemic microorganisms become pathogenic and his orienta-
tion to the social determinants of the application of scientific knowl-
edge, vague as it is, are important contributions to our thinking
about disease. Illich's concern with undermining the legitimacy of
concentration of power in the hands of the disease care establish-
ment opens the way for a useful skepticism. And certainly Mc-
Keown's rational, data-guided approach is in the best empirical
tradition of attacking concrete dangers.
But much as all four positions share the fundamental virtue
of going beyond the narrow emphasis on what doctors do today as
the key to hopes for improvement of the health quality of life, they
all are characterized by fundamental inadequacies. Thomas' po-
sition is most open to criticism, which has been effectively leveled
by Dubos. The notion of the conquest of disease after disease by

magic bullets is doubly illusory: first, because chronic diseases can
be understood only in terms of multiple causation; second, because

the bugs are smarter. Mich has no real answer, except that we
should learn to live with our suffering. At best, he urges us to stay
away from doctors, thus at least avoiding iatrogenic diseases. Mc-
Keown disregards not only the difficulties of changing behavior but
also the data that show that the behaviors he would change account
for only a small part of the variance in disease causation.
The basic inadequacy of Thomas and McKeown, however,
is conceptual. Theirs is a pathogenic orientation. Their central
question is that of the etiology of diseases, in Thomas' case, or of
disease as well as diseases, in McKeown's. Hence they inevitably
focus on the elimination of disease agents. Mich, curiously enough,
shares this medical bias. He differs only in identifying the major
disease agent as the medical establishment.
Dubos, however, opens the way for a radical reconsidera-
tion of the problem. Not only does he, with Mich, assume that the
bugs will always be with us. Not only does he draw our attention
to the social-cultural-economic context in which the bugs flourish
and become But he also raises the question of our ca-

virulent.
pacities for adaptation and the social determinants of these capac-
ities. This is the heart of what I have called the salutogenic
orientation. Dubos' profound insights, his humanity, and his moving

fluency, however, do not lead him to go beyond posing the question.
Clearly, my own thinking has been strongly influenced by Dubos'
work; I have tried to go beyond what he has done.
The sense of coherence, I have argued, is a central com-
ponent of the answer. The stronger the sense of coherence of
and groups, the more adequately will they cope with
individuals
immanent in life, and the more likely are they to
the stressors
maintain or improve their positions on the health ease/dis-ease
continuum. In Chapter Eight, I dealt with the possible impact
of the doctor-patient relationship and of the dominant modes of
disease care delivery on the sense of coherence. But, as noted there,
this realm is hardly the decisive arena of life experience.
The is whether the societies in which our children
real issue
grow up and which we live our daily lives facilitate or impede

in
the development and maintenance of a strong sense of coherence.
Do they provide us with the generalized resistance resources that
give us experiences of an appropriate underload-overload balance

of stimuli, of consistency, and of participation in the shaping of
outcome? For most of us, I would suggest, such life experiences at —
our paid work, in our housework, in our social relations, in relation
to our community —are not at all frequent.

Of course it matters very much where one is located in the
social structure. Social classes, ethnic and racial groups, men and
women do not all share the same set of life experiences. To gen-
eralize about national societies without awareness of the great
difference it makes whether one is an Appalachian coal miner, a
suburban upper-middle-class housewife, or a university professor
is to obscure any possible understanding of differences in the sense
of coherence.
If we wish to see and future soberly in our
the present
world, we must use words and totalitarianism. The
like capitalism
social structures in which most of humanity lives and the daily

experiences to which we are exposed in these structures are far
from conducive to a strong sense of coherence. This observation is
as true for the United States as for the Soviet Union, for China
as for India. Societies with a marketing mentality and a fetishism
of commodities, with terror and arbitrary recasting of history, with
grinding poverty and starvation cannot foster a view of the world
as one that provides information and music except for the fortu-
nate few.
It would take another book and an extensive research effort
to subject to serious analysis the concrete social structuresand social
positions that in our world foster a strong sense of coherence. Im-
provement in health status is contingent on such analysis and on a
program of social action that could follow. The analysis is one of
the crucial tasks of social epidemiological research. The how-
task,
ever, cannot be undertaken unless there isan adequate set of
conceptual tools at our disposal. To provide such a set of tools has
been the central purpose of this book.
References
Alexander, f. Psychosomatic Medicine: Its Principles and Ap-

plications. New York: Norton, 1950.
Andrews, f. m., and WTTHEY, s. b. "Developing Measures of Per-
ceived Life Quality: Results from Several National Surveys."
Social Indicators Research, 1974, /, 1-126.
Andrews, f. m., and wiTHEY, s. b. Social Indicators of Well-
Being: Americans' Perception of Life Quality. New York:
Plenum, 1976.
antonovsky, a. "The Ideologies of American Jews: A Study in
Definitions of a Marginal Situation." Unpublished doctoral dis-
sertation, Department of Sociology, Yale University, 1955.
antonovsky, a. "Toward a Refinement of the 'Marginal Man'
Concept." Social Forces, 1956, 35, 57-62.
antonovsky, a. (Ed.). The Early Jewish Labor Movement in the
United States. New York: YIVO Institute for Jewish Research,

1961.
229
230 References
antonovsky, a. "Social Class and Illness: A Reconsideration."

Sociological Inquiry, 1967a, 37, 311-322.
antonovsky, a. "Social Class, Life Expectancy and Overall Mor-
tality." Milbank Memorial Fund Quarterly, 1967b, 45, 31-73.
antonovsky, a. "Social Class and the Major Cardiovascular Dis-

eases." Journal of Chronic Diseases, 1968, 21, 65-106.
antonovsky, a. "Social and Cultural Factors in Coronary Disease:
An Israel-North America Sibling Study." Israel Journal of
Medical Sciences, 1971, 7, 1578-1583.
antonovsky, a. "Breakdown: A Needed Fourth Step in the Con-
ceptual Armamentarium of Modern Medicine." Social Science
and Medicine, 1972, 6, 537-544.
antonovsky, a. "The Utility of the Breakdown Concept." Social
Science and Medicine, 1973, 7, 605-612.
antonovsky, a. "Conceptual and Methodological Problems in the
Study of Resistance Resources and Stressful Life Events." In B.
Dohrenwend and B. Dohrenwend (Eds.), Stressful Life Events:
Their Nature and Effects. New York: Wiley, 1974.
antonovsky, a. (Ed.). Abraham David Katz: Jew, Man, and
Sociologist. Jerusalem: Israel Institute of Applied Social Re-
search, 1975.
antonovsky, a. "Letter to the Editor." American Scientist, 1978,
66, 272-274.
antonovsky, a., and ARiAN, a. Hopes and Fears of Israelis: Con-
sensus in a New Society. Jerusalem: Jerusalem Academic Press,
1972.
antonovsky, a., and bernstein,
j. "Social Class and Infant
and Medicine, 1977, //, 453-470.
Mortality." Social Science
antonovsky, a., and hartman, h. "Delay in the Detection of
Cancer." Health Education Monographs, 1974, 2, 98-128.
antonovsky, a., and kats, r. "The Life Crisis History as a Tool
in Epidemiologic Research." Journal of Health and Social Be-
havior, 1967, 8, 15-20.
antonovsky, a., and kats, r. "The Model Dental Patient: An
Empirical Study of Preventive Health Behavior." Social Science
and Medicine, 1970, 4, 367-380.
antonovsky, a., and lorwin, l. (Eds.). Discrimination and Low
Incomes. New York: New York State Commission Against Dis-
crimination, 1959.
References 231
antonovsky, a., and shoham, I. Social Resistance Resources and

Health in Middle Age. Jerusalem: Israel Institute of Applied
Social Research, 1978.
antonovsky, a., and others. "Epidemiologic Study of Multiple
Sclerosis in Israel. I: An Overall Review of Methods and Find-
ings." Archives of Neurology, 1965, 13, 183-193.
antonovsky, a., and others. "Twenty-Five Years Later: A
Limited Study of the Sequelae of the Concentration Camp
Experience." Social Psychiatry, 1971, 6, 186-193.
antonovsky, h., and antonovsky, a. "Commitment in an Israeli
Kibbutz." Human Relations, 1974, 27, 303-319.

bahnson, c. b. "Role of the Ego Defenses: Denial and Repression
in the Etiology of Malignant Neoplasm." Annals of the New
York Academy of Sciences, 1966, 125, 827-845.
bahnson, c. b. "Epistemological Perspective of Physical Disease
from the Psychodynamic Point of View." American Journal of
Public Health, 1974, 64, 1034-1040.
becker, e. "Review of Tiger and Fox' The Imperial Animal."
Transaction/ SOCIETY, 1972, 9, 40-43.
becker, h. Outsiders: Studies in the Sociology of Deviance. New
York: Free Press, 1963.
belloc, n. b., breslow, l., and hochstim, j. r. "Measurement of
Physical Health in a General Population Survey." American
Journal of Epidemiology, 1971, 93, 328-336.
ben-sira, z. "The Doctor-Patient Relationship: Collectivism or
Exchange?" Unpublished doctoral dissertation, Department of
Sociology, Hebrew University, 1972.
BEN-smA, z. "The Function of the Professional's Affective Behavior
in Client Satisfaction : A Revised Approach to Social Interaction
Theory." Journal of Health and Social Behavior, 1976, 17, 3-11.
bensman, j., and Rosenberg, b. Mass, Class, and Bureaucracy.
Englewood Cliffs, N.J.: Prentice-Hall, 1963.
berkman, l. f. "Social Networks, Host Resistance and Mortality:
A Follow-Up Study of Alameda County Residents." Unpublished
doctoral dissertation, Department of Epidemiology, School of
Public Health, University of California, Berkeley, 1977.
berkman, l. f. "Psychosocial Factors, Host Resistance and Mor-
tality." American Journal of Epidemiology, in press.
blum, h. l., and keranen, g. m. Control of Chronic Diseases in

:
232 References
Man. Washington, D.C.: American Public Health Association,

1966.
booth, a., and cowell, j. "Crowding and Health." Journal of
Health and Social Behavior, 1976, 17, 204-220.

bowlby, j. 'The Making and Breaking of Affectional Bonds."
British Journal of Psychiatry, 1977, 130, 201-210, 421-431.
breslow, l. "A Quantitative Approach to the World Health Or-
ganization Definition of Health: Physical, Mental, and Social
Weil-Being." International Journal of Epidemiology, 1972, 1,
347-355.
bridges-webb, c. "The Traralgon Health and Illness Survey."
International Journal of Epidemiology, 1973, 2, 63-71 ; 1974, 3,
37-46, 233-246.
brinton, c. The Anatomy of Revolution. Englewood Cliffs, N.J.
Prentice-Hall, 1965.
bronfenbrenner, u. (Ed.). Influences on Human Development.
Hinsdale, Dryden, 1972.
111.:
buckley, w. (Ed.). Modern Systems Research for the Behavioral

Scientist. Chicago: Aldine, 1968.
burnet, m. Natural History of Infectious Disease. Cambridge,
Great Britain: Cambridge University Press, 1953.
burnet, m. Genes, Dreams, and Realities. Aylesbury, Bucks, Great
Britain: Medical and Technical Publishing Co., 1971. (Also
New York Basic Books, 1971.)
:
burnet, m. Intrinsic Mutagenesis: A Genetic Approach to Aging.

New York: Wiley, 1974.
cannon, w. b. Bodily Changes in Pain, Hunger, Fear, and Rage.
New York: Appleton-Century-Crofts, 1929.
cannon, w. b. The Wisdom of the Body. New York: Norton, 1939.
(Revised and expanded edition, 1963.)
CAPLOvrrz, d. The Poor Pay More. New York: Free Press, 1963.
Caroline, n. and Schwartz, h. "Chicken Soup Rebound and
l.,
Relapse of Pneumonia." Chest, 1975, 67, 215-216.

"Case Records of the Massachusetts General Hospital." New Eng-
land Journal of Medicine, 1977, 297, 828-834.
cassel, j. "Psychosocial Processes and 'Stress': Theoretical Form-
ulation." International Journal of Health Services, 1974, 4, 471-
482.
References 233
cassel, j. "The Contribution of the Social Environment to Host

Resistance." American Journal of Epidemiology, 1976, 104, 107—
123.
cassel, j. "The Relation of the Urban Environment to Health:
Toward a Conceptual Frame and a Research Strategy." In L. E.

Hinkle, Jr., and W. C. Loring (Eds.), The Effect of the Man-
Made Environment on Health and Behavior. Atlanta Center for:
Disease Control, Public Health Service, 1977.

cassel, j., Patrick, r., and jenkins, d. "Epidemiological An-
alysis of the Health Implications of Culture Change: A Con-
ceptual Model." Annals of the New York Academy of Sciences,
1960,84,938-949.
cassel, j., and tyroler, h. a. "Epidemiological Studies of Culture
Change. I: Health Status and Recency of Industrialization."
Archives of Environmental Health, 1961, 3, 25-33.
cassell, e. j. The Healer's Art: A New Approach to the Doctor-
Patient Relationship. Philadelphia: Lippincott, 1976.
Center for Disease Control. Reported Morbidity and Mortality in
the United States: Annual Summary 1977. DHEW Publication
No. CDC78-8241. Atlanta: Public Health Service, 1978.
chan, k. b. "Individual Differences in Reactions to Stress and
Their Personality and Situational Determinants." Social Science
and Medicine, 1977, 11, 89-103.
cobb, s. "Social Support as a Moderator of Life Stress." Psychoso-
matic Medicine, 1976, 37, 300-314.
cobb, s., and rose, r. m. "Hypertension, Peptic Ulcer, and Dia-
betes in Air Traffic Controllers." Journal of the American Med-
ical Association, 1973, 224, 489-492.
Cochrane, a. l. Effectiveness and Efficiency: Random Reflections

on Health Services. London: Nuffield Provincial Hospitals Trust,
1972a.
Cochrane, a. l. "The History of the Measurement of 111 Health."
International Journal of Epidemiology, 1972b, /, 89-92.
collen, m. f. A Case Study of Multiphasic Health Testing. Oak-
land, Calif. : Permanente Medical Group and Kaiser Foundation
Research Institute, 1977.
Commission on Chronic Illness. Chronic Illness in a Large City:

234 References
The Baltimore Study. Cambridge, Mass.: Harvard University

Press, 1957.
coser, l. The Functions of Social Conflict. New York: Free Press,
1956.
coser, r. l. "The Complexity of Roles as a Seedbed of Individual
Autonomy." In L. A. Coser (Ed.), The Idea of Social Structure:

Papers in Honor of Robert K. Merton. New York: Harcourt
Brace Jovanovich, 1975.
cousins, n. "Anatomy of an Illness (as Perceived by the Patient)."
New England Journal of Medicine, 1976, 295, 145&-1463.

cross man, r. (Ed.). The God That Failed. New York: Harper &
Row, 1950.
cutler, s. j., and young, j. l. "Demographic Patterns of Cancer
Incidence in the United States." In J. F. Fraumeni (Ed.),
Persons at High Risk of Cancer. New York: Academic Press,
1975.
dahrendorf, r. Class and Class Consciousness in Industrial Society.
Stanford, Calif.: Stanford University Press, 1966.
datan, n. "Women's Attitudes Toward the Climacterium in Five
Israeli Sub-Cultures." Unpublished doctoral dissertation, Com-
mittee on Human Development, University of Chicago, 1971.
datan, n., antonovsky, a., and maoz, b. A Time to Reap: The
Middle Age of Women in Five Israeli Subcultures. Baltimore:
Johns Hopkins University Press, forthcoming.
datan, n., and ginsberg, l. h. (Eds.) Normative Life Crises. New
.
York: Academic Press, 1975.

dobzhansky, t. Evolution, Genetics, and Man. New York: Wiley,
1955.
dodge, d. l., and martdm, w. t. Social Stress and Chronic Illness:
Mortality Patterns in Industrial Society. Notre Dame, Ind.:
University of Notre Dame Press, 1970.
DOHRENWEND, b., and DOHRENWEND, b. (Eds.). Stressful Life
Events: Their Nature and Effects. New York: Wiley, 1974.
dorn, h. and cutler, s. j. Morbidity from Cancer
f., in the
United States. Monograph No. 56. Washington, D.C.: Public
Health Service, 1959.
dubos, r. j. The Mirage of Health. London: Allen & Unwin, 1960.
,c
dubos, r. j. The Evolution of Microbial Diseases." In R. J.
"
References 235
Dubos and J. G. Hirsch (Eds.), Bacterial and Mycotic Infections

of Man. (4th ed.) Philadelphia: Lippincott, 1965.
dubos, r. j. Man, Medicine, and Environment. New York:
Praeger, 1968.
dubos, r. j. A God Within. London: Angus and Robertson, 1973.
durkheim, e. Suicide: A Study in Sociology. New York: Free
Press, 1951. (Originally published 1897.)
engel, g. l. "The Psychosomatic Approach to Individual Suscepti-
bility to Disease." Gastroenterology, 1974, 67, 1085-1093.
engel, g. l. "The Need Medical Model: A Challenge
for a New
for Biomedicine." Science, 1977, 196, 129-136.
engel, g. l., and schmale, a. h. " Conservation- Withdrawal A :
Primary Regulatory Process for Organismic Homeostasis." In

Ciba Foundation, Physiology, Emotion, and Psychosomatic Ill-
ness.Symposium 8. Amsterdam: Elsevier, 1972.
erikson, e. h. "Growth and Crises of the 'Healthy Personality.'
In J. E. Senn (Ed.), Symposium on the Healthy Personality.

New York: Josiah Macy, Jr., Foundation, 1950.
erikson, e. h. "The Problem of Ego Identity." Journal of the
American Psychoanalytic Association, 1956, 4, 58-121. (Re-
printed in M. R. Stein, A. J. Vidich, and D. M. White (Eds.),
Identity and Anxiety. New York: Free Press, 1960.)
erikson, e. h. Young Man Luther: A Study in Psychoanalysis and
History. New York: Norton, 1958.
erikson, e. h. "Identity, Psychosocial." In D. L. Sills (Ed.), The
International Encyclopedia of the Social Sciences. Vol. 7. New
York: Macmillan, 1968.
eyer, j., and sterling, p. "Stress-Related Mortality and Social
Organization." Review of Radical Political Economy, 1977, 9,
1-44.
fanshel, s. "A Meaningful Measure of Health for Epidemiology."
International Journal of Epidemiology, 1972, 1, 319-337.
feldman, j. j. "The Household Interview Survey as a Technique
for the Collection of Morbidity Data." Journal of Chronic Dis-
eases, 1960, 11, 535-557.
frank, j. d. in Illness and Healing."
"Mind-Body Relationships
Preventive Medicine, 1975, 2, 46-59.
frankenhaueser, m., and gardell, b. "Underload and Overload
236 References
in Working Life: Outline of a Multidisciplinary Approach."

Journal of Human Stress, 1976, 2(3), 35-46.
freidson, e. Professional Dominance. New York: Atherton, 1970.
fromm, e. Escape from Freedom. New York: Holt, Rinchart and
Winston, 1941.
galdston, I. (Ed.). Beyond the Germ Theory: The Roles of De-
privation and Stress in Health and Disease. New York: Health
Education Council, 1954.
Gallagher, e. b. (Ed.). The Doctor-Patient Relationship in the
Changing Health Scene. DHEW
Publication No. NIMH 78-
183. Washington, D.C.: Public Health Service, 1978.
garfield, s. r. "The Delivery of Medical Care." Scientific Ameri-
can, 1970, 222, 15-23.
garfield, s. r., and others. "Evaluation of a New Ambulatory
Medical Care Delivery System." New England Journal of Med-
icine, 1976,294,426-431.
glass, d. c. "Stress, Behavior Patterns, and Coronary Disease."
American Scientist, 1977, 65, 177-187.
goffman, e. Asylums. New York: Doubleday, 1961.
gove, w. r. "Societal Reaction as an Explanation of Mental Ill-
ness." American Sociological Review, 1970, 35, 873—884.

gove, w. r. "Sex, Marital Status, and Mortality." American Journal
of Sociology, 1973, 79, 45-67.
gunderson, e. k. e., and rahe, r. h. (Eds.). Life Stress and Ill-
ness. Springfield, 111.: Thomas, 1974.

gutman, h. g. The Black Family in Slavery and Freedom. New
York: Pantheon, 1976.
GUTTMAN, l. "Measurement as Structural Theory." Psychometrika,
1974, 36, 329-347.
hinkle, "The Effect of Exposure to Culture Change,
l. e., jr.
Social Change, and Changes in Interpersonal Relationships on

Health." In B. Dohrenwend and B. Dohrenwend (Eds.), Stress-
ful Life Events: Their Nature and Effects. New York: Wiley,
1974.
hinkle, l. e., jr. "Measurement of the Effects of the Environment
upon the Health and Behavior of People." In L. E. Hinkle, Jr.,

and W. C. Loring ( Eds. ) The Effect of the Man-Made Environ-
,
References 237
ment on Health and Behavior. Atlanta : Center for Disease Con-

trol, Public Health Service, 1977.
hinkle, l. e., jr., and loring, w. c. (Eds.)\ The Effect of the
Man-Made Environment on Health and Behavior. Atlanta:
Center for Disease Control, Public Health Service, 1977.
hinkle, l. e., jr., and wolff, h. g. "Health and Social Environ-
ment: Experimental Investigations." In A. Leighton and others
( Eds. ) , Explorations in Social Psychiatry. New York : Basic Books,
1957.
hochschild, a. r. "Disengagement Theory: A Critique and Pro-
posal." American Sociological Review, 1975, 40, 553-569.
holmes, o. w. The Autocrat of the Breakfast-Table. (Rev. ed.)
Boston: Houghton Mifflin, 1884.
holmes, t. h., and masuda, m. "Life Change and Illness Suscepti-
bility." In B. Dohrenwend and B. Dohrenwend (Eds.), Stressful
Life Events: Their Nature and Effects. New York: Wiley, 1974.
holmes, t. h., and rahe, r. h. "The Social Readjustment Rating
Scale." Journal of Psychosomatic Research, 1967, 11, 213-218.
illich, i. New York: Harper & Row, 1973.
Tools for Conviviality.
illich, Medical Nemesis: The Expropriation of Health. New
i.
York: Bantam, 1976.

jacobs, s., and ostfeld, a. "An Epidemiological Review of the
Mortality of Bereavement." Psychosomatic Medicine, 1977, 39,
344-357.
jones, e. The Life and Work of Sigmund Freud. (Edited and
abridged in one volume by L. Trilling and S. Marcus.) New
York: Basic Books, 1961.
justice, b., and duncan, d. f. "Life Crises as a Precursor to Child
Abuse." Public Health Reports, 1976, 91, 1 10-1 15.
kagan, a. r., and levi, l. "Health and Environment. Psychosocial
Stimuli: A Review." Social Science and Medicine, 1974, 8, 225-
241.
ranter, r. m. "Commitment and Social Organization." American
Sociological Review, 1968, 33, 499-517.
kaplan, b. h., cassel, j. c, and gore, s. "Social Support and
Health." Medical Care, 1977, 25 (suppl.), 47-58.
kardiner, a. The Individual and His Society. New York: Co-
lumbia University Press, 1939.
238 References
kardiner, a. The Psychological Frontiers of Society. New York:

Columbia University Press, 1945.
kardiner, a., and ovesey, l. The Mark of Oppression. New

York: World, 1951.
KASL,s. v. "The Effects of the Residential Environment on Health
and Behavior: A Review." In L. E. Hinkle, Jr., and W. C.
Loring (Eds.), The Effect of the Man-Made Environment on
Health and Behavior. Atlanta: Center for Disease Control,
Public Health Service, 1977.
khutz, s., and moos, r. h. "Physiological Effects of the Social En-
vironment." In P. M. Insel and R. H. Moos (Eds.), Health and
the Social Environment. Lexington, Mass.: Heath, 1974.
kleinman, a., eisenberg, l., and good, b. "Culture, Illness and
Care: Clinical Lessons from Anthropological and Cross-Cultural
Research." Annals of Internal Medicine, 1978, 88, 251-258.
knupfer, g. "Portrait of the Underdog." Public Opinion Quar-
terly, 1947, /i, 103-1 14.
kohn, m. l. "Social Class and Schizophrenia: A Critical Review."
In D. Rosenthal and S. S. Kety (Eds.), The Transmission of
Schizophrenia. Oxford: Pergamon Press, 1968.
kohn, m. l. Class and Conformity: A Study in Values. (1st ed.)
Homewood, 111.: Dorsey Press, 1969.
kohn, M. l. "Social Class and Schizophrenia: A Critical Review

and a Reformulation." Schizophrenia Bulletin, 1973, 7, 60-79.
kohn, m. l. "The Interaction of Social Class and Other Factors
in the Etiology of Schizophrenia." American Journal of Psy-
chiatry, 1976a, 133, 177-180.
kohn, m. l. "Occupational Structure and Alienation." American
Journal of Sociology, 1976b, 82, 1 1 1-130.
kohn, m. l. "The Substantive Complexity of Work and Its Rela-
tionship to Adult Personality." In E. H. Erikson and N. Smelser
(Eds.), forthcoming.
kohn, m. l., and schooler, c. "The Reciprocal Effects of the
Substantive Complexity of Work and Intellectual Flexibility: A
Longitudinal Assessment." American Journal of Sociology, 1978,
84, 24^-52.
kosa, j., antonovsky, a., and Zola, i. k. (Eds.). Poverty and
References 239
Health: A Sociological Analysis. Cambridge, Mass.: Harvard

University Press, 1969.
kuhn, t. The Structure of Scientific Revolutions. Chicago: Uni-
versity of Chicago Press, 1962.
lazarus, r. s., and cohen, j. b. "Environmental Stress." In I.
Altaian and J. F. Wohlwill (Eds.), Human Behavior and En-

vironment. Vol. 2. New York:
Plenum, 1977.
lefcourt, h. m. Locus of Control: Current Trends in Theory and
Research. New York: Wiley, 1976.
lerner, L m. Heredity, Evolution and Society. San Francisco:
W. H. Freeman, 1968.
leshan, l. You Can Fight for Your Life: Emotional Factors in
the Causation of Cancer. New York: Harcourt Brace Jovan-
ovich, 1978.
levi, l. "Psychosocial Stress and Disease: A Conceptual Model."
In E. K. E. Gunderson and R. H. Rahe (Eds.), Life Stress and
Illness. Springfield, 111.: Thomas, 1974.
levin, d. l., and others. Cancer Rates and Risks. (2nd ed.)
Washington, D.C.: Public Health Service, 1974.
lewis, c. e., and lewis, m. a. "The Potential Impact of Sexual
Equality on Health." New England Journal of Medicine, 1977,
297, 863-869.
lewis, o. "The Culture of Poverty," Anthropological Essays. New
York: Random House, 1970.
liddell, h. s. "Sheep and Goats: The Psychological Effects of
Laboratory Experiences of Deprivation and Stress upon Certain
Experimental Animals." In I. Galdston (Ed.), Beyond the Germ
Theory. New York: Health Education Council, 1954.
lowenthal, m. f., and CHnuBOGA, d. "Social Stress and Adapta-
tion: Toward a Life Course Perspective." In C. Eisdorfer and
M. P. Lawton (Eds.), The Psychology of Adult Development
and Aging. Washington, D.C. American Psychological Associa-
:
tion, 1973.
MC ALPINE, D., LUMSDEN, C. E., AND ACHESON, E. D. Multiple
Sclerosis: A Reappraisal. (2nd ed.) Edinburgh and London:
Churchill Livingstone, 1972.
mc keown, t. The Role of Medicine: Dream, Mirage, or Nemesis?
London: Nuffield Provincial Hospitals Trust, 1976.

240 References
mc kinlay, j. b. "The Changing Political and Economic Context

of the Patient-Physician Encounter." In E. B. Gallagher (Ed.),
The Doctor-Patient Relationship in the Changing Health Scene.
Washington, D.C.: Public Health Service, 1978.
malinowski, b. "Culture." In E. R. A. Seligman (Ed.), The En-
cyclopedia of the Social Sciences. Vol. 4. New York: Macmillan,

1931.
maoz, b., and others. "Female Attitudes to Menopause." Social
Psychiatry, 1970, 5, 35-40.
mathews, k. a., and others. "Competitive Drive, Pattern A,
and Coronary Heart Disease: A Further Analysis of Some Data
from the Western Collaborative Group Study." Journal of
Chronic Diseases, 1977, 30, 489-498.
mechanic, d. Students Under Stress: A Study in the Social Psy-
chology of Adaptation. New York: Free Press, 1962.

mechanic, d. Medical Sociology: A Selective View. New York:
Free Press, 1968.
medawar, p. "Review of L. Leshan's You Can Fight for Your
Life." New York Review of Boohs, June 9, 1977.
mercer, j. r. "Who Is Normal? Two Perspectives on Mild Mental
Retardation." In E. G. Jaco (Ed.), Patients, Physicians and
Illness. (2nd ed.) New York: Free Press, 1972.
merton, r. k. "Social Structure and Anomie," Social Theory and
Social Structure. New York: Free Press, 1957.
merton, r. k. Social Theory and Social Structure. (Expanded ed.)
New York: Free Press, 1968.
merton, r. k. *The Ambivalence of Physicians," Sociological Am-
bivalence and Other Essays. New York: Free Press, 1976.
monjan, a. a., and collector, m. i. "Stress-Induced Modulation
of the Immune Response." Science, 1977, 196, 307-308.

morris, j. n. Uses of Epidemiology. Edinburgh and London:
Churchill Livingstone, 1957.
moss, g. e. Illness, Immunity and Social Interaction: The Dyna-
mics of Biosocial Resonation. New York: Wiley, 1973.
naditch, m. p. "Locus of Control, Relative Discontent and Hyper-
tension." Social Psychiatry, 1974, 9, 11 1-1 17.
National Center for Health Statistics. "Prevalence of Selected
References 241
Chronic Digestive Conditions, United States, July-December,

1968." Vital and Health Statistics, 1973a, Series 10, No. 83.
National Center for Health "Prevalence of Selected
Statistics.
Chronic Respiratory Conditions, United States, 1970." Vital

and Health Statistics, 1973b, Series 10, No. 84.
National Center for Health Statistics. "Prevalence of Chronic
Circulatory Conditions, United States, 1972." Vital and Health
Statistics, 1974, Series 10, No. 94.
National Center for Health Statistics. "Prevalence of Selected
Impairments, United States, 1971." Vital and Health Statistics,
1975, Series 10, No. 99.

National Center for Health Statistics. Health: United States, 1975.
DHEW Publication No. HRA 76-1232. Rockville, Md.: De-
partment of Health, Education, and Welfare, 1976.
National Center for Health Statistics. "Current Estimates from the
Health Interview Survey, United States, 1976." Vital and Health
Statistics, 1977a, Series 10, No. 119.
National Center for Health Statistics. "Profile of Chronic Illness in
Nursing Homes, United States: National Nursing Home
Survey." Vital and Health Statistics, 1977b, Series 13, No. 29.
National Center for Health Statistics. "Acute Conditions, Inci-
dence, and Associated Disability, United States, July 1975-June
1976." Vital and Health Statistics, 1978, Series 10, No. 120.
National Health Education Committee. The Killers and the Crip-
piers: Facts on Major Diseases in the United States Today. New
York: McKay, 1976.
nisbet, r. The Social Philosophers: Community and Conflict in
Western Thought. New York: Crowell, 1973.
NUCKOLLS, K. B., CASSEL, J., AND KAPLAN, B. H. "Psychosocial
Assets, Life Crisis and the Prognosis of Pregnancy." American
Journal of Epidemiology, 1972, 95, 431-441.
oakley, a. Woman's Work. New York: Pantheon, 1975.
obeyesekere, g. "Psycho-Cultural Exegesis of a Case of Spirit
Possession from Sri Lanka." Contributions to Asian Studies,
1973, 8, 41-89.
oettgen, h. f. "Immunotherapy of Cancer." New England Jour-
nal of Medicine, 1977, 297, 484-491.
ostfeld, a. m., and d'atri, d. a. "Psychophysiological Responses
242 References
to the Urban Environment." International Journal of Psychiatry

in Medicine, 1975, 6, 15-28.
parkes, c. m. "Psychosocial Transitions: A Field for Study." Social
Science and Medicine, 1971, 5, 101-115.
parkes, c. m. Bereavement: Studies of Grief in Adult Life. New
York: International Universities Press, 1972.
PARKES, C. BENJAMIN, B., AND FITZGERALD, R. G. "Broken
M.,
Heart: A Study of Increased Mortality Among Wid-
Statistical
owers." British Medical Journal, 1969, /, 740-743.

parsons, t. The Social System. New York: Free Press, 1951.
parsons, t. "Epilogue." In E. B. Gallagher (Ed.), The Doctor-
Patient Relationship in the Changing Health Scene. DHEW
Publication No. NIMH 78-183. Washington, D.C.: Public
Health Service, 1978.
pearlin, l. i. "Sex Roles and Depression." In N. Datan and L. H.
Ginsberg (Eds.), Normative Life Crises. New York: Academic
Press, 1975.
pearlin, L. i., and johnson, j. s. "Marital Status, Life-Strains
and Depression." American Sociological Review, 1977, 42, 704-
715.
pearlin, l. l, and schooler, c. "The Structure of Coping."
Journal of Health and Social Behavior, 1978, 19, 2-21.
perlmuter, l. c, and monty, r. a. "The Importance of Per-
ceived Control: Fact or Fantasy?" American Scientist, 1977, 65,
759-765.
petrie, a. Individuality in Pain and Suffering. Chicago: University
of Chicago Press, 1967.
Phillips, d. p., and feldman, k. a. "A Dip in Deaths Before
Ceremonial Occasions: Some New
Relationships Between Social
Integration and Mortality." American Sociological Review,
1973,55,678-696.
powles, j. "On the Limitations of Modern Medicine." Science,
Medicine and Man, 1973, /, 1-30.
radloff, r., and helmreich, r. Groups Under Stress: Psycholog-
ical Research in SEALAB II. New York: Appleton-Century-
Crofts, 1968.
rahe, r. h. "The Pathway Between Subjects' Recent Life Changes
and Their Near-Future Illness Reports." In B. Dohrenwend and

References 243
B. Dohrenwend (Eds.), Stressful Life Events: Their Nature and

Effects. New York: Wiley, 1974.
rahe, r. h. "Editorial: Life Change Measurement Clarification."
Psychosomatic Medicine, 1978, 40, 95-98.
rapoport, "The Promise and Pitfalls of Information Theory."
a.
In W. Buckley (Ed.), Modern Systems Research for the Be-
havioral Scientist. Chicago: Aldine, 1968.
roazen, p. Freud: Political and Social Thought. New York:
Knopf, 1968.
roeske, n. a. "The Emotional Response to Hysterectomy." Psy-
chiatric Opinion, 1978, 15 (2), 11-20.
rosenstock, i. m. "What Research in Motivation Suggests for
Public Health." American Journal of Public Health, 1960, 50,
295-302.
rotter, j. b. Generalized Expectancies for Internal Versus Ex-
ternal Control of Reinforcement. Psychological Monographs
No. 80, 1966.
roueche, b. "Annals of Medicine: Antipathies." New Yorker,
March 13, 1978, pp. 61-84.
ruesch, j. Duodenal Ulcer. Berkeley: University of California
Press, 1948.
rutstein, d. d., and others. "Measuring the Quality of Medical
Care." New England Journal of Medicine, 1976, 294, 582-588.
schachtel, e. "On Alienated Concepts of Identity."
g. In E.
Josephson and M. Josephson (Eds.), Man Alone. New York:
Dell, 1962.
scheff, t. j. Being Mentally III. Chicago: Aldine, 1966.
schmale, a. h. "Giving Up as a Final Common Pathway to
Changes in Health." In Z. J. Lipowski (Ed.), Advances in
Psychosomatic Medicine. Vol. 8. Basel, Switzerland: S. Karger,
1972.
schrodinger, e. "Quote from What Is Life?" In W. Buckley
(Ed.), Modern Systems Research for the Behavioral Scientist.
Chicago: Aldine, 1968.
sghur, m. Freud: Living and Dying. New York: International
Universities Press, 1972.
Scott, r. a. The Making of Blind Men: A Study of Adult Social-
ization. New York: Russell Sage Foundation, 1969.
244 References
scott, r. a., and Howard, a. "Models of Stress." In S. Levine and

N. A. Scotch (Eds.), Social Stress. Chicago: Aldine, 1970.
seligman, m. e. p. Helplessness: On Depression, Development, and
Death. San Francisco: W. H. Freeman, 1975.
selye, h. The Stress of Life. New York: McGraw-Hill, 1956.
selye, h. "Confusion and Controversy in the Stress Field." Journal
of Human Stress, 1975, /, 37-44.
shaut, b. "Structural Ambiguity and Limits to Coping." Journal
of Human Stress, 1977, 3 (4), 32-45.
sheehy, g. Passages: Predictable Crises of Adult Life. New York:
Dutton, 1974.
shuval, j. t. "Some Persistent Effects of Trauma: Five Years
After the Nazi Concentration Camps." Social Problems, 1957—
1958,5,230-243.
shuval, j. t., antonovsky, a., and davdzs, a. m. Social Functions
of Medical Practice: Doctor-Patient Relationships in Israel. San
Francisco: Jossey-Bass, 1970.
sim mel, c Conflict and the Web of Group Affiliations. (K. H.
Wolff and R. Bendix, Trans.) New York: Free Press, 1955.
simon, w., and gagnon, j. h. "The Anomie of Affluence: A Post-
Mertonian Conception." American Journal of Sociology, 1976,
82, 356-378.
solomon, g. f., and am kraut, a. a. "Emotions, Stress and Im-
munity." In P. M. Insel and R. H. Moos (Eds.), Health and

the Social Environment. Lexington, Mass.: Heath, 1974.
sorochan, w. d. "Health Concepts as a Basis for Orthobiosis." In
E. Hart and W. Sechrist (Eds.), Dynamics of Wellness. Belmont,
Calif.: Wadsworth, 1970.
srole, l., and others. Mental Health in the Metropolis: The
Midtown Study. Vol. 1. New York: McGraw-Hill, 1962.
starr, p. "Review of W. Gaylin and others' The Limits of Benev-
olence/* New York Times Booh Review, May 21, 1978, p. 9.
susser, m. Community Psychiatry: Epidemiologic and Social
Themes. New York: Random House, 1968.
susser, m. "Ethical Components in the Definition of Health."
International Journal of Health Services, 1974, 4, 539-548.
sweeney, d. r., tinung, d. c., and schmale, a. h., jr. "Differ-
entiation of the 'Giving Up' —Helplessness and Hopeless-
Affects
ness." Archives of General Psychiatry, 1970, 23, 378-382.
References 245
syme, s. l., and berkman, l. f. "Social Class, Susceptibility, and

Sickness." American Journal of Epidemiology, 1976, 104, 1-8.
syme, s. l., and torfs, c. p. "Epidemiologic Research in Hyper-
tension: A Critical Appraisal." Journal of Human Stress, 1978,
4, 43-48.
szasz, t. The Myth of Mental Illness. New York: Hoeber Harper,
1963.
szasz, t., and hollender, m. h. "A Contribution to the Philos-
ophy of Medicine: The Basic Models of the Doctor-Patient Re-

lationship." Archives of Internal Medicine, 1956, 97, 585-592.
taeuber, c. (Ed.). America in the Seventies: Some Social In-
dicators. Philadelphia: American Academy of Political and
Social Sciences, 1978.
thacker, s. b., and salber, e. j. "Hospitalizations
green, s. b.,
in a Southern Rural Community: An Application of the 'Ecol-

ogy Model.' " International Journal of Epidemiology, 1977, 6,
55-63.
thomas, l. The Lives of a Cell. Toronto: Bantam Books of
Canada, 1974.
toffler, a. Future Shock. New York: Random House, 1970.
tyroler, h. and cassel, j. c. "Health Consequences of Cul-
a.,
ture Change: The Effect of Urbanization on Coronary Heart

Mortality in Rural Residents of North Carolina." Journal of
Chronic Diseases, 1964, 17, 167-177.
vickers, g. "The Concept of Stress in Relation to the Disorganiza-
tion of Human Behavior." In W. Buckley (Ed.), Modern Sys-
tems Research for the Behavioral Scientist. Chicago: Aldine,
1968.
weisenberg, m. "Pain and Pain Control." Psychological Bulletin,
1977, 84, 1008-1044.

weissman, m. m., and paykel, e. s. "Moving and Depression in
Women." Transaction/ SOCIETY, 1972, 9 (9), 24-28.
wertheim, e. s. "Positive Mental Health, Western Society and the
Family." International Journal of Social Psychiatry, 1974, 21,
982/1-5.
wertheim, e. s. "Person-Environment Interaction: The Epi-
genesis of Autonomy and Competence." Journal of Med-
British
ical Psychology, 1975, 48, 1-8, 95-111, 237-256, 391-402.

wertheim, e. s. "Developmental Genesis of Human Vulnerability:
246 References
Conceptual Re-Evaluation." In Anthony and £. Koupernik

J. £.
(Eds.), Vulnerable Children. Yearbook of the International
Association for Child Psychiatry and Allied Professions, Vol. 4.
New York: Wiley, 1978.
WHITE, K. L., WILLIAMS, T. F., AND GREENBERG, B. G. "The Ecology
of Medical Care." New England Journal of Medicine, 1961,
265, 885.
winkelstein, w. "Epidemiological Considerations Underlying the
Allocation of Health and Disease Care Resources." International
Journal of Epidemiology, 1972, /, 69-74.
woHLwnx, j. f. "The Problem for a
Physical Environment: A
Psychology of Stimulation." In H. C. Lindgren and others
(Eds.), Current Research in Psychology. New York: Wiley,
1971.
wolf, s., and goodell, h. Harold G. Wolff's Stress and Disease.
(2nded.) Springfield, m.: Thomas, 1968.
young, a. "Internalizing and Externalizing Medical Belief Systems:
An Ethiopian Example." Social Science and Medicine, 1976a,
10, 147-156.
young, a. "Some and Practices for
Implications of Medical Beliefs
Social Anthropology." American Anthropologist, 1976b, 78, 5-
24.
young, a. "Order, Analogy and Efficacy in Ethiopian Medical
and Psychiatry, 1977, /, 183-199.
Divination." Culture, Medicine
Zola, i. and Symptoms: An Analysis of Patients' Pre-
k. "Culture
senting Complaints." American Sociological Review, 1966, 31,
615-630.
Index
62, 64, 68, 69, 74, 77, 93, 96, 99,

101, 106, 108, 115, 116, 117, 153,
Accidents and survivors, as psycho- 162, 163, 217, 229-231, 234, 238-
social stressor, 80-81 239, 244
acheson, e. d., 34, 239 antonovsky, h., 116, 117, 231
Action implication: as breakdown arian, a., 68, 230
facet, 62-64; concept of, 63 Arthritis, data on, 22
Acute morbidity condition: defined, Artifactual-material factors, as gen-
28; prevalence of, 27-28, 29 eralized resistance resources, 106-
Aggression, as psychosocial stressor, 107
82 Autonomy, sense of coherence related
Alcohol abuse, data on, 22 to, 140-141
ALEXANDER, F., 105, 229
Allergy, data on, 19, 20, 22, 26
Alorese culture, and sense of coher- B
ence, 150
American Cancer Society, 23 BAHNSON, C. B., 173, 231
AMKRAUT, A. A., 181, 244 BECKER, E., 82, 231
ANDREWS, F. M., 68, 229 BECKER, H., 25, 231
Anemia, data on, 19 BELLOC, N. B., 31, 54-55, 231
Animal studies, and health linked to BENJAMIN, B., 189, 242
sense of coherence, 179 ben-ska, z., 207, 231
antonovskv, a., 2, 3, 4, 5, 6, 43, 56, BENSMAN, J., 156, 231
247
248 Index
Bereavement, and health linked to 187, 191; sense of coherence related

sense of coherence, 166 to,139-142, 146-147
BERKMAN, L. F., 66, 115, 163-165, cHnuBOOA, d., 114, 239
168, 231, 245 Chronic conditions, substantial, de-
BERNSTEIN, j., 3, 108, 163, 230 fined, 18
Bias, cultural, danger of, 152-158 Circulatory conditions, chronic, data
Biochemical factors, as generalized on, 23
resistance resources, 103-106
Clearinghouse on Health Indices, 50n
Birth defects, data on, 21
cobb, s., 115, 168,233
blum, h. l., 18-19, 24, 231-232
COCHRANE, a. i>, 47-48, 101, 233
booth, a., 178, 232
Cognitive appraisal, coping related
bowlby, j., 139, 232
to, 111
Breakdown: concept of, 4-5; con-
Cognitive factors, as generalized re-
tinuum of, 55-64; mapping sen-
sistance resources, 107-110
tence of, 64-65; and negative life
cohen, j. b., 72, 74, 80, 87, 111, 239
events, 177; profile of, 64-67
breslow, l., 31, 54-^5, 59, 231, 232 Coherence, sense of: case histories of,
128-136; continuum for, 158; and
bridges- webb, c, 16, 232
cultural bias, 156-158; defined,
BRiNTON, c, 134-136, 157, 232
123-128; direct evidence of link
Bronchitis, chronic, data on, 19, 22
with health by, 161-163; ego iden-
BRONFENBRENNER, U., 85, 87, 232
tity distinct from, 109-110; emer-
BUCKLEY, w., 78, 232
gence of, 187-188; evolution of
Bureaucratic modality, of doctor-
perspective on, 1-11; fake, 158-
patient relationship, 209, 210, 213-
159; and generalized resistance re-
214, 215
sources, 122; global orientation of,
burnet, m., 14, 78-79, 103, 220, 232
180; and health care system, 198-
BUTLER, S., 102
219; and health outlook, 220-227;
health related to, 160-181; indirect
evidence of link with health by,
163-179; modification of, 188;
Cancer, psychological factors in, 173 operationalizing, 158-159; percep-
cannon, w. b., 71, 129, 165, 176, 232 tions of, 123-159; in salutogenic
CAPLOvrrz, d., 108, 232 model, 183, 186-187; sources of,
Cardiovascular diseases, data on, 19 136-152, 190-192
CAROLINE, N. l., 193n, 232 COLLECTOR, M. L, 96, 240
cassel, j. c, 83, 166-167, 168, 171, Collegial modality, of doctor-patient
178-179, 180, 232-233, 237, 241, relationship, 208, 209, 21 1-212, 215
245 collen, m. r., 31-32, 233
cassell, e. j., 36-37, 43-44, 204- Comanche culture, and sense of co-
206,211-212,213,233 herence, 150
Cataclysmic events, as stressors, 74, Commission on Chronic Illness, 17-
80, 87, 188
22, 233-234
Center for Disease Control, 26, 27n,
Commitment, concept of, 116-117
233
Competence, sense of coherence re-
CHAN, k. b., 122, 233
lated to, 140-141
Change: and health linked to sense
Complexity: sense of coherence re-
of coherence, 169-171, 189; as psy-
chosocial stressor, 83 lated to, 148; substantive, of work,
Childrearing: in salutogenic model, 144-145, 147, 148

Index 249
Conformity/self-direction, sense of Dis-ease, concept of, 5

coherence related to, 144, 146-147 Disease: illness distinct from, 43,
Conservation-withdrawal, sense of 204-206; notifiable, prevalence of,
coherence related to, 138-139, 166 26-27
Consistency, sense of coherence re- Disease care system: clinical model
lated to, 187, 189, 200, 208 of, 39-45; number of people in, 42;
Control: locus of, 152-154, 158, 171— public health model of, 45-47. See
172; personal, 154—155; sense of, also Health care system
related to sense of coherence, 127- dobzhansky, t., 104, 234
128 dodge, d. l., 166, 168, 234
Coping: for, 162-163; con-
ability dohrenwend, b., 176, 177, 234
cept 110-111, 112; functions of,
of, dohrenwend, b., 176, 177, 234
194; as generalized resistance re- DORN, H. F., 23, 234
source, 110-114; sex differences in, DOSTOEVSKY, F., 124
148 Drive reduction, tension related to,
Coronary heart disease, and Type A 95-96

behavior pattern, 174—175 dubos, r. j., 5, 11, 13, 16, 53, 79,
coser, l., 157, 234 104, 105, 193, 221, 222-223, 224,
coser, r. l., 137, 145-149, 234 225, 226, 234-235
cousins, n., 128-132, 195, 234 DUNCAN, D. F., 143, 237
cowell, j., 178, 232 DURKHEIM, E., 91, 235
CROSSMAN, R., 127, 234 Dynamic, related to sense of coher-
Crowding, and health linked to sense ence, 125
of coherence, 178-179
Cultural-historical sources, for sense
of coherence, 149-152
Cultural variables, and health linked Ego identity: in salutogenic model,
to sense of coherence, 168-171 187-188, 191; sense of coherence

cutler, s. j., 23, 234 distinct from, 109-110; and tension
management, 108-110
eisenberg,l., 43, 212, 238
D Emotional affect, and coping strate-
gies, 114
d'atri, d. a., 169,241-242 Emotional factors, as generalized re-
dahrendorf, 234
r., 88, sistance resources, 107-110
DATAN, N., 6, 86, 115, 234 Emphysema, pulmonary, data on, 19
DAVTES, A. M., 2, 43, 244 engel, g. l., 107, 137, 138, 139, 141,
Deafness, data on, 18, 20-21 152, 162, 166, 173, 179, 235
Democratic modality, of doctor- Entropy: negative, and generalized
patient relationship, 209, 210-211, resistance resources, 120-122; stres-
214 sors related to, 77-78
Dental caries, data on, 24—25 Epidemiology, and health care, 45-4-7
Depression, helplessness related to, erikson, e. h., 85, 109, 151, 235
172 Expectant faith, and health linked to
Destruction, as psychosocial stressor, sense of coherence, 172-173
82 eyer, j., 39n, 235
Diabetes, data on, 18, 22
Digestive diseases, data on, 21, 24
Disability: causes of, 22; prevalence
of, 28, 30 Facet theory, and breakdown, 57
250 Index
fanshel, s., 47, 50-51, 59, 235 goffman, e., 25, 149n, 236
Farsightedness, as coping strategy, good, b., 43, 212, 238
113, 170 goodell, h., 75-76, 105, 169, 170,
FELDMAN, J. J., 23, 235 246
FELDMAN, K. 242
A., 1 14, gore, s., 166, 167, 237
FITZGERALD, R. G., 189, 242 cove, w. *., 25, 114, 165, 236
Flexibility, as coping strategy, 113, GREEN, S. B., 42, 245
170 GREEN BERG, B. G., 16-17, 246
FLEXNER, A., 214 Group Health Insurance, 43
FOX, R., 82 GUNDERSON, E. K. E., 176, 236
FRANK, J. D., 172-173, 235 GUTMAN, H. G., 8, 236
FRANKENHAUESER, M., 87, 148, 235- GUTTMAN, L,, 8, 57, 236
236
FREIDSON, E., 213, 236 H
freud, s., 81-82, 132-134, 195
FRIED, M., 3 HARTMAN, H., 101, 230
FRIEDMAN, M., 173-174 Health: clinical model of, 39-45; as
FROMM, E., 151, 236 a continuum, 38-69; continuum
Functional limitation: as breakdown model of, 47-52; defined, 52, 53;
facet, 59-61; concept of, 60 dichotomous model of, implica-
Functioning, concept of, 50 tions of, 43-45; outlook for, 220-
227; in salutogenic model, 196-
197; salutogenic model of, 182-
197; sense of coherence related to,
gagnon, j. h., 88-89, 244 160-181; study of, rather than dis-
GALDSTON, i., 80, 87, 236 ease, 12-37; and value systems, 48-
gallagher, e. b., 214, 236 50
gardell, b., 87, 148, 235-236 Health care: availability of, 217; pre-
garfield, s. r., 46, 236 ventive, 217-218
GAVLTN, W., 218 Health care system: and high-risk
Generalized resistance resources groups, 218-219; implications of
(GRRs) : absence of, as stressor, sense of coherence for, 198—219;
119-120; analysis of, 102-119; con- modalities of relationships in, 207-
cept of, 5, 99; dual function of, 214; patients in relationship with,
194-195; general adaptation syn- 199-214; routine encounters in,
drome distinct from, 105; mapping 202-207; and social structure, 214-
sentence of, 102-103, 120-121; 219; traumatic situations in, 200-
meaning 99-100; as negative
of, 202; trust in practitioners of, 218.
entropy 120-122; in
providers, See also Disease care system
salutogenic model, 189-192; signifi- Health ease/dis-ease. See Breakdown
cance of, 122; sources of, 137, 152, Health Insurance Plan, 43
190-192; and tension management, HELMREICH, R-, 95, 242
98-122 Helplessness: depression related to,
Genetics, stressors related to, 78-79 172; sense of coherence related to,
GINSBERG, L. H., 86, 234 138, 139-140; and Type A behav-
Giving-up syndrome, sense of coher- ior, 175
ence related to, 137-138, 162 HINKLE, L. E., JR., 76, 79, 169, 170-
glass, d. c., 175, 180, 236 171, 236-237
Goal achievement, as psychosocial Historical sources, for sense of co-
stressor, 88-89 herence, 149-152
Index 251
History: horrors of, as psychosocial Kaiser-Permanente, 31-32, 43, 46

stressor, 82; immediacy of, as psy- kanter, r. m., 116, 237
chosocial stressor, 83 kaplan, b., 166, 167, 237, 241
HOCHSCHILD, A. R., 114-115, 237 kardiner, a., 137, 149-152, 237-238
hochstim, j. r., 31, 54-^55, 231 kasl, s. v., 169, 238
HOLLENDER, M. H., 205, 245 kats, r., 2, 4, 74, 101, 230
holmes, o. w., 83-84, 195, 237 KERANEN, G. M., 18-19, 24, 231-232
holmes, t. h., 75, 81, 86, 93-94, 143, KnuTz, s., 168, 238
176-177, 237 KLEINMAN, A., 43, 212, 238
Homeostasis, stressors related to, 71- Knowledge-intelligence, and tension
72,80 management, 107-108
Hopelessness, sense of coherence re- knupfer, g., 108, 238
lated to, 138 kohn, m. l., 107, 137, 142-149, 161-
HOWARD, 72, 112, 244
a., 162, 238
Human Ecology Study Program, kosa, j., 3, 77, 238-239
75-76, 170, 176 kuhn, t., 38, 239
Hyperuricemia, data on, 19
Lawfulness, related to sense of co-

illich, l, 11, 203, 221, 223-224, 225,
herence, 127
226, 237
lazarus, r. s., 72, 73, 74, 80, 87,
Illness: chronic, prevalence of, 17-
110,111,239
204-
26; disease distinct from, 43,
LEFCOURT, H. M., 152, 158-159, 171,
206
239
Immunological surveillance, concept
LERNER, I. M., 104, 239
of, 103
LESHAN, L., 173, 239
Immunopotentiators, and tension
levi, l., 180, 237, 239
management, 104 LEVIN, D. L., 23, 239
Intelligence, and tension manage-
lewis, c. e., 156-157, 239
ment, 107-108
lewis, m. a., 156-157, 239
Interpersonal-relational factors, as
lewis, o., 143, 239
generalized resistance resources,
LIDDELL, H. S., 179, 239
114-117
Life cycle crises, normative, as psy-
Intrapsychic conflict, as psychosocial
chosocial stressors, 84-86
stressor, 81-82 and health linked to
Life events:
sense of coherence, 176-179; in
salutogenic model, 187-189
Limitation: functional, as breakdown
jacobs, s., 166, 237 facet, 59-61; prevalence of, 28, 30
JENKINS, d., 171, 233 Load balance: and psychosocial
johnson, j. s., 172, 242 86-87, 95; sense of coher-
stressors,
jones, e., 132n, 237 ence related to, 187, 189, 200, 205,
justice, b., 143, 237 208,211,216
Locus-of-control concept, sense of
coherence related to, 152-154,
158, 171-172
KAFKA, F., 127 loring, w. c., 169, 237
KAGAN, A. R., 180, 237 LORWTN, L., 2, 230
252 Index
LOWENTHAL, M. F., 1 14, 239 moss, o. e., 72, 240

LUMSDEN, C. E., 34, 239 Mutilation, as psychosocial stressor,
82
N
MC ALPINE, D., 34, 239
mc keown, t., 11, 33, 39n, 101, 102, NAMTCH, M. P., 153, 240
221, 222, 223, 224-225, 226, 239 National Cancer Study, 23
mc kinlay, j. b., 210, 214, 240 National Center for Health Statistics,
Macrosociocultural factors, as gen- 17, 22, 23-24, 27, 28, 29n, 30, 32,
eralized resistance resources, 117— 34, 240-241
119 National Health Education Commit-
Magic, and tension management, 118 tee, 19-22, 32, 33, 34, 241
malinowski, b., 117-118, 126, 240 Neoplasms, malignant, data on, 22-
malraux, a., 80 23
MANN, T., 124 Networks, social, 81, 164-165
maoz, b., 6, 85, 115, 234, 240 NEUGARTEN, 85
B.,
Marital status, and health linked to NISBET, R., 88, 241

sense of coherence, 165 Nonidentity, and tension manage-
Market modality, of doctor-patient ment, 108-109
relationship, 209-210, 215 Nonpatient, concept of, 40-41
martin, w. t., 166, 168, 234 Normality, concepts of, 49
masuda, m., 75, 93-94, 176-177, 237 Notifiable diseases, prevalence of,
mathews, k. a., 174, 240 26-27
mechanic, d., 94, 95, 101, 113, 240 NUCKOLLS, K. B., 166, 167, 241
meoawar, p., 173, 240
Medical education, and modalities of
relationships, 214-215
Medicine, technological levels of,
221-222. See also Health care OAKLEY, A, 149n, 241
system OBEYESEKERE, G., 212, 241
Menopause: as psychosocial stressor, OETTGEN, H. F., 104, 241
84-85; and social supports, 1 15— ORWELL, C, 73
116 ostfeld, a. m., 166, 169, 237, 241-
Mental illness, data on, 25-26 242
Mental retardation, data on, 19, 25 Overestimation, by vested interests,
mercer, j. r., 49, 240 and morbidity data, 33-34
merton, r. k., 88-89, 119, 203, 240 Overlap, of morbidity data, 32
MEYER, A., 176 Overload. See Load balance
Mobility, as stressor, 94—95 OVESEY, l., 151, 238
Modalities, impact of, on sense of
coherence, 207-214
MONJAN, A. A., 96, 240
MONTY, R- A., 152-153, 242
moos, r. h., 168, 238 Pain: as breakdown facet, 57-59;
Morbidity: data on, 16-35; hypoth- concept of, 58
esis of, 15-16; methodological is- parkes, c. m., 166, 189, 202, 242
sues in, 32-35; overall, 30-32 parsons, t., 50*, 203, 208, 211, 212,
morris, j. n., 4, 15, 240 242
Index 253
Participation, sense of coherence re- ROAZEN, p., 96, 243

lated to, 155-156, 187, 189, 200, ROESKE, N. A., Ill, 243
205, 208, 212, 216 Role-set complex, sense of coherence
Pathogenesis: drawbacks of, 36-37; related to, 146, 147, 191
salutogenesis essential to, 45; ubiq- rose, r. m., 168, 233
uity of, 12-14 rosenberg, b., 156, 231
Patient: concept of, 40-41; psycho- rosenman, r., 173-174
logical needs of, 43-44; in rela- rosenstock, m., 100-101, 243
i.
tionship with health care system, rotter, j. b., 152-153, 155, 158, 171,
199-214 243
PATRICK, R., 171, 233 ROUECHE, B., 26, 243
PAYKEL, E. s., 94-95, 245 Routine encounters, impact of, on

PEARLIN, L. I., 172, 194, 242 sense of coherence, 202-207
PERLMUTER, L. c, 152-153, 242 ruesch, j., 171, 243
Personality structure, basic, concept RUTSTEIN, D. D., 51, 243
of, 149
petrie, a., 113, 242
Phase-specific crises, as psychosocial
stressors, 84
PHILLIPS, d. p., 114, 242 Sacred modality, of doctor-patient
Physical factors, as generalized re- relationship, 209, 212-213, 214,
sistance resources, 103-106 215
Plasticity, and tension management, SALBER, E. J., 42, 245
104 Salutogenesis: concept of, 12-13; as
Potentiation, tension related to, 96 essential to pathogenesis, 45; evo-
powLES, j., 39n, 242 lution of perspective on, 1-11; im-
Predictability, sense coherence
of plications of, for health care
related to, 126-127, 152, 187 system, 198-219; model of, 182-
Prognostic implication: as breakdown 197; significance of, 35-37, 226-
facet, 61-62; concept of, 62 227
Psychological variables: and health SCHACHTEL, E. G., 108-109, 243
linked to sense of coherence, 171- SCHEFF, T. J., 25, 243
175; impact of, 201; for sense of Schizophrenia, social-structural
coherence, 137-142 sources of, 142, 161-162
schmale, a. h., 137-138, 139, 162,
166, 235, 243, 244
schooler, c., 144, 194, 238, 242
radloff, r., 95, 242 schrodinger, e., 78, 120, 243
rahe, r. h., 75, 81, 86, 110, 143, schur, m., 81, 132, 243
176, 177, 178, 236, 237, 242-243 schwartz, h., 193n, 232
rapoport, a., 122, 243 scott, r. a., 59, 72, 112, 243-244
Rationality, as coping strategy, 112— SEGALL, A., 3
113, 170 Self-direction, sense of coherence
Religion, and tension management, related to, 144, 145, 146-147
118 SELIGMAN, M. E. P., 107, 137, 139-
Respiratory conditions, chronic, data 140, 141, 153, 154, 165-166, 172,
on, 24 175, 179, 244
Risk factor confusion, and morbidity selye, h., 5, 74, 93, 94, 99, 105,
data, 32 129, 180, 194, 196, 244
254 Index
Sense of coherence. See Coherence, avoidance of, 100-102; character-

sense of istics of, 71-73; defined, 72; on
shaijt, b., 110, 148, 244 group 91-92; as objective or
level,
SHAW, C. B., 157 subjective, perception of,
73-76;
SHEEHY, o., 88, 244 72-73; psychosocial, 79-51; in
SHOHAM, L, 8, 162, 231 salutogenic model, 192-193; ten-
shuval, j. t., 2, 43, 95, 244 sion and stress related to, 70-97;
SIMMEL, c, 88, 244 ubiquity of, 70, 76-91
smoN, w., 88-89, 244 Structural variables, and sense of
Situational variables, and health coherence, 142-148, 163-168
linked to sense of coherence, 175— Substitutability, and tension man-
179 agement, 119
Social environment, and health susser, m., 25-26, 47, 48-50, 59,
linked to sense of coherence, 168 244
Social networks: experiences in, as Sweden, job complexity and control
psychosocial stressors, 81; and in, 148
health linked to sense of coherence, SWEENEY, D. R., 137-138, 244

164-165 syme, s. l., 66, 163-164, 167, 245
Social relations, conflictual, as psy- szasz, t., 25, 205, 245
chosocial stressor, 87-88
Social-role complexes, in salutogenic
model, 191
Social-structural variablesand health :
linked to sense of coherence, 163— TAEUBER, c, 68, 245

168; for sense of coherence, 142— Tanala culture, and sense of co-
148 herence, 150-151
Social supports: as generalized resis- Tension: character of, 92-97; de-
tance resources, 114-117; and fined, 94; stress distinct from, 3,
health linked to sense of coherence, 10, 96,180
166-168 Tension management and avoidance
:
Socialization, inadequate, as psycho- of stressors, 100-102; concept of,

social stressor, 86, and generalized
3, 96; resistance
SOLOMON, g. f., 181, 244 resources,98-122; in salutogenic
SOROCHAN, W. D., 68, 244 model, 193-195
Specific resistance resources (SRRs), thacker, s. b., 42, 245
in salutogenic model, 193, 194 thomas, l., 11, 121, 221-222, 223,
spitz, r., 140, 179
224, 225, 226, 245
srole, l., 25, 244 THOMAS, W. L, 1 13
starr, p., 218, 244 TIGER, 82
L.,
Status integration, and health linked ttnuno, d. c, 137-138, 244
to sense of coherence, 166 TOFFLER, A., 83, 245
sterling, p., 39n, 235 TORFs, c. p., 66, 245
STERN BACH, R. A., 57 Traumatic situations, impact of, on
Stress: defined, 93; in salutogenic sense of coherence, 200-202
model, 195-196; tension distinct TRILLTNO, L., 132-134
from, 3, 10, 96, 180 Type A
behavior partem, as psycho-
Stressors: as absence of generalized logical variable, 174-175
resistance resources, 119-120; TYROLER, H. a., 83, 171, 233, 245
Index 255
U WHITE, K. L., 16-17, 246

WILLIAMS, 246
T. F., 16-17,
Underestimation, and morbidity data, WINKELSTEIN, w., 15-16, 34, 246
33 Withdrawal, sense of coherence re-
Underload. See Load balance lated to, 138-139, 166
Urban environment, and health WITHEY, S. B., 68, 229
linked to sense of coherence, 169 WOHLWILL, J. F., 87, 246
U.S. Department of Health, Educa- wolf, s., 75-76, 105, 169, 170, 246
tion and Welfare, 34
wolff, h., 75, 79, 105, 169, 170, 237
Work, substantive complexity of, and
sense of coherence, 144—145, 147,
148
Valuative-attitudinal factors, as gen-
World Health Organization (WHO),
eralized resistance resources, 110—
deficiencies of health definition by,
114 52-55
VICKERS, G., 97, 245
Visual impairment, data on, 20-21
Voodoo death, and sense of coher-
ence, 165-166
young, a., 212, 213, 246
W young, j. l., 23, 234
WEISENBERG, M., 57, 245

WEISSMAN, M. M., 94-95, 245
Well-being, health related to, 67-69
WERTHEIM, e. s., 107, 137, 140-141, zetzel, e. r., 96
157, 245-246 zola, l k., 3, 17, 77, 238-239, 246

Antonovsky (1979)

Uploaded by

Copyright:

Available Formats

Antonovsky (1979)

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

Antonovsky (1979)

Uploaded by

Copyright:

Available Formats

:

Copyright © 1979 by: Jossey-Bass Inc., Publishers

Copyright under International, Pan American, and

Library of Congress Catalogue Card Number LC 79-83566

International Standard Book Number ISBN 0-87589-412-7

Manufactured in the United States of America

JACKET DESIGN BY WILLI BAUM

This book poses the question of salutogenesis, of the origins (genesis)

taken up by colleagues in other life sciences, so that some day more

no attempt to simplify a highly complex problem or to avoid techni-

and personally, are committed to understanding and enhancing the

surelyborrowed from some studies without adequate acknowledg-

a most provocative research assistant and colleague from 1968 to

tant. Together with her I moved from the trees of generalized

School of Public Health, University of California, Berkeley. My

pleasurable to have spent a year down the hall from him.

Allen, Margaret Boyd, Lynda Brodsky, Shoshanna Churgin, Ber-

Over the years, I have been the recipient of a number of

Beersheba, Israel Aaron Antonovsky

The Author xiii

Prologue : Evolution of a New Perspective 1

One Studying Health Instead of Disease 12

Two Measuring Health on a Continuum 38

Three Stressors, Tension, and Stress 70

Four Tension Management and Resources for

Five Perceiving the World as Coherent 123

Six Relation of the Sense of Coherence to Health 160

Seven The Salutogenic Model of Health 182

Eight Implications for an Improved Health Care

Epilogue: Outlook for Human Health 220

Aaron Antonovsky is Kunin-Luncnfeld Professor of Medical

With his wife Helen, a developmental psychologist, he emi-

of Social Medicine, and he started a research project with Judith

New Perspectives on Mental

It is said that Oriental scholars find it appropriate to introduce their

this approach appealing. I would like to think that, by pointing out

My doctoral dissertation (Antonovsky, 1955, 1956) was not

asked how members of (what at that time was called) a minority

group defined their marginal social situation. In fact, I was investi-

gating cognitive coping responses to a socially structured psycho-

Jewish labor movement in the United States (Antonovsky, 1961),

nineteenth- and twentieth-century stressor, immigration. Then, as re-

I began teaching in a school of public health. Both these projects

were preliminaries to my direct entry into the field of stress research.

industrialized society —rare birds, indeed —have lower rates of heart

There was, however, a complication. Having abandoned

rate of a particular disease) . But, second, there is the problem of why

At this point, then, the dependent variable that concerned

in the bind of categorizing people as healthy or was clear to

scale study (some 1,150 women y was based on representative sam-

sons for predicting a direct correlation between traditionalism and

come clear, particularly in Chapter Four, this insight became a cen-

selected sample study group compared with a control group. And

saluto genesis. A statistically significant difference between groups

human by Displaced Persons camps, illegal

to the 1950s, when I taught courses on "The Negroes in the United

concentration camp survivors, the poor, Afro-American slaves, or

was dealing with concentration camp survivors or poor people, I

The central concerns of this book are, I trust, clear by now.