http://www.coedu.usf.edu/bostow/rtorres/Skinner/BAChapters.

htm
About Behaviorism

Introduction -About Behaviorism (pg. 3)
Chapter 1 -The Causes of Behavior (pg. 10)
Chapter -The !or"d !ithin the #$in (pg. %)
Chapter 3 -Innate Behavior (pg. 3&)
Chapter % -'perant Behavior (pg. (1)
Chapter ( -)erceiving (pg. *0)
Chapter + -,erba" Behavior (pg. -*)
Chapter & -Thin$ing (pg. 113)
Chapter * -Causes and .easons (pg. 13)
Chapter - -/no0ing (pg. 1(1)
Chapter 10 -The Inner !or"d of 1otivation and 2motion (pg. 1+3)
Chapter 11 -The #e"f and 'thers (pg. 1*%)
Chapter 1 The 3uestion of Contro" (pg. 0*)
Chapter 13 -!hat Is Inside the #$in4 (pg. *)
Chapter 1% -#tumming 5p (pg. %1)
Introduction
Go Back

Behaviorism is not the science of human behavior; it is the philosophy of that
science. Some of the questions it asks are these: Is such a science really possible? Can
it account for every aspect of human behavior? What methos can it use? !re its la"s
as vali as those of physics an biolo#y? Will it lea to a technolo#y$ an if so$ "hat
role "ill it play in human affairs? %articularly important is its bearin# on earlier treat&
ments of the same sub'ect. (uman behavior is the most familiar feature of the "orl
in "hich people live$ an more must have been sai about it than about any other
thin#; ho" much of "hat has been sai is "orth savin#? Some of these questions "ill
eventually be ans"ere by the success or failure of scientific an technolo#ical
enterprises$ but current issues are raise$ an provisional ans"ers are neee no". !
#reat many intelli#ent people believe that ans"ers have alreay been foun an that
they are all unpromisin#. (ere$ for e)ample$ are some of the thin#s commonly sai
about behaviorism or the science of behavior. *hey are a$ I believe$ "ron#.

1. 1. It i#nores consciousness$ feelin#s$ an states of min.
2. +. It ne#lects innate eno"ment an ar#ues that all behavior is acquire
urin# the lifetime of the iniviual.
3. 3. It formulates behavior simply as a set of responses to stimuli$ thus
representin# a person as an automaton$ robot$ puppet$ or machine.
4. %. It oes not attempt to account for co#nitive processes.
5. (. It has no place for intention or purpose.
. +. It cannot e)plain creative achievements&in art$ for e)ample$ or in
music$ literature$ science$ or mathematics.
!. &. It assi#ns no role to a self or sense of self.
". *. It is necessarily superficial an cannot eal "ith the epths of the min
or personality.
#. -. It limits itself to the preiction an control of behavior an misses the
essential nature of bein# a man.
1$. 10. It "orks "ith animals$ particularly "ith "hite rats$ but not "ith people$
an its picture of human behavior is therefore confine to those features
"hich human bein#s share "ith animals.
11. 11. Its achievements uner laboratory control cannot be uplicate in aily
life$ an "hat it has to say about human behavior in the "orl at lar#e is
therefore unsupporte metascience.
12. 1. It is oversimplifie an na,ve an its facts are either trivial or alreay
"ell kno"n.
13. 13. It is scientistic rather than scientific. It merely emulates the sciences.
14. 1%. Its technolo#ical achievements coul have come about throu#h the use
of common sense.
15. 1(. If its contentions are vali$ they must apply to the behavioral scientist
himself$ an "hat he says is therefore only "hat he has been conitione to
say an cannot be true.
1. 1+. It ehumani-es man; it is reuctionistic an estroys man qua man.
1!. 1&. It is concerne only "ith #eneral principles an therefore ne#lects the
uniqueness of the iniviual.
1". 1*. It is necessarily antiemocratic because the relation bet"een
e)perimenter an sub'ect is manipulative$ an its results can therefore be
use by ictators but not by men of #oo "ill.
1#. 1-. It re#ars abstract ieas such as morality or 'ustice as fictions.
2$. 0. It is inifferent to the "armth an richness of human life$ an it is
incompatible "ith the creation an en'oyment of ail$ music$ an literature
an "ith love for one.s fello" men.

*hese contentions represent$ I believe$ an e)traorinary misunerstanin# of
the achievements an si#nificance of a scientific enterprise. (o" can it be e)plaine?
*he early history of the movement may have cause trouble. *he first e)plicit
behaviorist "as /ohn B. Watson$ "ho in 0102 issue a kin of manifesto calle
%sycholo#y as the Behaviorist Views It. !s the title sho"s$ he "as not proposin# a ne"
science but ar#uin# that psycholo#y shoul be reefine as the stuy of behavior. *his
may have been a strate#ic mistake. 3ost of the psycholo#ists at the time believe
they "ere stuyin# mental processes in a mental "orl of consciousness$ an they
"ere naturally not incline to a#ree "ith Watson. 4arly behaviorists "aste a #oo
eal of time$ an confuse an important central issue$ by attackin# the introspective
stuy of mental life.

Watson himself ha mae important observations of instinctive behavior an
"as$ inee$ one of the first etholo#ists in the moern spirit$ but he "as #reatly im&
presse by ne" evience of "hat an or#anism coul learn to o$ an he mae some
rather e)treme claims about the potential of a ne"born human infant. (e himself
calle them e)a##erations$ but they have been use to iscreit him ever since. (is
ne" science "as also$ so to speak$ born prematurely. 5ery fe" scientific facts about
behavior&particularly human behavior "ere available. ! shorta#e of facts is al"ays a
problem in a ne" science$ but in Watson.s a##ressive pro#ram in a fiel as vast as
human behavior it "as especially ama#in#. (e neee more factual support than he
coul fin$ an it is not surprisin# that much of "hat he sai seeme oversimplifie
an naive.

!mon# the behavioral facts at han "ere refle)es an conitione refle)es$
an Watson mae the most of them$ but the refle) su##este a push&pull type of
causality not incompatible "ith the nineteenth&century conception of a machine. *he
same impression "as #iven by the "ork of the 6ussian physiolo#ist %avlov$ publishe
at about the same time$ an it "as not correcte by the stimulus&response psycholo#y
"hich emer#e urin# the ne)t three or four ecaes.

Watson naturally emphasi-e the most reproucible results he coul fin$ an
most of them ha been obtaine from animals&the "hite rats of animal psycholo#y
an %avlov.s o#s. It seeme to be implie that human behavior ha no istin#uishin#
characteristics. !n to bolster this claim that psycholo#y "as a science$ an to fill out
his te)tbook$ he borro"e from anatomy an physiolo#y$ an %avlov took the same
line by sistin# that his e)periments on behavior "ere really 7an investi#ation of the
physiolo#ical activity of the cerebral corte)$8 althou#h neither man coul point to any
irect observations of the nervous system "hich thre" li#ht on behavior. *hey "ere
also force into hasty interpretations of comple) behavior$ Watson ar#uin# that
thinkin# "as merely subvocal speech an %avlov that lan#ua#e "as simply a 8secon
si#nal system.8 Watson ha little or nothin# to say about intention or purpose or
creativity. (e emphasi-e the technolo#ical promise of a science of behavior$ but his
e)amples "ere not incompatible "ith a manipulative control.

3ore than si)ty years have passe since Watson issue his manifesto$ an a
#reat eal has happene in that thne. *he scientific analysis of behavior has mae
ramatic pro#ress$ an the shortcomin#s in Watson9s$ account are no"$ I believe$
chiefly of historical interest. :evertheless$ criticism has not #reatly chan#e. !ll the
misunerstanin#s liste above are to be foun in current publications by
philosophers$ theolo#ians$ social scientists$ historians$ men an "omen of letters$
psycholo#ists$ an many others. *he va#aries of the early history of the movement
can harly suffice as an e)planation.

Some trouble no oubt arises from the fact that human behavior is a sensitive
fiel. 3uch is at stake in the "ay in "hich "e look at ourselves$ an a behavioristic
formulation certainly calls for some isturbin# chan#es. 3oreover$ terms ori#inatin# in
earlier formulations are eeply imbee in our lan#ua#e$ an they have ha a place
in both technical an nontechnical literature for centuries. :evertheless$ it "oul be
unfair to ar#ue that the critic has not been able to free himself from these historical
pre'uices. *here must be some other reason "hy behaviorism as the philosophy of a
science of behavior is still so seriously misunerstoo.

I believe the e)planation is this: the science itself is misunerstoo. *here are
many ifferent kins of behavioral science$ an some of them$ as I shall sho" later$
formulate the fiel in "ays "hich o not raise important behavioristic issues. *he
criticisms liste above are most &effectively ans"ere by a special iscipline$ "hich
has come to be calle the e)perimental analysis of behavior. *he behavior of
iniviual or#anisms is stuie in carefully controlle environments$ an the relation
bet"een behavior an environment then formulate$ ;nfortunately$ very little is
kno"n about this analysis outsie the fiel. Its most active investi#ators$ an there
are hunres of them$ selom make any effort to e)plain themselves to
nonspecialists. !s a result$ fe" people are familiar "ith$ the scientific unerpinnin#s
of "hat$ I believe$ is the: most co#ent statement of the behavioristic position.

*he behaviorism I present in this book is the philosophy of this special version
of a science of behavior. *he reaer shoul kno" that not all behaviorists "ill a#ree
"ith everythin# I say. Watson spoke for 8the behaviorist$8 an in his time he "as the
behaviorist$ but no one can assume that mantle toay. What follo"s is
amittely&an$ as a behaviorist$ I must say necessarily&a personal vie". I believe$
ho"ever$ that it$ is a consistent an coherent account "hich satisfactorily ans"ers
the criticisms liste above.

I also believe in its importance. *he ma'or problems facin# the "orl toay can
lie solve only if "e improve our unerstanin# of human behavior. *raitional vie"s
have been aroun for centuries$ an I think it is fair to say that they have prove to
be inaequate. *hey are lar#ely responsible for the situation in "hich "e no" fin
ourselves. Behaviorism offers a promisin# alternative$ an I have "ritten this book in
an effort to make its position clear.
%Chapter 1 eksik&
2 The World Within the Skin
Go Back

! small part of the universe is containe "ithin the skin of each of us. *here is
no reason "hy it shoul have any special physical status because it lies "ithin this
bounary$ an eventually "e shoul have a complete account of it from anatomy an
physiolo#y. :o very #oo account is no" available$ ho"ever$ an it therefore seerris
all the more important that "e shoul be in touch "ith it in other "ays. We feel it
an in some sense observe it$ an it "oul seem foolish to ne#lect this source of
information 'ust because no more than one person can make contact "ith one inner
"orl. :evertheless$ our behavior in makin# that contact nees to be e)amine.

We respon to our o"n boy "ith three nervous systems$ t"o of "hich are
particularly concerne "ith internal features. *he so&calle interoceptive system
carries stimulation from or#ans like the blaer an

The World Within the Skin 25

alimentary tract$ from #lans an their ucts$ an from bloo vessels. It is primarily
important for the internal economy of the or#anism. *he so calle proprioceptive
system carries stimulation from the muscles$ 'oints$ an tenons of the skeletal frame
an from other or#ans involve in the maintenance of posture an the e)ecution of
movement. We use the verb 8feel8 in escribin# our contact "ith these t"o kins of
stimulation. ! thir nervous system$ the e)teroceptive$ is primarily concerne "ith
seein#$ hearin#$ tastin#$ smellin#$ an feelin# thin#s in the "orl aroun us$ but it
also plays an important part in observin# our o"n boy.

Observing and Describing
The World Within the Skin

!ll three nervous systems presumably evolve to their present conition
because they serve important biolo#ical functions$ but they came to serve another
function "ith the appearance of verbal behavior. %eople eventually aske questions
of each other$ the ans"ers to "hich calle for a ifferent kin of responin# to the
boy. <uestions such as 8!re you hun#ry?8$ 8=oes your hea ache?8.$ 8What are you
oin#?8$ 8What o you plan to o tomorro"?8$ 8What i you o yesteray?8$ an 8Why
are$ you oin# that?8 evoke ans"ers "hich are useful >In preictin# an preparin# for
"hat a person "ill o$ an they seem to #ive information about a "orl beyon the
reach of other people.

We mi#ht e)pect that because a person is in such intimate contact "ith his
o"n boy he shoul be able to escribe its conitions an processes particularly "ell$
but the very privacy "hich seems to confer a special privile#e on the iniviual makes
it ifficult for the community to teach him to make istinctions. *he

ABOUT BEHAVIORISM2

community can teach a chil to name colors in various "ays. ?or e!a"#le$ it can sho"
him colore ob'ects$ ask him to respon "ith color "ors$ an commen or correct
him "hen his responses correspon or fail to correspon "ith the colors of the
ob'ects. If the chil has normal color vision$ "e e)pect him to learn to ientify colors
accurately. *he community cannot$ ho"ever$ follo" the Sam@$ practice in teachin#
him to escribe the states of his o"n boy because it lacks the information it nees to
commen or correct him.

Reporting Things Felt

?ortunately$ it oes not follo" that no one can learn to escribe some of the states of
his o"n boy$ because the verbal community can to some e)tent solve the problem of
privacy. ?or e)ample$ it can teach responses escriptive of internal conitions by
usin# associate public conitions. Somethin# of the sort happens "hen a blin person
is tau#ht to name the ob'ects he feels by a teacher "ho merely sees the ob'ects. *he
teacher can commen or correct him because the visual an tactual stimuli are
almost perfectly correlate. *he verbal community follo"s a rather similar practice
"hen it teaches a chil such an e)pression as 8*hat hurts.8 When the chil has
receive a sharp blo" or cut$ the public blo" or cut is fairly reliably associate "ith
the private stimuli #enerate by it. *he verbal community uses the public
information$ but the chil may eventually say 8*hat hurts8 "hile responin# only to
the private event. (e has learne to escribe a private stimulus "ith an accuracy
"hich epcns only upon ho" "ell the public an private events a#ree.

*he practice e)plains "hy terms "hich escribe pains

The World Within the Skin 2%

almost al"ays escribe their public causes. 8%ain8 itself comes from the Greek an
Aatin for punishment. ! sharp pain is the$ pain prouce by a sharp ob'ect; a ull pain
by a ull ob'ect. %ains can be "renchin# or piercin#; a heaache may poun; an
8e)cruciatin#. is relate to crucifi)ion. We often ask about feelin#s by askin#$ 8What
oes it feel like&' an the ans"er usually refers to a public conition "hich often
prouces a similar private effect. *hus$ a person "ho has ha a stroke of luck may
say$ 8I feel as if I. "on a million ollars.8 ! stanar literary practice is to escribe
feelin#s by escribin# conitions "hich are likely to arouse similar feelin#s. Beats
reporte "hat it felt like "hen he first looke into Chapman.s translation of (omer$ in
the follo"in# "ay:

*hen felt I like some "atcher of the skies
When a =e" planet s"ims into his ken(
Cr like stout Corte-$ "hen "ith ea#le eyes
(e star. at the %acific

*he verbal community may also circumvent the restrictions impose by privacy
by usin# collateral responses to the stimuli "hich a person is to learn to ientify or
escribe. ?or e)ample$ it may observe not only that a chil receives a painful blo"$
but that he cries. *he private stimuli "hich come to control the response 8*hai. hurts8
are then less likely to be escribe "ith terms first escriptive of public stimuli.
Similarly$ althou#h the community may teach a chil to say$ 8I am hun#ry$8 because it
kno"s that the chil has not eaten for a lon# time$ it is much more likely to take
avanta#e of collateral behavior: it observes that the chil respons quickly or eats
ravenously "hen #iven foo. It then tDls him that he is hun#ry$ an the chil

ABOUT BEHAVIORISM 2)

may acquire the e)pression 8I am hun#ry8 "ith respect to collateral private stimuli to
"hich the verbal community ba no access.

*erms referrin# to emotional or motivational states often sho" some
connection "ith the e)ternal circumstances responsible for them. ?or e)ample$ "e
feel sa in the ori#inal sense of sated$ or e!*ited in the sense of stirred +#$ but these
e)pressions may be little$ more than metaphors. We are not tense in the literal sense
of bein# stretche$ or epresse in the literal sense of "ei#he o"n. We ma> have
acquire these "ors uner circumstances "hich have no connection "ith behavior or
feelin#s. !lmost all terms escriptive of emotions "hich o not carry a. irect
reference to incitin# conitions "ere ori#inally metaphors.

!lthou#h the verbal community solves the problem of privacy in this "ay an
succees in teachin# a person to escribe many states of his boy$ the escriptions
are never completely accurate. *he physician allo"s for consierable latitue "hen
his patient escribes his aches an pains. *he ifficulty is not that the patient is not
bein# stimulate in a perfectly clear "ay$ it is simply that he has never been e)pose
to instructional conitions uner "hich he has learne to escribe the stimuli
aequately. 3oreover&an this is a point of the #reatest importance$ to "hich I shall
return later&the ori#inal biolo#ical functions responsible for the evolution of the
nervous system have not prouce the system the verbal community nees. !s a
result$ "e are particularly likely to istrust reports of private stimulation$ especially
"hen a escription has other consequences&as$ for e)ample$ in malin#erin#.

The World Within the Skin 2,

Reporting Behavior

-+rrent Behavior. *he question 8What are you oin#?8 asks for information "hich
may be quite public but "hich is at the moment out of reach of the questioner$ "ho
may be speakin# over the telephone$ for e)ample$ or in the ark$ or aroun a comer.
*he vocabulary in "hich the ans"er is #iven can be acquire "hen the behavior is
visible to all parties$ an the verbal community therefore suffers no limitation.
=escriptions may be confine to topo#raphy @8I am "avin# my han8.E or may inclue
effects on the environment @8I am$ rinkin# a #lass of "ater8 or 8I am se"in# a button
on my shirt8E. %roprioceptive stimuli are ominant "hen a person escribes his o"n
behavior in the ark$ but they are closely relate to the public stimuli use in
instruction. by the verbal community. <uestions of this sort are aske because the
ans"ers are important to the community$ but$ as "e shall see later$ they also become
important to the speaker himself an in "ays "hich are likely to maintain their ac&
curacy.

.ro/a/le Behavior. 8What are you incline to o?8 is a metaphorical question$ to
"hich a metaphorical ans"er mi#ht be 8I lean to"ar #oin#.8 *o tend to o somethin#
is also a metaphor$ su##estin# bein# pulle or stretche. !ns"ers presurnably epen
upon stimulation #enerate by conitions associate "ith a marke probability of
action. When somethin# funny happens on a solemn occasion$ "e may report$ 8I felt
like lau#hin#8 or$ 8I "ante to lau#h8 or$ 8I coul scarcely keep from lau#hin#.8 *he
stimulation thus escribe presumably accompanie earlier instances "hen

!BC;* B4(!5IC6IS32D

lau#hter occurre an a suitable vocabulary "as acquire.

.er*e#t+al Behavior. A person may be aske$ 8=o you see that?8 or$ less
iiomatically$ 8!re you seein# that?8 an the ans"er may be checke by askin# for the
name or a escription of "hat is seen.

.ast Behavior. !ns"ers to such questions as 8What i you o yesteray?8 or
8Whom i you see?8 can use a vocabulary acquire in connection "ith current
behavior. ! person simply speaks from a special vanta#e point: lie "as necessarily
there. Such questions are scarcely ifferent from$ say$ 8What happene yesteray?8
@Whether it is easier to escribe yesteray.s behavior if one also escribe it
yesteray is a matter of some importance. It has been su##este$ for e)ample$ that
"e. o not remember "hat happene in infancy because ."e "ere not able to
escribe it at that time F8infant8 once meant 8incapable of speechGH$ but "e o not
constantly escribe the behavior "e are en#a#in# in althou#h "e can usually escribe
it later. :evertheless$ the quick for#ettin# of reams an of passin# thou#hts "hich
have not been clearly 8note8 su##ests that a current runnin# account is the best "ay
to make sure that behavior can be escribe at a later ate.E

-overt Behavior. A much more ifficult question is 8What are you thinkin#?8
"here 8thinkin#, refers to behavior e)ecute on such a small scale that it is not
visible to others. @Cther uses of the "or 8think8 are iscusse in Chapter ,.E In
escribin# covert behavior "e may be escribin# public behavior in miniature$ but it
is more likely that "e are escribin# pri

The World Within the Skin 01

vate conitions associate "ith public behavior but not necessarily #enerate by it.
5erbal behavior can easily become covert because it oes not require environmental
support. 8I sai to myself . . .8 is use synonymously "ith 8I thou#ht . . but "e o not
say$ 8I s"am to myself.8

Covert perceptual behavior is especially pu--lin#. Ima#inin# or fantasyin#$ as
"ays of 7seein#G somethin# in the absence of the thin# seen$ are presumably a matter
of oin# "hat one oes "hen "hat is seen is present. I shall return to this point in
Chapter I.

*he verbal community may resort to instrumental amplification$ as of the
activity of muscles$ an thus in a sense make covert behavior public$ an encoura#e a
return to the overt level as by askin# a person to 8think out lou$8 but it cannot
maintain the accuracy of covert behavior. *here is no problem$ ho"ever$ in the
provenance of the vocabulary. *he "ors use to escribe covert behavior are the
"ors acquire "hen behavin# publicly.

2+t+re Behavior. !nother ifficult question is 8What are you #oin# to o?8 *he
ans"er is$ of course$ not a escription of the future behavior itself. It may be a report
of stron# covert behavior likely to be emitte public$ "hen the occasion arises @8When
I see him$ I shall remin him that he o"es me ten ollars8E. It may be a preiction of
behavior base on current conitions "ith "hich the behavior is often associate
@8When thin#s are like this$ I #enerally #ive up8 or 8I.m hun#ry an I am #oin# to #et
somethin# to eat8E. It may tic a report of a stron# probability of behavin# in a #iven
"ay.
Statements about future behavior often involve the "or 8feel.8 %erhaps 8I feel like
playin# cars8 may be

ABOUT BEHAVIORISM02

translate as 8I feel as I often feel "hen I have starte to play cars.8 8.What o 3o+
want to o?8 may refer to the future in the sense of askin# about the probability of
behavior.

!n attitue @8.=o you really "ant to o "hat you are oin#?8 or 8=o you really
"ant to #o to the beach for your vacation?8.E may be part of the metaphor of in&
clination or tenency.

In #eneral the. verbal community can check the accuracy of statements
re#arin# inclinations an tenencies$ at least in a statistical "ay$ by lookin# at "hat
happens$ an the accuracy of the control maintaine by private stimuli is thus to
some e)tent assure. We shall see that self&escriptive behavior also serves the in&
iviual himself an that "hen it oes so$ it tens to remain accurate.

M+lti#le Translations. Conitions relevant to behavior are reporte accorin# to
the circumstances in "hich they have been acquire$ an this means that an
e)pression may be translate in several "ays. Consier the report 8I am$ "as$ or "ill
be hun#ry.8 8I am hun#ry8 may be equivalent to 8I have hun#er pan#s$8 an if the
verbal community ha some means of observin# the contractions of the stomach
associate "ith pan#s$ it coul pin the: response to these stimuli alone. It may also be
equivalent to 8I am eatin# actively.8 ! person "ho observes that he is eatin#
voraciously may say$ 8I really am hun#ry$8 or$ in retrospect$ 8I "as hun#rier than I
thou#ht$8 ismissin# other evience as unreliable. 8I am hun#ry8 may also be
equivalent to 8It has been a lon# time since I have ha anythin# to eat$8 althou#h the
e)pression is most likely to be use in escribin# future behavior$ 8If I miss my inner$
I shall be hun#ry.8 8I am hun#ry8 may also be equivalent to

The World Within the Skin 00

8I feel like eatin#. in the sense of 8I have felt this "ay before "hen I have starte to
eat.8 It may be equivalent to 8I am covertly en#a#in# in behavior similar to that
involve in #ettin# an consumin# foo8 or 8I am fantasyin# eatin#8 or 8I am thinkin# of
thin#s I like to eat8 or 8I am .eatin# to myself.. 8 *o say$ 8I am hun#ry$8 may be to
report several or all of these conitions.

Identi!ing the "auses o One#s Behavior

8What are you oin#?8 is frequently a request for further information. *he question
mi#ht be aske of someone "ho is rumma#in# a bo) of small ob'ects$ an a
characteristic response mi#ht be 8I am lookin# for my ol pocketknife$8 *he "or
8rumma#in#8 escribes a particular kin of behavior; in aition to a particular
topo#raphy$ it implies a reason. ! person "ho is rumma#in# is lookin# for somethin#$
an the rumma#in# "ill cease "hen it is foun. ! ifferent question$ 8What are you
lookin# for?8 narro"s the fiel$ an 83y ol pocketknife8 ientifies the ob'ect sou#ht$
the finin# of "hich "ill brin# the behavior to an en. ! further question$ 8Why are
you lookin# for your knife?8 mi#ht call out the ans"er 8Because I "ant it$8 "hich
usually means more than 8Because it is "antin#.8

! more irect question about causes is 8Why are you oin# that?8 an the
ans"er is usually a escription of feelin#s: 8Because I feel like oin# it.8 Such an
ans"er is often acceptable$ but if the verbal community insists upon somethin# else$
it may ask$ 8Why o you feel like oin# it?8 an the ans"er "ill then be either a
reference to other feelin#s or @at lon# lastE to e)ternal circumstances. *hus$ in reply
to 8Why are you movin#

ABOUT BEHAVIORISM 04

your chair?8$ a person may say$ 8*he li#ht "as ba8 or 8*o #et a better li#ht on my
book.8

<uestions of this kin are not al"ays correctly ans"ere$ since "e often o
not kno" "hy "e behave as "e o. In spite of the apparent intimacy of the "orl
"ithin the skin$ an in spite of the avanta#e a person en'oys as an observer of his
personal history$ another person may kno" more about "hy he behaves. *he
psychotherapist "ho attempts to #ive his patient insi#ht is presumably emphasi-in#
causal relationships of "hich his patient is not yet a"are.

When "e o not kno" "hy "e behave$ "e are likely to invent causes: 8I i it$
so I must have thou#ht it "oul help.8 It is possible that many myths are little more
than invente causes of the superstitious behavior$ seemin#ly uncause$ to be
iscusse in Chapter J.

4)planations of behavior vary "ith the kins of ans"ers accepte by the
verbal community. If a simple 8I feel like it8 suffices$ nothin# else "ill. appear. ?reu
"as influential in chan#in# the kins of ans"ers often #iven to 8Why$ are you oin#
that?8 (e emphasi-e feelin#s but allo"e for references to personal history. *he
e)perimental analysis of behavior #oes irectly to the$ anteceent causes in the
environment.

Sel$%no&ledge

I have been emphasi-in# a ifference bet"een feelin#s an reportin# "hat
one feels. We may take feelin# to be simply responin# to stimuli$ but reportin# is the
prouct of the special verbal contin#encies arran#e by a community. *here is a
similar ifference bet"een behavin# an reportin# that one is behavin# or reportin#
the causes of one.s behavior. In arran#in# conitions uner "hich a person escribes
the public or private

The World Within the Skin 05

"orl in "hich lie lives$ a community #enerates that very special form of behavior
calle kno"in#. 6esponin# to an empty stomach by #ettin# an in#estin# foo is one
thin#; kno"in# that one is hun#ry is another. Walkin# over rou#h terrain is one thin#;
kno"in# that one is oin# so is another.

Self&kno"le#e is of social ori#in. It is only "hen a person.s private "orl
becomes important to others that it is mae important to him. It then enters into the
control of the behavior calle kno"in#. But self kno"le#e has a special value to the
iniviual himself. ! person "ho has been 8mae a"are of himself8 by the questions
he has been aske is in a better position to preict an control his o"n behavior.

! behavioristic analysis oes not question the practical usefulness of reports of
the inner "orl that is felt an introspectively observe. *hey are clues @0E to past
behavior an the conitions affectin# it$ @+E to current behavior an the conitions
affectin# it$ an @2E to conitions relate to future behavior. :evertheless$ the
private "orl "ithin the skin is not clearly observe or kno"n. I have mentione t"o
reasons$ to "hich I shall have many occasions to return: in teachin# self&kno"le#e
@0E the verbal community must make o "ith rather primitive nervous systems$ an
@+E it cannot fully solve the problem of privacy. *here is an ol principle that nothin#
is ifferent until it makes a ifference$ an "ith respect to events in the "orl "ithin
the skin the verbal community has not been able to make thin#s ifferent enou#h. !s
a result$ there is room for speculation$ "hich over the centuries has sho"n the most
e)traorinary iversity.

%lato is sai to have iscovere the min$ but it "oul be more accurate to
say that he invente one version of it. Aon# before his time$ the Greeks ha

ABOUT BEHAVIORISM 0

constructe an elaborate e)planatory system$ a stran#e mi)ture of physiolo#y an
metaphysics. ! pure mentalism "as not lon# in makin# its appearance$ an it has
ominate Western thinkin# for more than t"o thousan years. !lmost all versions
conten that the min is a nonphysical space in "hich events obey nonphysical la"s.
*he 8consciousness8 of "hich a person is sai to be a"are has become such a staple of
Western thinkin# that 8everyone kno"s "hat it means to be conscious$8 an the
behaviorist "ho raises a question is calle isin#enuous$ as if he "ere refusin# to a&
mit the. evience of his senses.

4ven those "ho insist upon the reality of mental life "ill usually a#ree that
little or no pro#ress has been mae since %lato.s ay. 3entalistic theories are sub'ect
to chan#es in fashion an$ as in the history of clothin# or architecture$ one has only to
"ait lon# enou#h to fin. an earlier vie" back in style. We have ha !ristotelian
revivals an are no" sai to be returnin# to %lato. 3oem psycholo#y can claim to be
far beyon %lato in controllin# the environments of "hich people are sai to be
conscious$ but it has not #reatly improve their access to consciousness itself$
because it has not been able to improve the verbal contin#encies uner "hich feelin#s
an states of min are escribe an kno"n. Cne has only to look at any half&o-en
current mentalistic theories to see ho" much variety is still possible.

Behaviorism$ on the other ban$ has move for"ar. %rofitin# from recent
avances in the e)perimental analysis of behavior$ it has looke more closely at the
conitions uner "hich people respon to the "orl "ithin their skin$ an it can no"
analy-e$ one by one$ the key terms in the mentalistic armamentarium. What follo"s
is offere as an e)ample.
3 Innate Behavior
'o Back

(he human species) *ike a** other species) is the product of nattua* se*ection. +ach of its
members is an e,treme*- comp*e, or.anism) a *i/in. s-stem) the sub0ect of anatom- and
ph-sio*o.-. 1ie*ds such as respiration) di.estion) circu*ation) and immuni2ation ha/e been set
apart for specia* stud-) and amon. them is the fie*d we ca** beha/ior.

3t usua**- in/o*/es the en/ironment. (he newborn infant is so constructed that it takes in
air and food and puts out wastes. Breathin.) suck*in.) urination) and defecation are thin.s the
newborn infant does, but so, of course) are a** its other ph-sio*o.ica* acti/ities)

4hen we 5know enou.h about the anatom- and ph-sio*o.- of the newborn) we sha** be
ab*e to sa- wh- it breathes) suck*es) urinates) and defecates) but at the moment we must be
content with describin. the beha/ior itse*f and in/esti.atin. the conditions under which it
occurs6such as e,terna* or interna* stimu*ation) a.e) or *e/e* of depri/ation.
ABOUT BEHAVIORISM 38

Reflexes and Released Behaviors

7ne kind of re*ation between beha/ior and stimu*ation is ca**ed a ref*e,. As soon as the word
was coined) it was taken to refer to the under*-in. anatom- and ph-sio*o.-) but these are sti**
on*- rou.h*- known. At the moment a ref*e, has on*- a descripti/e force8 it is not an
e,p*anation. (o Sa- that a bab- breathes or suck*es because it possesses appropriate ref*e,es is
simp*- to sa- that it breathes or suck*es) presumab*- because it has e/o*/ed in such a wa- that it
does so. Breathin. and suck*in. in/o*/e responses to the en/ironment) but in no other wa- are
the- to be distin.uished from the rest of respiration and di.estion.

4hen ref*e,es first be.an to be studied in iso*ated parts of the or.anism) the resu*ts were
fe*t to cha**en.e the ro*e of inner determiners of conduct. Some ref*e,es) for e,amp*e) seemed to
disp*ace the Riichenmarkseee 6the sou*) or mind) of the spina* cord6the defense of which was an
ear*- attack on an en/ironmenta* ana*-sis.

Beha/ior usua**- in/o*/es the en/ironment in a more comp*e, wa-. 4e**6known e,amp*es
are found in *ower species. Courtin.) matin.) bui*din. nests) and carin. for -oun. are thin.s
or.anisms do) and a.ain presumab*- because of the wa- the- ha/e e/o*/ed. Beha/ior of this sort
is usua**- ca**ed instincti/e rather than ref*e,i/e) and the etho*o.ist speaks of the en/ironment as
9re*easin.9 beha/ior) a *ess compe**in. action than e*icitin. a ref*e, response. :e*eased) or
instincti/e) beha/ior is a*so more f*e,ib*e than ref*e,i/e in adaptin. to ad/entitious features of
the en/ironment. But to sa- that a bird bui*ds a nest because it possesses a nest6bui*din. instinct)
or because certain conditions re*ease nest bui*din.) is mere*- to describe the fact) not to e,p*ain
it.
Innate Beha!ior 3"

3nstincti/e beha/ior presents a more comp*e, assi.nment for the ph-sio*o.ist than ref*e,)
and at the moment we ha/e few re*e/ant facts and can on*- specu*ate about the kinds of s-stems
which ma- be in/o*/ed.

4hen we sa- that a .ood prose st-*ist has an 9instinct9 which permits him to 0ud.e without
ref*ection that a sentence is we** written) we mean nothin. more than that he possesses certain
deep*- in.rained beha/ior of uncertain pro/enance. 4e often mean *itt*e more in speakin. of
instincts in .enera*) and there is perhaps no harm in usin. the word in this wa-) but much more is
often read into the term. A ref*e, has been described b- sa-in. that 9stimu*i initiate a state of
tension that seeks dischar.e) brin.in. about re*a,ation.9 9+/er- instance of instincti/e beha/ior)9
said 4i**iam ;c<ou.a**) 9in/o*/es the knowin. of some thin. or ob0ect) a fee*in. in re.ard to it)
and a stri/in. towards or awa- from that ob0ect.9 1ee*in.s are ascribed to the beha/in. or.anism
when it is said that the moth *ikes the *i.ht it f*ies toward or bees the appearance and odor of the
f*owers the- fre=uent. The difficu*ties raised b- the ke- terms in sentences of that sort6tension)
dischar.e) re*a,ation) knowin.) fee*in.) stri/in.) and *ikin.6wi** be considered in *ater chapters.

3nstincts as <ri/in. #orces$ A more serious mistake is made in con/ertin. an instinct into
a force. 4e are not *ike*- to speak of a force in e,p*ainin. the fact than an or.anism di.ests its
food or de/e*ops immunit- to a disease) but the notion often appears in discussin. the or.anism5s
re*ation to its en/ironment. >erbert Spencer5s 9*ife force)9 Schopenhauer5s 9b*ind wi** to e,ist)9
and Ber.son?s %lan vital' were ear*- e,amp*es of the con/ersion of bio*o.ica* processes into
more ener.etic or substantia* forms. (he an !ita) for

ABOUT BEHAVIORISM &'

e,amp*e) was said to be 9a tire*ess power continua**- dri/in. onward and upward.9 (he 1reudian
instincts were a*so treated as dri/in. forces8 beha/ior which *ed to dan.er) i** hea*th) or death
was said to show a death instinct) whi*e beha/ior said to be 9in the ser/ice of *ife9 showed a *ife
instinct) a*thou.h the obser/ed fact was simp*- that beha/ior mi.ht ha/e sustainin. or de6
structi/e conse=uences.

(wo e,amp*es which ha/e recent*- attracted a .ood dea* of attention ma- be noted: %1&
4hen in0ured or threatened) an or.anism is *ike*- to attack6for e,amp*e) b- strikin. or
bitin.6and) as 3 sha** ar.ue in a moment) some 5beha/ior of this sort ma- be as much a part of the
.enetic endowment as respiration or di.estion) but we ha/e no reason to sa- that an or.anism
attacks because, it possesses an a..ressi/e instinct. (he attack is the on*- e/idence we ha/e of
the tendenc- to attack. %2& Some species defend the territories in which the- *i/e) and some of the
beha/ior seems to be due to a .enetic endowment) but to sa- that an or.anism defends its
territor- because o( a territoria* imperati/e or an- other kind. of instinct is simp*- to sa- that it is
the kind of or.anism which defends its territor-. %(he e,pression 9.enetic endowment9 is itse*f
dan.erous. @ike ref*e,es and instincts) it tends to ac=uire properties not warranted b- the
e/idence and to be.in to ser!e as a cause rather than as representin. the current effects of natura*
se*ection) from which attention is then def*ected.&
<arwin5s theor- of natura* se*ection came /er- *ate in the histor- of thou.ht. 4as it de*a-ed
because it opposed re/ea*ed truth) because it was an entire*- new sub0ect in the histor- of
science) because it was characteristic on*- of *i/in. thin.s) or because it dea*t with purpose and
fina* causes without postu*atin. an act of creationA 3 think not. <arwin simp*- disco/ered the
ro*e
Innate Beha!ior &)

of se*ection) a kind of causa*it- /er- different from the push6pu** mechanisms of science up to
that time. (he ori.in of a fantastic /ariet- of *i/in. thin.s cou*d be e,p*ained b- the contribution
which no/e* features) possib*- of random pro/enance) made to sur/i/a*. (here was *itt*e or
nothin. in ph-sica* or bio*o.ica* science) that foreshadowed se*ection as a causa* princip*e.

A*thou.h we stiff do not know much about the anatom- and ph-sio*o.- under*-in.
beha/ior) we can specu*ate about the process of se*ection which made them part of a .enetic
endowment. Sur/i/a* ma- be said to be contin*ent u+on certain kinds of beha/ior. 1or e,amp*e)
if members of a species did not mate) care for their -oun.) or defend themse*/es a.ainst
predators) the species wou*d not sur/i/e. 3t is not eas- to stud- these 9contin.encies of sur/i/a*9
e,perimenta**- because se*ection is a s*ow process) but some effects ma- be shown b- stud-in.
species which =uick*- mature to breedin. a.e and b- carefu**- arran.in. conditions of se*ection.

Contin.encies of sur/i/a* are often described with terms which suu..est a different kind
of causa* action. 9Se*ection pressure9 is an e,amp*e. Se*ection is a specia* kind of causa*it-
which is not proper*- represented as a force or pressure. (o sa- that there is 9no ob/ious
se*ection pressure on mamma*s that e,p*ains the hi.h *e/e* of inte**i.ence reached b- primates9
is simp*- to sa- that it is hard to ima.ine conditions under which s*i.ht*- more inte**i.ent
members of a species wou*d be more *ike*- to sur/i/e. %4hat is wron.) b- the wa-) is the
su..estion that 9pressure9 is5 e,erted primari*- b- other species. Sur/i/a* ma- depend a*most
who**- on 9competin. with9 the ph-sica* en/ironment) when inte**i.ent beha/ior is c*ear*-
fa/ored.&

ABOUT BEHAVIORISM &,

Contin.encies of sur/i/a* are more easi*- ima.ined if the beha/ior makes it more probab*e
that indi/idua*s wi** sur/i/e and breed and if the contin.encies pre/ai* o/er *on. periods of time.
Conditions within the bod- ha/e usua**- satisfied both these re=uirements) and some features of
the e,terna* en/ironment) such as the c-c*es of da- and ni.ht) or the seasons) or temperature) or
the .ra/itationa* fie*d) are *on.6*astin.. And so are other members of the same species) a fact
which e,p*ains the prominence .i/en b- etho*o.ists to courtship) se,) parenta* care) socia*
beha/ior) p*a-) imitation) and a..ression. But p*ausib*e conditions of se*ection are hard to find in
support of such an assertion as that 9princip*es of .rammar are present in the 3rtind at birth)9
since .rammatica* beha/ior can hard*- ha/e been sufficient*- important to sur/i/a*) for a *on.
enou.h time) to e,p*ain its se*ection. As 3 sha** note a.ain *ater) /erba* beha/ior cou*d arise on*-
when the necessar- in.redients bad a*read- e/o*/ed for other reasons.

Preparation for New Environments
3: :+SB7C<+C( C7C<3(37C3C'

Contin.encies of sur/i/a* cannot produce usefu* beha/ior if the en/ironment chan.es
substantia**- from .eneration to .eneration) but certain mechanisms ha/e e/o*/ed b- /irtue of
which the indi/idua* ac=uires beha/ior appropriate to a no/e* en/ironment durin. his *ifetime.
(he conditioned ref*e, is a re*ati/e*- simp*e e,amp*e. Certain cardiac ref*e,es support stron.
e,ertion) as in runnin. awa- from or6 stru..*in. with a predator8 and there is presumab*- an
ad/anta.e if the heart responds before runnin. or stru..*in. be.ins8 but predators /ar- in
appearance) and it is on*- throu.h respondent conditionin. that a particu*ar appearance can e*icit
Innate Beha!ior &3

appropriate cardiac beha/ior in ad/ance of runnin. or stru..*in..

A conditioned ref*e,) as a thin. a person possesses) has no more e,p*anator- force than
an unconditioned or innate ref*e,. (he heart of the runner does not be.in to beat stron.*- and
rapid*- 0ust before a race because of the conditioned cardiac ref*e,8 the ref*e, is simp*- a wa- of
identif-in. the fact that it be.ins to beat rapid*-. (he runner has been chan.ed when situations at
the start of a race ha/e been fo**owed b- stron. e,ertion) and as a chan.ed. or.anism he beha/es
in a different wa-. 3t is mere*- con/enient to identif- the chan.es as the 9ac=uisition of a
conditioned ref*e,.9

Dust as we point to contin.encies of sur/i/a* to e,p*ain an unconditioned ref*e,) so we
can point to 9Contin.encies of reinforcementE to e,p*ain a conditioned ref*e,. :ef*e,
phenomena) conditioned and unconditioned) ha/e) of course) been known for centuries) but it is
on*- recent*- that contin.encies of sur/i/a* and contin.encies of 6reinforcement ha/e been
in/esti.ated.
Inner Su++ements$ (he conditioned ref*e, is a simp*e princip*e of *imited scope describin.
certain simp*e facts) but man- interna* states and acti/ities) comparab*e with the dri/in. force of
instincts) ha/e been in/ented to e,p*ain it. (he runner5s heart is said to beat fast before the start
of the race because he Fassociates9 the situation with the e,ertion which fo**ows. But it is the
en/ironment) not the runner) that 9associates9 the two features) in the et-mo*o.ica* sense of
0oinin. or unitin. them. Cor does the runner 9form a connection9 between the two thin.s8 the
connection is made in the e,terna* wor*d. Conditioned responses are a*so said to occur in
9anticipation9 of) or in 9e,pectation9 of) customar- conse=uences) and the conditioned

ABOUT BEHAVIORISM &&

stimu*us is said to function as a 9si.n)E Fsi.na*)9 or 9s-mbo*.9 3 sha** return to these e,pressions
*ater.

Preparation for New Environments
11: 7B+:AC( C7C<3(37C3C'

A /er- different process) throu.h which a person comes to dea* effecti/e*- with a new
en/ironment) is operant conditionin.. ;an- thin.s in the en/ironment) such as food and water)
se,ua* contact) and escape from harm) are crucia* for the sur/i/a* of the indi/idua* and the
species) and an- beha/ior which produces them therefore has sur/i/a* /a*ue. (hrou.h the
process of operant conditionin.) beha/ior ha/in. this kind of conse=uence becomes more *ike*-
to occur. (he beha/ior is said to be stren*thened b- its conse=uences) and for that reason the
conse=uences themse*/es are ca**ed 9reinforcers.9 (hus) when a hun.r- or.anism e,hibits
beha/ior that +roduces food) the beha/ior is reinforced b- that conse=uence and is therefore
more *ike*- to 6recur. Beha/ior that reduces a potentia**- dama.in. condition) such as an
e,treme of temperature) is reinforced b- that conse=uence and therefore tends to recur on simi*ar
occasions. (he process and its effects ha/e .i/en rise to a *ar.e number of menta*istic concepts)
man- of which wi** be e,amined in the fo**owin. chapters.

(he standard distinction between operant and ref*e, beha/ior is that one is /o*untar- and
the other in/o*untar-. 7perant beha/ior is fe*t to be under the contro* of the beha/in. person and
has traditiona**- been attributed to an act of wi**. :ef*e, beha/ior) on the other hand) is not under
comparab*e contro* and has e/en been attributed to in/adin. wi**s) such as those of possessin.
spirits. Snee2in.) hiccuppin.) and other ref*e, acts were once attributed to the <e/i*) from whom
we
Innate Beha!ior &.

sti** protect a friend who has snee2ed b- sa-in.) 9'od b*ess -ouG9 %;ontai.ne said *ie crossed
himse*f e/en when he -awned.& 4hen no in/ader is assumed) the beha/ior is simp*- ca**ed
automatic.

Intermingling of Contingencies of Survival and Reinforcement

(here are certain remarkab*e simi*arities between contin.encies of sur/i/a* and
contin.encies of reinforcement. Both e,emp*if-) as 3 ha/e noted) a kind of causa*it- which was
disco/ered /er- *ate in the histor- of human thou.ht. Both account for purpose b- mo/in. it
after the fact) and both are re*e/ant to the =uestion of a creati/e desi.n. 4hen we ha/e re/iewed
the contin.encies which .enerate new forms of beha/ior in the indi/idua*) we sha** be in a better
position to e/a*uate those which .enerate innate beha/ior in the species. ;eanwhi*e we ma- note
the importance of insistin. upon the distinction.

3mprintin.. 7perant conditionin. and natura* se*ection are combined in the so6ca**ed
imprintin. of a new*- hatched duck*in.. 3n its natura* en/ironment) the -oun. duck*in. mo/es
toward its mother and fo**ows her as she mo/es about. (he beha/ior has ob/ious sur/i/a* /a*ue.
4hen no duck is present) the duck*in. beha/es in much the same wa- with respect to other ob6
0ects. %3n Htopia) (homas ;ore reported) the chicks hatched in an incubator fo**owed those who
fed and cared for them))& :ecent*- it has been shown that a -oun. duck*in. wi** come to
approach and fo**ow an- mo/in. ob0ect) particu*ar*- if it is about the same si2e as a duck6for
e,amp*e) a shoe bo,. +/ident*- sur/i/a* is sufficient*- we** ser/ed e/en if the beha/ior is not
ABOUT BEHAVIORISM&/

under the contro* of the specific /isua* features of a duck. ;ere*- approachin. and fo**owin. is
enou.h.
+/en so) that is not a correct statement of what happens. 4hat the duck*in. inherits is the
capacit- to be reinforced b- maintainin. or reducin. the distance between itse*f and a mo/in.
ob0ect. 3n the natura* en/ironment) and in the *aborator- in which imprintin. is studied)
approachin. and fo**owin. ha/e these conse=uences) but the contin.encies can be chan.ed. A
mechanica* s-stem can be constructed in which mo/ement toward an ob0ect causes the ob0ect to
mo/e rapid*- awa-) whi*e mo/ement awa- from the ob0ect causes it to come c*oser. Hnder these
conditions) the duck*in. wi** mo/e awa- from the ob0ect rather than approach or fo**ow it. A
duck*in. wi** *earn to peck a spot on the wa** if peckin. brin.s the ob0ect c*oser. 7n*- b- know6
in. what and how the duck*in. *earns durin. its *ifetime can we be sure of what it is e=uipped to
do at birth.

Imitation and the Instinct o( the Herd$ Catura* se*ection and operant conditionin. are often
confused when the- produce beha/iors ha/in. simi*ar topo.raphies. (he sur/i/a* /a*ue of
beha/in. as others beha/e seems ob/ious. 3f one member of a .roup responds to an approachin.
predator b- f*-in.) runnin.) or swimmin. awa-) and the. rest of the .roup then does the same) a**
ma- reach safet- a*thou.h on*- one has made direct contact with the predator. (he conditions are
suitab*e for natura* se*ection because other members are an) endurin. part of the en/ironment of
a species. Ce/erthe*ess) /er- simi*ar beha/ior is produced b- contin.encies of reinforcement. 3n
.enera* when a person is beha/in. in a .i/en wa-) he is doin. so because of pre/ai*in.
contin.encies) and simi*ar beha/ior
Innate Beha!ior &0

on the part of another person in the same situation is *ike*- to be sub0ect to the same
contin.encies. 3f one obser/es peop*e runnin. down a street) one ma- respond indirect*- to the
same contin.encies b- runnin. with them) thereb- possib*- escapin. dan.er or disco/erin.
somethin. interestin.. (o speak of an instinct of 9imitation9 or an 9instinct of the herd9 is
ambi.uous8 it ma- refer to contin.encies of sur/i/a* or contin.encies of reinforcement.

Territoriait- and A..ression. (hese terms do not refer to specific forms of beha/ior. An
or.anism ma- defend its territor- or attack others in man- different wa-s. ;odern warfare is
often said to e,emp*if- territoria*it- and a..ression) but it wou*d be hard to find an- act of a
so*dier that cou*d ha/e been se*ected b- contin.encies of sur/i/a*. At best) war*ike beha/ior is
ac=uired because of an inherent capacit- to be reinforced b- .ains in territor- or dama.e
inf*icted upon others.

A..ressi/e beha/ior ma- be innate and re*eased b- specific circumstances in which
sur/i/a* /a*ue is p*ausib*e. An infant or chi*d ma- bite) scratch) or strike if ph-sica**- restrained
when it cou*d not ha/e *earned to do so. 7r the beha/ior ma- be shaped and maintained because
peop*e are susceptib*e to reinforcement b- si.ns of dama.e to others. (he capacit- to be
reinforced when an opponent cries out or runs awa- wou*d ha/e sur/i/a* /a*ue because a person
so endowed wou*d =uick*- *earn to) defend himse*f) 7r) third) the beha/ior ma- be reinforced b-
conse=uences not e,p*icit*- re*ated to a..ression. 1ood and se,ua* contact) reinforcin. for other
reasons) ma- reinforce an attack on a competitor if food or a se,ua* partner is thus obtained.
(he intermin.*in. of contin.encies of sur/i/a* and
ABOUT BEHAVIORISM &8

reinforcement causes troub*e) and it is not surprisin. that nati/ists and en/ironmenta*ists often
disa.ree and sometimes rather a..ressi/e*- defend their respecti/e territories.

S+ecies1S+eci(ic %Uni!ersas$% (he term 9instinct9 is sometimes a/oided b- referrin.
instead to species6specific beha/ior on the. theor- that somethin. characteristic of a** members
of a species is probab*- part of its .enetic endowment. But contin.encies of reinforcement arc8
species6specific too. 4e ha/e seen an e,amp*e in the beha/ior of the duck*in. that fo**ows its
mother because of the 9uni/ersa*9 fact that mo/in. in the direction of an ob0ect norma**- brin.s
it c*oser) Hni/ersa* features of *an.ua.e do not imp*- a uni/ersa* innate endowment) because the
contin.encies of reinforcement arran.ed b- /erba* communities ha/e uni/ersa* features.
Bs-choana*-sts ha/e made a .reat dea* of the uni/ersa*it- of the 7edipus comp*e,) but the
contin.encies of persona* reinforcement in the fami*- in a .i/en cu*ture ma- be e=ua**-
uni/ersa*.

The Im+ortance o( Maintainin* the 2istinction$ 3t is no doubt true that ear*- beha/iorists
were undu*- enthusiastic about the *earnin. processes the- were disco/erin. and ne.*ected the.
ro*e of beha/iora* .enetics) but reactions to the beha/iorist position ha/e a*so been marked b-
e,a..eration. (here is no *on.er an- need for contro/ers-) e/en thou.h we are sti** a *on. wa-
from understandin. a** the interactions between contin.encies of sur/i/a* and contin.encies of
reinforcement.

3n an important sense a** beha/ior is inherited) since the or.anism that beha/es is the
product of natura* se*ection. 7perant conditionin. is as much a part of the

Innate Beha!ior &"

.enetic endowment as di.estion or .estation. (he =uestion is not whether the human species has
a .enetic endowment but how it is to be ana*-2ed. 3t be.ins and remains a bio*o.ica* s-stem) and
the beha/ioristic position is that it is nothin. more than that.

Iuite apart from the detai*s 7f the resu*tin. beha/ior) there are .ood reasons for
distin.uishin. between the two kinds of contin.encies. (he- differ .reat*- in their bearin. on the
=uestion with which we be.an: 4h- do peop*e beha/e as 6the- doA Contin.encies of reinforce6
ment ha/e the ed.e with respect to prediction and contro*. (he conditions under which a person
ac=uires beha/ior are re*ati/e*- accessib*e and can often be manipu*ated8 the conditions under
which a species ac=uires beha/ior are /er- near*- out of reach. 7ne unfortunate conse=uence is
that .enetic sources sometimes become a kind of dumpin. .round: an- aspect of beha/ior which
at the moment escapes ana*-sis in terms of contin.encies of reinforcement is *ike*- to be
assi.ned to .enetic endowment) and we are *ike*- to accept the e,p*anation because we are so
accustomed to .oin. no further than a state of the or.anism.

he Evolution of !ind"

(he concept of mind had been thorou.h*- e*aborated before the ad/ent of e/o*utionar-
theor-) and some accommodation was :needed. 4hen and how did mind e/o*/eA 4hat kind of
mutation cou*d ha/e .i/en rise to the first menta* state or process which) in contributin. to the
sur/i/a* of the person in whom it occurred) became part of the human .enetic endowmentA (he
=uestion is not un*ike that raised b- the con/ersion of rea*it- into e,perience or of thou.ht into
action. 4hat sort of ph-sica* .ene cou*d carr- the potentia* of mind)
ABOUT BEHAVIORISM .'

and how cou*d mind satisf- ph-sica* contin.encies of sur/i/a*A 3f mind is nothin. more than a
manifestation of ph-sio*o.-) such =uestions can be answered) or at *east postponed without
an,iet- unti* ph-sio*o.- can answer them) but not a** who subscribe to menta*ism accept that
position. ;ind has been said b- some66Teihard de Chardin) for e,amp*e6to be the end and
purpose of e/o*ution) if not somethin. be-ond it. (he) distin.uished scientist Janne/ar Bush has
put it this wa-:

4e seem) thus) to ha/e arri/ed at a concept of how the ph-sica* uni/erse about us6a** the
*ife that inhabits the speck we occup- in this uni/erse6bas e/o*/ed o/er the eons of time b-
simp*e materia* processes) the sort of processes we e,amine e,perimenta**-) which we describe
b- e=uations) and ca** the 9*aws of nature.9 +,cept for one thin.G ;an is conscious of his
e,istence. ;an a*so possesses) so most of us be*ie/e) what he ca**s his free wi**. <id
consciousness and free wi** too arise mere*- out of 9natura*9 processesA (he =uestion is centra* to
the contention between those who see nothin. be-ond a new materia*ism and those who
sec6Somethin..

(he beha/iorist has a simp*er answer. 4hat has e/o*/ed is an or.anism) part of the
beha/ior of which has been tentati/e*- e,p*ained b- the in/ention of the concept of mind. Co
specia* e/o*utionar- process is needed when the facts are considered in their own ri.ht.
4 Operant Behavior
'o Back

(he process of operant conditionin. described in the precedin. chapter is simp*e enou.h.
4hen a bit of beha/ior has the kind of conse=uence ca**ed reinforcin.) it is more *ike*- to
occur a.ain. A positi/e reinforcer stren.thens an- beha/ior that produces it: a .*ass of water is
positi/e*- reinforcin. when we are thirst-) and if we then draw and drink a .*ass of water) we
are more *ike*- to do so a.ain on simi*ar occasions. A ne.ati/e reinforcer stren.thens an-
beha/ior that reduces or terminates it: when we take off a shoe that is pinchin.) the reduction
in pressure is ne.ati/e*- reinforcin.) and we are more *ike*- to do so a.ain when a shoe
pinches.

The process supp"ements natura" se"ection. Important conse6uences of behavior
0hich cou"d not p"a7 a ro"e in evo"ution because the7 0ere not sufficient"7 stab"e
features of the environment are made effective through operant conditioning during the
"ifetime of the

ABOUT BEHAVIORISM 52

individua"8 0hose po0er in dea"ing 0ith his 0or"d is thus vast"7 increased.

The Feelings of Reinforcers

The fact that operant conditioning8 "i$e a"" ph7sio"ogi9 ca" processes8 is a product of
natura" se"ection thro0s "ight 'n the 6uestion of 0hat $inds of conse6uences are
reinforcing and 0h7. It is common"7 said that a thing is reinforcing because it fee"s8
"oo$s8 sounds8 sme""s8 or tastes good8 but from the point of vie0 of evo"utionar7 theor7 a
susceptibi"it7 to reinforcement is due to its surviva" va"ue and not to an7 associated
fee"ings.

The point ma7 be made for the reinforcers 0hich p"a7 a part in the conditioning of
ref"e:es. #a"ivation is e"icited b7 certain chemica" stimu"i on the tongue (as other
secretions are e"icited b7 other stimu"i in "ater stages of digestion) because the effect
has contributed to the surviva" of the species. A person ma7 report that a substance
tastes good8 but it does not e"icit sa"ivation because it tastes good. #imi"ar"78 0e pu"" our
hand a0a7 from a hot ob;ect8 but not because the ob;ect feels painfu". The behavior
occurs because appropriate mechanisms have bi-:n se"ected in the course of evo"ution.
The fee"ings are mere"7 co""atera" products of the conditions responsib"e for the
behavior.

The same ma7 be said of operant reinforcers. #a"t and sugar are critica"
re6uirements8 and individua"s 0ho 0ere especia""7 "i$e"7 to be reinforced b7 them have
more effective"7 "earned and remembered 0here and ho0 to get them and have
therefore been more "i$e"7 to survive and transmit this susceptibi"it7 to the species. It
has often been pointed out that competition for a mate tends to se"ect the more s$i""fu"
and po0erfu" members of a species8 but it a"so se"ects those more

OPERANT BEHAVIOR 53

susceptib"e to se:ua" reinforcement. As a resu"t8 the human species8 "i$e other species8
is po0erfu""7 reinforced b7 sugar8 sa"t8 and se:ua" contact. This is ver7 different from sa-
ing that these things reinforce because the7 taste or feet good.

<ee"ings have dominated the discussion of re0ards and punishments for
centuries. 'ne reason is that the conditions 0e report 0hen 0e sa7 that a taste8 odor8
sound8 pic"ure8 or piece of music is de"icious8 p"easant8 or beautifu" are part of the
immediate situation8 0hereas the effect the7 ma7 have in changing our behavior is
much "ess sa"ient-and much "ess "i$e"7 to be =seen8> because the verba" environment
cannot estab"ish good contingencies. According to the phi"osoph7 of hedonism8 peop"e
act to achieve p"easure and escape from or avoid pain8 and the effects referred to in 2d-
0ard ?. Thorndi$es famous ?a0 of 2ffect 0ere fee" ings@ >satisf7ing> or >anno7ing.> The
verb >to "i$e> is a s7non7m of >to be p"eased 0ith>A 0e sa7 >if 7ou "i$e> and >If 7ou
p"ease4B more or "ess interchangeab"7.

#ome of these terms refer to other effects of reinforcers-satisf7ing8 for e:amp"e8
is re"ated to satiation -but most refer to the bodi"7 states generated b7 reinforcers. It is
sometimes possib"e to discover 0hat reinforces a person simp"7 b7 as$ing him 0hat he
"i$es or ho0 he fee"s about things. !hat 0e "earn is simi"ar to 0hat 0e "earn b7 testing
the effect of a reinforcer@ he is ta"$ing about 0hat has reinforced him in the past or 0hat
he sees himse"f >going for.> But this does not mean that his fee"ings are causa""7
effectiveA his ans0er reports a co""atera" effect.

The e:pressions >I "i$e Brahms8C =I "ove Brahms.C >I en;o7 Brahms8> and >Brahms
p"eases me> ma7 easi"7 be ta$en to refer to fee"ings8 but the7 can be regarded as
statements that the music of Brahms is reinforcing. A

ABOUT BEHAVIORISM 54

person of 0hora the e:pressions are true 0i"" "isten to the radio 0hen it p"a7s Brahms
rather than turn it off8 bu7 and p"a7 records of Brahms8 and go to concerts 0here
Brahms is p"a7ed. The e:pressions have anton7ms (>I dis"i$e Brahms >I hate Brahms8>
>I detest Brahms8> and >Brahms bores me>)8 and a person for 0hom Brahms is thus
aversive 0i"" act to avoid or escape from hearing him. These e:pressions do not refer to
instances of reinforcement but rather to a genera" susceptibi"it7 or the "ac$ of it.

The a""usion to 0hat is fe"t needs to be carefu""7 e:amined. <ee"ings are especia""7
p"ausib"e 0hen the e:perience is directed to0ard a "iving person. The statement >I "ove
m7 0ife> seems to be a report of fee"ings8 but it a"so invo"ves a probabi"it7 of action. !e
are disposed to do to a person 0e "ove the things he8 "i$es or "oves to have done. !e
are not disposed to do to a person 0e dis"i$e (or especia""7 to a person 0e hate) the
things he "i$es or "oves to have doneA on the contrar7 0e are disposed to do the things
he. dis"i$es or hates to have done. !ith respect to a person 0ith 0hom 0e interact8
then8 to >"ove> is to behave in 0a7s having certain $inds of effects8 possib"7 0ith
accompan7ing conditions 0hich ma7 be fe"t.

Wants, Needs, Desires and Wishes

#ome menta"istic terms refer to conditions 0hich affect both the susceptibi"it7 to
reinforcement and the strength of a"read7 reinforced behavior. !e use >0ant> to de-
scribe a shortage@ a hungr7 man 0ants food in the simp"e sense that food is 0anting.
>Deeds> origina""7 meant vio"ent force8 restraint8 or compu"sion8 and 0e sti"" ma$e a
distinction bet0een 0anting to act (because of positive"7 reinforcing conse6uences) and
needing to act

OPERANT BEHAVIOR 55

(because not acting 0i"" have aversive conse6uences)8 but for most purposes the terms
are interchangeab"e. !e sa7 that a car needs gaso"ine and8 much "ess idiomatica""78
that gaso"ine is 0antiEg8 but to sa7 that a person >0ants to get out> suggests aversive
contro". The significant fact is that a person 0ho needs or 0ants food is particu"ar"7
"i$e"7 to be reinforced b7 food and that he is particu"ar"7 "i$e"7 to engage in an7 behavior
0hich has previous"7 been reinforced 0ith food. A person under aversive contro" is
particu"ar"7 "i$e"7 to be reinforced if I"e C#32?pes and to engage in an7 behavior 0hich
has "ed to escape.

If 0e $no0 the "eve" of deprivation or aversive stimu"ation8 0e can more
accurate"7 prcdict8 ho0 reinforceing a given event 0i"" be and ho0 "i$e"7 it is that a per-
son 0i"" engage in re"evant behavior. The $no0"edge has "ong been used for purposes
of contro". )eop"e have been made hungr7 so that the7 0i"" >0or$ for food> and so that
the7 can be reinforced 0ith food8 as the7 have been made miserab"e so that the7 0i""
act in 0a7s 0hich reduce their miser7.

An event is not reinforcing because it reduces a need. <ood is reinforcing even
0hen it does Eot satiate8 and deprivation can be changed in 0a7s 0hich are not rein-
forcing. The re"ation bet0een a state of deprivation and the strength of appropriate
behavior is presumab"7 due to surviva" va"ue. It behavior "eading to ingestion 0ere
strong at a"" times8 a person 0ou"d gross"7 overeat and use his energies inefficient"78

It is a mista$e to sa7 that food is reinforcing because 0e fee" bungr7or because
0e fee" the need for food8 or that 0e are more "i$e"7 to engage in food-reinforced
behavior because 0e fee" hungr7. It is the cond!on fe"t as hunger 0hich 0ou"d have
been se"ected in the

ABOUT BEHAVIORISM 5"

evo"ution of the species as most immediate"7 invo"ved in operant reinforcement.

The stites associated 0ith 0anting and needing are more "i$e"7 to be fe"t if no
re"evant behavior is at the moment possib"e. The "over 0rites >I 0ant 7ou> or >I need
7ou> 0hen nothing e"se can be done8 and if he is doing an7thing e"se8 aside from
0riting8 it must be a matter of c:istirig in the state 0hich "ie describes 0ith these
e:pressions. If behavior then becomes possib"e8 it is eas7 to sa7 that it 0as caused b7
the 0ant or need8 rather than b7 the deprivation or aversive stimu"ation responsib"e for
both the behavior and the state fe"t.

Eesiring8 "onging8 hoping8 and 7earning are more c"ose"7 re"ated to a current
absence of appropriate behavior because the7 terminate 0hen action begins. >I miss
7ou> cou"d a"most be thought of as a metaphor based on target practice8 e6uiva"ent to
>17 behavior 0ith respect to 7ou as a person cannot reach its mar$> or >I "oo$ for 7ou
and fai" to find 7ou.> The "over in the arms of his be"oved is not instant"7 free of 0anting
and needing her8 but be is no "onger missing her or "onging or 7earning for her. !ishing
is perhaps most e:c"usive"7 a reference to a heightened state of deprivation or aversive
stimu"ation 0hen no behavior is possib"e. A person ma7 0ish that he cou"d act (>I 0ish I
cou"d go>) or he ma7 0ish for the conse6uences (>I 0ish I 0ere there>).

The effects of operant reinforcement are often represented as inner states or
possessions. !hen 0e reinforce a person 0e are said to give him a motive or incentive8
but 0e infer the motive or the incentive from the bebavior. !e ca"" a person high"7
motivated 0hen a"" 0e $no0 is that he behaves energetica""7.

Eepriving a. person of something he needs or 0ants is not a forcefu" act8 and the
effect bui"ds up s"o0"78
OPERANT BEHAVIOR 5#

but states of deprivation are given a more dramatic ro"e 0hen the7 are ca""ed drives or
urges. <reud sa0 men merci"ess"7 >driven b7 po0erfu" bio"ogica" forces d0e"" iEg in the
depths of the mind or persona"it7.> !e are said to be at the merc7 of se:8 hunger8 and
hatred8 even though the7 are said to supp"7 the ps7chic energ7 needed for action.
<reudBs "ibido has been defined as =emotiona" or ps7chic energ7 derived from primitive
bio"ogica" urges.> These metaphors are based on aversive contro". The coachman does
d$%e his horses b7 0hipping them unti" the7 move for0ard8 and8 in the case of hunger at
"east8 strong interna" stimu"ation ma7 have a simi"ar function8 but deprivation as such is
not a driving force.

1enta"istic terms associated 0ith reinforcers and 0ith the states in 0hich
reinforcers are effective ma$e it difficu"t to spot functiona" re"ations. <or e:amp"e8 the
statement >The term BaggressionB shou"d be restricted to behavior motivated b7 the 0ish
to in;ure> is intended to ma$e a usefu" distinction bet0een behavior 0hich is mere"7
aggressive in form and an7 part of such behavior 0hich is emitted because it in;ures
another person8 but nothing is gained b7 spea$ing of the 0ish to in;ure or8 in particu"ar8
of being motivated b7 a 0ish. !hen the 5ti"itarians he"d that p"easure and pain 0ere the
>motives inf"uencing human behavior> the7 0ere referring to fee"ings associated 0ith
conse6uences rather than motives. The e:perimenta" ana"7sis of contingencies of
reinforcement puts these matters in better order.

Idea and Will

The conse6uences 0hich shape and maintain the behavior ca""ed an operant are not
present in the setting in 0hich a respon-se occursA the7 have become part of

ABOUT BEHAVIORISM 5&

the histor7 of the organism. The current setting ma7 affect the probabi"it7 of a response8
as 0e sha"" see in the ne:t chapter8 but it is not the on"7 thing that does so. To a"ter a
probabi"it7 is not to e"icit a response8 as in a ref"e:.

A person ma7 fee" or other0ise observe some of the conditions associated 0ith
the probabi"it7 that he 0i"" behave ia a given 0a7. <or e:amp"e8 he ma7 sa7 that he
>fee"s "i$e going8> that he >0ants to go8> that he >shou"d "i$e to go8> or that he >0ishes to
go.> The same terms are used to identif7 reinforeers-as in sa7ing8 >I fee" "i$e a drin$8> >I
0ant a drin$8> >I shou"d "i$e a drin$8> or >I 0ish I had a drin$.> It is possib"e that the report
>I fee" "i$e going> is c"ose to >I fee" no0 as I have fe"t in the past 0hen I have gone>A and
>I 0ant to go> ma7 be a report of deprivation or a shortage. >I 0ish> is8 is 0e have seen8
probab"7 c"oser to a report of a sheer probabi"it7 of action. !hether or not a person fee"s
or other0ise observes the "i$e"ihood of a response8 the simp"e fact is that at some point
a response occurs.

To distinguish an operant from an e"icited ref"e:8 0e sa7 that the operant response
is >emitted.> (It inight be better to sa7 simp"7 that it appears8 since emission ma7 imp"7
that behavior e:ists inside the organism and then comes oti". But the 0ord need not
mean e;ectionA "ight is not in. the hot fi"ament before it is emitted.) The principa" feature
is that there seems to bt no necessar7 prior causa" event. !e recogniFe this 0hen 0e
sa7 that >it occurred to him to go> as if to sa7 that >the act of going occurred to him.>
>Idea> is used to represent behavior in this sense (0e sa7 >the idea occurred to him>B)8
but in e:pressions "i$e >to get an idea8> or >to borro0 an idea> the 0ord suggests an
independent entit7. Deverthe"ess8 0hen 0e sa78 >I have an ideaA "etBs
OPERANT BEHAVIOR 5'

tr7 the rear doorA it ma7 be un"oc$ed8> 0hat is >had> is the behavior of tr7ing the rear
door. !hen a person successfu""7 imitates a dancing teacher8 he ma7 be said to >get
the idea8> a"though 0hat he gets is nothing more than behavior simi"ar to that of the
teacher. Dor need 0e refer to more than behavior 0hen 0e sa7 that a person 0ho
"aughs at a ;o$e has >got the point8> or that a person 0ho responds appropriate"7 to a
passage in a boo$ has >got its meaning.>

The apparent "ac$ of an immediate cause in operant behavior has "ed to the
invention of an initiating event. Behavior is said to be put into p"a7 0hen a person 0i""s
to act. The tetm has a confusing histor7. The simp"e future8 as in >Ge 0i"" go8> ta$es on
an additiona" meaning 0hen 0e sa78 >Ge 0i"" go in spite of the danger.> !i""ing is c"ose
to choosing8 particu"ar"7 0hen the choice is bet0een acting or not actingA to 0i"" or to
choose is evident"7 as unhera"ded as to act. B7 attributing other0ise une:p"ained
behavior to an act of 0i"" or choice8 one seems to reso"ve puFF"ement. That is perhaps
the principa" raison d!$e of the conceptA behavior is satisfactori"7 accounted for as "ong
as 0e have no reason to e:p"ain the act of 0i"". But the conditions 0hich determine the
form of probabi"it7 of an operant are in a personBs histor7. #ince the7 are not conspicu-
ous"7 represented in the current setting8 the7 are easi"7 over"oo$ed. It is then eas7 to
be"ieve that the 0i"" is free and that the person is free to choose. The issue is
determinism. The spontaneous generation of behavior has reached the same stage as
the spontaneous generation of maggots and micro-organisms in )asteurBs da7.

><reedom> usua""7 means the absence of restraint or coercion8 but more
comprehensive"7 it means a "ac$ of an7 prior deten-nination@ >A"" things that come to be8
e:cept acts of 0i""8 have causes.> #ome theo"ogians have

ABOUT BEHAVIORISM "(

been concerned for the freedom needed in order to ho"d a person responsi"@)1e8 and
the7 bave not been so easi"7 satisfiedA so-ca""ed Arminian doctrine he"d that a person
acts free"7 on"7 if he has chosen to act and onl) f !*e c*oosn+ !o ac! ,as b$ou+*!
abou! b) ano!*e$ ns!ance of c*oosn+-

The conspic?iousness of the causes is at issue 0hen ref"e: behavior is ca""ed
invo"untar7-one is not free to sneeFe or not to sneeFeA the initiating cause is the pepper.
'perant behavior is ca""ed vo"untar78 but it is not rea""7 uncausedA the cause is simp"7
harder to spot. The critica" condition for the apparent e:ercise of free 0i"" is positive
reinforcement8 as the resu"t of 0hich a persun fee"s free and ca""s himse"f free and sa7s
he does as he l.es or 0hat he ,an!s or is p"eased to do. (As 0e sha"" see n Chapter
18 a more important point is that positive"7 reinforcing conse6uences do not generate
avoidance or escape or an7 behavior designed to change the conditions in 0hich it
occurs.)

?i$e >idea8> >0i""> is used a"most interchangeab"7 0ith behavior or at "east with the
probability of behaving8 A 0i""ingness is a readiness or "i$e"ihood. A hea"th authorit7
has said that the important thing in maintaining a @regimen of e:ercise or diet is 0i""
po0erA a"" he means is that the important thing is that a person continue to e:ercise or
diet. A Icadees >0i"" to po0er> suggests behavior reinforced b7 economic8 re"igious8 or
governmenta" accretions in po0er. The statement that %1 some peop"e do not 0i""
because the7 are afraid> seems to refer to nothing more than the fact that the7 do riot
be*a%e because the7 are afraid. The biographica" statement that >the gir" he 0as
infatuated 0ith H0hom he never metI 0as a destructive agent8 para"7Fing his 0i"">
presumab"7 means that she para"7Fed some parts of his behavior.

OPERANT BEHAVIOR "/

A ver7 different ro"e of the 0i"" fo""o0s from its seeming spontaneit7 and m7ster78
0hich suggest that conse6uences ma7 be produced 0ithout ph7sica" action. >It 0as 0ith
the magic of his o0n 0i"" that Brahma created 0hatcveir is.> It is b7 an act of 0i"" that a
person is supposed to inf"uence the fa"" of dice in ps7cho$incsis.

!rpose and Intention

)ossib"7 no charge is more often "eve"ed against behaviorism or a science of
behavior than that it cannot dea" 0ith purpose or intention. A stimu"us-response formu"a
has no ans0er8 but operant behavior is the ver7 fie"d of purpose and intention. B7 its
nature it is directed to0ard the future@ a person acts in o$de$ !*a! something 0i"" happen8
and the order is tempora". >)urpose> 0as once common"7 used as a verb8 as 0e no0
use >propose.> >I propose to go> is simi"ar to >I intend to go.> If instead 0e spea$ of our
purpose or intention in going8 it is eas7 to suppose that the nouns refer to things.

A good dea" of misunderstanding has arisen from the fact that ear"7
representations of purpose 0ere spatia". The racerBs purpose is to reach the goa"8 and
0e p"a7 parcheesi 0ith the purpose of bringing our pieces home. In the maFes in 0hich
purposive behavior 0as once studied8 organisms moved to0ard the p"ace 0here rein-
forcement 0as to occur. To use goa" for purpose (>D>at is his goa" in "ife4>) is to identif7
it 0ith a terminus. But it is meaning"ess8 for e:amp"e8 to sa7 that the goa" -"et a"oneJ the
purpose-of "ife is death8 even though the u"timate termination is death. 'ne does not "ive
in ordcr to dic or 0ith the purpose of d7ing8
ABOUT BEHAVIORISM "2

0hether 0e are spea$ing in terms of natura" se"ection or operant conditioning.

Koa"s and purposes are confused in spea$ing of purpose in a homing device. A
missi"e reaches its target 0hen its course is appropriate"7 contro""ed8 in part b7
information coming from the target during its f"ight. #uch a device is sometimes said to
>have purpose bui"t into it8> but the feedbac$ used in guidance (the heart of c7bernetics)
is not reinforcement8 and the missi"e has no purpose in the present sense. (<eedbac$
ma7 be used in a $ind of e:p"icit goa"-see$ing behavior to be discussed in Chapter *.)

Dot a"" conse6uences are reinforcing8 and much of the effect of those 0hich are
depends upon the contingencies. )s7choana"7sts have often said that the gamb"erBs
true purpose is to punish himse"f b7 "osing. It is a"most a"0a7s the case that the gamb"er
eventua""7 "oses8 and the behavior therefore has that conse6uence8 but it is not
therefore reinforcing. Kamb"ing can be demonstrated in man7 other species and is
e:p"ained b7 a specia" schedu"e of reinforcement to be noted in a moment. The u"timate
"oss (the >negative uti"it7>) does not offset the effect of the schedu"e.

The 5ti"itarians supposed that it might be possib"e to measure 6uantities of
p"easure and pain in such a 0a7 that the p"easure generated b7 socia""7 ob;ectionab"e
behavior cou"d be offset b7 a ca"cu"ated amount of pain in the forra of punishment.
5nfortunate"78 the condition generated b7 a reinforcer and fe"t as p"easure is re"ative"7
insignificant in determining the 6uantit7 of behavior produced compared 0ith the
schedu"e of reinforcement.

A va"id distinction "ies bac$ of the statement >1otives and purposes are in the
brain and heart of man8 0hereas conse6uences are in the 0or"d of fact.> .e

OPERANT BEHAVIOR "3

move the gratuitous ph7sio"ogiFing8 and the point is made that motives and purposes
are in peop"e 0hi"e contingencies of reinforcement are in the environment8 but motives
and purposes are at best the effects of reinforcements. The change 0rought b7
reinforcement is often spo$en of as >the ac6uisition of purpose or intention8> and 0e are
said to >give a person a purpose> b7 reinforcing him in a given 0a7. These are conven-
ient e:pressions8 but the basic fact is that 0hen a person is >a0are of his purpose8> he
is fee"ing or observing introspective"7 a condition produced b7 reinforcement.

#ee$ing or "oo$ing for something seems to have a particu"ar"7 strong orientation
to0ard the future. !e "earn to "oo$ for an ob;ect 0hen 0e ac6uire behavior 0hich
common"7 has the conse6uence of discovering it. Thus8 to "oo$ for a match is to "oo$ in
a manner previous"7 reinforced b7 finding matches. To see$ he"p is to act in 0a7s 0hich
have in the past "ed to he"p. If past conse6uences have not been ver7 e:p"icit8 0e are
"i$e"7 to "oo$ in vague and unproductive 0a7s. )eop"e can usua""7 sa7 0hat the7 are
"oo$ing for and 0h7 the7 are "oo$ing in a given p"ace8 but "i$e other species the7 a"so
ma7 not be ab"e to do so.

1an7 features of the debate about purpose in human behavior are reminiscent of
the debate about purpose in evo"ution. As the Co"umbia Enc)clo0eda 0u!s it@

A sti"" preva"ent misunderstanding of evo"ution is the be"ief that an anima" or p"ant
changes in order to better adapt to its environmentA e.g.8 that it deve"ops an e7e for the
purpose of seeing. #ince mutation is a random process and since most mutations are
harmfu" rather than neutra" or beneficia" to the organism8 it s evident that the occur-
rence of a variation is itse"f a matter of chance8 and that one cannot spea$ of a 0i"" or
purpose on the part of the individua" to deve"op a ne0 structure or trait that might prove
he"pfu".

ABOUT BEHAVIORISM "4

Feelings "ssociated With
#ched!les of Reinforce$ent

The probabi"it7 that a person 0i"" respond in a given 0a7 because of a histor7 of
operant reinforcement changes as the contingencies change. Associated bodi"7
conditions can be fe"t or observed introspective"78 and the7 are often cited as the
causes of the states or changes in probabi"it7.

!hen a given act is a"most a"0a7s reinforced8 a person is said to have a fee"ing
of confidence. A tennis p"a7er reports that he practices a particu"ar shot >unti" he fee"s
confident>A the basic fact is that he practices unti" a certain proportion of his shots are
good. <re6uent reinforcement a"so bui"ds faith. A person fee"s sure8 or certain8 that he
0i"" be successfu". Ge en;o7s a sense of master78 po0er8 or potenc7. The infant is said
to ac6uire a sense of infanti"e omnipotence. <re6uent reinforcement a"so bui"ds and
maintains an interest in 0hat a person is doing. In a"" this the behavior is erroneous"7
attributed to the fee"ings rather than to the contingencies responsib"e for 0hat is fe"t.

!hen reinforcement is no "onger forthcoming8 behavior undergoes >e:tinction>
and appears rare"78 if at a"". A person is then said to suffer a "oss of confidence8
certaint78 or sense of po0er. Instead8 his fee"ings range from a "ac$ of interest through
disappointment8 discouragement8 and a sense of impotence to a possib"7 deep
depression8 and these fee"ings are then said-erroneous"7-to e:p"ain the absence of the
behavior. <or e1a20le3 a person is said to be unab"e to go to 0or$ because he is
discouraged or depressed8 a"though his not going8 together 0ith 0hat he fee"s8 is due to
a "ac$ of reinforcerneiat-eit"ier in his 0or$ or in some other part of his "ife.
OPERANT BEHAVIOR "5

<rustration is a rather different condition8 0hich inc"udes a tendenc78 often
characteristic of a fai"ure to be reinforced8 to attac$ the s7stem. Thus8 a person 0ho
$ic$s the8 vending machine 0hich has fai"ed to de"iver cigarettes or ba0"s out his 0ife
0ho has forgotten to bu7 them is said to do so because of frustration. The e:pression
>frustrated e:pectations> refers specifica""7 to a condition produced b7 the termination of
accustomed reinforcement.

A different $ind of fee"ing is associated 0ith the "ac$ of an appropriate occasion
for behavior8 the archet7pa" pattern of 0hich is homesic$ness. !hen a person has "eft
home for the first time8 much of the behavior appropriate to that environment can no
"onger be emitted. The condition fe"t ma7 be simi"ar to depression8 0hich is said to be
common in peop"e 0ho have moved from one cit7 to another. It is ca""ed
>nosta"giaC--2tera""78 the pain generated b7 a strong tendenc7 to return home 0hen
return is impossib"e. A simi"ar condition prevai"s 0hen one is simp"78 "ost8 and the 0ord
is then >for"orn.> A >"ove"orn> person is unab"e to emit behavior directed to0ard the
person he "oves. A person 0ho is a"one ma7 fee" "onesomeA the essentia" condition is
that there is no one 0ith 0hom he can ta"$ or behave in other 0a7s. The behavior of the
homesic$8 for"orn8 "ove"orn8 or "one"7 is common"7 attributed to the fee"ings e:peri enced
rather than to the absence of a fami"iar environment.

1ost reinforcements occur intermittent"78 and the schedu"es on 0hich the7 are
programmed generate conditions 0hich are described 0ith a 0ide range of terms. The
so-ca""ed ratio schedu"es supp"7 man7 good e:amp"es. !hen the ratio of responses to
reinforcements is favorab"e8 the behavior is common"7 attributed to (1) di"igence8
industr78 or ambition8 () determination8

ABOUT BEHAVIORISM ""

stubbornness8 sta7ing po0er8 or perseverance (continuing to respond over "ong periods
of time 0ithout resu"ts)8 (3) e:citement or enthusiasm8 or (%) dedication or compu"sion.

The ratio of responses to reinforcements ma7 be =stretched> unti" it becomes
6uite unfavorab"e. This has happened in man7 incentive s7stems8 such as the
piece-rate pa7 of home industries in the nineteenth centur7. The schedu"e generates a
dangerous"7 high "eve" of activit78 and those interested in the 0e"fare of 0or$ers usua""7
oppose it. It is not un$no0n8 ho0ever8 in dai"7 "ife. A 0riter 0ho ma$es his "iving b7
0riting one artic"e or stor7 after another is on a $ind of fi:ed-ratio schedu"e8 and he is
often a0are of one resu"t@ the comp"etion of one artic"e is often fo""o0ed b7 a period re-
semb"ing e:tinction during 0hich "ie is unab"e to start a ne0 one. The condition is
sometimes ca""ed >abu"ia8> defined as a "ac$ of 0i"" po0er8 or a neurotic inabi"it7 to act8
and this is often cited as the source of the troub"e8 in spite of the fact that the schedu"e
produces a simi"ar effect in a 0ide range of species.

,ariab"e-ratio schedu"es8 in 0hich reinforcement occurs after a given average
number of responses but in 0hich the ne:tresponse to be reinforced cannot be pre-
dicted8 are particu"ar"7 interesting. A favorab"e histor7 in 0hic"i the average is s"o0"7
en"arged is said to generate 0i"" po0er8 together 0ith "arge amounts of ps7chic energ78
or "ibido. It is said that Git"er pro"onged the #econd !or"d !ar for near"7 a 7ear >b7 an
incredib"e e:ercise of 0i"" po0er 0hich a"" the others in Kerman7 "ac$ed8> but his
behavior (and hence his >0i"" po0er>) can be p"ausib"7 attributed to an e:traordi nari"7
favorab"e program (favorab"e for Git"er8 disastrous for the 0or"d) in 0hich each of a
series of reinforcing successes re6uired an increasing"7 greater

OPERANT BEHAVIOR "#

amount of effort. (This $ind of interpretation of a historica" event can never be more than
p"ausib"e8 but it is a better e:p"anation than 0i"" po0er.)

A"" gamb"ing s7stems are based on variab"e-ratio schedu"es of reinforcement8
a"though their effects are usua""7 attributed to fee"ings. It is fre6uent"7 said8 for e:amp"e8
that peop"e gamb"e because of the e:citement8 but the e:citement is c"ear"7 a co""atera"
product. It is a"so sometimes said that peop"e gamb"e >to satisf7 their sense of master78
to dominate8 to 0in>-in spite of the fact that gamb"ers a"most a"0a7s eventua""7 "ose. The
inconsistenc7 is e:p"ained b7 ca""ing the gamb"er 0ho ruins himse"f and his fami"7
>compu"sive> or >patho"ogica"8> his >iffationa"> behavior thus being attributed to an
i""ness. Gis behavior is >abnorma"> in the sense that not ever7one responds 0ith simi"ar
dedication to the prevai"ing contingencies8 but the fact is simp"7 that not ever7one bias
been e:posed to a program through 0hich a high"7 unfavorab"e8 ratio is made effective.
The same variab"e-ratio schedu"e affects those 0ho e:p"ore8 prospect8 invent8 conduct
scientific research8 and compose 0or$s of art8 music8 or "iterature8 and in these fie"ds a
high "eve" of activit7 is usua""7 attributed to dedication rather -than compu"sion or
irrationa"it7.

It is characteristic of intermittent reinforcement that behavior ma7 be sustained
over "ong periods of time 0ith ver7 "itt"e return. This has been e:p"ained b7 sa7ing8
>Guman beings are creatures of hope and not genetica""7 designed to resign
themse"ves8> but there is nothing essentia""7 human about the effects8 and it is not hope
or resignation but the contingencies 0hich are the conspicuous and accessib"e cause.

AB'5T B2GA,I'.I#1 "&

Aversive Stimufi and Punishment



Aversive stimu"i8 0hich generate a host of bodi"7 conditions fe"t or introspective"7
observed8 are the stimu"i 0hich function as reinforcers 0hen the7 are reduced or
terminated. The7 have different effects 0hen re"ated to behavior in other 0a7s. In
respondent conditioning8 if a previous"7 neutra" stimu"us8 such as a be""8 is fre6uent"7
fo""o0ed after an interva" b7 a no:ious stimu"us8 such as an e"ectric shoc$8 the be""
comes to e"icit reactions8 primari"7 in the autonomic nervous s7stem8 0hich are fe"t as
an:iet7. The be"" has become a conditioned aversive stimu"us8 0hich ma7 then have the
effect of changing the probabi"it7 of an7 positive"7 reinforced behavior in progress. Thus8
a person engaged in a "ive"7 conversation ma7 begin to spea$ "ess energetica""7 or
more erratica""7 or ma7 stop spea$ing a"together at the approach of someone 0ho has
treated him aversive"7. 'n the other hand8 his negative"7 reinforced behavior ma7 be
strengthened8 and he ma7 act more compu"sive"7 or aggressive"7 or move to escape.
Gis behavior does not change because he fee"s an:iousA it changes because of the
aversive contingencies 0hich generate the condition fe"t as an:iet7. The change in
fee"ing and the change in behavior have a common cause.

)unishment is easi"7 confused 0ith negative reinforcement8 sometimes ca""ed
>aversive contro".> The same stimu"i are used8 and negative reinforcement might be
defined as the punishment of not behaving8 but punishment is designed to remove
behavior from a repertoire8 0hereas negative reinforcement generates behavior.

)unishing contingencies are ;ust the reverse of reinforcing. !hen a person
span$s a chi"d or threatens to

O0e$an! Be*a%o$ "&

span$ him because he has misbehaved8 he is 0$esen!n+ a negative reinforcer rather
than removing one8 and 0hen a government fines an offender or puts him in prison8 it is
-removing a positive reinforcer (or a situation in 0hich behavior has occasiona""7 been
positive"7 reinforced) rather than presenting a negative one. If the effect 0ere simp"7 the
reverse of the effect of reinforcement8 a great dea" of behavior cou"d be easi"7
e:p"ainedA but 0hen behavior is punished8 various stimu"i generated b7 the behavior or
the occasion are conditioned in the respondent pattern8 and the punished behavior is
then disp"aced b7 incompatib"e behavior conditioned as escape or avoidance. A
punished person remains >inc"ined> to behave in a punishab"e 0a78 but he avoids
punishment b7 doing something e"se instead8 possib"7 nothing more than stubborn"7
doing nothing.

!hat a person fee"s 0hen he is in a situation in 0hich he has bcen punished or
0hen he has engaged in previous"7 punished behavior depends upon the t7pe of
punishment8 and this often depends in turn upon the punishing agent or institution. If he
has been punished b7 his peers8 "ie is said to fee" shameA if he has been punished b7 a
re"igious agenc78 he is said to fee" a sense of sinA and if be has been punished b7 a
governmenta" agenc78 he is said to fee" gui"t. If he acts to avoid further punishment8 he
ma7 moderate the condition fe"t as shame8 sin88 or gui"t8 but he does not act because of
his fee"ings or because his fee"ings are then changedA he acts because of the punishing
contingencies to 0hich he has been e:posed8

The condition fe"t as shame8 gui"t8 or a sense of sin is not due simp"7 to an ear"ier
occurrence of an aversive stimu"us. A thunderstorm ma7 set up conditions fe"t as
an:iet78 and during a storm positive"7 reinforced

ABOUT BEHVIORISM #(

behavior ma7 be 0ea$ened8 and negative"7 reinforced (such as f"ight or concea"ment)
strengthened8 but this condition is not fe"t as gui"t. The point has been made b7 sa7ing
that >a person cannot fee" gui"t7 if he has no ob;ect-directed impu"ses to fee" gui"t7
about.> 1ore e:act"78 he fee"s gui"t7 on"7 0hen he behaves8 or tends to behave8 in a
punishab"e 0a7.

A 0riter 0ho sa7s8 >The more I read of the ear"7 and mid-,ictorians8 the more I
see an:iet7 and 0orr7 as the "eading e"ite to understanding them8> is suggesting an
e:p"anation of behavior in terms of fee"ings generated b7 punishing circumstances8
0here the fee"ings are inferred from the behavior the7 are used to e:p"ain8 Ge is not
c"aiming to have an7 direct information about fee"ings8 and presumab"7 means
understanding 0hat the7 said and did8 but an:iet7 and 0orr7 are usefu" c"ues on"7 if the7
can be e:p"ained in turn. The 0riter attempts to do this 0hen he continues- >The7 0ere
tr7ing to ho"d together incompatib"e opposites8 and the7 0orried because the7 fai"ed. . . .
The7 0orried about immorta"it78 the7 0orried about se:8 the7 0orried about po"itics and
mone7.> These 0ere the e:terna" circumstances responsib"e for their behavior and for
the conditions fe"t as 0orr7.

The fre6uenc78 severit78 and schedu"e of punishment generate other aspects of
behavior often attributed to fee"ings or traits of character. In man7 fami"iar instances8
behavior has both punishing and reinforcing conse6uences. If behavior sti"" occurs but
in a 0ea$ened form8 it ma7 be said to sho0 inhibition8 timidit78 embarrassment8 fear8 or
caution. 2:cessive punishment is said to ma$e a shortage of positive reinforcement
more critica" and "eave a person >more vu"nerab"e to severe depression and to giving
up.> !e treat 0hat is fe"t not b7 changing the fee"ings but b7 changing the contingencies

OPERANT BEHAVIOR #/

-for e:amp"e8 b7 evo$ing the behavior 0ithout punishing it8 so that conditioned aversive
stimu"i ma7 undergo e:tinction.

Behavior 0hich is strong in spite of punishing conse6uences is said to sho0
braver78 courage8 or possib"7 audacit7. !e encourage a person not b7 ma$ing him fee"
more courageous but b7 emphasiFing reinforcing conse6uences and. minimiFing
punishing. A foo" rushes into a dangerous situation not because "ie fee"s rec$"ess but
because reinforcing conse6uences have comp"ete"7 offset punishingA and 0e ma7
attempt to correct his behavior b7 supp"7ing other (possib"7 verba") punishments.

!hen punishment is particu"ar"7 severe8 the se"f$no0"edge discussed in Chapter
ma7 be affected. The behavior suppressed ma7 inc"ude the behavior invo"ved in
$no0ing about associated bodi"7 conditions. The resu"t is 0hat <reud ca""ed
>repression.> <or <reud8 ho0ever8 the process invo"ved fee"ings rather than behavior8
and it too$ p"ace in the depths of the mind. <ee"ings 0ere repressed b7 other fee"ings
and po"iced b7 a censor from 0horn. the7 sometimes escaped in devious 0a7s. The7
cou"d continue to be troub"esome8 ho0ever8 and man 0as said to be >haunted b7 his
repressed "ongings.> I sba"" discuss some behaviora" aspects "ater.

#tr!ct!ralis$

2ar"7 studies of behavior 0ere often said to confine themse"ves to form or
structure-to treat behavior8 for e:amp"e8 as nothing more than >musc"e t0itches.> The
refusa" to accept fee"ings and states of mind as causes and an abiding concern for
>ob;ectivit7> seemed to support such a vie0. Gabit formation 0as a structura"ist
princip"e-. to ac6uire a habit 0as mere"7 to become ac

ABOUT BEHAVIORISM #2

customed to behaving in a given 0a7. The contingencies of reinforcement 0hich
generated the behavior8 "i$e the contingencies of surviva" 0hich produced an instinct8
0ere neg"ected.

<re6uenc7 theories of "earning 0ere a"so structura". The7 asserted simp"7 that
0hat has happened once 0i"" happen again8 that an organism 0i"" tend to do 0hat it has
done most often in the past. As I have noted8 behaviora"ism confined itse"f to the
topograph7 of po"itica" behavior8 and structura"ism in anthropo"og7 is often not far
be7ond the position that customs are fo""o0ed simp"7 because it is customar7 to fo""o0
them. 2ar"7 Kree$ and )ersian ;ustice 0as simp"e and s0ift because it 0as based
entire"7 on the topograph7 of a crime@ a person 0ho $i""ed another 0as gui"t7 of murder
regard"ess of the circumstances. I sha"" note "ater the significance of the fact that
support for the structura"ist position has conic from both phenomeno"og7 and e:-
istentia"ism8 0ith their neg"ect of past and future in the search for the essentia" features
of the here and no0.

If behaviorism had not rep"aced the fee"ings and states of mind8 0hich it discarded
as e:p"anations8 it cou"d indeed be ca""ed a $ind of structura"ism8 but it found
rep"acements in the environment. As 0e "earn more about the ro"e of contingencies of
reinforcement8 0e are more "i$e"7 to move be7ond forma" properties. The point ma7 be
i""ustrated 0ith the concept of imitation. In a pure"7 forma"istic definition8 one organism
might be said to be imitating another 0hen it behaves as the other behaves8 but8 as 0e
sa0 in Chapter 38 contingencies of both surviva" and reinforcement must be considered.
The patrons of a restaurant are behaving in rough"7 the same 0a7 0ith respect to their
dinners8 but the7 are not imitating each otherA the7 are behaving in simi"ar 0a7s
because the7 are e:posed to simi"ar

OPERANT BEHAVIOR #3

contingencies. The man 0ho runs after a thief is not imitating him8 though both are
running.

#tructura"ism is invo"ved in the distinction often dra0n bet0een "earning or
competence and performance. The distinc?ion 0as usefu" in ear"7 studies of "earning
because the change-8 in performance then observed 0ere rather erratic. #ince it 0as
assumed that "earning 0as an order"7 process8 there appeared to be a discrepanc78 but
it 0as reso"ved b7 supposing that "earning 0its not ver7 accurate"7 revea"ed in the be-
havior the organism disp"a7ed. )erformance 0as c"ear"7 a structura"ist termA it referred
to 0hat an organism did 0ithout referring to 0h7 it did it. Improved techni6ues have
revea"ed an order"7 re"ation bet0een performance and contingencies and have
e"iminated the need to appea" to a separate inner "earning process or to competence.

The same confusion ma7 be seen in the contention that operant and respondent
conditioning represent a sing"e process8 a contention said to be opposed to the vie0
that the t0o $inds of conditioning affect different s7stems of behavior8 respondent
conditioning being appropriate to the autonomic nervous s7stem and operant
conditioning to the s$e"eta" muscu"ature. It is true that much of the activit7 of the
autonomic nervous s7stem does not have natura" conse6uences 0hich cou"d easi"7
have become part of operant contingencies8 but such conse6uences can be arranged.
(In Chapter 11 1 sha"" report an effort to bring the vascu"ar s7stem of the arm under
operant contro" b7 instrumenta""7 amp"if7ing a measure of the vo"ume of the arm.) But
the basic difference is not in the topograph7 of response s7stems but in the
contingencies. The environmenta" arrangements 0hich produce a conditioned ref"e: are
6uite different from those 0hich produce operant behavior8 regard"ess
ANOUT BEHAVIORISM #4

of the respective s7stems. (The fact that both processes ma7 go on in a given situation
a"so does not mean that the7 are the same process. A chi"d ac6uiring operant behavior
no doubt a"so ac6uires conditioned ref"e:es8 and )av"ovBs dog8 though restrained b7 the
e:perimenta" stand8 0as operant"7 reinforced in adventitious 0a7s b7 the occasiona"
presentation of food.) !e must 0ait to see 0hat "earning processes the ph7sio"ogist 0i""
eventua""7 discover through direct observation8 rather than through inferencesA
mean0hi"e8 the contingencies permit a usefu" and important distinction.

#tructura"ism often goes be7ond mere description8 and one of its strategies has
had a ver7 "ong histor7. !hen the notion of a functiona" re"ation 0as not 7et fu""7
understood8 e:p"anations of phenomena 0ere sought in their structures. )"atoBs doctrine
of forms 0as an effort to e:p"ain events 0ith princip"es derived from the same or simi"ar
events. It has been said that from )"ato to /ep"er mathematics 0as not regarded as de-
scribing ce"estia" motion but as e:p"aining it. The search for e:p"anation in form or
structure goes on8 Kesta"t ps7cho"og7 tried to supp"ement the structura" notion of habit
formation 0ith organiFationa" princip"es. 1athematica" properties ho"d their o"d
e:p"anator7 forceA it has been said8 for e:amp"e8 that for one anthropo"ogist >the
re"ations of $inship do not evo"ve as much as the7 tend to e:press a"geb raic re"ations.>

As I noted in Chapter 18 a mere"7 structura" account ma7 be supp"emented b7
invo$ing time as an independent variab"e. The gro0th of the embr7o from a ferti"iFed
egg to a fetus at term is a remar$ab"e e:amp"e of deve"opment8 and it has been
suggested that simi"ar se6uences in the gro0th >of a s$i""8 of an art8 of a concept in the
mind> ma7 be important. The behavior of a person or a cu"ture is said to pass through
various
OPERANT BEHAVIOR #5

stages unti" it reaches maturit7. The ps7chopatho"og7 of the drug addict has been said
to be due to >arrested infanti"e ps7chic deve"opment.> As these e:amp"es suggest8 0hat
gro0s is said to be something in the mind8 as 0ith )iaget8 or in the persona"it78 as 0ith
<reud. But if a chi"d no "onger behaves as be behaved a 7ear before8 it is not on"7
because he has gro0n but because he has had the time to ac6uire a much bigger
repertoire through e:posure to ne0 contingencies of reinforce8 ment8 and particu"ar"7
because the contingencies affecting chi"dren at different ages are different. A chi"dBs
0or"d >deve"ops8> too.

Compared 0ith the e:perimenta" ana"7sis of behavior8 Lve"opmenta" ps7cho"og7
stands in the position of evo"utionar7 theor7 before Ear0in. B7 the ear"7 nineteenth
centur7 it 0as 0e"" $no0n that species had undergone progressive changes to0ard
more adaptive forms. The7 0ere deve"oping or maturing8 and improved adaptation to
the environment suggested a $ind of purpose. I"e 6uestion 0as not 0hether evo"utionar7
changes occurred but 0h7. Both ?amarc$ and Buffon appea"ed to the purpose
supposed"7 sho0n b7 the individua" in adapting to his environment -a purpose someho0
transmitted to the species. It remained for Ear0in to discover the se"ective action of the
environment8 as it remains for us to supp"ement deve"opmenta"ism in behaviora"
science 0ith an ana"7sis of the se"ective action of the environment.

The %ind in Operant Behavior

In most of this chapter I have been concerned 0ith fee" ings or states of mind 0hich ma7
be interpreted as co""atera" products of the contingencies 0hich generate behavior. It
remains for us to consider other men

ABOUT BEHAVIORISM #"

ta"istic processes 0hich are said to be needed if operant conditioning is to ta$e p"ace.
The mind is not merc"7 a spectatorA it is said to p"a7 an active ro"e in the determination
of behavior.

1an7 2ng"ish idioms containing the 0ord >mind> suggest a probabi"it7 of action8
as in >I have a mind to go.> 1ind is often represented as an agent8 scarce"7 to be
distinguished from the person 0ho has the mind. >It crossed m7 nnd that I shou"d go> is
scarce"7 more than BIt occurred to ne that I shou"d go.> !hen responses of g"ands or
smooth musc"e (under contro" of the autonomic nervous s7stem) are brought under op-
erant contro" b7 ma$ing reinforcement contingent upon them8 the re#?H"t is said to
demonstrate the contro" of %B mind over matter>A but 0hat it demonstrates is that a
person ma7 respond 0ith his g"ands or his smooth musc"es under operant
contingencies. A mechanica" arm designed to be operated b7 musc"es norma""7 op-
erating some other part of the bod7 is said to be >thought-operated> or >operated b7 the
mind8> a"though it is operated b7 the person 0ho origina""7 moved some other part of his
bod7. !hen peop"e shoot other peop"e8 it is said that >minds $i""8 not guns8> and that >a
manBs mind 0as the instrument direct"7 responsib"e for the assassination of Mohn <.
/enned7 and 1artin ?uther /ing8> but peop"e are shot b7 peop"e8 not b7 minds.

The vie0 that menta" activit7 is essentia" to operant behavior is an e:amp"e of the
vie0 that fee"ings or introspective"7 observed states are causa""7 effective. !hen a
person rep"ies to the 6uestion >!i"" 7ou go tomorro04> b7 sa7ing8 >I donBt $no08 I never
$no0 ho0 I 0i"" fee"8> the assumption is that 0hat is in doubt is the fee"ing rather than
the behavior-that the per


O0e$an! Be*a%o$ ##

son 0i"" go if heA fee"s "i$e going rather than that he 0i"" fee" "i$e going if he goes.
Deither statement is8 of course8 an e:p"anation.

There are other 0ords referring to menta" activities said to be more #pCcirica""7
re6uired b7 behavior. )eop"e must >;udge> 0hat 0i"" or 0i"" not occur if the7 do or do not
act in certain 0a7s. The dog in the )av"ovian e:periment sa"ivates in anticipation of food
or because it >e:pects> food. In operant e:periments a rat presses a "ever because it
>anticipates> that food 0iMI be de"ivered or e:pects food to be de"ivered 0hen it does so.
>In socia" "earning theor7 the potentia" of the occurrence of a behavior is considered to
be a function of the e:pectanc7 that the behavior 0i"" "ead to a particu"ar reinforcement
or reinforcements and the va"ue of these reinforcements in a given situation.> !e shou"d
have to trans"ate these statements in some such 0a7 as this@ >Ti"e probabi"it7 of
behavior depends upon the $ind of fre6uenc7 of reinforcement in simi"ar situations in the
past.> A person ma7 0e"" fee" conditions associated 0ith >;udging8> >anticipating8> and
>e:pecting8> but he does not need to do so.

'perant behavior is a"so said to re6uire the >association> of ideas. The fact that a
bab7 "earns to avoid a hot stove is said to imp"7 that >the bab7 has the abitit7 to
associate his act . . . 0ith getting burned.> But8 as in a conditioned ref"e:8 touching and
burning are associated in the contingencies. .einforcement is a"so said to >supp"7
information>@ >!ith other than ver7 7oung chi"dren 0e can never sa7 that the ma;or
effect of reinforcement is other than a source of information used b7 the chi"d to confirm
or change his e:pectations and to deve"op ne0 and tentative so"utions.> Increasing the
probabi"it7 that peop"e 0i"" re

ABOUT B4HAVIORISH #&

spond in certain 0a7s is sometimes said to be a matter of >raising consciousness.> Go0
fast a rat 0i"" run in a maFe is said to depend upon 0hether it >$no0s that food is an7
"onger avai"ab"e in the end bo:.> 1 shaft return to $no0"edge8 information8 and
consciousness in "ater chapters.

Another supposed menta" process said to be needed in operant conditioning is
understanding. )eop"e must 8Bunderstand the regu"arities upon 0hich the7 can count.>
Their action must be >grounded on the understanding of ho0 things behave.> Another
state said to be needed is be"ief. )eop"e must be"ieve that 0hat the7 are doing has
some chance of obtaining 0hat the7 0ant or avoiding something to 0hich the7 are
averse. But the chances are in the contingencies. The re"ation of be"iefs to other
conditions8 such as 0ants and needs8 can be easi"7 stated@ to sa7 that >desires enter
into the causation of be"iefs> is simp"7 to sa7 that the probabi"it7 of behavior 0ith 0hich a
be"ief is associated depends not on"7 upon reinforcement but upon a state of deprivation
or aversive stimu"ation.

It is sometimes said that operant conditioning is simp"7 one aspect of the pursuit
of happiness8 and the e:pression 0i"" he"p to summariFe severa" points in this chapter.
Gappiness is a fee"ing8 a b7-product of operant reinforcement. Tbe things 0hich ma$e
us happ7 are the things 0hich reinforce us8 but it is the things8 not the fee"ings8 0hich
must be identified and used in prediction8 contro"8 and interpretation. )ursuit suggests
purpose@ 0e act to achieve happiness. But pursuit8 "i$e search8 is simp"7 behavior 0hich
has been reinforced b7 achieving something. Behavior becomes pursuit on"7 after
reinforcement. It has been said that the pursuit of happiness cannot be an e:p"anation
of behavior because >nothing proves that
OPERANT BEHAVIOR #'

men in modern societies are happier than men in archaic societies8> but operant
reinforcement is effective 6uite apart from an7 u"timate gain8 as the negative uti"it7 of
gamb"ing8 abundant"7 demonstrates.
$5 Bercei/in.
'o Back
Berhaps the most difficu*t prob*em faced b- beha/iorism has been the treatment of conscious
content. Are we not a** fami*iar with co*ors) sounds) tastes) and swe**s which ha/e no
counterparts in the ph-sica* wor*dA 4hat is their p*ace in a beha/ioristic accountA 3 be*ie/e the
answer is to be found in the specia* ro*e assi.ned to stimu*i in an operant ana*-sis. 3t ca**s for a
certain amount of technica* detai*) and 3 sha** treat it in some depth.
Bercei/er or :ecei/erA
3n the traditiona* /iew a person responds to the wor*d around him in the sense of actin. upon
it. +t-mo*o.ica**-) to e,perience the wor*d is to test it) and to percei/e it is to capture it6to take it
in and possess it. 1or Krhc 'r66eks) to know was to be intimate with. A person cou*d not) of
course) capture and possess the rea* wor*d) but be cou*d make copies of it) and these
B+:C+3J3C' "1
were the so6ca**ed data6tbe .i/ens6with which) in *ieu of rea*it-) he worked. 1*e cou*d store them
in his memor- and *ater retrie/e and act upon them more or *ess as *ie mi.ht ha/e done when
the- were first .i/en.
(he opposin. /iew 666 common) 3 be*ie/e) to a** /ersions of beha/iorism6is that the initiatin.
action is taken b- the en/ironment rather than b- the percei/er. (he ref*e, was a conspicuous
e,amp*e) and a stimu*us6response /ersion of beha/iorism kept to the same pattern) as did
information theor- and some computer mode*s. A part of the en/ironment entered the bod-) was
transformed there) perhaps was stored) and e/en tua**- emer.ed as a response. Curious*- enou.h)
this differed from the menta*istic picture on*- with respect to the initiator of action. 3n both
theories the en/ironment penetrated the bod-: in the menta*istic /iew) it was taken in b- the
percei/er8 in the stimu*us6response /iew) it battered its wa- in. (he two formu*ations cou*d be
combined69an ima.e of the outer wor*d strikin. the retina of the e-e acti/ates a most intricate
process that resu*ts in /ision6. the transformation of the retina* ima.e into a perception.9 Both
formu*ations directed attention to the inner representation of rea*it- in its /arious
transformations. A basic =uestion cou*d be put this wa-: 4hat becomes of the stimu*usA
3n an operant ana*-sis) and in the radica* beha/iorism bui*t upon it) the en/ironment sta-s
where it is and where it has a*wa-s been66outside the bod-.
(he Stimu*us Contro* of 7perant Beha/ior
(he en/ironment affects an or.anism after) as we** as before) it responds. (o stimu*us and
response we add
AB7H( B+>AJ37:3S; "2
conse=uence) and it is not 0ust a third term in a se=uence. (he occasion upon which beha/ior
occurs) the beha/ior itse*f) and its conse=uences are interre*ated in the contin.encies of
reinforcement we ha/e a*read- e,amined. As the resu*t of its p*ace in these contin.encies) a
stimu*us present when a response is reinforced ac=uires some contro* o/er the response. 3t does
not then e*icit the response as in a ref*e,8 it simp*- makes it more probab*e that it wi** occur
a.ain) and it ma- do so in combination with other conditions affectin. probabi*it-) such as those
discussed in the precedin. chapter. A response reinforced upon a .i/en occasion is most *ike*- to
occur on a /er- simi*ar occasion) but because of a process ca**ed .enera*i2ation it ma- appear on
occasions sharin. on*- some of the same properties. 3f) howe/er) it is reinforced on*- when a
particu*ar propert- is present) that propert- ac=uires e,c*usi/e contro* throu.h a process ca**ed
discrimination.
(he ro*e of the stimu*us .i/es operant beha/ior a specia* character. (he beha/ior is not
dominated b- the current settin.) as it appeared to be in stimu*us response ps-cho*o.-8 it is not
9stimu*us6bound.9 Ce/erthe*ess) the en/ironmenta* histor- is sti** in contro*8 the .enetic
endowment of the species p*us the contin.encies to which the indi/idua* has been e,posed sti**
determentsK what he wi** percei/e.
Conditions A;ectin. 4iat 3s Seen
;an- of the issues discussed in the precedin. chapter e,tend to the stimu*us contro* of operant
beha/ior. 1or e,amp*e) perception is in a sense purposi/e or intentiona*. A person is not an
indifferent spectator soakin. up the wor*d *ike a spon.e. An ear*- ob0ection to
B+:C+3J3C' "3
Dohn @ocke5s theor- of human understandin. was that stimu*ation seemed to be co*d*-
en.ra/ed on the tabida rasa of the mind) and efforts were made to supp*ement the theor- b-
sa-in. that a person 9behe*d thin.s as *iked or dis*iked) appro/ed or disappro/ed) or p*easin. or
disp*easin.)9 or that a person 90ud.ed9 the wor*d as he percei/ed it) But e,pressions of that sort
simp*- assi.n to fancifu* inner processes what is to be found in .enetic endowment and persona*
histor-. 4e are not mere*- 9mindfu*9 of the wor*d about us8 we respond to it in idios-ncratic
wa-s because of what has happened when we ha/e been in contact with it. And 0ust as operant
conditionin. does not mean that a person 9infers what wi** happen when he acts)9 so the contro*
e,erted b- stimu*i does not mean that he 9infers what e,ists in the wor*d around him.9
3t is often pointed out that a person who has been dri/en o/er a route as a passen.er cannot
find his wa- as we** as one who has himse*f dri/en the route an e=ua* number of times. Anima*s
carried about in a .i/en settin. do not then mo/e about in it as we** as anima*s who ha/e a*read-
mo/ed about. Both ha/e been e,posed to the same /isua* stimu*i) but the contin.encies ha/e
been different. (o ask wh- the passen.er and the anima* carried about ha/e not 9ac=uired
know*ed.e of the settin.9 is to miss the point. (he- ha/e not ac=uired beha/ior under the contro*
of the settin..
4hen a stimu*us is weak or /a.ue) it is often c*ear that other conditions are affectin. the
probabi*it- that a person wi** see a thin. in a .i/en wa-. (he *o/er 9thinks he sees9 his be*o/ed
in a crowd but on*- if the /isua* stimu*us is f*eetin. or obscure. %(he effect is studied in the
*aborator- b- e,posin. a stimu*us for a /er- short time) sa-) or <ear the ed6c of the /isua*
AB7H( B+>AJ37:3S; "4
fie*d) or in faint *i.ht.& A s*i.ht noise at ni.ht is beard as a bur60ar or a mouse b- those who
respond /i.orous*- to bur.*ars or mice. @e/e* of depri/ation makes a difference8 or mistaken*-
9hears the te*ephone9 if a ca** is important).it) and the se,ua**- depri/ed see pha**uses or /a.inas
in ob0ects bearin. *itt*e .eometrica* simi*arit- to those or.ans. 3n other words) a person sees one
thin. as somethin. e*se when the probabi*it- of seein. the *atte* is hi.h and the contro* e,erted
b- the former is *ow.
(he importance of the histor- of the percei/er is c*ear when a chess master *ooks at a .ame in
pro.ress. 4hat he sees is /er- different from what is seen b- one who does not p*a- chess or
who has not p*a-ed it *on.. 1or the master) the settin. is an occasion upon which man- different
mo/es ha/e been made with .ood or bad resu*ts in .ames with which he is fami*iar) (o the
person who is 0ust *earnin. to p*a-) the settin. ma- be an occasion for a number of mo/es but
mo/es which ha/e not been much affected b- conse=uences. (o the comp*ete*- naLf/e) the
board and its pieces are a /isua* settin. to be described on*- throu.h possib*e resemb*ances to
situations in his non6chess6p*a-in. histor-) 4e reco.ni2e the importance of a histor- of
reinforcement when we undertake to make it more *ike*- that a person wi** see a particu*ar
thin.66or) in other words) that he wi** en.a.e in a particu*ar kind of seein.. 4e can present a
thin. sudden*- or conspicuous*- or in a no/e* and hence surprisin. wa-) and we can point to it if
our sub0ect has *earned to fo**ow a point 6that is) if he has *earned to beha/e effecti/e*- under
contin.encies in which a thin. indicated p*a-s an important part. But we can a*so arran.e that a
particu*ar ob0ect wi** be seen b- estab*ishin. contin.encies which can be met on*- b- respondin.
to it. (raffic si.ns are
B+:C+3J3C' "5
desi.ned to be easi*- seen) but we see them or i.nore them *ar.e*- because of the contin.ent
conse=uences. ;easures of this sort are often said to increase a person5s awareness) or to e,pand
his mind or consciousness) but the- simp*- brin. him under more effecti/e contro* of his
en/ironment.
(he structura*ists ha/e tried to e,p*ain perception in terms of the form) or confi.uration) of
what is percei/ed. 'esta*t ps-cho*o.ists ma- be said to ha/e ar.ued that certain kinds of patterns
force the or.anism to percei/e them in certain wa-s. Some i**usions) for e,amp*e) scurn
irresistib*e8 we see what we know is not rea**- there. Some e,amp*es seem to be reasonab*-
e,p*ained in terms of natura* se*ection: it is not surprisin. that when we see a bird f*- behind a
tree trunk) we beha/e as if it continued to e,ist when out of si.ht) and e/en see it mo/e from one
side to the other as we see a traffic *i.ht 0ump from red to .reen. Sma** .aps in order*- patterns
are ne.*ected with profit as we 9ne.*ect9 the b*ind spots in our e-es. 4e do not need to postu*ate
structura* princip*es to e,p*ain these characteristics. Contin.encies of reinforcement a*so
contribute to irresistib*e perceptions: a rotatin. trape2oid which refuses to appear to .o around is
made more effecti/e b- representin. it as a window frame.
+,perience Jersus :ea*it-
(he .reat differences in what is seen at different times in a .i/en settin. su..est that a
stimu*us cannot be described in pure*- ph-sica* terms. Beha/iorism is said to be at fau*t in
fai*in. to reco.ni2e that what is important is 9how the situation *ooks to a person9 or 9how a
person interprets a situation9 or 9what meanin. a situation has for a person.9 But to in/esti.ate
how a
AB7H( B+>AJ37:3S; "
situation *ooks to a person) or how he interprets it) or what meanin. it has for him) we must
e,amine his beha/ior with respect to it) inc*udin. his descriptions of it) and we can. do this on*-
in terms of his .enetic and en/ironmenta* histories. (o e,p*ain how the rea* wor*d is con/erted
into an interna* iconic representation) one authorit- has su..ested the fo**owin.: 91or perception
to .o be-ond the e/idence of the senses the brain must ha/e stored information) a**owin. it to
use a/ai*ab*e sensor- data to choose between possibi*ities deri/ed from past situations. Beha/ior
is not contro**ed direct*- b- stimu*i . . . but b- the brain5s h-potheses of what probab*- *ies in
outside space and in the immediate future.9 %(his is an e,amp*e) b- the wa-) of a current practice
of a/oidin. dua*ism b- substitutin. 9brain9 for 9mind.9 (he brain is said to use data) make
h-potheses) make choices) and so on) as the mind was once said to ha/e done. 3n a beha/ioristic
account it is the person who does these thin.s.& But we obser/e simp*- that a person responds to
a current settin. %9the e/idence of his senses9& because of his e,posure to contin.encies of which
the settin. has been a part. 4e ha/e no reason to sa- that he has stored information which be
now retrie/es in order to interpret the e/idence of his senses.
Some of the histor- re*e/ant to perception ma- ha/e occurred durin. the e/o*ution of the
species. 4hat is seen seems to 9depart from the ob0ect wor*d)9 for e,amp*e) in the i**usions
mentioned abo/e) in some of which the mind is said to 9infer and predict rea*it- from incomp*ete
data)9 but we shou*d sa- instead that because of his .enetic endowment a person responds in a
possib*- effecti/e wa- to what seem to be fra.mentar- stimu*i.
(he ps-choph-sicists ha/e most ri.orous*- e,p*ored
B+:C+3J3C' "!
the correspondence between e,perience and rea*it-. +ar*- ps-cho*o.ists) *ike 4undt and
(itchener) tried to disco/er what a person saw %or heard) fe*t) and so on& under the pure contro*
of current stimu*i) free of the effects of pre/ious e,posure. A trained obser/er was to describe his
sensations without makin. the 9stimu*us error96that is) to describe what he was *ookin. at as if
he had ne/er seen it before or cou*d ne/er ha/e *earned an-thin. about it. >e was to see a 9patch
of co*or9 rather than an ob0ect8 *ie was to ha/e a sa*t- taste rather than taste sa*t8 he was to fee*
warm rather than the warmth of the sun on his skin.
3n doin. so) he was to see the irreducib*e e*ements of menta* *ife) but e/en so) sensation
seemed different from rea*it- because chan.es in stimu*i did not produce comparab*e chan.es in
what was seen. A ps-choph-sica* function was said to represent the re*ation between the two
wor*ds8 but we cou*d sa- instead that it represents facts about the discriminati/e contro* of
stimu*i. (he position of conscious content .rew weaker when methodo*o.ica* beha/iorism)
to.ether with operationism. and *o.ica* positi/ism) =uestioned the usefu*ness) of sensations as
scientific data) and ps-cho ph-sicians then turned to the process of discrimination) as we ha/e
seen. But it was possib*e to stud- discrimination whi*e be*ie/in. in the e,istence of a wor*d of
e,perience.
1urther studies of discrimination) particu*ar*- research on the sensor- processes of anima*s)
were responsib*e for further pro.ress. 3n 1"5 C*aude Bernard had contended that 9e,perimenta*
studies of sense or.ans must be made on man because anima*s cannot direct*- account to us for
the sensations the- e,perience)9 but there is now an e*aborate) 9anima* ps-choph-sics)9 in which
stimu*us 6contro* is ana*-2ed with .reat pru
AB7H( B+>AJ37:3S; ""
cision. 3t is sti** *ike*- to be said that the e,perimenter has 9tau.ht the anima* to report what it
sees)9 but the resu*ts can be much more consistent*- formu*ated in terms of the contro* set up b-
specific contin.encies of reinforcement. 7f a** the .reat menta*istic e,p*anations) the
9understandin.9 or 9know*ed.e9 of the British empiricists has suffered the most i.nominious
fate: it has been reduced to the ph-sio*o.- of the e-e and car.
(he distinction between a ph-sica* and a menta* wor*d) most often found in 4estern cu*tures)
presumab*- arose) as in B*ato5s supposed disco/er- of the mind) in the effort to so*/e the
dimensiona* prob*em of menta* *ife8 there was not enou.h room in the bod- for the copies of the
wor*d a person seemed to possess. @ater) with the rise of science) a different kind of discrepanc-
appeared. 4ere the =ua*ities of ima.es and ideas to be found in nature at a**A (o use a we**6worn
e,amp*e) did a fa**in. tree make a noise if no one beard itA @i.ht mi.ht be a matter of corpusc*es
or wa/es) but it certain*- did not seem to be a matter of co*ors8 .reen was not a wa/e *en.th of
*i.ht. (his was not a serious prob*em for ear*- phi*osophers) who had no reason to =uestion the
fact that the- *i/ed in a wor*d of co*ors) sounds) and so on. Cor is it a prob*em to mi**ions of
peop*e toda-) who a*so be*ie/e that the- do so. Cot is it a prob*em for a beha/iorist.
(o ar.ue that *a-man and scientist are simp*- *ookin. at two aspects of the same thin. is to
miss the point) because aspect is what causes troub*e: peop*e see different thin.s when the- ha/e
been e,posed to different contin.encies of reinforcement. @ike e/er-one e*se) the scientist sees
.reen) but he a*so responds in other wa-s to the same settin.. 3t is a mistake) howe/er) to sa- that
the concepts of science are constructed from persona* sensor- e,perience. Both *a-
B+:C+3J3C' "#
man and scientist respond6in simi*ar or different wa-s) dependin. upon the contin.encies6to the
features of a .i/en settin.. %3 sha** return to the persona* know*ed.e of the scientist in Chapter #.&
(he stimu*us contro* of beha/ior is sub0ect to se/ere *imitations. 7ur .enetic endowment
restricts contro* to e*ectroma.netic radiation in the /isib*e ran.e) for e,amp*e) and to sonic
sounds) and e/en within these ran.es the e-e and car ha/e their defects. (heir fau*ts are not)
howe/er) a matter of fau*t- inferences. (he discrepancies are not in a correspondence between
e,perience and rea*it- but in stimu*us contro*. 3t is easier to make the point when rea*it- is more
comp*e,. 4hen an unfortunate war is attributed to misperception9 or a seminar is de/oted to
the 9discrepanc- between the rea*it- and perception of techno*o.ica* chan.e)9 trans*ation is
mandator-. >ow are we to percei/e the rea*it- of the war or the techno*o.ica* chan.e in order to
disco/er that it has indeed been mispercei/edA 4e are a*wa-s 9dea*in. with rea*it-)9 a*thou.h
the term must be taken to inc*ude more than a current presentation. (he important differences are
amon. beha/iors) and these in turn are e,p*ained b- differences in past contin.encies.
(he Cop- (heor-
(hose who be*ie/e that we see copies of the wor*d ma- contend that we ne/er see the wor*d
itse*f) but it is at *east e=ua**- p*ausib*e to sa- that we ne/er see an-thin. e*se. (he cop- theor-
of perception is most con/incin. with respect to /isua* stimu*i. (he- are fre=uent*- copied in
4orks of art as we** as in optica* s-stems of mirrors and *enses) and hence it is not difficu*t to
ima.ine some p*ausib*e s-stem of stora.e. 3t is much
AB7H( B+>AJ37:3S; #$
*ess con/incin. to sa- that we do not hear the sounds made b- an orchestra but rather some inner
reproduction. ;usic has tempora* patterns) and on*- recent*- ha/e copies been a/ai*ab*e which
mi.ht *end themse*/es to a menta* metaphor. (he ar.ument is who**- uncon/incin. in the fie*d
of taste and odor) where it is not eas- to ima.ine copies distin.uishab*e from the rea* thin.) and it
is se*dom if e/er made in the case of fee*in.. 4hen we fee* the te,ture of a sheet of paper) we
feet the paper) not some interna* representation. Bossib*- we do not need copies of tastes) odors)
and fee*in.s) since we are a*read- ph-sica**- intimate with them) and for presumab*- the same
reason we are said to feet interna* states *ike hun.er or an.er rather than copies.
(he troub*e is that the notion of an inner cop- makes no pro.ress whatsoe/er in e,p*ainin.
either sensor- contro* or the ps-cho*o.- or ph-sio*o.- of perception. (he basic difficu*t- was
formu*ated b- (heophrastus more than two thousand -ears a.o:
. . . with re.ard to hearin.) it is stran.e of him M+mpedoc*esN to ima.ine that he has rea**-
e,p*ained how creatures hear) when he has ascribed the process to interna* sounds and assumed
that the ear produces a sound within) *ike a be**) B- means of this interna* sound we mi.ht hear
sounds without) but how shou*d we hear this interna* sound itse*fA (he o*d prob*em wou*d sti**
confront us.
Simi*ar*-) as a modern authorit- has pointed out) it is as difficu*t to e,p*ain how we see a
picture in the occipita* corte, of the brain as to e,p*ain how we see the outside wor*d) which it is
said to represent. (he beha/ior of seein. is ne.*ected in a** such formu*ations. 3t can take its
proper p*ace on*- if attention is .i/en to other terms in the contin.encies responsib*e for stimu*us
contro*.
B+:C+3J3C' #1
Seein. in the Absence of the (hin. Seen
4hen a person reca**s somethin. he once saw) or en.a.es in fantas-) or dreams a dream)
sure*- he is not under the contro* of a current stimu*us. 3s he not then seein. a cop-A A.ain) we
must turn to his en/ironmenta* histor- for an answer. After hearin. a piece of music se/era*
times) a person ma- hear it when it is not bein. p*a-ed) thou.h probab*- not as rich*- or as
c*ear*-. So far as we know) he is simp*- doin. in the absence of the music some of the thin.s he
did in its presence. Simi*ar*-) when a person sees a person or p*ace in his ima.ination) fie ma-
simp*- be doin. what he does in the presence of the person or p*ace. Both 9reminiscin.9 and
9rememberin.9 once meant 9bein. mindfu* of a.ain9 or 9brin.in. a.ain to mind96in other words)
seein. a.ain as one once saw. +,p*icit techni=ues of 9ca**in. to mind9 are techni=ues of
stren.thenin. perceptua* beha/ior) as we sha** see in Chapter !.
Beha/iorism has been accused of 9re*e.atin. one of the paramount concerns of the ear*ier
ps-cho*o.ists the stud- of the ima.e6to a position of not 0ust ne.*ect) but dis.race.9 3 be*ie/e) on
the contrar-) that it offers the on*- wa- in which the sub0ect of ima.in. or ima.inin. can be put
in .ood order.
Seein. in the absence of the thin. seen is fami*iar to a*most e/er-one) but the traditiona*
formu*ation is a metaphor) 4e tend to act to produce stimu*i which are reinforcin. when seen. 3f
we ha/e found the cit- of Jenice reinforcin. %we refer to one reinforcin. effect when we ca** it
beautifu*&) we ma- .o to Jenice in order to be thus reinforced. 3f we cannot .o) we ma- bu-
pictures of Jenice6rea*istic pictures in co*or of
AB7H( B+>AJ37:3S; #2
its most beautifu* aspects) a*thou.h a b*ack6and6white sketch ma- be enou.h. 7r we ma- see
Jenice b- readin. about it if we ha/e ac=uired the capacit- to /isua*i2e whi*e readin..
%(echno*o.- has made it much easier to see reinforcin. thin.s in their presence and hence has
reduced the chance to see them in their absence. (wo or three .enerations a.o a chi*d read) or
was read to) from books with few or no i**ustrations8 toda- he watches te*e/ision or reads books
with co*ored pictures on e/er- pa.e) and he is therefore much *ess *ike*- to ac=uire a repertoire
of seein. under the contro* of /erba* stimu*i.& 4ith no e,terna* support whatsoe/er) we ma-
simp*- 9see Jenice9 because we are reinforced when we do so. 4e sa- that we da-dream about
Jenice. (he mistake is to suppose that because we create ph-sica* stimu*i which enab*e us to see
Jenice more effecti/e*- b- .oin. to Jenice or bu-in. a picture) we must therefore create menta*
stimu*i to be seen in memor-. A** we need to sa- is that if we are reinforced for seein. Jenice)
we are *ike*- to en.a.e in that beha/ior 6that is) the beha/ior of seein. Jenice6e/en when there
is /er- *itt*e in the immediate settin. which bears a resemb*ance to the cit-. Accordin. to one
dictionar-) fantas- is defined as 9the. act or function of formin. ima.es or representations in
direct perception or in memor-)9 but we cou*d sa- as we** that it is the act or function of seein.
in direct perception or in memor-.
4e ma- a*so see a thin. in its absence) not because we are immediate*- reinforced when we
do so) but because we are then ab*e to en.a.e in beha/ior which is subse=uent*- reinforced.
(hus) we ma- see Jenice in order to te** a friend how to find his wa- to a particu*ar part of the
cit-. 3f we were to.ether in the cit- itse*f) we mi.ht take him a*on. a .i/en route) but we
B+:C+3J3C' #3
can 9take ourse*/es a*on. the route /isua**-9 when we are not there and describe it to him. 4e
can do so more effecti/e*- b- pointin. to a map or a sketch of the route) but we do not consu*t a
9co.niti/e map9 when we describe what we see in 9ca**in. the cit- to mind.9 Onowin. a cit-
means possessin. the beha/ior of .ettin. about in it8 it does not mean possessin. a map to be
fo**owed in .ettin. about. 7ne ma- construct such a map from the actua* cit- or b- seein. the
cit- when absent from it) but /isua*i2in. a route throu.h a cit- in order to describe it to a friend
is seein. as %not what& one sees in .oin. throu.h the cit-.
C*aude Bernard mi.ht a*so ha/e said that it is impossib*e) to .et anima*s to report the thin.s
the- are ima.inin.) but there is no reason wh- the contin.encies under which a person sees
thin.s which are not there shou*d not be effecti/e with other species. 3t is possib*e to .et anima*s
to respond to after6ima.es) and b- increasin. depri/ation we can induce a pi.eon to respond to a
s=uare 9as if it were a trian.*e.9 (here is no reason wh-) with such measures) we cou*d not .et it
to respond to a b*ank surface when it has pre/ious*- been reinforced on*- when the surface had a
trian.*e pro0ected upon it. (he desi.n of 9/erba*9 contin.encies which wou*d permit it to te** us
that it 9saw9 a trian.*e wou*d be an interestin. e,ercise.
A person is chan.ed b- the contin.encies of reinforcement under which he beha/es8 be does
not store the contin.encies) 3n particu*ar) he doesn5t store copies of the stimu*i which ha/e p*a-ed
a part in the contin.encies. (here are no 9iconic representations9 in his mind8 there are no 9data
structures stored in his memor-98 he has no 9co.niti/e map9 of the wor*d in which fie has *i/ed.
>e has simp*- been chan.ed in such a
AB7H( B+>AJ37:3S; #4
wa- that stimu*i now contro* particu*ar kinds of perceptua* beha/ior.
Seein. in the absence of the thin. seen is most dramatica**- e,emp*ified in dreamin. when
as*eep. Current stimu*ation is then minima**- in contro*) and a person5s histor- and resu*tin.
states of depri/ation and emotion .et their chance. 1reud emphasi2ed the si.nificance of wishes
and fears p*ausib*- inferred from dreamin.) but unfortunate*- he was responsib*e for
emphasi2in. the distinction between seein. and what is seen. (he dreamer en.a.ed in dream
work8 he sta.ed the dream as a theatrica* producer sta.es a p*a- and then took his p*ace in the
audience and watched it. But dreamin. is perceptua* beha/ior) and the difference between
beha/ior when as*eep and when awake) either in or out of a re*e/ant settin.) is simp*- a
difference in the contro**in. conditions.
:apid e-e mo/ements durin. steep seem to confirm this interpretation. 4hen most acti/e*-
dreamin.) peop*e mo/e their e-es about as if the- were obser/in. a /isua* presentation. %(he
midd*e6car musc*es a*so seem to mo/e durin. dreams in/o*/in. auditor- perception.& 3t has been
ar.ued that e-e mo/ement) as we** as car musc*e mo/ement) show that 9ph-sio*o.ica* input9
affects dreamin.) but such beha/ior is =uite c*ear*- a ph-sio*o.ica* output. 4e can scarce*-
suppose that the iconic representations obser/ed in dreamin. are under the e-e*ids or in the outer
ear.
(here are man- wa-s of .ettin. a person to see when there is nothin. to be seen) and the- can
a** be ana*-2ed as the arran.ement of contin.encies which stren.then perceptua* beha/ior.
Certain practices in beha/ior therap-) in which the patient is asked to ima.ine /arious conditions
or e/ents) ha/e been critici2ed as not .enuine*- beha/iora* because the- make use of
B+:C+3J3C' #5
ima.es. But there are no ima.es in the sense of pri/ate copies) there is perceptua* beha/ior8 and
the measures taken b- the ps-chotherapist are desi.ned to stren.then it. A chan.e takes p*ace in
the patient5s beha/ior if what he sees %hears) fee*s) and so on& has the same positi/e*- or
ne.ati/e*- reinforcin. effect as if he were seein. the thin.s themse*/es. 3t is se*dom if e/er
enou.h simp*- to instruct the patient to 9ha/e fee*in.s)9 to ask him to fee* se,ua**- e,cited or
nauseated) but he ma- be shown porno.raphic or nauseatin. materia* or be asked to 9/isua*i2e as
c*ear*- as possib*e9 a se,ua* or dis.ustin. episode.
(hat a person ma- see thin.s when there is nothin. to be seen must ha/e been a stron. reason
wh- the wor*d of the mind was in/ented. 3t was hard enou.h to ima.ine bow a cop- of the
current en/ironment cou*d .et into the head where it cou*d be 9known)9 but there was at *east a
wor*d outside which mi.ht account for it. But pure ima.es seem to indicate a pure mind stuff. 3t
is on*- when we ask how either the wor*d or a cop- of the wor*d is seen that we *ose interest in
copies. Seein. does not re=uire a thin. seen.
;ind and Stimu*us Contro*
4e saw in Chapter 4 that the word 9mind9 is sometimes a mere s-non-m for the person who
acts. 3t can a*so stand for the person who percei/es. 4hen a person is out of touch with rea*it-)
his mind is said to be wanderin.) or possib*- absent. (he /erb 9to mind9 often means simp*- to
respond. 4e warn someone to mind the *ow cei*in.) meanin. simp*- that he shou*d see and
respond to it. 3n this sense we ask someone to mind the chi*dren) and he ma- comp*ain that the
chi*dren do not mind him.
AB7H( B+>AJ37:3S; #
;ind is a*so sometimes simp*- the p*ace in which thin.s are seen) (hin.s 9come to mind9 or
are 9ca**ed to mind)9 and one who is sufferin. a de*usion ma- be to*d that 9it5s a** in -our mind)9
as distinct from bein. in the rea* wor*d. As the p*ace in which thin.s are percei/ed) mind is
c*ose*- associated with cop- theor- and was an important part of the ps-cho*o.- of conscious
content. 4hen operationism *ed to the stud- of the process of discrimination rather than of
sensations) a person was re.arded as *ookin. at or *istenin. to the rea* wor*d. >e was no *on.er
reportin. his perceptions or sensations8 he was reportin. stimu*i. (he wor*d was back where it
be*on.ed.
(he issue is critica* when we turn to the difference between seein. a thin. and seein. that one
is seein. it. 3f there) are no copies of thin.s inside the bod- at an- time) then A that can be seen
introspecti/e*- is the act of seein.) and this is what one reports when asked) 9<o -ou see thatA9 3t
is S7 possib*e) howe/er) to discriminate between thin.s which are there or not there to be seen. 3
cou*d be said to know that this sheet of paper is rea**- there because 3 pick up a pen and write on
it) and that the bri.ht after6ima.e which bothers me is not there because 3 do not tr- to brush it
awa-. 3 ha/e *earned the difference between two kinds of seein.. (he thirst- man does not reach
for the fantasi2ed .*ass of water) but the dreamer does not know that what he is seein. is 9not
rea**- there)9 and he responds as fu**- as a person who is as*eep can. %3ntrospecti/e know*ed.e of
dreamin. is weak or *ackin. because the conditions needed for se*f6obser/ation are *ackin.) and
when such se*f6know*ed.e sur/i/es) into the wakin. state) it usua**- disappears =uick*- as one
for.ets one5s dreams.& 3t is a*so possib*e to know that -ou ha/e seen somethin. before. 4e re6
co.ni2e what we ha/e once
B+:C+3J3C' #!
co.ni2ed. 3n a de0a /u this feature of se*f6know*ed.e is defecti/e.
7ther kinds of se*f6know*ed.e about stimu*us contro* become a/ai*ab*e when we ana*-2e the
contin.encies which contro* our beha/ior.
K 5erbal Behavior
Go Back

6elatively late in its history$ the human species uner"ent a remarkable
chan#e: its vocal musculature came$ uner operant control. Aike other species$ it ha
up to that point isplaye "arnin# cries$ threatenin# shouts$ an other innate
responses$ but vocal operant behavior mae a #reat ifference because it e)tene
the scope of the social environment$ Aan#ua#e "as born$ an "ith it many important
characteristics of human behavior for "hich a host of mentalistic e)planations have
been invente.

*he very ifference bet"een 8lan#ua#e8 an 8verbal behavior8 is an e)ample.
Aan#ua#e has the character of a thin#$ somethin# a person acquires an possesses.
%sycholo#ists speak of the 8acquisition of lan#ua#e8 in the chil. *he "ors an
sentences of "hich a. lan#ua#e is compose are sai to be tools use to e)press
meanin#s$ thou#hts$ ieas$ propositions$ emotions$ nees$ esires$ an many other
thin#s in or on the speaker.s min. ! much more prouctive vie" is that

VERBA5 BEHAVIOR ,,

verbal behavior is behavior. It has a special character only because it is reinforce by
its effects on people&&at first other people$ but eventually the speaker himself. !s a
result$ it is free of the spatial$ temporal$ an mechanical relations "hich prevail
bet"een operant behavior an nonsocial consequences. If the openin# of a oor "ill
be reinforcin#$ a person may #rasp the knob$ turn it$ an push or pull in a #iven "ay$
but if$ instea$ fie says$ 8%lease open the oor$8 an a listener respons appropriately$
the same reinforcin# consequence follo"s. *he contin#encies are ifferent$ an they
#enerate many important ifferences in the behavior "hich have lon# been obscure
by mentalistic e)planations.

(o" a person speaks epens upon the practices of the verbal community of
"hich he is a member. ! verbal repertoire may be ruimentary or it may isplay an
elaborate topo#raphy uner many subtle kins of stimulus control. *he contin#encies
"hich shape it "ay be inul#ent @as "hen parents respon to their chilren.s crue
appro)imations to stanar formsE or emanin# @as in the teachin# of ictionE.
=ifferent verbal communities shape an maintain ifferent lan#ua#es in the same
speaker$ "ho then possesses ifferent repertoires havin# similar effects upon
ifferent listeners. 5erbal responses are classifie as requests$ commans$ per&
missions$ an so on$ epenin# upon the reasons "hy the listener respons$ the
reasons often bein# attribute to the speaker.s intentions or moos. *he fact that the
ener#y of a. response is not proportional to the ma#nitue of the result has
contribute to the belief in verbal ma#ic @the ma#ician.s 8%resto chan#o8 converts a
hankerchief into a rabbitE. Stron# responses appear in the absence of an appropriate
auience$ as 6ichar III emonstrate "hen he crie$ 8! horseL a horseL 3y

ABOUT BEHAVIORISM166

kin#om for a horseL8 althou#h there "as no one to hear him.

!part from an occasional relevant auience$ verbal behavior requires no
environmental support. Cne nees a bicycle to rie a bicycle but not to say 8bicycle.8
!s a result$ verbal behavior can occur on almost any occasion. !n important
consequence is that most people fin it easier to say 8bicycle8 silently than to 8rie a
bicycle silently.8!notber important consequence is that the speaker also becomes a
listener an may richly reiaforce his o"n behavior.

'eaning and Reerence

*he term 8meanin#$8 thou#h closely associate "ith verbal behavior$ has been use to
make some of the istinctions alreay iscusse. *hose "ho have confuse
behaviorism "ith structuralism$ in its emphasis on form or topo#raphy$ have
complaine that it i#nores meanin#. What is important$ they conten$ is not "bat a
person is oin# but "hat his behavior means to him; his behavior has a eeper
property not unrelate to the purpose$ intention$ or e)pectation iscusse in Chapter
M. But the meanin# of a response is not in its topo#raphy or form @that is the mistake
of the structuralist$ not the behavioristE; it is to be foun in its anteceent history.
*he behaviorist is also accuse of escribin# the environmental settin# in physical
terms an overlookin# "hat it means to the responin# person$ but here a#ain the
meanin# is not in the current settin# but in a history of e)posure to contin#encies in
"hich similar settin#s have playe a part.

In other "ors$ meanin# is not properly re#are as a propcrty either of a
response or a situation but rather of the contin#encies responsible for both the topo#

VERBA5 BEHAVIOR 161

raphy of behavior an the control e)erte by stimuli. *o take a primitive e)ample$ if
one rat presses a lever to obtain foo "hen hun#ry "hile another oes so to obtain
"ater "hen thirsty$ the topo#raphies of their behaviors may be inistin#uishable$ but
they may be sai to iffer in meanin#: to one rat pressin# the lever 8means8 foo; to
the other it 8means8 "ater. But these are aspects of the contin#encies "hich have
brou#ht behavior uner the control of the current occasion. Simila*ly$ if a rat is
reinforce "ith foo "hen it presses the lever in the presence of a flashin# li#ht but
"ith "ater "hen the li#ht is steay$ then it coul be sai that the flashin# li#ht
means foo an the steay li#ht means "ater$ but a#ain these are references not to
some property of the li#ht but to the contin#encies of "hich the li#hts have been
parts.

*he same point may be mae$ but "ith many more implications$ in speakin# of
the meanin# of verbal behavior. *he over&all function of the behavior is crucial. In an
archetypal pattern a speaker is in contact "ith a situation to "hich a listener is
ispose to respon but "ith "hich lie is not in contact. ! verbal response on the part
of the speaker makes it possible for the listener to respon appropriately. ?or
e)ample$ let us suppose that a person has an appointment$ "hich he "ill keep by
consWtin# a clock or a "atch. If none is available$ he may ask someone to tell him the
time$ an the response perinits him to respon effectively. *he speaker sees the clock
an announces the time; the listener hears the announcement an keeps his
appointment. *he three terms "hich appear in the contin#encies of reinforcement
#eneratin# an operant arc ivie bet"een t"o people: the speaker respons to the
settin#$ an the listener en#a#es in the behavior an is affecte by the consequences.
*his "ill happen only if the be

ABOUT BEHAVIORISM 162

haviors of speaker an listener are supporte by aitional contin#encies arran#e by
the verbal community.

*he listencr.s belief in "hat the speaker says is like the belief "hich unorlies
the probability of any response @8I believe this "ill "ork8E or the control e)erte by
any stimulus @8I believe this is the ri#ht place8E. It epens on past contin#encies$ an
nothin# is #aine by internali-in# them. *o efine interpersonal trust as 8$in
e)pectancy hel by an iniviual or a #roup that the "or$ promise$ verbal or "rittcn
statement of another iniviual or #roup can be relie on8 is to complicate matters
unnecessarily.

*he "eanin7 o8 a res#onse 8or the s#eaker inclues the stimulus "hich controls
it @in the e)ample above$ the settin# on the face of a clock or "atchE an possibly
aversive aspects of the question$ from "hich a responsc brin#s release. *he "eanin7
8or the listener is close to the meanin# the clock face "oul have if it "ere visible to
him$ but it also inclues the contin#encies involvin# the appointment$ "hich make a
response to the clock face or the verbal response probable at such a time. ! person
"ho "ill leave for an appointment upon seein# a certain position of the hans of a
clock "ill also leave upon hearin# a response mae by a person "hose responses in
the past have been accu*ately controlle by the position of the hans an "hich for
that reason control stron# responses no".

Cne of the unfortunate h&uplications of communication theory is that the
meanin#s for speaker an listener are the same$ that somethin# is mae common to
both of them$ that the speaker conveys an iea or meanin#$ transmits information$ or
imparts kno"le#e$ as if his mental possessions then become the mental possessions
of the listener. *here are no meanin#s "hich are the

VERBA5 BEHAVIOR 160

same in the speaker an listener. 3eanin#s are not inepenent entities. We may look
for the meanin# of a "or in the ictionary$ but ictionaries o not #ive meanin#s; at
best they #ive other "ors havin# the same meanin#s. We must come to a ictionary
alreay 8provie "ith meanin#s.8

! referent mi#ht be efine as that aspect of the cavironment "hich e)erts
control over the response of "hich it is sai to be the referent. It oes so because of
the reinforcin# practices of a verbal community. In traitional terms$ meanin#s an
referem&s are not to be foun in "ors but in the circumstances uner "hich "ors
are use by speakers an unerstoo by listeners$ but 8use8 an 8unerstanin#8 nee
further analysis.

5erbal responses are often sai to be taken by the listener as si#ns$ or symbols$
of the situations they escribe$ an a #reat eal has been mae of the symbolic
process$ some e)amples of "hich "e shall consier in the follo"in# chapter. Certain
atmospheric conitions may be a 8si#rt of rain$8 an "e respon to them to avoi
#ettin# "et. We usually respon in a sli#htly ifferent "ay in escapin# from the rain
ii.self if "e have ha no si#n of it in avance. We can say the same thin# about the
"eatherman.s verbal responses$ "hich are no more a si#n or symbol of rain than the
atmospheric chan#e.

Meta#hor. We have seen that a stimulus present "hen a response is reinforce
acquires some control ever the probability that that response "ill occur$ an that this
effect #enerali-es: stimuli shaiin# some of its properties also acquire some control. In
verbal behavior one kin of response evoke by a merely similar stimulus is calle a
metaphor. *he response is not transferre from one situation to another$ as the
etymolo#y

ABOUT BEHAVIORISM 164

mi#ht su##est; it simply occurs because of a similarity in stimuli. (avin# come to say
8e)ploe8 in connection "ith firecrackers or bombs$ a person may escribe a frien
"ho suenly behaves in a violent manner as 8e)ploin#y in an#er.8 Cther fi#ures of
speech illustrate other behavioral processes.

!bstraction. ! characteristic feature of verbal behavior$ irectly attributable to
special contin#encies of reinforcement$ is abstraction. It is the listener$ not the
speaker$ "ho takes practical action "ith respect to the stimuli controllin# a verbal
response$ an as a result the behavior of the speaker may come uner the control of
properties of a stimulus to "hich no practical response is appropriate. ! person learns
to react to re thin#s uner the nonsocial contin#encies of his environmcnt$ but he
oes so only by emittin# a practical response for each re thin#. *he contin#encies
cannot brin# a sin#le response uner the control of the property of reness alone. But
a sin#le property "ay be important to the listener "ho takes many kins of practical
action on many ifferent occasions because of it an "ho therefore reinforces
appropriately "hen a #iven ob'ect is calle re. *he referent for re can never be
ientifie in any one settin#. If "e sho" a person a re pencil an say$ 8What is that?8
an he says$ 86e$8 "e cannot tell "hat property evoke his response$ but if "e sho"
him many re ob'ects an he al"ays says$ 86e$8 "e can o so&an "ith increasin#
accuracy as "e multiply cases. *he speaker is al"ays responin# to a physical ob'ect$
not to 8reness8 as an abstract entity$ an he respons 8re8 not because he possesses
a concept of reness but because special contin#encies have brou#ht that response
uner the control of that property of stimuli.
VERBA5 BEHAVIOR 165

*here is no point in askin# ho" a person can 8kno" the abstract entity calle
reness.8 *he contin#encies e)plain the behavior$ an "e nee not be isturbe
because it is impossible to iscover the referent in any sin#le instance. We nee not$
"ith William of Cckham an the :ominalists$ eny that abstract entities e)ist an
insist that such responses are merely "ors. What e)ist arc the contin#encies "hich
brin# behavior uner the control of properties or of classes of ob'ects efine by
properties. @We can etermine that a sin#le response is uner the control of one
property by namin# it. ?or e)ample$ if "e sho" a person a pencil an say$ 8What color
is this?8 he "ill then respon to the property specifie as color&provie he has been
sub'ect to an appropriate history of reinforcement.E

-on*e#ts. When a class is efine by more than one property$ the referent is
usually calle a concept rather than an abstract entity. *hat concepts have real
referents has been pointe out by sayin# that 8they are iscoveries rather than
inventions&they represent reality.8 In other "ors$ they e)ist in the "orl before
anyone ientifies them. But iscovery @as "ell as inventionE su##ests mental action in
the prouction of a concept. ! concept is simply a feature of a set of contin#encies
"hich e)ist in the "orl$ an it is iscovere simply in the sense that the
contin#encies brin# behavior uner its control$ *he statement 8Scientific concepts
enable certain aspects of the enormous comple)ity of the "orl to be hanle by
men.s mins8 is vastly improve by substitutin# 8human bein#s8 for 8men.s mins.8

ABOUT BEHAVIORISM 16

Sentences and (ropositions

*he traitional notion of meanin# an referent runs into trouble "hen "e be#in
to analy-e lar#er verbal responses uner the control of more comple) environmental
circumstances. What are the referents of sentences&not to mention para#raphs$
chapters$ or books? ! sentence surely means more than its separate "ors mean.
Sentences o more than refer to thin#s; they say thin#s. But "hat are the thin#s they
say? ! traitional ans"er is 8%ropositions.8 But propositions are as elusive as meanin#s.
Bertran 6ussell.s vie" has been paraphrasc as follo"s: 8*he si#nificance of a
sentence is that "hich is common to a sentence in one lan#ua#e an its translation
into another lan#ua#e. ?or e)ample$ .I am hun#ry. an ./9ai faim9 have in common
elements "hich constitute the si#nificance of a sentence. *his common element is the
proposition.8 But "hat is this common element? Where is it to be foun? ! ictionary
that #ave the meanin#s of sentences "oul simply contain other sentences havin# the
same meanin#s.

! translation can bcst be efine as a verbal stimulus that has the same. effect
as the ori#inal @or as much of the same effect as possibleE on a ifferent verbal com&
munity. ! ?rench translation of an 4n#lish book is not another statement of a set of
propositions; it is another sample of verbat behavior havin# an effect upon a ?rench
reaer similar to the effect of the 4n#lish vevsion on an 4n#lish reaer. *he same
interpretation may be mae of a translation from one meium into another. It has
been sai that the %relue to *ristan and Isolde is 8an astonishin#ly intense an
faithful translation into music of the emotions whi*h a**o"#an3 the union of a pair of
lovers.8 6ather than try to ientify the feelin#$ let alone the proposition$ "hich is thus
translate$ "e

VERBA5 BEHAVIOR 16%

may say simply that the music has somethin# of the same effect as physical union.

*he concepts of e)pression an communication may be treate in a shnilar
"ay. ! speaker or a listener respons to conitions of his boy "hich he has learne
to call feelin#s$ but "hat he says or hears is behavior$ ue to contin#encies of "hich
the felt conitions may be by&proucts. *o say that music e)presses 8"hat is
ine)pressible in co#nitive$ an especially in scientific$ lan#ua#e8 is to say that it has
an effect that verbal behavior cannot have. 5erbal behavior oes not communicate
feelin#s$ thou#h it may result in Conitions similarly felt. It oes not communicate
propositions or instructions. *o 8instruct8 a mother cat to esert her youn# by
eliverin# an electric shock to a part of her brain oes not communicate an
instruction that "as first hel in the min of the scientist; the shock simply has an
effect @a ash of col "ater "oul have prouce the same resultE. 5on ?risch.s
account of the lan#ua#e of bees @an account "hich is becomin# increasin#ly suspectE
i not make him a Champollion$ reain# a 6osetta stone.

*he concept of stimulus control replaces the notion of referent "ith respect
not only to responses "hich occur in isolation an are calle "ors @such as nouns an
a'ectivesE but also to those comple) responses calle sentences. %ossibly 8fact8 coul
be sai to escribe a referent of the latter$ althou#h its su##estion of truth versus
falsity raises ifficulties. *he chil respons in sentences to events in his
environment&&events$ involvin# more than one property or thin#$ or relations amon#
thin#s$ or relations of actor an acte upon$ an so on$ an his responses contain
elements "hich he never has any occasion to emit alone. *he lin#uist assi#ns these
elements to synta) or N*ammar. (e oes so

ABOUT BEHAVIORISM 16)

as part of an analysis of the practices of a #iven verbal community$ from "hich be
e)tracts rules "hich may be use in the construction of ne" sentences$ as "e shall
see in Chapter J.

The 'anipulation o Words and Sentences

Structuralism has been stron#ly encoura#e in lin#uistics because verbal
behavior often seems to have an inepenent status. We are incline to #ive special
attention to its form because "e can report it easily$ an rather accurately$ simply by
moelin# it$ as in a irect quotation. *he report 8(e sai$ .hammer. 8 #ives a much
more complete escription of the topo#raphy of his behavior than 8Ile "as
hammerin#.8 In teachin# a chil to talk$ or an ault to pronounce a ifficult "or$ "e
prouce a moel&that is$ "e say the "or an arran#e contin#encies uner "hich a
response havin# similar properties "ill be reinforce. *here is nothin# especially
verbal about moelin# @in teachin# sports or the ance$ the instructor 8sho"s a person
"hat to o8 in the sense of oin# it hirnselfE$ but "ith the invention of the alphabet$
it became possible to recor verbal behavior$ an the recors$ free of any supportin#
cnvironment$ seeme to have an inepenent e)istence. ! speaker is sai to 8kno"8 a
poem or an oath or a prayer. 4arly eucation in China an Greece "as lar#ely a
matter of memori-in# literary "orks. *he stuent seeme to kno" the "isom
e)presse by the "ork$ even thou#h his behavior "as not necessarily uner the
control of the conitions "hich inuce the ori#inal speaker or "riter$ or an infon&ne
listener$ to respon in a #iven "ay.

5erbal behavior has this kin of
inepenent status

VERBA5 BEHAVIOR 16,

"hen it is in transmission bet"een speaker an listener &for e)ample$ "hen it is the
8information8 passin# over a telephone "ire or bet"een "riter an reaer in the form
of a te)t. ;ntil fairly recently$ lin#uistics an literary criticism confine themselves
almost e)clusively to the analyses of "ritten recors. If these ha any meanin#$ it "as
the meanin# for the reaer$ since the circumstances uner "hich the behavior ha
been prouce by the "riter ha been for#otten$ if they "ere ever kno"n.

*he availability of verbal behavior in this apparently ob'ective form has cause
a #reat eal of trouble. By iviin# such recors into "ors an sentences "ithout
re#ar to the conitions uner "hich the behavior "as emitte$ "e ne#lect the
meanin# for the speaker or "riter$ an almost half the fiel of verbal behavior
therefore escapes attention. Worse still$ bits of recore speech are move about to
compose ne" 8sentences$8 "hich are then analy-e for their truth or falsity @in terms
of their effect on a reaer or listenerE$ althou#h they "ere never #enerate by a
speaker. Both lo#ician an lin#uist ten to create ne" sentences in this "ay$ "hich
they then treat as if they "ere the recors of emitte verbal behavior. If "e take the
sentence 8*he sun is a star8 an put the "or 8not8 in the proper place$ "e transform
it into 8*he sun is not a star8 but no one has emitte this instance of a verbal
response$ an it oes not escribe a fact or e)press a proposition. It is simply the
result of a mechanical process.

%erhaps there is no harm in playin# "ith sentences in this "ay or in analy-in#
the kins of transformations "hich o or o not make senter0ces acceptable to the
orinary reaer$ but it is still a "aste of time$ particularly "hen the sentences thus
#enerate coul not have been emitte as verbal behavior. ! classical e)ample is

ABOUT BEHAVIORISM 116

a parao)$ such as 8*his sentence is false$8 "hich appears to be true if false an false
if true. *he important thin# to consier is that no one coul ever have emitte the
sentence as verbal behavior. ! sentence must be in e)istence before a speaker can
say$ 8*his sentence is false$8 an the response itself "ill not serve$ since it i not
e)ist until it "as emitte. What the lo#ician or lin#uist calls a sentence is not
necessarily verbal behavior in any sense "hich calls for a behavioral analysis.

*he transformational rules "hich #enerate sentences acceptable to a listener
may be of interest$ but even so it is a mistake to suppose that verbal behavior is #en&
erate by them. *hus$ "e may analy-e the behavior of small chilren an iscover
that$ for e)ample$ part of their speech consists of a small class of 8moifiers8 an a
lar#er class of 8nouns$8 @*his fact about verbal behavior is ue to the contin#encies of
reinforcement arran#e by most verbal communities.E It oes not follo" that the chil
8forms a noun phrase of a #iven type8 by 8selectin# first one "or from the small class
of moifiers an selectin# secon one "or from the lar#e class of nouns.8 *his is a
lin#uist.s reconstruction after the fact.

*he analysis of verbal behavior$ particularly the socalle iscovery of #rammar$
came very late. ?or thousans of years no one coul have kno"n he "as speakin#
accorin# to rule. What happens "hen rules are iscovere "ill be consiere in
Chapter J.

9evelo#"ent. !n unue concern for the structure of verbal behavior has
encoura#e the metaphor of evelopment or #i&o"th. Aen#th of utterance is plotte
as a function of a#e$ an semantic an #rammatical features are observe as they
8evelop$8 *he #ro"th of lan#ua#e in a chil is easily compare "ith the #ro"th

VERBA5 BEHAVIOR 111

of an embryo$ an #rammar can then be attribute to rules possesse by the chil at
birth. ! pro#ram in the form of a #enetic coe is sai to 8initiate an #uie early
learnin# . . . as a chil acquires lan#ua#e.8 But the human species i not evolve
because of an inbuilt esi#n& it evolve throu#h selection uner contin#encies of
survival$ as the chil.s verbal behavior evolves uner the selective action of
contin#encies of reinforcement. !s I have note$ the "orl of a chil evelops$ too.

! chil oes seem to acquire a verbal repertoire at an ama-in# spee$ but "e
shoul not overestimate the accomplishment or attribute it to invente lin#uistic
capacities. ! chil may 8learn to use a ne" "or8 as the effect of a sin#le
reinforcement$ but it teams to o nonverbal thin#s "ith comparable spee. *he
verbal behavior is impressive in part because the topo#raphy is conspicuous an easily
ientifie an in part because it su##ests hien meanin#s.
If the structuralists an evelopmentalists ha not confine themselves so narro"ly
to the topo#raphy of behavior at the e)pense of the other parts of the contin#cncies
of reinforcement$ "e shoul kno" much more about bo" a chil learns to speak. We
kno" the "ors a chil first uses an the characteristic orers in "hich they ten to
be use. We kno" the len#th of utterances at #iven a#es$ an so on. If structure "ere
enou#h$ that "oul be the "hole story. But a recor of topo#raphy nees to be
supplemente by an equally etaile recor of the conitions uner "hich it "as
acquire. What speech has the chil hear? ;ner "hat circumstances has he hear
it? What cffects has he achieve "hen be has uttere similar responses? ;ntil "e have
this kin of information$ the success or failure of any analysis of verbal behavior
cannot be 'u#e.

ABOUT BEHAVIORISM 112

"reative )erbal Behavior

in verbal behavior$ as in all operant behavior$ ori#inal forms of response are evokc
by situations to "hich a person has not previously been e)pose. *he ori#in of
behavior is not unlike. the ori#in of species. :e" combinations of stimuli appear in
ne" settin#s$ an responses "hich escribe them may never have been mae by the
speaker before$ or hear or rea by him in the speech of others. *here are many
behavioral processes #eneratin# 8mutations$8 "hich are then sub'ect to the selective
action of contin#encies of reinforcement. "e all prouce novel forms&for e)ample$ in
neolo#isms$ blens$ portmanteau "ors$ "itty remarks involvin# istortion$ an the
mistakes of hasty speech.

! #reat eal has been mae of the fact that a chil "ill 8invent8 a "eak past
tense for a stron# verb$ as in sayin# 8he #oe8 instea of 8he "ent.8 If he has never
hear the form 8#oe8 @that is$ if he has associate only "ith aultsE$ he must have
create a ne" form. But "e o not speak of 8creation8 if$ havin# acquire a list of
color "ors an a list of ob'ect "ors$ he for the first time says 8purple automobile.8
*he fact that the terminal 8&e8 su##ests 8#rammar8 is unnecessarily e)citin#. It is
quite possible that it is a separable operant$ as a separate inicator of the past tense
or of complete action in another lan#ua#e mi#ht be$ an that 7#o8 an a terminal
8&e8 are put to#ether$ as 8purple8 an 8automobile8 are put to#ether$ on a novel
occasion. *he so&calle creativc aspect of verbal behavior "ill be mentione a#ain
later.
& Thin'ing
'o Back

In menta"istic formu"ations the ph7sica" environment is moved into the mind and
becomes e:perience. Behavior is moved into the mind as purpose8 intention8 ideas8 and
acts of 0i"".. )erceiving the 0or"d and profiting from e:perience become
>genera"-purpose cognitive activities8> and abstract and conceptua" thin$ing has
sometimes been said to have no e:terna" reference at a"". Kiven such 0e"" -estab"ished
precedents8 it is not surprising that certain remaining behaviora" functions shou"d a"so
be moved inside. Tota" interna"iFation 0as recent"7 announced b7 three cognitive
ps7cho"ogists 0ho8 upon comp"eting a boo$8 are said to have dec"ared themse"ves
>sub;ective behaviorists.>

In this chapter I consider a number of behaviora" processes 0hich have given rise
to the invention of 0hat are -usua""7 ca""ed higher menta" processes. The7 compose one
great part of the fie"d of thinIdng. It is a difficu"t fie"d8 and no one8 so far as I $no08
c"aims to give a definitive account. The present ana"7sis is short

ABOUT BEHAVIORISM //4

of perfection for another reason@ it must be brief. But if a behavioristic interpretation of
thin$ing is not a"" 0e shou"d "i$e to have8 it must be remembered that menta" or
cognitive e:p"anations are not e:p"anations at a"".

>Thin$ing> often means >behaving 0ea$"78> 0here the 0ea$ness ma7 be due8 for
e:amp"e8 to defective stimu"us contro". #ho0n an ob;ect 0ith 0hich 0e are not ver7
fami"iar8 0e ma7 sa78 >I thin$ it is a $ind of 0rench8> 0here >I thin$> is c"ear"7 opposed to
>I $no0.> !e report a "o0 probabi"i?7 for a different reason 0hen 0e sa78 >I thin$ I sha""
go8> rather than >I sha"" go> or >I $no0 I sha"" go.>

There are more important uses of the term. !atching a chess game8 0e ma7
0onder >0hat a p"a7er is thin$ing of> 0hen he ma$es a move. !e ma7 mean that 0e
0onder 0hat he 0i"" do ne:t. In other 0ords8 0e 0onder about his incipient or inchoate
behavior. To sa78 >Ge 0as thin$ing of moving his roo$8> is perhaps to sa78 >Ge 0as on
the point of moving it8> 5sua""78 ho0ever8 the term refers to comp"eted behavior 0hich
occurs on a sca"e so sma"" that it cannot be detected b7 others8 #uch behavior is ca""ed
covert. The commonest e:amp"es are verba"8 because verba" behavior re6uires no
environmenta" support and because8 as both spea$er and "istener8 a person can ta"$ to
himse"f effective"7A but nonverba" behavior ma7 a"so be covert. Thus8 0hat a chess
p"a7er has in mind ma7 be other moves he has made as he has p"a7ed the game
covert"7 to test the conse6uences.

Covert behavior has the advantage that 0e can act 0ithout committing ourse"vesA
0e can revo$e the behaviorand tr7 again if private conse6uences are not reinforcing. (It
is usua""7 on"7 0hen behavior has been emitted8 b7 the 0a78 that one spea$s of an act
of 0i""A

THIN5IN6 //5

the term suggests ta$ing a stand and accepting the irrevocab"e conse6uences.) Covert
behavior is a"most a"0a7s ac6uired in overt form8 and no one has ever sho0n that the
covert form achieves an7thing 0hich is out of reach of the overt. Covert behavior is a"so
easi"7 observed and b7 no means unimportant8 and it 0as a mista$e for methodo"ogica"
behaviorism and certain versions of "ogica" positivism and structura"ism to neg"ect it
simp"7 because it 0as not >ob;ective.> It 0ou"d a"so be a mista$e not to recogniFe its
"imitations. It is far from an ade6uate substitute for traditiona" vie0s of thin$ing. It does
not e:p"ain overt behavior@ it is simp"7 more behavior to be e:p"ained.

The present argument is this@ menta" "ife and the 0or"d in 0hich it is "ived are
inventions. The7 have been invented on the ana"og7 of e:terna" behavior occurring
under e:terna" contingencies. Thin$ing is behaving. The mista$e is in a""ocating the
behavior to the rnind. #evera" e:amp"es sho0ing ho0 this has been done ma7 be
considered.

The ()ognitive( )ontrol of #ti$!li

The ancient vie0 that perception is a $ind of capturing or ta$ing possession of the
0or"d is encouraged b7 the rea" distinction 0e ria$e bet0een seeing and "oo$ing at8
hearing and "istening to8 sme""ing and sniffing8 tasting and savoring8 and fee"ing and
fee"ing of8 0here the second term in each pair does indeed refer to an act. It is an act
0hich ma$es a stimu"us more effective. B7 sniffing8 for e:amp"e8 0e thro0 air against
the surfaces containing the sense organs of sme""8 and as a resu"t 0e can detect an
odor 0e might other0ise miss8 !e a"so act to reduce stimu"ationA 0e s6uint or shut our
e7es8

ABOUT BEHAVIORISM //"

p"ug our cars8 spit8 ho"d our breath8 or pu"" our hand a0a7 from a painfu" ob;ect. #ome of
these >precurrent8> or preparator78 behaviors are part of our genetic endo0mentA others
are produced b7 contingencies of reinforcement.

A rather simi"ar process can be demonstrated as fo""o0s@ A hungr7 pigeon is
occasiona""7 reinforced 0ith food 0hen it pec$s a circu"ar dis$ on the 0a"" of an e:-
perimenta" chamber. If it is reinforced 0hen the dis$ is red but not 0hen it is green8 it
eventua""7 stops pec$ing 0hen the dis$ is green. 5nfortunate"7 for the pigeon8 the co"or
0ashes out and becomes difficu"t or impossib"e to detect. The pigeon can strengthen
the co"or b7 pec$ing another dis$8 ho0ever8 and it 0i"" do so as "ong as the co"or
remains important. The production of additiona" stimu"i favoring a discriminative
response is a fami"iar part of science. In testing the acidit7 of a so"ution8 for e:amp"e8
another so"ution is added8 and if the co"or changes in a specified 0a78 the acidit7 can be
determined.

Ana"ogous menta" or cognitive activities have been invented. !e attend to a
stimu"us or ignore it 0ithout changing an7 ph7sica" condition (for e:amp"e8 0e can "isten
to a particu"ar instrument in recorded music8 in part b7 suppressing our responses to the
other instruments)8 and 0e are said to do so 0ith various menta" mechanisms. .adio
and te"evision are presumab"7 responsib"e for the current metaphor of >tuning the 0or"d
in or out.> An o"der metaphor8 resemb"ing 1a:0e""Bs Eemon in the second "a0 of
thermod7namics8 portra7s a $ind of gate$eeper-a "o7a" servant 0ho admits 0anted
stimu"i and defends his master against un0anted. It has been said to be >conceivab"e
that the nervous s7stem actua""7 s0itches off one eat in order to "isten to the ot"ier.> !e
have not e:p"ained an7thing8

THIN5IN6 //#

of course8 unti" 0e have e:p"ained the behavior of the gate$eeper8 and an7 effort to do
so 0i"" suffice to e:p"ain the change in stimu"us contro".

!hat is invo"ved in attention is not a change of stimu"us or of receptors but the
contingencies under"7ing the process of discrimination. !e pa7 attention or fai" to pa7
attention to a "ecturer or a traffic sign depending upon 0hat has happened in the past
under simi"ar circumstances. Eiscrimination is a behaviora" process@ the contingencies8
not the mind8 ma$e discriminations. !e sa7 that a person discerns or >ma$es out> an
ob;ect in a fog or at a great distance in the sense that he eventua""7 responds to it
correct"7. Eiscern8 "i$e discriminate8 ma7 mean an act favoring a response (it ma7 be
c"oser to >"oo$ at> than to >see>)8 but it need not be. !e discern the important things in a
given setting because of past contingencies in 0hich the7 have been important.

Abstracting and forming concepts are "i$e"7 to be ca""ed cognitive8 but the7 a"so
invo"ve contingencies of reinforcement. !e do not need to suppose that an abstract
entit7 or concept is he"d in the mindA a subt"e and comp"e: histor7 of reinforcement has
generated a specia" $ind of stimu"us contro". It is common"7 said that concepts >unif7 our
thoughts8> but the evidence seems to be that the7 simp"7 enab"e us to ta"$ about
features of the 0or"d common to a "arge assortment of instances8 'ne scientist has said
that >there is e:ce""ent reason to be"ieve that the 0ho"e of chemistr7 is e:p"icab"e in
terms of e"ectrons and the 0ave functions 0hich describe their "ocation. This is an
enormous simp"ification of thought.> It is certain"7 an enormous simp"i fication-or 0ou"d
be8 if feasib"e-but it is the simp"ification of verba" and practica" behavior rather than of
thought. The same 0riter has said that concepts are

ABOUT BEHAVIORISM //&

>discoveries as 0e"" as-indeed8 more than-inventions> and that the7 are >an e:ercise of
the human mind 0hich represents rea"it78> but be confesses that the nature of the
re"ationship is a m7ster7. It is the m7ster7 of the abstract entit7 rather than of the
avai"ab"e facts. The referents of concepts are in the rea" 0or"dA the7 are not ideas in the
mind of the scientist. The7 are discoveries or inventions simp"7 in the sense that a
verba" environment has evo"ved in 0hich obscure properties of nature are brought into
the contro" of human behavior. It is probab"7 too "ate to trace the emergence of concepts
such as mass8 energ78 or temperature8 even 0ith the he"p of the historian of science8
and their current use is perhaps as difficu"t to ana"7FeA but nothing is gained b7 putting
them in the mind of the scientist.

An e:amp"e from a popu"ar artic"e on p"ace "earning sho0s ho0 troub"esome it is
to e:p"ain behavior b7 inventing a concept instead of b7 pursuing contingencies.
Chi"dren 0ho have been taught to comp"ete the e:pressions >3 N +> b7 sa7ing >-C are
then sho0n >+ N 3> >'ne chi"d is hope"ess"7 puFF"ed8 another readi"7 ans0ers B-.B It is
c"ear that the t0o pupi"s have "earned different things@ the first chi"d has "earned a
specific ans0er to a specific 6uestionA the second has "earned an arithmetica" concept.>
But 0hat does this te"" us4 Can 0e be sure that the second chi"d has not a"so been
taught to sa7 >-> to >+ N 3> at some other time4 Gas he perhaps "earned a "arge number
of instances such as >" N O N I> and >I N 3 O 3 N 1>4 Gas he "earned to state the ru"e
of commutation and to e:emp"if7 it4 If 0e are content to spea$ of an arithmetica"
concept8 0e sha"" never find out 0hat the chi"d has actua""7 "earned.

THIN5IN6 //'

#earch and Recall

Another so-ca""ed cognitive activit7 0hich affects a personBs contact 0ith contro""ing
stimu"i is search. To "oo$ for something is to behave in 0a7s 0hich have been re-
inforced 0hen something has turned up. !e sa7 that a hungr7 anima" moves about
"oo$ing for food. The fact that it is active8 and even the fact that it is active in particu"ar
0a7s8 ma7 be part of its genetic endo0ment8 e:p"ained in turn b7 the surviva" va"ue of
the behavior8 but the 0a7 in 0hich an organism "oo$s for food in a fami"iar environment
is c"ear"7 dependent upon its past successes. !e te"" a chi"d to find his shoe8 and the
chi"d starts to "oo$ in p"aces 0here shoes have been found.

There are8 ho0ever8 more specia"iFed strategies of "oo$ing for things. !hat does
one do to find an ob;ect in a bo: of rubbish (>scrutiniFe> comes from an e:pression
having to do 0ith the sorting out of trash) or on the she"ves of a 0arehouse4 Go0 does
one go about finding a 0ord on a page or findirig and crossing out a"" the aBs in a co"umn
of print4 The s$i""fu" searcher moves about8 sorts out materia"s8 and moves his e7es in
0a7s 0hich ma:imiFe the chances of finding things and mini miFe the chances of
missing8 and "ie does so because of past contingencies. !e have no reason to ca"" the
behavior cognitive8 but a rather simi"ar process is said to ta$e p"ace in the 0or"d. of the
mind.

<or various reasons8 suggested b7 such terms as >memorandum... memento8>
>souvenir8> and >memoria"8> peop"e have made copies of the 0or"d around them8 as 0e""
as records about 0hat has happened in that 0or"d8 and have stored them for future use.
<ami"iar e:amp"es are scratches on c"a7 tab"ets8 engraved "egends on monuments8
boo$s8 paintings8 photographs8 phonographic recordings8 and the magnetic
ABOUT BEHAVIORISM /2(

stores of computers. 'n a future occasion such a record can evo$e behavior
appropriate to an ear"ier occasion and ma7 permit a person to respond more effective"7.
The practice has "ed to the e"aboration of a cognitive metaphor8 no doubt antedating b7
centuries an7 ps7cho"ogica" s7stem-ma$ing8 in 0hich e:periences are said to be stored
in memor78 "ater to be retrieved or reca""ed and used in order to behave more effective"7
in a current setting.

4hat is said to be stored are copies of stimu*ifaces) names) dates) te,ts) p*aces) and so
on6which when retrie/ed ha/e some of the effect of the ori.ina*s. (he copies cannot ha/e the
dimensions of the ori.ina*s8 the- must be transduced and encoded6possib*- as en.rams)
re/erberatin. circuits) or e*ectrica* fie*ds. Stora.e is particu*ar*- hard to ima.ine for the memor-
of a musica* composition or a stor-) which has tempora* properties. Ce/erthe*ess) a** these thin.s
are said to 9reside9 in memor-.

But 0hat is the menta" para""e" of ph7sica" search4 Go0 are 0e to go about finding
an item in the storehouse of memor74 )"ato raised a fundamenta" 6uestion@ >A man
cannot in6uire either about that 0hich "ie $no0s or about that 0hich be does not $no0A
for if he $no0s he has no need to in6uireA and it not8 he cannot8 for he does not $no0
the ver7 sub;ect about 0hich he is to in6uire.> <or >in6uire> reach >search.> If 0e can
remember a name8 0e have no need to search our memor7A if 0e cannot remember it8
ho0 do 0e go about "oo$ing for it4 The cognitive ps7cho"ogist ta"$s about various
s7stems of access borro0ed from the fi"ing s7stems of "ibraries8 computers8
0arehouses8 posta" s7stems8 and so on. Thus8 the superior retrieva" of certain $inds of
items is attributed to >an addressing s7stem that a""o0s immediate access to itcms>-as it
certain"7 shou"dJ
THIN5IN6 /2/

In a behaviora" ana"7sis probabi"it7 is substituted for accessibi"it7. The
contingencies 0hich affect an organism are not stored b7 it. The7 are never inside itA
the7 simp"7 change it. As a resu"t8 the organism behaves in specia" 0a7s under specia"
$inds of stimu"us contro". <uture stimu"i are effective if the7 resemb"e the stimu"i 0hich
have been part of ear"ier contingenciesA an incidenta" stimu"us ma7 >remind> us of a
person8 p"ace8 or event if it has some resemb"ance to that person8 p"ace8 or event.
BBeing reminded means being made "i$e"7 to respond8 possib"7 perceptua""7. A name
ma7 remind us of a person in the sense that 0e no0 see him. This does not mean
con;uring up a cop7 of the person 0hich 0e then "oo$ at- it simp"7 means behaving as
0e behaved in his presence upon some ear"ier occasion. There 0as no cop7 of his
visua" appearance inside us then8 as there is none no0. The incidenta" stimu"us does
not send us off in search of a stored cop78 0hich 0e perceive ane0 0hen 0e find it.

The e:tensive e:periments b7 cognitive ps7cho"ogists on accessibi"it7 can a"" be
reinterpreted in terms of probabi"it7. If fami"iar 0ords are more 6uic$"7 reca""ed than
strange ones8 it is because the7 have a greater initia" probabi"it78 due to the histor7
a""uded to b7 the 0ord >fami"iar.> !e do not need to conc"ude that >the 0ord store has a
form of organiFation 0hich a""o0s 6uic$er access to the more common"7 re6uired items
than to the rarer ones8>

Techni6ues of reca"" are not concerned 0ith searching a storehouse of memor7
but 0ith increasing the probabi"it7 of responses. 1nemonics are pre-"earned or easi"7
"earned behaviors 0hich prompt or other0ise strengthen the behavior to be reca""ed. If
0e have forgotten the ne:t part of a piece of music 0e are p"a7ing or a poem 0e are
reciting8 0e go bac$ for a running
ABOUT BEHAVIOMSM /22

start8 not because the music or poem has been stored as a unit of memor78 so that one
part he"ps us find the other part8 but because the e:tra stimu"ation 0e generate in the
running start is sufficient to evo$e the forgotten passage. In reca""ing a name it is usefu"
to go through the a"phabet8 not because 0e have stored a"" the names 0e $no0 in
a"phabetica" order but because pronouncing the sound of a "etter is pronouncing part of
the nameA 0e prompt the response in ourse"ves as 0e prompt it in someone e"se 0hom
0e are he"ping to reca"" it. !hen8 in reca""ing a name8 0e find a 0rong name too
po0erfu"8 it is not because the 0rong name a8 mas$s the target> in our storehouse of
memor7 but because it is repeated"7 emitted to the e:c"usion of the name 0e are
reca""ing. Techni6ues of "earning to observe in such a 0a7 that one remembers more
readi"7 are not techni6ues of storage but rather of generating effective perception. The
artist "oo$ing at a scene 0hich he 0i"" "ater s$etch 0i"" to some e:tent s$etch it as he
"oo$s8 thus strengthening the $ind of behavior 0hich 0i"" be important to him "ater.

The metaphor of storage in memor78 0hich has seemed to be so dramatica""7
confirmed b7 the computer8 has caused a great dea" of troub"e. The computer is a bad
mode"-as bad as the c"a7 tab"ets on 0hich the metaphor 0as probab"7 first based. !e
do ma$e e:terna" records for future use8 to supp"ement defective contingencies of
reinforcement8 but the assumption of a para""e" inner record-$eeping process adds
nothing to our understanding of this $ind of thin$ing. (It is not the behaviorist8
incidenta""78 but the cognitive ps7cho"ogist8 0ith his computer-mode" of the mind8 0ho
represents man as a machine.)

THIN5IN6 /23

#olving roble$s

'ther so-ca""ed cognitive processes have to do 0ith so"ving prob"ems. It is a fie"d
mar$ed b7 a great dea" of m7ster78 part of it due to the 0a7 in 0hich it has been
formu"ated. )rob"ems need to be so"ved8 0e are to"d8 because a person needs >to orient
himse"f in an infinite"7 comp"e: rea"it78 to order the end"ess particu"arit7 of e:perience8 to
find essences behind facts8 to attach meaning to being-in-the-0or"d.> <ortunate"78 a
much simp"er statement is possib"e. A person has a prob"em 0hen some condition 0i""
be reinforcing but he "ac$s a response that 0i"" produce it. Ge 0i"" so"ve the prob"em
0hen he emits such a response. <or e:amp"e8 introducing someone 0hose name one8
has forgotten is a prob"em 0hich is so"ved b7 reca""ing or other0ise "earning the name.
An a"gebraic e6uation is so"ved b7 finding the va"ue of :. The prob"em of a sta""ed car is
so"ved b7 starting the car. The prob"em of an i""ness is so"ved b7 finding an effective
treatment. #o"ving a prob"em is8 ho0ever8 more than emitting the response 0hich is the
so"utionA it is a matter of ta$ing steps to ma$e that response more probab"e8 usua""7 b7
changing the environment. Thus8 if the prob"em is to sa7 0hether t0o things are the
same or different8 0e ma7 put them side b7 side to faci"itate a comparisonA if it is to
ma$e sure that 0e sha"" treat them as different8 0e separate them. !e group simi"ar
things in c"asses in order to treat them in the same 0a7. !e put things in order if the
so"ution re6uires a series of steps. !e restate a verba" response b7 trans"ating it from
0ords into s7mbo"s. !e represent the premises of a s7""ogism 0ith over"apping circ"es.
!e c"arif7 6uantities b7 counting and measuring. !e confirm a so"ution b7 so"ving a
prob"em a second time8 possib"7 in a different 0a7.
ABOUT BEHAVIORISM /24

!e "earn some of these strategies from the prob"ematic contingencies to 0hich 0e
are e:posed8 but not much can be "earned in a sing"e "ifetime8 and an important function
of a cu"ture is to transmit 0hat others have "earned. !hether prob"em so"ving arises
from ra0 contingencies or from instruction b7 others8 it is ac6uired in overt form (0ith
the possib"e e:ception of a strateg7 "earned at the covert "eve" from private conse-
6uences) and can a"0a7s be carried out at the overt "eve". The covert case8 to 0hich the
term >thin$ing> is most "i$e"7 to be app"ied8 en;o7s no specia" advantage be7ond that of
speed or confidentia"it7.

7*oce- A prob"em to 0hich a good dea" of attention has been given arises 0hen
t0o or more responses appear to be possib"e and a person chooses or decides among
them. The prob"em is to escape from indecision rather than to discover an effective
response. !e faci"itate choosing or ma$ing a decision in various 0a7s-for e:amp"e8 b7
>revie0ing the facts.> If 0e are 0or$ing 0ith e:terna" materia"s8 verba" or other0ise8 0e
ma7 indeed re-vie0 them in the sense of "oo$ing at them again. If8 ho0ever8 0e are
0or$ing covert"78 0e do not recover the facts8 as if 0e 0ere pu""ing papers out of a fi"eA
0e mere"7 see them again. In revie0ing an argument 0e simp"7 argue again.
.e-vie0ing is not re-ca""ing8 since a"" the facts to be used are avai"ab"e8

It is said that a person has made a choice 0hen he has ta$en one of t0o or more
seeming"7 possib"e courses of action. The troub"e "ies in the 0ord possib"e. #imp"7 to
ma$e one of severa" >possib"e> responses-as in 0a"$ing aim"ess"7 through a
par$-re6uires no serious act of decision8 but 0hen conse6uences are important and the
probabi"ities of t0o or more responses are near"7 e6ua"8 a prob"em must be so"ved. A
person
THIN5IN6 /25

usua""7 so"ves it and escapes from indecision b7 changing the setting.

To sa7 that >humans can ma$e choices and desire to do so> simp"7 means that a
situation in 0hich t0o or more responses are about e6ua""7 probab"e ma7 be aversive8
and that an7 decision-ma$ing behavior 0hich strengthens one response and ma$es the
other un"i$e"7 is reinforced. To sa7 that >humans re6uire freedom to e:ercise the
choices the7 are capab"e of ma$ing> adds further comp"ications. To e:ercise a choice is
simp"7 to act8 and the choice a person is capab"e of ma$ing is the act itse"f. The person
re6uires freedom to ma$e it simp"7 in the sense that he can ma$e it on"7 if there are no
restraints--either in the ph7sica" situation or in other conditions affecting his behavior.

It is eas7 to over"oo$ the behavior 0hich actua""7 so"ves a prob"em. In one
c"assica" account8 a chimpanFee seemed to have fitted t0o stic$s together in order to
ra$e in a banana 0hich 0as other0ise out of reach through the bars of his cage. To sa7
that the chimpanFee sho0ed >inte""igent behavior based on a perception of 0hat 0as
re6uired to so"ve the prob"em. some 0a7 of overcoming the distance barrier> is to ma$e
it a"most impossib"e to discover 0hat happened. To so"ve such a prob"em a chimpanFee
must have "earned at "east the fo""o0ing@ to stop reaching for a banana out of reachA to
stop reaching 0ith short stic$sA to discriminate bet0een "ong and short stic$s8 as b7
using "ong stic$s to ra$e in bananas successfu""7A to pic$ up t0o stic$s in separate
handsA and to thrust stic$s into ho"es. !ith this preparation8 it is not impossib"e that in
that rare (but poor"7 authenticated) instance the chimpanFee stuc$ one stic$ into the
ho"e at the end of another and used the resu"ting "ong stic$ to ra$e in the banana8

The importance of the behaviora" ana"7sis is c"ear

ABOUT BEHAVIORISM /2"

0henever 0e underta$e to do an7thing about prob"em so"ving. To teach comparab"e
behavior to a chi"d8 for e:amp"e8 0e shou"d at some time or other have to emphasiFe a""
these ingredients. It is doubtfu" 0hether 0e cou"d ma$e much progress b7 impressing
the chi"d 0ith >the need to overcome a distance barrier.>

)reative Behavior

The creative mind has never been 0ithout its prob"ems8 as the c"assica"
discussion in )"atoBs Meno suggests. It 0as an inso"ub"e prob"em for stimu"us-response
ps7cho"og7 because if behavior 0ere nothing but responses to stimu"i8 the stimu"i might
be nove" but not the behavior. 'perant conditioning so"ves the prob"em more or "ess as
natura" se"ection so"ved a simi"ar prob"em in evo"utionar7 theor7. As accidenta" traits8
arising from mutations8 are se"ected b7 their contribution to surviva"8 so accidenta"
variations in behavior are se"ected b7 their reinforcing conse6uences.

That chance can p"a7 a part in the production of an7thing as important as
mathematics8 science8 or art has often been 6uestioned. 1oreover8 at first g"ance8 there
seems to be no room for chance in an7 comp"ete"7 determined s7stem8 The Church8 in
its be"ief in a predestined master p"an8 censured 1ontaigne for using 0ords "i$e fortune
and nature8 and if #aint Augustine sought heaven"7 counse" b7 opening his Bib"e and
reading the first 0ords that met his e7es8 it 0as on"7 because the7 did not meet his e7es
b7 chance. Another deterministic s7stem8 ps7choana"7sis8 has initiated another age in
0hich chance is tabooA for the strict <reudian8 no one can forget an appointment or ca""
a person b7 the 0rong name or ma$e a s"ip of the tongue b7 chance.

THIN5IN6 /2#

Pet the biographies of 0riters8 composers8 artists8 scientists8 mathematicians8 and
inventors a"" revea" the importance of happ7 accidents in the production of origina"
behavior.

The concept of se"ection is again the $e7. The mutations in genetic and
evo"utionar7 theor7 are random8 and the topographies of response se"ected b7
reinforcement are8 if not random8 at "east not necessari"7 re"ated to the contingencies
under 0hich the7 0i"" be se"ected. And creative thin$ing is "arge"7 concerned 0ith the
production of >mutations.> 2:p"icit 0a7s of ma$ing it more "i$e"7 that origina" behavior
0i"" occur b7 introducing =mutations> are fami"iar to 0riters8 artists8 composers8
mathematicians8 scientists8 and inventors. 2ither the setting or the topograph7 of
behavior ma7 be de"iberate"7 varied. The painter varies his co"ors8 brushes8 and
surfaces to produce ne0 te:tures and forms. The composer generates ne0 rh7thms8
sca"es8 me"odies8 and harmonic se6uences8 sometimes through the s7stematic
permutation of o"der forms8 possib"7 0ith the he"p of mathematica" or mechanica"
devices. The mathematician e:p"ores the resu"ts of changing a set of a:ioms. The
resu"ts ma7 be reinforcing in the sense that the7 are beautifu" or8 in most of
mathematics and in science and invention8 successfu".

Dove" verba" responses are "i$e"7 to be generated b7 discussion8 not on"7
because more than one histor7 of reinforcement is then active but a"so because
different histories ma7 b7 accident or design "ead to nove" settings. The so-ca""ed histor7
of ideas offers man7 e:amp"es. In the eighteenth centur7 in <rance the "eaders of the
2n"ightenment borro0ed a good dea" from 2ng"ish 0riters-in particu"ar8 Bacon8 ?oc$e8
and De0ton. As one author has put it8 >2ng"ish thoughts in <rench
ABOUT BENAVIORISM /2&

heads produced in the "ong run some astonishing and e:p"osive conse6uences.> The
sentence is intentiona""7 metaphorica"8 of course8 and mi:es the menta" (>thoughts>) 0ith
the anatomica" (>heads>)8 but it ma$es the va"id point that trans"ations from 2ng"ish into
<rench that are then read b7 peop"e 0ith ver7 differeut verba" histories ma7 generate
nove" responses.

The #tr!ct!re of %ind

The structure of thought and the deve"opment of the mind have8 of course8 been
popu"ar themes for centuries. As 0e sha"" see in the ne:t t0o chapters8 there are certain
ob;ective states of $no0"edge8 but thought processes are behaviora"8 and a structura"ist
account is necessari"7 incomp"ete if it neg"ects genetic and persona" histories. The
deve"opment of thin$ing has been most often described 0ith horticu"tura" metaphors.
The gro0th of the mind is a centra" figure. The teacher is to cu"tivate the mind as a
farmer cu"tivates his fie"ds8 and the8 inte""ect is to be trained as a vine is trained in a
vine7ard. 1ean0hi"e the deve"opment of the 0or"d to 0hich a thin$ing person is
e:posed is over"oo$ed.

!e have noted that those 0ho stud7 the >deve"opment of "anguage> in the chi"d
te"" us much about vocabu"ar78 grammar8 and "ength of sentences but ver7 "itt"e about
the hundreds of thousands of occasions upon 0hich a chi"d hears 0ords and sentences
spo$en or the man7 thousands of times he himse"f spea$s them 0ith resu"ts8 and that
no ade6uate account of the >deve"opment of "anguage> is therefore possib"e. !e ma7
sa7 the same thing for the gro0th of the mind. The behavior 0hich is said to indicate the
possession of the concept of inertia and the age at 0hich it norma""7 appears are no
doubt important facts8 but 0e shou"d a"so

THIN5IN6 /2'

$no0 something about the man7 thousands of occasions upon 0hich a chi"d has
pushed8 pu""ed8 t0isted8 and turned things in >deve"oping> that concept.

In the absence of an7 ade6uate account of the deve"opment or gro0th of a
personBs e:posure to an environment8 the a"most inevitab"e resu"t is that important
aspects of thin$ing are assigned to genetic endo0ment. Dot on"7 is verba" behavior said
to sho0 the operation of innate ru"es of grammar8 but >innate ideas such as siFe8 shape8
motion8 position8 number8 and duration> are said to >give form and meaning to the
confused fragmentar7 data that 0e e:perience ever7 da7 in our "ives.> #iFe8 shape8
motion8 position8 number8 and duration are features of the en%$on2en!- The7 have pre-
vai"ed "ong enough and behavior 0ith respect to them has been crucia" enough to ma$e
the evo"ution of appropriate behavior possib"e8 but contingencies of reinforcement are at
0or$ ever7 da7 in the "ife of the individua" to generate supp"ementar7 behavior under
the contro" of the same features. The greatest achievements of the human species (not
of the human mind) have occurred too recent"7 to ma$e a genetic e:p"anation de-
fensib"e8 but 0hether 0e appea" to contingencies of surviva" or contingencies of
reinforcement 0e can at "east dispense 0ith the appea" to innate ideas. It ma7 be true
that there is no structure 0ithout construction8 but 0e must "oo$ to the constructing
environment8 not to a constructing mind.

The Thin'ing %ind

1ind is said to )"a7 an important ro"e in thin$ing. it is sometimes spo$en of as the p"ace
0here thin$ing occurs8 0here one image8 memor78 or idea "eads to another in a >stream
of consciousness.> It can be empt7
ABOUT BEHAVIORISM /3(

or fi""ed 0ith factsA it can be ordered or chaotic. >1athematics8> sa7s a prestige
advertisement of a te"ephone compan78 >happens in the mind. . . 8 It is essentia""7 a
thing of the mind8 for it 0or$s through concepts8 s7mbo"s8 and re"ationships.> #ometimes
the mind appears to be the instrument of thin$ingA it can be $een or du""8 mudd"ed b7
a"coho" or c"eared b7 a bris$ 0a"$. But usua""7 it is the thin$ing agent. It is the mind
0hich is said to e:amine sensor7 data and ma$e inferences about the outside 0or"d8 to
store and retrieve records8 to fi"ter incoming information8 to put bits of information in
pigeonho"es8 to ma$e decisions8 and to 0i"" to act.

In a"" these ro"es it has been possib"e to avoid the prob"ems of dua"ism b7
substituting >brainC for >mind.> The brain is the p"ace 0here thin$ing is said to ta$e
p"aceA it is the instrument of thin$ing and ma7 be $een or du""A and it is the agent 0hich
processes incoming data and stores them in the form of data structures. Both the mind
and the brain are not far from the ancient notion of a homuncu"us-an inner person 0ho
behaves in precise"7 the 0a7s necessar7 to e:p"ain the behavior of the outer person in
0hom he d0e""s.

A much simp"er so"ution is to identif7 the mind 0ith the person. Guman thought is
human behavior. The histor7 of human thought is 0hat peop"e have said and done.
1athematica" s7mbo"s are the products of 0ritten and spo$en verba" behavior8 and the
concepts and re"ationships of 0hich the7 are s7mbo"s are in the environment. Thin$ing
has the dimensions of behavior8 not of a fancied inner process 0hich finds e:pression in
behavior.

!e are on"7 ;ust beginning to understand the effects of comp"e: contingencies of
reinforcement8 but if our ana"7sis of the behavior ca""ed thin$ing is sti"" defective8 the
facts to be treated are neverthe"ess re"ative"7 c"ear

THIN5IN6 /3/

cut and accessib"e. In contrast8 the 0or"d of the mind is as remote toda7 as it 0as 0hen
)"ato is said to have discovered it. B7 attempting to move human behavior into a 0or"d
of nonph7sica" dimensions8 menta"istic or cognitive ps7cho"ogists have cast the basic
issues in inso"ub"e forms. The7 have a"so probab"7 cost us much usefu" evidence8
because great thin$ers (0ho presumab"7 $no0 0hat thin$ing is) have been "ed to report
their activities in sub;ective terms8 focusing on their fee"ings and 0hat the7
introspective"7 observe 0hi"e thin$ing8 and as a resu"t the7 have fai"ed to report
significant facts about their ear"ier histories.
* "auses and Reasons
Go Back

Some important kins of thinkin# remain to be consiere. *he behavior
iscusse in the last chapter is the prouct of contin#encies of reinforcement; it is
"hat happens "hen$ in a #iven environmental settin#$ behavior has certain kins of
consequences. *he so&calle intellectual life of the min uner"ent an important
chan#e "ith the avent of verbal behavior. %eople be#an to talk about "hat they
"ere oin# an "hy they "ere oin# it$ they escribe their behavior$ the settin# in
"hich it occurre$ an the consequences. In other "ors$ in aition to bein#
affecte by contin#encies of reinforcement$ they be#an to analy-e them.

"ommands Advice+ and Warnings

Cne of the first verbal practices of this sort must have been #ivin# orers or
commans. 83ove over8 e

-AUSES A:9 REASO:S 100

scribes an act an implies a consequence: the listener is to move over&or elseL *he
speaker tells the listener "hat he is to o an arran#es aversive consequences uner
"hich he learns to o it an to o it a#ain "henever the orer is repeate. ! "arnin#
usually iffers from an orer or comman because the aversive consequences are not
arran#e by the person "ho issues it: 8Watch outL8 escribes an act an implies a
consequence$ such as escape from a fallin# rock$ but the latter is a natural result of
the behavior rather than one contrive by the speaker$ !vice @8<o West$ youn#
manL8E specifies behavior an implies positively reinforcin# consequences "hich are
also riot contrive by the aviser @8. . . an you "ill make your fortune8E. ! person
hees "arnin#s an takes avice epenin# upon "hat has happene in similar
circumstances in the past. !s in Chapters I an K$ the probability that he "ill respon
coul be calle a measure of his trust or belief in the speaker or in "hat the speaker
says.

Directions and Instructions

Cne person #ives another irections by notin# or implyin# a reinforcin# consequence$
by escribin# behavior havin# that consequence$ an especially by escribin# the
controllin# environment: 8*o #et to Boston$ follo" 6oute 12 to the intersection "ith
6oute M1I$ turn left on 6oute 1D . . .8 =irections for operatin# a venin# machine
escribe$ a series of acts to be unertaken in orer: 8*o operate$ place coin in slot
an pull plun#er beneath item "ante.8 =irections o not impart kno"le#e or convey
information: they escribe behavior to be e)ecute an state or imply consequences.

Instructions are esi#ne to make
further irection

ABOUT BEHAVIORISM 104

unnecessary. ! person learnin# to rive a car respons to the verbal behavior of the
person sittin# besie him; he starts$ stops$ shifts$ si#nals$ an so on "hen tol to o
so. *hese verbal stimuli may at first be irections$ but they become instruction if
verbal help is #iven only as neee. *he river.s behavior is then eventually taken
over by the natural$ nonverbal contin#encies of rivin# a car. *o learn to rive simply
throu#h e)posure to those contin#encies "oul take a very lon# time. *he "oul&be
river "oul have to iscover "hat happens "hen he moves the shift lever$ turns the
"heel$ presses on the accelerator$ applies the brake$ an so on$ an all "ith #reat
an#er to himself. By follo"in# instructions$ he avois e)posure to many of these con&
tin#encies an eventually behaves as the instructor himself behaves.

*he instructor has not 8communicate8 his kno"le#e or his e)perience to the
learner. *he final uninstructe behavior is shape an maintaine by the natural
contin#encies of car an hi#h"ay. *he instructor has mae it possible for the learner
to come uner their control speeily an "ithout harm.

3uch of eucation is instruction in verbal behavior. *he stuent is tol ho" to
8use "ors8 rather than ho" to use an accelerator or brake. In neither case is he #iven
kno"le#e; he is tol ho" to behave. *he instruction #iven by a labele picture often
"orks very quickly; the vie"er kno"s at once "hat an ob'ect is calle an "hat the
label means. ! efinition is a seemin#ly more internal form of instruction$ but its
effect is simply that one verbal response is no" use interchan#eably "ith another.

-AUSES A:9 REASO:S 105

Folklore+ 'a,ims+ and (roverbs

Some$ forms of instruction can be transmitte from #eneration to #eneration because
the contin#encies they escribe are IonL &lastin#. Such a ma)im as 8*o lose a frien$
len him money8 escribes behavior @lenin# moneyE an a consequence @losin# a
frienE. We mi#ht translate Aa 6ochefoucaul9s 8Self&esteem is the #reatest of
flatterers8 in this "ay& 8We are more likely to say #oo thin#s about ourselves than
about others$ an "hat "e say is more likely to be a matter of pleasin# the listener
than of reportin# the facts.8 *he craftsman.s rules of thumb are part of folklore an
may become permanent features of a culture if they make it easier to teach or
remember the behavior they escribe. ?olklore$ ma)ims$ an proverbs are often
especially effective because many of the avanta#es of the behavior they stren#then
are lon# eferre an o not function "ell as reinforcers.

When social contin#encies characteristic of a small$ slo"ly chan#in# community
are isturbe$ formal irections nee to be invoke "hich "ere once unnecessary.
Cne "riter has note that until a fe" ecaes a#o 8instinct about the rhythm of the
mother ton#ue serve instea of principles. :o" e)plicit ones arc neee to fill the
place of instinct8 @"here instinct presumably means behavior irectly shape by the
verbal communityE.

-overnmental and Religious .a&s

When people be#an to live to#ether in #roups$ a social environment arose$ an it "as
marke by certain practices. *hose "ho behave in "ays in'urious to others$ for
e)ample$ "ere punishe by those they in'ure. Stan

ABOUT BEHAVIORISM 10

ar "arnin#s coul be formulate$ as the behavior came to be calle ba an
punishe accorin#ly$ even by those "ho "ere not in'ure by a specific instance. *he
contin#encies became more po"erful "hen they "ere coifie in the reli#ious an
#overnmental "arnin#s$ irections$ an instructions calle la"s. By obeyin# the la" a
person avois punishment.

Special contin#encies are arran#e to enforce #overnmental an reli#ious la"s$
but uncontrive social contin#encies maintaine by the #roup may have the same
effect. Where a la" coifies pre&e)istin# social sanctions oppose to stealin#$ for
e)ample$ a person may be#in by obeyin# the commanment 8*hou shalt not steal$8
but he may eventually refrain from stealin# to avoi the isapproval an criticism of
his friens. In oin# so he comes uner the control of the unmoifie social sanctions
from "hich the la" "as ori#inally erive.

The .a&s o Science

?rancis Bacon$ a la"yer$ seems to have been the first to speak of the la"s of science.
!s a "ell&#overne state o"e its orer to its la"s$ so it mi#ht be possible to iscover
the la"s responsible for the orer in the physical "orl. Scientific la"s probably
emer#e from the lore of craftsmen$ an a simple e)ample "ill illustrate the
ifference bet"een behavior shape by natural contin#encies an behavior #enerate
by a rule. In the for#e of a meieval blacksmith a lar#e bello"s provie the force
raft neee for a hot fire. *he bello"s "as most efficient if one opene it fully
before closin# it an opene it quickly an close it slo"ly. *he blacksmith learne to
operate the bello"s in this "ay because of the reinforcin# result of a steay$ hot fire.
(e coul

-AUSES A:9 REASO:S 10%

have learne to$ o so "ithout escribin# his behavior$ but a escription may have
been helpful in operatin# the bello"s properly or in rememberin# ho" to o so after
an interval. ! short verse serve this function;

;p hi#h$
=o"n lo"$
;p quick$
=o"n slo"$
!n that.s the "ay to blo".

*he verse "as helpful for a ifferent reason "hen the blacksmith hire an
apprentice; he coul tell him ho" to operate the bello"s by teachin# him the verse as
a rule. *he apprentice follo"e the rule$ not because the fire "as then steaily hot$
but because he "as pai for oin# so. (e nee never have seen the effect on the fire.
(is behavior "as entirely rule&#overne; the blacksmith.s "as both
contin#ency&shape an to some e)tent rule&#overne after he ha iscovere the
rule.

4arly scientific la"s supplemente the natural contin#encies of the physical
"orl. ! farmer spain# the soil or a mason pri-in# a stone "ith a pole "as controlle
by contin#encies involvin# levers: the soil or stone move most reaily if force "as
applie as far as possible from the fulcrum. Spaes an poles "ere mae lon# for that
reason$ an some lore$ similar to the blacksmith.s rule$ may have been use to teach
ne" "orkers ho" to choose an "here to #rasp spaes or poles. ! more formal
statement of the la" of the lever permitte the principle to be use in situations
"here contin#ency&shape behavior "as unlikely or impossible.

=ifferences in thou#ht processes have been attribute to the apparent
ifferences bet"een the la"s of reli#ion or #overnment an the la"s of science. *he
first are sai to be 8mae$8 the secon merely iscovere$ but the ifference is not in
the la"s but in the contin#encies

ABOUT BEHAVIORISM10)

the la"s escribe. *he la"s of reli#ions an #overnments coify contin#encies of
reinforcement maintaine by social environments. *he la"s of science escribe
contin#encies "hich prevail in the environment quite apart from any eliberate
human action.

By learnin# the la"s of science$ a person is able to behave effectively uner the
contin#encies of an e)traorinarily comple) "orl. Science carries him beyon
personal e)perience an beyon the efective samplin# of nature inevitable in a
sin#le lifetime. It also brin#s him uner the control of conitions "hich coul play no
part in shapin# an maintainin# his behavior. (e may stop smokin# because of a rule
erive from a statistical stuy of the consequences$ althou#h the consequences
themselves are too eferre to have any reinforcin# effect.

"ontingenc!$Shaped )ersus
Rule$-overned Behavior

6ules can usually be learne more quickly than the behavior shape by the
contin#encies they escribe. 3ost people can learn the instruction 8%ush o"n on the
#earshift lever before movin# it into the reverse position8 more reaily than the
actual shiftin# movement$ especially if the lever oes not move easily or if$ in other
cars "ith "hich the river is familiar$ it oes not nee to be pushe o"n. 6ules
make it easier to profit from similarities bet"een contin#encies: 8*his #earshift
operates like that in a B3W.8 6ules are particularly valuable "hen contin#encies are
comple) or unclear or for any other reason not very effective.

! person may use the rules of a lan#ua#e to speak correctly "hen he has not
been aequately e)pose to a verbal community. In learnin# a secon lan#ua#e$ for

-AUSES A:9 REASO:S 10,

e)ample$ he can iscover appropriate responses in a t"o&lan#ua#e ictionary an
appropriate rules in a #rammar. If these ais are aequate$ he can presumably speak
correctly$ but he "oul be helpless "ithout the ictionary an the #rammar$ an even
if he memori-e both of them$ lie "oul still not know the lan#ua#e in the sense to
be iscusse in the follo"in# chapter.

! person "ho is follo"in# irections$ takin# avice$ heein# "arnin#s$ or
obeyin# rules or la"s oes not behave precisely as one "ho has been irectly e)pose
to the contin#encies$ because a escription of the contin#encies is never complete or
e)act @it is usually simplifie in orer to be easily tau#ht or unerstooE an because
the supportin# contin#encies are selom fully maintaine. *he apprentice "ho
operates the bello"s simply because he is pai to o so oes not operate it as if he
"ere irectly affecte by the conition of the fire. =rivin# a car by follo"in#
instructions iffers from the behavior finally shape by the movement of the car on a
hi#h"ay. Speakin# a lan#ua#e "ith the help of a ictionary an a #rammar is not like
speakin# it throu#h e)posure to a verbal community. *he feelin#s associate "ith the
t"o kins of behavior are also ifferent$ but they o not e)plain the ifference in the
behaviors.

*he control e)erte by irections$ avice$ rules$ an la"s is more conspicuous
than that e)erte by the contin#encies themselves$ in part because it is less subtle$
an the latter has therefore seeme to mean a #reater personal contribution an
inner "orth. =oin# #oo because one is reinforce by the #oo of others is more
hi#hly honore than oin# #oo because the la" emans it. In the first case$ the
person feels "ell ispose; in the secon$ he may feel little more than a fear of
punishment. Civic virtue an piety are reserve for

ABOUT B2HAVIORISM 146

those "ho are not merely follo"in# rules. *his is necessarily the case "hen the
contin#encies have never been analy-e&"hen$ as in poetry or mysticism$ they are
sai to be ineffable.

6ule&follo"in# behavior is sai to be the veneer of civili-ation$ "hereas
behavior shape by natural contin#encies comes from the epths of the personality or
min. !rtists$ composers$ an poets sometimes follo" rules @imitatin# the "ork of
others$ for e)ample$ is a version of rule follo"in#E$ but #reater merit attaches to
behavior "hich is ue to a personal e)posure to an environment. ;nlike those "ho
submit to contin#encies arran#e to support rules$ a 8natural8 artist$ composer$ or
poet "ill behave in iiosyncratic "ays an "ill be more likely to feel the boily
conitions$ calle e)citement or 'oy$ associate "ith 8natural8 reinforcers.

*he planne or "ell&mae "ork may suffer from the suspicion "hich attaches
to any calculate behavior. *he. intuitive mathematician seems superior to one "ho
must procee step by step. We naturally ob'ect to the calculatin# frien "ho has
learne ho" to make friens an influence people. %ossibly that is "hy contin#encies
sometimes #o une)amine or unreporte; a escription "oul estroy some of their
effect. *here are those "ho en'oy music an on.t "ant to kno" "hy$8 an Stenhal$
notin# in his /ournal the 8loveliest evenin#8 he ha ever spent$ as 8I kno" very "ell
the secret of the pleasure I have en'oye but I "ill not "rite it o"n in orer not to
tarnish it.8

It is a mistake$ as I pointe out in Chapter I$ to say that the "orl escribe by
science is someho" or other closer to 8"hat As really there$8 but it is also a mistake to
say that the personal e)perience of artist$ composer$ or poet is closer to 8"hat is
really there.8 !ll behavior is etermine$ irectly or inirectly$ by consequences$

-AUSES A:9 REASO:S 141

an the behaviors of both scientist an nonscientist are shape by "hat is really there
but in ifferent "ays.

Are the Rules in the "ontingencies/

I have evote a #oo eal of space to rule&#overne an contin#ency&shape
behavior for several reasons. Cne has to o "ith the problem of kno"le#e$ "hich is
iscusse in the ne)t chapter but about "hich somethin# shoul be sai here. We o
not nee to escribe contin#encies of reinforcement in orer to be affecte by them.
Ao"er or#anisms presumably o not o so$ nor i the human species before it
acquire verbal behavior. ! person chan#e by operant reinforcement has not
8learne a probability8; he has learne to respon at a #iven rate because of a #iven
frequency of reinforcement. We o not nee to say that 8rules are constructe by the
min in the course of the acquisition of kno"le#e.8 *he mason uses a lever
efficiently "ithout kno"in# the la"$ an a chil or o# learns to catch a ball "ithout
8in some sense e)tractin# the rules #overnin# tra'ectories.8

*he so&calle rules of #rammar have recently been the sub'ect of a #oo eal
of controversy. It is sai that there are rules an instructions "hich #overn the use of
lan#ua#e an "hich "e obey "ithout bein# a"are of them. Certainly for thousans of
years people spoke #rammatically "ithout kno"in# that there "ere rules of #rammar.
Grammatical behavior "as shape$ then as no"$ by the reinforcin# practices of verbal
communities in "hich some behaviors "ere more effective than others$ an sentences
"ere #enerate by the 'oint action of past reinforcements an current settin#s. But it
is the contin#encies "hich 8#overn the use of lan#ua#e$8 not rules$ "hether or not
they are e)tracte.

ABOUT BEHAVIORISM142

Reason and Reasons

%ossibly the most amire co#nitive or mental process is reason. It is sai to be a thin
. & of the min "hich istin#uishes man from the brutes. It "as once thou#ht of as a
possession$ 8an essence of innate ieas$ #rante anterior to e)perience$ by "hich the
absolute bein# of thin#s is isclose to us.8 But by the ei#hteenth century$ accorin#
to Cassirer$ reason 8is much less a possession than it is a moe of acquisition. 6eason
is not the area$ the treasury of the min$ in "hich truth$ like a minte coin$ lies
protecte. 6eason is rather the principle an ori#inal force of the min$ "hich impels
to the iscovery of truth an to the efinin# an assurin# of it.8 *he reference to an
impellin# force su##ests that "e are still a lon# "ay from a behavioral efinition.

We often speak of the consequences of behavior as reasons. We cite them in
e)plainin# our o"n behavior: 8*he reason I "ent to the bank "as to #et some money.8
*he term seems more suitable than cause$ especially if "e have not fully unerstoo
the process of selection$ because anythin# "hich follo"s behavior oes not seem to
be in the ri#ht place to be the cause of it. :evertheless$ a reason "hich lies in the
future is no more effective than any other future event. It oes not become effective
because a person 8keeps it in min8 or 8thinks of it8 or 8kno"s the probability that it
"ill occur$8 for e)pressions of this sort merely reflect an effort to fin a prior
representative of a future consequence.

*he consequences escribe or implie in avice$ "arnin#s$ instructions$ an
la"s are the reasons "hy a person takes avice$ hees "arnin#$.;$ follo"s instruc&
tions$ an obeys la"s. %eople are not born "ith a reainess to follo" avice or hee
"arnin#s. Stimuli havin#

-AUSES A:9 REASO:S 140


the status of avice an "arnin#s must play a part in a lon# history of conitionin#
before a person can be inuce to behave by bein# #iven reasons. *o #ive a stuent
reasons "hy somethin# is "orth learnin# is to point to possibly reinforcin#
consequences$ but they may be lon# eferre$ an the stuent.s behavior "ill chan#e
as a result of the pointin# only if the teacher has been part of effective contin#encies
in the past. When a therapist points to reasons "hy his patient.s behavior is costin#
him friens$ be can be sai to 8clarify a relation bet"een behavior an certain
aversive consequences$8 but the patient "ill chan#e only if the therapist makes
remarks effective in other "ays>&not by 8builin# trust or belief8 but by makin# his
behavior a part of contin#encies in "hich the patient has been reinforce. @:either
the teacher nor the therapist is resortin# to 8co#nitive input8 in such e)amples.E

Reasoning

0: I:=;C*IC:

Inuction has been efine as reasonin# from part to "hole$ from particulars to
#enerals. %ossibly "e may translate by sayin# that in analy-in# instances "e can
e)tract rules "hich apply to classes of events. We have seen that operant
conitionin# has been sai to inicate such a process; an or#anism reinforce on one
or more occasions is sai to 8infer or 'u#e that similar consequences "ill follo" upon
other occasions.8 ?ortunately operant conitionin# is effective even "hen this oes
not occur$ but somethin# like it may occur "hen a person analy-es the circumstances
in "hich he is livin#. Inuction is not the process by "hich behavior is stren#thene by
reinforcement; it is an analysis of the conitions uner "hich behavior is reinforce.
*he

ABOUT BEHAVIORISM 144

analysis may lea to escriptions "hich$ as "e have 'ust seen$ can evoke behavior
appropriate to the contin#encies "ithout irect e)posure to them.

! person may solve a problem by chan#in# the settin# in "hich it appears$ an
a fe" problem&solvin# strate#ies "ere note in the last chapter. ! person can acquire
them as he acquires any behavior$ but usually from an instructional social
environment$ (e may also solve a problem by analy-in# it in the present sense$
because in oin# so he arrives at a rule "hich$ "hen follo"e$ solves the problem.
6easonin# about behavior is a matter of analy-in# the reasons for behavior$ an
reasonin# about a problem is a matter of lookin# at the problematical contin#encies
rather than merely alterin# them throu#h establishe problem&solvin# proceures.
6easonin# in this sense steps in "hen routine methos of problem solvin# leave off$
but it is not that "e then move from noncreative to creative measures. *he is&
tinction is bet"een the practical manipulation of a settin# an the analysis of that
settin#. 6easonin# tells us "hy stanar problem&solvin# proceures "ork$ 'ust as a
statement of contin#encies of reinforcement tells us "hy a person behaves as he
oes.

%sychoanalytic "riters sometimes confuse rational an irrational "ith conscious
an unconscious. @Irrational$ like unreasonable$ has unfortunate overtones; irrational
behavior is not appropriate to current circumstances; it appears to be emitte for the
"ron# reasons. But this has little if anythin# to o "ith the present istinction.E !ll
behavior$ effective or not$ is at first nonrational in the sense that the contin#encies
responsible for it have not been analy-e. !ll behavior is at first unconscious$ but it
may become conscious "ithout becomin# rational: a person may kno" "hat he is
oin# "ithout kno"in# "hy he is oin# it.

-AUSES A:9 REASO:S 145

%eople o not behave irrationally simply because they are not a"are of all the
variables at$ issue. It is a step for"ar to iscover that "e carry ba ne"s in part
because "e are reinforce by the iscomfiture of our friens an that "e mention the
name ofL a person because there is someone in the room "ho resembles him$
althou#h "e have not up to this point 8seen8 him. We may ob'ect "hen this is pointe
out$ because "e may not "ant to believe$ as one "riter has put it$ that 8there is more
to human personality than immeiate consciousness tells us there is$8 but "hat is left
out is not to be foun in the 8transrational re#ion of the min.8 We cannot$ of course$
analy-e contin#encies "hich "e o not observe$ but "e can observe them "ithout
analy-in# them. *o act by takin# reasons for action into account an to moify one.s
behavior in terms of that account is more than bein# a"are of "hat one is oin#.

Several aspects of the life of reason
eserve comment.

2oll3 and Reason. 4rasmus in his In .raise o8 2oll3 pointe out that one cannot
be#in "ith reason. *he life of reason is no oubt amirable; but there "oul be
nothin# to be reasonable about if it "ere not for the effect of foo$ se)$ an other
basic reinforcers the thin#s 4rasmus calle folly. 8*he persistence of human folly in
the face of heroic efforts to enli#hten it8 "ith reason may be the tra#ey of our
times$ but if "e are to take effective action$ reason "ill consist of an analysis of the
contin#encies represente by folly an of the uses "hich may be mae of them. *o
say that the irrational is a 8rich spectrum of life&enhancin# human possibilities8 is to
point irectly to rcinforcers. *hey nee not be suppresse by reason; on the contrary$
they may be mae vastly more effective.

ABOUT BEHAVIORISM 14

Int+ition and Reason. It has been sai that 8uner behaviorist assumptions$
"hich insistc that lan#ua#e "as behavior$ such concepts as intuition "ere re#are
as bein# as unfit for scientific stuy as #hosts or reams$8 but behavin# intuitively$ in
the sense of behavin# as the effect of unanaly-e contin#encies$ is the very startin#
point of a behavioristic analysis. ! person is sai to behave intuitively "hen he oes
not use reason. Instinct is sometimes a synonym: it is sai to be a mistake to
8attribute to lo#ical esiNm "hat is the result of blin instinct$8 but the reference is
simply to behavior shape by unanaly-e contin#encies of reinforcement. *he blin
instinct of the artist is the effect of the iiosyncratic consequences of his "ork. It is
no 8betrayal of reason8 to accept "hat artists teach us about life$ nature$ an society$
since not to accept it "oul be to assert that contin#encies are effective only "hen
they have been escribe or formulate as rules.

It is also a mistake to reserve intuition for the effect of contin#encies from
"hich it oes not seem to be possible to erive rules. *o say that "e 8intuitively8
reco#ni-e that a sentence such as 8/ohn is "eak to please. is ill&forme is to imply that
no rule of #rammar "ill permit us to call the sentence "ell&forme$ but "hat "e
intuitively reco#ni-e is that the behavior "e possess by virtue of the practices of our
verbal community oes not inclue a sentence of this form$ nor shoul "e as a
member of such a community respon to it in an effective "ay.

! person may iscriminate bet"een t"o ob'ects "ithout bein# able to ientify
the istin#uishin# property. *he intuitive ia#noses of a physician$ the intuition "ith
"hich an art critic ientifies a school or artist$ an the intuitive skill "ith "hich some
people quickly learn to fin their "ay about a city illustrate behaviors for

-AUSES A:9 REASO:S14%

"hich no rule has yet been formulate. Science often arrives very late in analy-in#
contin#encies. It is sai$ for e)ample$ that the concept of torque require nearly t"o
hunre years to be formulate$ althou#h skillful behavior "ith respect to systems
involvin# torque ha lon# e)iste an coul be acquire "ithout benefit of rule in a
very short time.

2aith and Reason. ?aith is a matter of the stren#th of behavior resultin# from
contin#encies "hich have not been analy-e. Crthoo) crecal behavior$ in the sense
of behavior conformin# to la"s$ is very ifferent from the e)periential result of a
mystical e)perience. It has often been sai$ in fact$ that proofs of the e)istence of
Go are etrimental to faith$ because they supply reasons for a belief that "oul
other"ise be more hi#hly value as intuitive.

I"#+lse and 9eli/eration. 8I conceive$8 sai *homas (obbes$ 8that "hen a man
eliberates "hether he shall o a thin# or not o it$ he oes nothin# else but consier
"hether it be better for himself to o it or not to o it8&"hether$ in short$ he "oul
be reinforce by the consequences. =eliberate behavior procees throu#h an analysis
of reasons; impulsive behavior is the irect effect of contin#encies. Impulsive "orks
"ere once calle ecstatic$ an carefully esi#ne "orks euplastic. ?or the Greeks a
pruent or reasonable person possesse s&#hros3n&( it "as the mark of a temperate
person&that is$ of a person "hose behavior has been tempere by an analysis of its
consequences.

Invented Reasons. *he avanta#es #aine from e)aminin# the reasons for one.s
behavior are perhaps responsible for the tenency to construct reasons "hen

ABOUT BEHAVIORISM14)

none$ can be foun. Superstitious behavior$ for e)ample$ is the prouct of
aventitious contin#encies of reinforcement whi*h are in no orinary sense
reasonable. :o rule can be erive from the contin#encies. :evertheless$ the
behavior may be stron#. !ske$ 8Why are you oin# that?8 a superstitious person is
likely to invent an ans"er. Ile ritualistic practices of a "hole culture have le to the
elaborate ans"ers foun in mytbs. In many countries rain is a reinforcin# event$ an it
leas to a "ie ran#e of superstitious behaviors$ incluin# rain ancin#. *he
e)planation #iven for a rain ance mi#ht be that it pleases the person$ force$ or spirit
that brin#s rain.

@Superstitious behavior has$ ho"ever$ its reasons. ! reinforcer has an effect
even thou#h the behavior it follo"s oes not prouce it. It has been sho"n in lo"er
or#anisms that the intermittent presentation of a noncontin#ent reinforcer selects
an maintains a response 8for acciental reasons.8 *he history of mytholo#y supplies
many comparable e)amples in human sub'ects.E

Reasoning

(: =4=;C*IC:

It is not the ob'ect of a behavioral analysis to say "hat inuction is. Aike reasonin# or
inference$ the term oes not usefully escribe any sin#le behavioral process.
:evertheless$ finin#$ offerin#$ or inventin# reasons loosely efines a fiel "hich may
be profitably analy-e. ! pi#eon pecks a isk an is reinforce "hen the isk is re
but not "hen it is #reen; it then stops peckin# "hen the isk is #reen. We o not nee
to say that it has ra"n the inference that #reen isks are not "orth peckin#$ !
baseball aficionao #oes to the hall park on clear ays but not "hen it is rainin#
heavily. We o

-AUSES A:9 REASO: 14,

not nee to say that on a #iven rainy ay he infers that no #ame "ill be playe. If he
is plannin# to "atch a televise #ame playe in another city an hears that it is
rainin# there$ he may not turn on his television set$ but "e have no reason to say that
he has inferre that a #ame "ill not be playe. We nee a separate term only to
escribe the erivin# of a rule from the contin#encies. *he pi#eon cannot o this$ but
the aficionao can 8reason from particulars to #enerals8 in sayin#$ 8Baseball is not
playe in heavy rain.8

=euction$ as reasonin# from #enerals to particulars$ is also not a process
requirin# a behavioral analysis$ but there is a fiel havin# to o "ith the control
e)erte by rules "hich nees attention. If someone "ho kno"s nothin# about baseball
is tol that #ames are never playe in heavy rain an that it is rainin# heavily$ "hat
behavioral process or processes "ill keep him from #oin# to the ball park or "ill lea
him to say that no #ame "ill be playe? It is temptin# to make a rou#h istinction
bet"een inuction as the erivin# of rules an euction as the applyin# of rules$ but
this "oul be to overlook the fact that euction is often a matter of erivin# ne"
rules from ol$ particularly from a consieration of certain key terms$ such as 8all$8
8some$8 i#no8t 8if$8 an 8or$8 "here the iscovery of rules for erivin# ne" rules from
ol "oul seem to be an e)ample of inuction.

*his is not the place for a survey of reasonin#. I am simply tryin# to su##est the
kins of behavioral processes to be foun in these traitional fiels. It has often been
pointe out that a lo#ical cor mathematical formulation follo"s a #reat intellectual
achievement rather than prouces it. It has been sai that

:e"ton coul hol a problem in his min for hours an ays an "eeks until it
surrenere to hirn its secret. *hen$

ABOUT BEHAVIORISM 156

bein# a supreme mathematical technician$ be coul ress it up$ ho" you "ill$ for the
purposes of e)position$ but it "as his intuition "hich "as preeminently
e)traorinary&&&&7so happy in his con'ectures$8 sai e 3or#an 8as to seem to kno"
more than he coul possibly have any means of provin#.8

*he e)traction of rules "as eviently a seconary sta#e. It is a much more
e)plicit sta#e$ ho"ever$ an therefore more likely to be analy-e by lo#icians or
mathematicians. *he initial 8intuitive8 sta#e$ "hich falls to the lot of the behaviorist$
is far more refractory. *here is nothin# to be one about this; it is amittely a
ifficult fiel. ! first step$ ho"ever$ is to ackno"le#e its nature.. We$ #ain nothin#
by attributin# :e"ton.s achievement to intuition or happy con'ecture.

Tr+th. *he truth of a statement of fact is limite by the sources of the behavior
of the speaker$ the control e)erte by the current settin#$ the effects of similar
settin#s in the past$ the effects upon the listener leain# to precision or to
e)a##eration or falsification$ an so on. *here is no "ay in "hich a verbal escription
of a settin# can be absolutely true$ ! scientific la" is erive from possibly many
episoes of this sort$ but it is similarly limite by the repertoires of the scientists
involve. *he verbal community of the scientist maintains special sanctions in an
effort to #uarantee valiity an ob'ectivity$ but$ a#ain$ there can be no absolute. :o
euction from a rule or la" can therefore be absolutely true. !bsolute truth can be
foun$ if at all$ only in rules erive from rules$ an here it is mere tautolo#y.
# $nowing
'o Back

4e sa- that a newborn bab- knows how to cr-) suck*e) and snee2e. 4e sa- that a chi*d
knows how to wa*k and how to ride a tric-c*e. (he e/idence is simp*- that the bab- and chi*d
e,hibit the beha/ior specified. ;o/in. from /erb to noun) we sa- that the- possess know*ed.e)
and the e/idence is that the- possess beha/ior. 3t is in this sense that we sa- that peop*e thirst for)
pursue) and ac=uire know*ed.e.

But this brin.s us at once to the =uestion of what it means to possess beha/ior. 4e saw in
Chapter 4 that to sa- that a response is emitted does not imp*- that it has been inside the
or.anism. Beha/ior e,ists on*- when it is bein. e,ecuted. 3ts e,ecution re=uires a ph-sio*o.ica*
s-stem) inc*udin. effectors and receptors) ner/es) and a brain. (he s-stem was chan.ed when the
beha/ior was ac=uired) and it is the chan.ed s-stem which is 9possessed.9 (he beha/ior it
mediates ma- or ma- not be /isib*e at an- .i/en moment. (here are para**e*s in other parts of
bio*o.-. An or.anism 9pos
ABOUT BEHAVIORISM ).,

sesses9 a s-stem of immune reactions in the sense that it responds to in/adin. or.anisms in a
specia* wa-) but its responses are not in e,istence) unti* it is bein. in/aded. 3t is often usefu* to
speak of a repertoire of beha/ior which) *ike the repertoire of a musician or a compan- of
p*a-ers) is what a person or compan- is capab*e of doin.) .i/en the ri.ht circumstances. Onow*6
ed.e is possessed as a repertoire in this sense.

$inds of $nowing

7ne meanin. of 9to know9 is simp*- to be in contact with) to be intimate with. 3t is in this sense
that a person is said to know sin) beaut-) or sorrow) or a man to know a woman in the bib*ica*
sense of ha/in. carna* know*ed.e of her. (here is an imp*ication) of course) that beha/ior is
chan.ed b- the contact.

4e are said to know how to do somethin.6open a window) spe** 9anaco*uthon)9 so*/e a
prob*em6if we can do it. 3f we can .et from here to there) we are said to know the wa-. 3f we can
recite a poem or p*a- a piece of music without readin. it) we are said to know it 9b- heart)9 a
curious bit of ph-sio*o.i2in..

4e are a*so said to know about thin.s. 4e know a*.ebra) Baris) Shakespeare) or @atin)
not on*- in the sense of ha/in. had contact with a fie*d) a p*ace) a poet) or a *an.ua.e but in the
sense of possessin. /arious forms of beha/ior with respect to them. 4e know about e*ectricit- if
we can work successfu**-) /erba**- or otherwise) with e*ectrica* thin.s.

A** these forms of knowin. depend on a pre/ious e,posure to contin.encies of
reinforcement) but we are a*so said to ha/e a specia* kind of know*ed.e if we can simp*- state
instructions) directions) ru*es) or *aws. A person ma- know how to operate a piece of e=uip
34O5I46 ).3

ment because he has read the instructions) or how to .et about in a cit- because he has studied a
map) or how to beha/e *e.a**- because he knows the *aw) a*thou.h he ma- ne/er ha/e operated
the e=uipment) /isited the cit-) or fe*t the hand of the *aw himse*f. Onow*ed.e which permits a
person to describe contin.encies is =uite different from the know*ed.e identified with the
beha/ior shaped b- the contin.encies. Ceither form imp*ies the other.

Ba/*o/5s do.s ha/e been said to know 9when to sa*i/ate)9 but the- did not sa*i/ate because
the- knew that the be** wou*d be fo**owed b- food. A rat cou*d be said to know when to press a
*e/er to .et food) but it does not press because it knows that food wi** be de*i/ered. A ta,i dri/er
cou*d be said to know a cit- we**) but he does not .et around because he possesses a co.niti/e
map.

%oes $nowledge Come &rom Experience'

Dohn @ocke and other British empiricists emphasi2ed mere contact with a stimu*atin.
en/ironment. (he- did not e,p*ain wh- a person shoud attend to the wor*d around him) wh- he
shoud connect %associate& two features which occurred to.ether so that one then reminded him
of the other) or wh- he shoud think about them at a**. 4e saw in Chapter 5 that some of @ocke5s
successors introduced an e*ement of be*ief or wi** into the empirica* position) but know*ed.e
about the wor*d is due to more than contact with a .i/en settin.) because it is due to the
contin.encies of reinforcement of which that settin. is a part. (he 9e,perience9 from which
know*ed.e is deri/ed consists of the fu** contin.encies.

ABOUT BEHAVIORISM ).&

$nowledge as Power and as Contemplation

4e do not act b- puttin. know*ed.e to use8 our know*ed.e is action) or at *east ru*es for action.
As such it is power) as 1rancis Bacon pointed out in re0ectin. scho*asticism and its emphasis on
knowin. for the) sake of knowin.. 7perant beha/ior is essentia**- the e,ercise of power: it has
an effect on the en/ironment. (he ad/ancement or au.mentation of *earnin. proposed b- Bacon
was the furtherin. of human beha/ior in the interests of the human condition) and the
achie/ements of modern science show that he correct*- foresaw its character. Ce/erthe*ess) the
concern for power has recent*- been cha**en.ed. (he 4est is said to ha/e made a fetish of the
contro* of nature. 3t is certain*- not difficu*t to point to the unhapp- conse=uences of man-
ad/ances in science) but it is not c*ear how the- can be corrected e,cept throu.h a further
e,ercise of scientific power.

(here is room in a beha/ioristic ana*-sis for a kind of knowin. short of action and hence
short of power) 7ne need not be acti/e*- beha/in. in order to fee* or to introspecti/e*- obser/e
certain states norma**- associated with beha/ior. (o sa-) 93 know a sea *ion when 3 see one)9 is to
report that one can identif- a sea *ion but not that one is now doin. so. A response temporari*-
for.otten ma- sti** be c*aimed as know*ed.e) as when we sa-) 93 can5t think of it at the moment
but 3 know it as we** as 3 know m- own name.9

4e a*so use 9know9 to mean 9bein. under the contro* of)9 a condition which is not the
on*- determiner of our beha/ior. 4hen we sa-) 93 went to the meetin. knowin. that P wou*d be
speakin.9 %where knowin.

34O5I46 )..

cou*d be rep*aced b- be*ie/in.) e,pectin.) rea*i2in.) or understandin.&) we report that our
beha/ior was affected b- some prior indication that P wou*d be at the meetin.) but the beha/ior
itse*f cou*d not be ca**ed knowin. that fact. (o sa-) 93 went thinkin* 7 wou*d be there)9 su..ests
a *ess c*ear or *ess re*iab*e prior indication) a distinction between thinkin. and knowin.
mentioned in Chapter !. 3t has been said that 9a** knowin. consists of h-potheses ... re.arded as
pro/en or he*d /er- tentati/e*-)9 but we are more *ike*- to sa- 93 think9 with re.ard to a tentati/e
h-pothesis and to reser/e 93 know9 for the pro/en case. (he difference is not critica*) howe/er.
(he assertion 93 know someone is hidin. in this room9 imp*ies weak e/idence but is ne/erthe*ess
a stron. response) presumab*- for other reasons. Simi*ar conditions ma- pre/ai* e/en thou.h a
remark is not made.
;uch of what is ca**ed contemp*ati/e know*ed.e is associated with /erba* beha/ior and with the
fact that it is the *istener rather than the. speaker who takes action. 4e ma- speak of the power of
words in affectin. a *istener) but the beha/ior of a speaker in identif-in. or describin. somethin.
su..ests a kind of know*ed.e di/orced from practica* action. Jerba* beha/ior p*a-s a principa*
ro*e in contemp*ati/e know*ed.e) howe/er) because it is we** adapted for automatic
reinforcement: the speaker ma- be his own *istener. (here are non/erba* beha/iors ha/in. the
same effect. Berceptua* responses which c*arif- stimu*i and reso*/e pu22*ement ma- be
automatica**- reinforcin.. 9'ettin. the meanin.9 of a difficu*t passa.e is simi*ar. (he who*e
wor*d of fantas- is perceptua* beha/ior which is automatica**- reinforcin.) and some parts fa**
within the fie*d of know*ed.e. Contemp*ation of this kind wou*d be impossib"e8

ABOUT BEHAVIORISM )./

howe/er) without a pre/ious e,posure to contin.encies in which action is taken and differentia**-
reinforced.

(nderstanding

3n a simp*e sense of the word) 3 ha/e understood what a person sa-s if 3 can repeat it correct*-. 3n
a somewhat more comp*e, sense) 3 understand it if 3 respond appropriate*-. 3 ma- do so 9without
understandin. wh- he sa-s it.9 (o understand wh-) 3 must know somethin. about the contro**in.
/ariab*es) about the circumstances under which 3 shou*d ha/e said it m-se*f. 3 come to
understand a difficu*t te,t in this sense when) b- readin. and rereadin. it) 3 ac=uire a stron.er
and stron.er tendenc- to sa- what the te,t sa-s.

Hnderstandin. sometimes means knowin. reasons. 3f 3 throw a switch to put a piece of
apparatus into operation and nothin. happens) 3 ma- tr- the switch a.ain) but m- beha/ior
=uick*- under.oes e,tinction) and 3 ma- then *ook to see whether the apparatus is connected with
the power source) or whether a fuse is b*own) or whether the startin. switch is broken. 3n doin.
so) 3 ma- come to understand wh- it has not worked) in the sense of disco/erin. the. reasons. 3
ha/e ac=uired understandin. b- ana*-2in. the pre/ai*in. contin.encies. (eachers are sometimes
ur.ed to .i/e their students a deeper understandin. of what the- are *earnin. b- showin. them
that the ru*es the- ha/e memori2ed are descriptions of rea* contin.encies. (he- are not to teach
the commutati/e *aw a*one8 the- are to show the reasons wh- it works.

4e ourse*/es often ac=uire a deeper understandin. of a ru*e in this sense throu.h e,posure
to the natura* contin.encies it describes. (hus) if we ha/e memori2ed

34O5I46 ).0

a ma,im and obser/ed it) we ma- a.ain be.in to be modified b- the natura* conse=uences. 4e
disco/er) for e,amp*e) that 9it rea**- is true9 that procrastination is the thief of time) and we then
understand the ma,im in a different sense. (he understandin. .ained b- mo/in. from
ru*e6.o/erned to contin.enc-6shaped beha/ior is usua**- reinforcin.) in part because the
reinforcers in the *atter case are *ess *ike*- to be contri/ed and hence *ess *ike*- to work in the
interest of others.

4e a*so find it reinforcin. when a ru*e) as a description of contin.encies) makes them *ess
pu22*in. or more effecti/e. 3f a .i/en situation has not e/oked an- /er- usefu* /erba* beha/ior)
we ma- be reinforced b- what a writer sa-s about it if we can then respond in the same wa-. 4e
understand what *ie sa-s in the sense that we can now formu*ate the contin.encies he describes
more e,act*- or respond to them more successfu**-.

$nowing as Possessing Information

3nformation theor- arose from the ana*-sis of transmitted si.na*s) as in a te*ephone *ine. 3n the
fie*d of /erba* beha/ior it cou*d be app*ied to the. sound stream of speech between speaker and
*istener or the marks in a *etter sent from writer to reader. (he messa.e has) as 3 ha/e said) an
apparent*- ob0ecti/e status.

3nformation is used in a /er- different wa- in describin. indi/idua* beha/ior. Dust as the
e,terna* practice of storin. and *ater consu*tin. memoranda is used metaphorica**- to represent a
supposed menta* process of storin. and retrie/in. memories) so the transmission of information
from one person to another has been used metaphorica**- to represent the transmission of input
to output %or of stimu*us to response&. (he
ABOUT BEHAVIORSM ).8

metaphor is at home in theories deri/ed historica**- from the ref*e, are) in which the
en/ironment enters %or is taken in b-& the bod- and is processed and con/erted into beha/ior.
@ike stored memories or data structures) information be.ins as input %necessari*- coded& but
chan.es pro.ressi/e*- unti* it becomes a predisposition to act. 3n an operant ana*-sis) as 3 ha/e
pointed out) we do not need to fo**ow the stimu*us throu.h the bod- or to see how it becomes a
response. Ceither the stimu*us nor the response is e/er in the bod- in an- *itera* sense. As a form
of know*ed.e) information can be treated more effecti/e*- as a beha/iora* repertoire.

3t is often said that reinforcement con/e-s information) but this is simp*- to sa- that it
makes a response not on*- more probab*e but more probab*e on a specific occasion. 3t brin.s a
response under the contro* of re*ated depri/ations or a/ersi/e) stimu*ation as we** as of stimu*i
present at the time it occurs) 3nformation in this sense refers to the contro* e,ercised b- en/iron6
menta* conditions.

3nformation theor-) with respect to. the beha/ior of the indi/idua*) is mere*- a
sophisticated /ersion of cop- theor-. (he e,terna* wor*d is interna*i2ed) not as a photo.raphic or
phono.raphic reproduction) but sufficient*- transduced) encoded) or otherwise modified to be
more p*ausib*- re.arded as stored within the bod-.

he Personal $nowledge of the Scientist

the centra* =uestion of scientific know*ed.e is not 4hat is known b- scientistsA but 4hat does
knowin. meanA (he facts and *aws of science are descriptions of the wor*d6that is) of pre/ai*in.
contin.encies of

34O5I46 )."

reinforcement. (he- make it possib*e for a person to act more successfu**- than be cou*d *earn to
do in one short *ifetime or e/er throu.h direct e,posure to man- kinds of contin.encies.

(he ob0ecti/it- which distin.uishes ru*e6.o/erned beha/ior from beha/ior .enerated b-
direct e,posure to contin.encies is furthered b- tests of /a*idit-) proof) practices minimi2in.
persona* inf*uences) and other parts of scientific method. Ce/erthe*ess) the corpus of science6the
tab*es of constants) the .raphs) the e=uations) the *aws6ha/e no power of their own. (he- e,ist
on*- because of their effects on peop*e. 7n*- a *i/in. person knows science in the sense of actin.
under its contro* with respect to nature. But this is not to sa- that 9e/er- instance of knowin.
in/o*/es comin. to terms in some wa- with the sub0ecti/e and phenomeno*o.ica*.9 Onow*ed.e
is sub0ecti/e in the 6tri/ia* sense of bein. the beha/ior of a sub0ect) but the en/ironment) past or
present) which determines the beha/ior *ies outside the beha/in. person.

3f action were determined b- fee*in.s or introspecti/e*- obser/ed states of mind) it wou*d
be true) as ;ichae* Bo*an-i and Berc- 4. Brid.man ha/e insisted) that science is ine,orab*-
persona*. As Brid.man once put it) 93 must describe thin.s as the- seem to me$ I cannot .et awa-
from m-se*f.9 (his is true in the sense that a scientist must beha/e as an indi/idua*. But if he
ana*-2es the wor*d around him) and if) as a resu*t) he states facts or *aws which make it possib*e
for others to respond effecti/e*- without persona* e,posure to that wor*d) then he produces
somethin. in which he himse*f is no *on.er in/o*/ed. 4hen man- other scientists arri/e at the
same facts or *aws) an- persona* contribution or persona* participation is reduced to a minimum.
4hat is fe*t or introspecti/e*- obser/ed b- those
ABOUT BEHAVIORISM )/'

whose beha/ior is .o/erned b- scientific *aws is /er- different from what is fe*t or
introspecti/e*- obser/ed as the resu*t of e,posure to the ori.ina* contin.encies.

3t is absurd to suppose that science is what a scientist fee*s or introspecti/e*- obser/es. Co
one person can respond to more than a minuscu*e part of the contin.encies pre/ai*in. in the
wor*d around him. 3f it is said instead that science is a kind of .roup consciousness) then we must
*ook at how this is he*d to.ether) and we sha** find that what are communicated amon. scientists
are statements of facts and ru*es and *aws) not fee*in.s. %(he persona* ro*e of the scientist some6
times seems to be emphasi2ed because of the apparent co*dness of ob0ecti/e know*ed.e) as some
re*i.ious works ha/e continued to be transmitted b- word of mouth) in spite of the in/ention of
writin. and printin.) because the written form seems de/oid of fee*in.. Spoken /erba* beha/ior
has a brief period of ob0ecti/it- between speaker and *istener) but it is Jer- brief) and the 0oint
presence of two parties .i/es ora* communication an apparent warmth and depth which is
missin. from a book.&

Isms

A phi*osoph-) a mora* c*imate) a c*ass consciousness) and a spirit of the times are other
inte**ectua* possessions which fa** within the fie*d of know*ed.e and account for some of the
*ar.e patterns of beha/ior characteristic of a peop*e) a c*ass) a period) or a cu*ture. A person is
said to act or speak as he does because he is a pra.matist) a member of ti*e pro*etariat) a practi6
tioner of the work ethic) or a beha/iorist. (erms of this sort c*assif- beha/ior ha/in. identifiab*e
conse=uences under .i/en circumstances. Conf*icts) such as
34O5I46 )/)

that between empiricism and rationa*ism) are conf*icts between contin.encies) and if the histor-
of ideas seems to show the de!eo+ment of human thou.ht) it is not because) for e,amp*e)
romanticism *eads to c*assicism and /ice /ersa) but because the practices characteristic of one
ism e/entua**- produce conditions. under which a different pattern of beha/ior is .enerated and
for a time maintained.

3n #i!e Sta*es o( 6reek Rei*ion, 'i*bert ;urra- described the chan.e in the :oman
+mpire under Christianit- as 9a rise of asceticism) of m-sticism) in a sense) of pessimism8 a *oss
of se*f6confidence) of hope in this *ife and faith in norma* human effort8 a despair of patient
in=uir-) a cr- for infa**ib*e re/e*ation8 an indifference to the we*fare of the state) a con/ersion of
the sou* to 'od.9 Accordin. to Beter 'a-) 9>e christened it a 5fai*ure of ner/e.95 9Christener is
possib*- a pun) but the fai*ure of ner/e is a rather characteristic appea* to pseudoph-sio*o.-) a
comin. down to earth after a sustained :i.ht of menta*ism. (he e/idence which 0ustifies
ascribin. the beha/ior of :omans to asceticism) m-sticism) pessimism) and so on shou*d ser/e as
we** in makin. a few .uesses about the pre/ai*in. contin.encies. (he ascetic is no *ess
reinforced b- de*icious food) se,) and so on than others) %indeed) his asceticism wou*d scarce*-
be admired if he were&) but his beha/ior is c*ear*- under the contro* of other conse=uences6most
of them probab*- the puniti/e sanctions of ear*- Christianit-. Bessimism and a *oss of
se*f6confidence) hope) and faith are) as we saw in Chapter 4) associated with a *ack of stron.
positi/e reinforcement) A despair of patient in=uir- su..ests defecti/e schedu*es of
reinforcement) and a cr- for infa**ib*e re/e*ation a search for ru*es in *ieu of contin.encies which
mi.ht shape beha/ior direct*-. An in

ABOUT B#HAVIORISM )/,

ifference to the "elfare of the state an a conversion of the soul to Go su##est a
shift from #overnmental to reli#ious sanctions. (o" much more "e shoul kno" if the
prevailin# contin#encies ha been escribe rather than the feelin#s an isms
#enerate by themL
01 The Inner World o 'otivation and 2motion
Go Back

We have been lookin# at "hat mi#ht be calle the intellectual sie of the life of
the min one.s e)periences in the "orl in "hich one lives$ one.s inferences about the
structure of that "orl$ one.s plans for ealin# "ith it$ one.s intentions$ purposes$
ieas$ an so on. I have interprete the facts to "hich these e)pressions seem to
refer as aspects of human behavior attributable to contin#encies of reinforcement or$
if I may repeat$ to the subtle an comple) relations amon# three thin#s: the situation
in "hich behavior occurs$ the behavior itself$ an its consequences.

!nother sie of the life of the min is sai to be concerne "ith instincts$
rives$ nees$ emotions$ an impulsive or efensive activities$ an it has attracte
attention mainly for psychotherapeutic reasons. *o
ABOUT BEHAVIORISM 14

mark this istinction the "or 8psyche$8 once applie to the intellect$ no" tens to be
reserve for the emotional an motivational life. *he t"o sies are not entirely
unrelate. *o take a very simple e)ample$ operant reinforcement brin#s behavior
uner the control of particular kins of eprivation an aversive stimulation; in
traitional terms$ nees or feelin#s fin satisfaction or e)pression throu#h action
upon the e)ternal environment. *he intellect is sometimes sai to control nees an
emotions$ althou#h it may fail to o so from time to time.


(ersonalities

We have seen that the intellectual life of the min has been fabricate on the
pattern of life in the e)ternal "orl. 3ove insie$ the environment is converte into
e)perience$ an action into ieas$ purposes$ an "ill. 3akin#$ storin#$ an consultin#
memorana set the pattern for the processin# of memories$ *echniques of solvin#
problems become co#nitive strate#ies. *he thinkin# person is thus converte into the
thinkin# min. Somethin# of the same sort has happene in the invention of an inner
"orl of motivation an emotion. *he person is replace by a self or personality$
perhaps by more than one. ! paper on the activist youth of the 01KDs$ for e)ample$
calls attention to a 8moal personality8 of activists. It escribes "hat youn# people
say an o "hen in the company of their families$ peers$ an teachers$ as "ell as
"hen they are bein# 8active.8 It is an analysis of the moal activist$ not the moal
personality.

! self or personality is at best a repertoire of behavior imparte by an
or#ani-e set of contin#encies. *he behavior a youn# person acquires in the bosom of
THE I::ER WOR59 O2 MOTIVATIO: A:9 EMOTIO: 15

his family composes one self; the behavior he acquires in$ say$ the arme services
composes another. *he t"o selves may e)ist in the same skin "ithout conflict until
the contin#encies conflict as they may$ for e)ample$ if his friens from the services
visit him in his home. !s 3ar) an many others have pointe out$ the iniviual is
born of society$ an his inivisibility epens upon the coherence of the society "hich
#ives birth to him. 8?ra#mentation of a life8 is sai to follo" 8social isor#ani-ation in
"hich a person has been rippe apart$8 fra#mentation bein# efine as an 8arran#e&
ment consciousness makes in response to an environment "here respect is not
forthcomin# as a matter of course.8 But it is behavior$ not consciousness$ that is
fra#mente an rippe apart$ an respect is only one of the isor#ani-e reniforcers.

Conflictin# contin#encies lea to conflictin# repertoires of behavior$ but they
are all e)hibite by one boy$ by one member of the human species. *he boy that
behaves in a consierate "ay most of the time is the same boy that is occasionally
callous or cruel; the boy that behaves heterose)ually most of the time is the same
boy that is occasionally homose)ual. What a person is really like coul mean "hat he
"oul have been like if "e coul have seen him before his behavior "as sub'ecte to
the action of an environment. We shoul then have kno"n his 8human nature.8 But
#enetic eno"ment is nothin# until it has been e)pose to the environment$ an the
e)posure immeiately chan#es it. Within limits$ "e may istin#uish bet"een the
contributions of survival an reinforcement. When %ascal sai that nature is only first
habit$ as habit is secon nature$ he coul be sai to have anticipate current
reco#nition that the species acquires behavior @instinctsE uner contin#encies of
survival "hile the in

ABOUT BEHAVIORISM1

iviual acquires behaviors @habitsE uner contin#encies of reinforcement.

In ?reu.s #reat triumvirate$ the e#o$ supere#o$ an i represent three sets of
contin#encies "hich are almost inevitable "hen a person lives in a #roup. *he i is
the /ueo Christian 8Cl !am8Oman9s 8unre#enerate nature$8 erive from his innate
susceptibilities to reinforcement$ most of them almost necessarily in conflict "ith the
interests of others. *he supere#o the /ueo Christian conscience speaks in the 8still
small voice8 of a @usuallyE punitive a#ent representin# the interests of other people. It
is efine in Webster.s Third International 9i*tionar3 as a ma'or sector of the psyche
that is mostly unconscious but partly conscious$ that evelops out o8 the e#o by in&
ternali-ation or intro'ection$ in response to avice$ threats$ "arnin#s$ an
punishment$ especially by parents but also by teachers an other authority$ that
reflects parental conscience an the rules of society$ an that serves as an ai in
character formation an as a protector for the e#o a#ainst over"helmin# i impulses.

But it is 8a ma'or sector of the psyche8 only in the sense of 8a ma'or part of
human behavior$8 an it is mostly unconscious only because the verbal community
oes not teach people to observe or escribe it. It is mainly the prouct of the
punitive practices of a society "hich attempts to suppress the selfish behavior
#enerate by biolo#ical reinforcers$ an it may take the form of imitatin# society
@8servin# as the vicar of society8E as the in'unctions of parents$ teachers$ an others
become part of its repertoire$ *he e#o is the prouct of the practical contin#encies in
dail3 life$ necessarily involvin# susceptibilities to reinforcement an the punitive
contin#encies arran#e by other people$ but isplayin# behavior shape an
maintaine by

THE I::ER WOR59 O2 MOTIVATIO: A:9 EMOTIO: 1%

a current environment. It is sai to satisfy the i if it achieves a certain amount of
biolo#ical reinforcement$ an the supere#o if it oes so "ithout riskin# too much
punishment. We o not nee to say that these three archetypal personalities are the
actors in an internal rama. *he actor is the or#anism$ "hich has become a person
"ith ifferent$ possibly conflictin#$ repertoires as the result of ifferent$ possibly
conflictin#$ contin#encies.

?reu.s analysis has seeme convincin# because of its universality$ but it is
the environmental contin#encies rather than the psyche "hich are invariant. *he
conflicts bet"een the supere#o an the i$ "hich the e#o so often fails to resolve$
sho" certain familiar patterns. In some cultures the fact that a son loves his mother
an vie"s his father as a rival is almost as characteristic of the human male as the
anatomy "hich efines his se)$ but a comparable universality is to be foun amon#
the social contin#encies of reinforcement maintaine by the kins of families in such
cultures. /un#.s archetypal patterns an collective unconscious can be trace to
either the evolution of the species or the evolution of cultural practices. 8*he
astonishin# sameness of the represse unconscious across all recore eras an
civili-ations8 is the sameness of the thin#s "hich reinforce people an of the
behaviors "hich prove in'urious to others. *he universal features sai to be
characteristic of all lan#ua#es are the result of universal characteristics of lan#ua#e
communities arisin# from the role playe by lan#ua#e in aily fife.

.ie in the (s!che

*he life of the min is sai to require an consume psychic ener#y. *his is
simply another "ay of representin#
ABOUT BEHAVIORISM1)

the probability of behavior erive from contin#encies of survival or of
reinforcement. Instinct is 7a sum of psychic ener#y "hich imparts irection to
psycholo#ical processes$8 in the sense that innate susceptibilities to reinforcement not
only Stren#then behavior but #ive it irection by shapin# an maintainin# its
topo#raphy. *he susceptibilities are to be trace to their survival value in the
evolution of the species. Some scheules of reinforcement create 8stores of ener#y.8
Cthers lea to its absence in abulia or epression. *he 8#reat positive forces8 sai to
8"ell in our epths8 are merely the #reat thin#s "e mi#ht o$ #iven favorable
circumstances.

*he "or 8epth$8 common in psychoanalysis$ often makes the un"arrante
su##estion that an analysis is profoun$ but it may also be taken to refer to certain
spatial features of the min. *he nineteenth century psycholo#ist treate
consciousness as the: place in "hich sensations coul be observe$ but the space
occupie by the e#o$ supere#o$ an i is more comple). *he min has ifferent parts
inferre from ifferent kins of behavior. *o be of t"o mins about somethin# is to
have ifferent thin#s to o about it. *he term 8schi-ophrenia8 ori#inally meant 8split
min8 an is still misuse in that sense. *o be besie oneself is to be$ for the
moment$ t"o people. =ifferent kins of behavior are sai to be kept in ifferent
compartments of the min. 8In most human bein#s there is a repository of violence$
but the brain thro"s up a barrier$ a fence$ to keep it in check. Secobarbital . . . may
rip o"n this mental fence an permit violence to rush forthG @another interestin#
mi)ture of matter an minE. 3usic$ to a "ell kno"n statesman$ is 8an outlet for
passionate emotion$8 as if 8opera suenly breaks into his poli
THE I::ER WOR59 O2 MOTIVATIO: A:9 EMOTIO: 1,

tical life an knocks o"n the neat compartments bet"een emotion an reason.8

*he best kno"n ivision of min is bet"een conscious an unconscious;
represse "ishes an fears resie in the unconscious$ but they may break into the
conscious min. It is often sai$ particularly by psychoanalysts$ that behaviorism
cannot eal "ith the unconscious. *he fact is that$ to be#in "ith$ it eals "ith
nothin# else. *he controllin# relations bet"een behavior an #enetic an
environmental variables are all unconscious so lon# as they are not observe$ an it
"as ?reu "ho emphasi-e that they nee not be observe @that is$ consciousE to be
effective. It requires a special verbal environment to impose consciousness on behav&
ior by inucin# a person to respon to his o"n boy "hile he is behavin#. If
consciousness seems to have a causal effect$ it is the effect of the special.
environment "hich inuces self&observation.

*o increase a person.s consciousness of the e)ternal "orl is simply to brin#
him uner more sensitive control of that "orl as a source of stimulation. 3ar) an
others have trie to 8thro" people into a hi#her level of consciousness8 in brin#in#
them uner the control of aspects of their environment "hich "ere previously in&
effective. =ru#s "hich alter the control are sometimes sai to e)pan consciousness.

What behaviorism re'ects is the unconscious as an a#ent$ an of course it
re'ects the conscious min as an a#ent$ too. ! bio#raphy of 3ohamme asserts that 8it
is obvious to non&3uslims that the "ors "hich 3ohamme hear ... "ere ictate to
him by his unconscious ... the voice of !llah "as in fact the voice of 3ohamme.s
unconscious.8 But if anyone spoke$ it "as 3ohamme himself$ even thou#h he i not
observe himself oin# so. It "as 3ohamme as a person$

ABOUT BEHAVIORISM 1%6

"ith a history responsible for his bein# 3ohamme$ not some fra#mentary inner
a#ent$ to "hom "e must turn to e)plain the behavior.

It is often sai that there is an intrapsychic life of the min$ totally
inepenent of the physical "orl$ in "hich memories evoke memories$ ieas su##est
ieas$ an so on. (ere are a fe" e)amples of the intrapsychic life of motivation an
emotion: ?eelin#s of frustration prouce a sense of po"erlessness$ or impotence$
"hich in turn leas to apathy or to feelin#s of a##ression. 6esentment of authority
turn$ into a represse murerous ra#e$ "hich masks a "ish to surrener. Weakene
faith in the future leas to an)iety an epression$ "hich isrupts thou#ht processes.
*he rive to conform prevents a person from kno"in# his o"n fears$ an#ers$ or sense
of hopelessness.

By turnin# to the facts on "hich these e)pressions are base$ it is usually
possible to ientify the contin#encies of reinforcement "hich account for the intra&
psychic activities. !mon# the relevant facts are these: ?rustration is #enerate by
e)tinction$ "hich is also often responsible for a##ressive behavior. *he controllin#
measures use by an authority make it more likely that a person "ill escape or
counterattack$ an relevant conitions may be felt as resentment; at the same time
the measures may #enerate compliant behavior$ "hich is "hy the authorities use
them. *he boily conitions associate "ith compliance may not be felt if the con&
itions associate "ith escape or counterattack are stron#.

Freud#s Deense 'echanisms

Aife in the inner "orl of emotion an motivation is ramatically illustrate
by the ?reuian ynamisms$ or
THE I::ER WOR59 O2 MOTIVATIO: A:9 EMOTIO: 1%1

efense mechanisms. *hey have been efine as 8personality reactions by means of
"hich an iniviual attempts to satisfy his emotional nees; e.#.$ to establish
harmony amon# conflictin# strivin#s& to reuce feelin#s of an)iety or #uilt arisin# from
"ishes$ thou#hts$ an emotions that are not acceptable.8 !lter&native efinitions may
be erive from the contin#encies responsible for the behavior from "hich the
ynamisms are inferre. I shall consier three e)amples$ usin# efinitions from
Webster.s Third International.

Re#ression; 'A process or mechanism of e#o efense "hereby "ishes or
impulses that arc incapable of fulfillment are kept from or mae inaccessible to
consciousness.8 ?or 8"ishes or impulses8 rea the 8probability of behavior8; for
8incapable of fulfillment8 rea 8e)tin#uishe or punishe8; an for 8kept from or mae
inaccessible to consciousness8 rea 8not introspectively observe8 in the sense of
Chapter +. We then have this: behavior "hich is punishe becomes aversive$ an by
not en#a#in# in it or not 8seein#8 it$ a person avois conitione aversive stimulation.
*here are feelin#s associate "ith this$ but the facts are accounte for by the
contin#encies.

*he "or 8repression8 is part of an elaborate metaphor "hich #ives a ynamic
character to the effect of punishment. When feelin#s cannot be e)presse$ pressure is
sai to buil up until an e)plosion occurs. ! ne"spaper asserts that 8the fri#htenin#
thin# about quiet people like Bremer an Sirhan an Cs"al is that there must be
millions of them in the ;nite States$ holin# their ra#e insie them until&lackin# the
safety valve most iniviuals have&they e)ploe$8 But "hat is happenin# "hen a
person 8hols his ra#e insie him$8 an "hat is the 8safety valve8 throu#h "hich most
people let off emotional steam? *he ans"ers are to be foun in

ABOUT BEHAVIORISM 1%2

the conitions uner "hich behavior becomes very stron# because it cannot be
emitte.

We are often a"are of a stron# tenency to o or say somethin# althou#h an
occasion is lackin#; "e may be 8burstin# "ith #oo ne"s8 but have no one to tell it to.
3ore often$ ho"ever$ "e o not respon because "e have been punishe; "e have
8represse our ra#e8 because "e have been punishe for 8e)pressin# it$8 It somethin#
happens suenly in the manner of an e)plosion$ it is because the situation chan#es.
We fin someone to talk to an 8talk a steay stream$8 or our behavior becomes
stron#er than the incompatible behaviors "hich have previously isplace it. If an
e)plosion has un"ante consequences for others$ appropriate steps may be taken to
prevent it. *he 8.pressure may be reuce8 by proviin# an environment in "hich be&
havior may be freely emitte or 8impulses may be channele into more useful
outlets.8 8*oy #uns$8 says a psychiatrist$ 8allo" chilren to "ork out conflicts an ven&
tilate some of their a##ressive ur#es.8 We shoul say instea that they permit chilren
to behave a##ressively in unpunishe "ays.

-onversion; ... 3e transformation of an unconscious conflict into a symbolically
equivalent somatic symptom.8 Cne of the more ramatic manifestations of the$
suppose po"er of mental life is the prouction of physical illness. !s an iea in the
min is sai to move the muscles "hich e)press it$ so nonsomatic activities in the
psyche are sai to affect the soma. ?or e)ample$ ulcers are sai to be prouce by an
8inner&irecte ra#e.8 We shoul say instea that the conition felt as ra#e is
meically relate to the ulcer$ an that a comple) social situation causes both.
Similarly$ "hen it is
sai that spontaneous miscarria#e is ue to a possibly
THE I::ER WOR59 O2 MOTIVATIO: A:9 EMOTIO: 1%0

unconscious hatre of the chil or of the father$ "e may say instea that the
conition felt as hatre is meically relate to miscarria#e$ an that it must be
attribute in turn to a comple) social situation. *he ulcer an the miscarria#e are
8symbolically equivalent8 to ra#e an hatre in that they are associate "ith a hi#h
probability of "orkin# harm. Conversion oes not emonstrate min over matter; the
psychic oes not chan#e the physical. %hysical conitions$ many of them relevant to
behavior an felt in various "ays$ have physical @meicalE effects.

S+/li"ation; 8=ischar#e of instinctual ener#y an especially that associate
"ith pre&#enital impulses throu#h socially approve activities.8 ?or 8ischar#e of
ener#y throu#h activities8 rea 8behavior$8 an for 8instinctual8 an 8associate "ith
pre&#enital impulses8 rea 8ue to certain biolo#ical reinforcers.8 If t"o forms of
behavior are both reinforce an if only one of them is punishe$ the other is more
likely to occur.

*he other ?reuian ynamisms or efense mechanisms may be treate in the
same "ay. *hey are not psychic processes takin# place in the epths of the min$
conscious or unconscious; they are the effects of contin#encies of reinforcement$
almost al"ays involvin# punishment. !t best "e may say that they are "ays in "hich a
person efens himself a#ainst punishment by acquirin# behavior effective in the
"orl in "hich he lives @as e#oE$ reinforce in part because of susceptibilities to
reinforcement "hich are part of his #enetic eno"ment @as iE$ an not punishe by
other persons or by himself @as supere#oE.

It has been sai that 8inhibitin# forces "hich oppose the ischar#e of tension
are the immeiate sub'ect of psycholo#y$8 an if this is true$ it is only because
inhibitin#

ABOUT BEHAVIORISM 1%4

forces an the ischar#e of tension are fi#ures of speech referrin# to punishment an
reinforcement$ respectively.

Inner "auses

!n an#ry person may have a rapi pulse an a flushe face; his behavior may
be stron#ly focuse on the ob'ect of his an#er an uncontrolle by other features of
the environment; he may sho" a stron# tenency to harm that ob'ect @8I coul have
kille him8E or may actually harm him. (e may feel much of the conition of his boy
at such a time an take it as the cause of his behavior$ but it is in fact part of the
effect for "hich a cause is sou#ht. Both the behavior an the collateral conitions felt
are to be e)plaine. !fter all$ "hy did he act and feel an#ry?

When an anteceent incitement is not easily spotte$ the felt conition is
likely to be assi#ne a more important role. ! person "ho is an#ry 8but oes not kno"
"hy8 is more likely to attribute his behavior to his feelin#s. *here oes not seem to be
anythin# else to attribute it to. ! mil emotion or moo is often particularly har to
e)plain$ an the moo itself is therefore sai to be causally effective @althou#h "e
must still look for the sources of the moo if "e are to e)plain the behaviorE.

Bruckner reporte the occasion of a creative musical act in the follo"in# "ay:
8Cne ay I came home an felt very sa. *he thou#ht ha crosse my min that
before lon# the master FWa#nerH "oul ie$ an then the C sharp minor theme Fof the
!a#io of the Seventh SymphonyH came to me.8 *his is a strai#htfor"ar statement.
*he thou#ht may have 8crosse his min8 as a verbal response or in some form much
less easily
THE I::ER WOR59 O2 MOTIVATIO: A:9 EMOTIO: 1%5

ientifie. @It is not usefully ientifie by callin# it a thou#ht or iea.E Bruckner oes
not say that he then 8conceive of8 or 8invente8 or 8create8 the theme to e)press his
saness; it 'ust 8came to him.8 lie may have thou#ht it covertly as he mi#ht have sun#
it alou or playe it on the or#an. We o not nee to say that the theme came to him
because he "as feelin# sa; certain circumstances @ne"s of Wa#nerE #rod+*ed the
conitions felt as saness an inuce him to behave musically in a special "ay.

On another occasion his bio#rapher "rites: 84late by the completion of his
Seventh Symphony$ Bruckner turne back to the *e =eum$8 but i he turn back
because he "as elate or because the completion of his Seventh Symphony "as a
hi#hly reinforcin# event stren#thenin# the behavior involve in musical composition
an proucin# the conition felt as$ elation? *he elation one feels upon completin# a
ifficult task is only one of the states associate "ith positive reinforcement. Cne is
also sai to feet pleasure @reinforcement is pleasin#E$ satisfaction @etymolo#ically
relate$ as "e have seen$ to satiationE$ 'oy$ or happiness. *he conitions thus felt can
scarcely be responsible for the behaviors of "hich they are consequences$ but they
are often taken to e)plain the behaviors "hich follo".

3any suppose inner causes of behavior$ such as attitues$ opinions$ traits of
character$ an philosophies$ remain almost entirely inferential. *hat a person is pro&
labor$ plannin# to vote for a #iven caniate$ intelli#ent$ liberal$ or pra#matic is
kno"n not from "hat he feels but from "hat lie says or oes. :evertheless$ terms
referrin# to traits of character are freely use in e)plainin# behavior. ! politician
continues to run for office because of 8ambition$8 makes shay eals because of
8#ree$8 opposes measures to eliminate is

ABOUT BEHAVIORISM 1%

crimination because of 8moral callousness$8 hols the support of his follo"ers because
of his 8leaership qualities$8 an so on$ "here no evience of the inner causes is
available e)cept the behavior attribute to them.

So calle mental measurement has been concerne "ith the statistical
treatment of some of these inner posessions. 6epertoires can be sample an a
person rate quantitatively "ith respect to other persons in a #roup. Certain traits
can be reuce to factors or vectors of min$ an it is then easy to suppose that
somethin# more than an invente cause has been iscovere. But many specialists in
the fiel have concee that factors are classificatory schemes rather than causes
an that "hat one can preict about behavior by measurin# a mental trait is
preicte from other behavior$ presumably because it has similar causes.

*he inner "orl of the psyche has lent itself to structuralist theories. *he
spatial features of the unconscious$ preconscious$ an conscious mins seem to
compose a kin of topo#raphy not unlike the #eo#raphy of the earth. ?actor analysis
has le to many imensional representations of the min or personality.$ !n "here
there is structure$ evelopmentalism cannot be far behin. *raits of character have
been sai to have 8hien propensities to #ro"th.8 ! person is sai to pass throu#h
various sta#es from infancy to maturity to senescence. 4rik 4rikson.s ei#ht
psychosocial sta#es of e#o evelopment are efine in terms of feelin#s an states of
min$ but the sta#es are in the contin#encies #eneratin# the conitions felt or
introspectively observe. *he chil of one or t"o may be sai to sho" trust versus
mistrust; his behavior is reinforce mainly throu#h the meiation of others; an
consistent contin#encies bree trust$ "hile inconsistent bree mistrust. When three
or four$ the

THE I::ER WOR59 O2 MOTIVATIO: A:9 EMOTIO: 1%%

chil sho"s autonomy versus oubt; he is no" actin# on the environment lar#ely by
himself an may or may not be successful. ?ailure may be milly punishe$ an mil
punishment #enerates a conition felt as shame. !t four or five$ the opposition is
bet"een initiative an #uilt; the chil moves into ne" contin#encies$ an punishment
for failure may be more e)plicit$ the conition therefore bein# felt as #uilt rather
than shame. ?rom si) to ten$ inustry is contraste "ith inferiority; scheules of
reinforcement buil hi#h or lo" levels of behavioral stren#th. !ccorin# to 4rikson$
rule&#overne behavior be#ins to be important at this point also. *he other four
sta#es may be analy-e in a similar "ay in terms of the prevailin# contin#encies. *hey
are all sta#es in the evelopment not of an e#o but of a "orl.

Wh! .ook Inside/

*he internali-ation of intellect is fully matche by that of the life of emotion
an motivation. *urnin# from observe behavior to a fanciful inner "orl continues
unabate. Sometimes it is little more than a lin#uistic practice. We ten to make
nouns of a'ectives an verbs an must then fin a place for the thin#s the nouns are
sai to represent. We say that a rope is stron#$ an before lon# "e are speakin# of its
stren#th. We call a particular kin of stren#th tensile$ an then e)plain that the rope
is stron# because it possesses tensile stren#th$ *he mistake is less obvious but more
troublesome "hen matters are more comple). *here is no harm in sayin# that a flui
possesses viscosity$ or in measurin# an comparin# ifferent fluis or the same flui
at ifferent temperatures on some convenient scale. But "hat oes viscosity mean? !
sticky stuff prepare to trap birs "as once mae from vis*+s$ Aatin for mistletoe.
ABOUT BEHAVIORISM 1%)

*he term came to mean 8havin# a ropy or #lutinous consistency$8 an viscosity 8the
state or quality of bein# ropy or #lutinous.8 *he term is useful in referrin# to a
characteristic of a flui$ but it is nevertheless a mistake to say that a flui flo"s
slo"ly because it is viscous or possesses a hi#h viscosity. ! state or quality inferre
from the behavior of a flui be#ins to be taken as a cause.

Consier no" a behavioral parallel. When a person has been sub'ecte to
milly punishin# consequences in "alkin# on a slippery surface$ he may "alk in a
manner "e escribe as cautious. It is then easy$ to say that he "alks "ith caution or
that he sho"s caution. *here is no harm in this until "e be#in to say that he "alks
carefully because of his caution. Some people may have been born cautious in the
sense that they learn very quickly to move cautiously or become e)cessively cautious
even "hen not e)cessively punishe$ but the behavior at issue can usually be trace
to a history of punishin# consequences.

*he e)traorinary appeal of inner causes an the accompanyin# ne#lect of
environmental histories an current settin# must be ue to more than a lin#uistic
practice. I su##est that it has the appeal of the arcane$ the occult$ the hermetic$ the
ma#ical&&those mysteries "hich have hel so important a position in the history of
human thou#ht. It is the appeal of an apparently ine)plicable po"er$ in a "orl "hich
seems to lie beyon the senses an the reach of reason. it is the appeal still en'oye
by astrolo#y$ numerolo#y$ parapsycholo#y$ an psychical research.

!bstract nouns lea the reaer into the epths. 8Aiberality amon# the rich$8
sai :iet-sche$ 8is often only a kin of timiity.8 *here is somethin# 8eep8 in

THE I::ER WOR59 O2 MOTIVATIO: A:9 EMOTIO: 1%,

that ma)im "hich is missin# in a simple escription of the behavior: 86ich people #ive
not to please but to appease.8 4)planations in epth are common in historical "ritin#.
*he 6omans conquere the 4truscans an "ere astonishe at their spoils of "ar. Aater
they #ot even more from Cartha#e. *his is sai to have ha the fol lo"in# effect:
8Covetousness an #ree$ restraine "ithin the 6oman community by ancient rules of
behavior$ havin# once been let loose upon the forei#ner$ coul no lon#er be
restraine at home.8 We mi#ht move from traits of character to contin#encies of
reinforcement by sayin# that the behavior of takin# the property of others$ stron#ly
reinforce an unpunishe in "arfare$ became too stron# to be seriously affecte by
the punitive sanctions implie in 8ancient rules of behavior.8 But the lettin# loose of
covetousness an #ree seems to #o to the heart of the problem$ "here the mere
contin#encies remain on the surface.

*he theater an the novel "oul probably not survive if the ramatist an
novelist staye out of the epths. In The .ortrait o8 a Aay$ the youn# 6alph *ouchett
evelops tuberculosis an must remain inactive for a lon# time. *his oes not bother
him$ ho"ever$ because he has never been stron#ly incline to o anythin#. But that is
too superficial a statement for (enry /ames$ "ho puts it this "ay: 8! secret hoar of
inifference ... came to his ai an helpe to reconcile him to sacrifice.8

!ske "hether he "as not concerne "ith his o"n safety urin# a tri# to the
moon$ an astronaut replie that 8astronauts o feel concern$ but lon# an aruous
trainin# pro#rams buil the confience neee to offset this concern.8 *he statement
that a feelin# of confience offsets the feelin# of concern is seemin#ly more pro

ABOUT BEHAVIORISM 1)6

foun than that one feels concern "hen one oes not kno" "hat to o an that one
learns "hat to o in a trainin# pro#ram.

*o say that the 8central patholo#y of our ay is a failure of "ill$ "hich
brou#ht psychoanalysis into bein#8 seems more profoun than to say that$ in the "orl
of our ay very little behavior is positively reinforce an much is punishe an that
psychoanalysis came into bein# to arran#e better contin#encies. *o say that the
Inustrial 6evolution in 4n#lan improve the material conition of the "orkin#
classes but 8estroye craftsmanship an the intelli#ent 'oy of man in his aily "ork8
by alienatin# @$separatin#E him from the en prouct of his labor seems more
profoun than to say that it estroye the naturally reinforcin# consequences of
makin# thin#s$ for "hich the contrive reinforcers of "a#es "ere a poor substitute.

*he reaction of a "orker to a "elfare chiseler appears to epen on a history
of social contin#encies$ common in Western cultures$ in "hich shirkers are punishe
by "orkers$ the latter possibly feelin# a conition calle resentment. In one analysis
of the effects of a chiseler on a "orker$ 8"ork8 becomes 8sacrifice$8 "hich is sai to
be a 8voluntary virtue$ a meanin# the sacrificer has create out of the material
circumstances of his life.8 *he refusal of the "elfare chiseler to make sacrifices then
8calls into question the meanin# of @the "orker.sE act of self abne#ation8 an makes
that 8"ille$ create meanin# vulnerable.8 !n elaborate psychic operation involvin#
sacrifice$ meanin#$ virtue$ volition$ self&abne#ation$ an "ill has the kin of presti#e
accore the meieval sorcerer$ a presti#e enie to the behaviorist$ "ho simply
reports a set of social contin#encies.

*o take one other e)ample$ the position of the black
THE I::ER WOR59 O2 MOTIVATIO: A:9 EMOTIO: 1)1

minority in !merica has been escribe this "ay: When a once 8lar#ely po"erless8
#roup acquires a sense of #ro"in# po"er$ 8its members e)perience an intensifie nee
for self&affirmation$ ;ner the circumstances$ collective self&#lorification$ foun in
some measure amon# all #roups$ becomes a frequent an intensifie counterresponse
to lon#&stanin# belittlement from "ithout.8 *he first step is to strike out such
e)pressions as 8sense of$8 8e)perience a nee$8. 8self&affirmation . 8 8self&#lorification$8
an 8belittlement.8 ! translation then reas: 8When a #roup of people acquire po"er$
they speak about their #oo qualities$ an in oin# so contraict "hat has lon# been
sai about them by others.8 :o oubt they also feel certain states of their boies as
they o so$ but they o not act because they have a sense of po"er; they act an
have a sense of po"er because of the chan#es "hich have taken place in their
environment. *hey o not speak "ell of themselves because of 8collective self&
#lorification8; they speak "ell of themselves because it is reinforcin# to hear
themselves "ell spoken of$ an they are especially likely to o so "hen they have
previously not been "ell spoken of. *he behavior at issue can be observe in a sin#le
person: 8When a person is able to o so$ he "ill speak of his #oo qualities in
contraictin# "hat others have sai of him.8 *here is nothin# very surprisin# about
this or very ifficult$ but it lacks the epth of the appeal to a nee for self&
affirmation an counterresponses of self&#lorification.

The 3selessness o Inner "auses

*here are$ of course$ reasons "hy a flui flo"s slo"ly$ an a molecular
e)planation of viscosity is a step for"ar. *here are physiolo#ical reasons "hy a
person
ABOUT 9EHAVIORISM1)2

behaves in a manner "e call cautious; an the physiolo#ist "ill$ "e assume$
eventually tell us "hat they are. I must ask the reaer to "ait until Chapter 02 to
consier "hether "hat are felt or introspectively observe are the thin#s "hich "ill
eventually be reporte an analy-e by the physiolo#ist$ but some comment on the
e)planations they are sai to supply may be appropriate here.

*he e)ploration of the emotional an motivational life of the min has been
escribe as one of the #reat achievements in the history of human thou#ht$ but it is
possible that it has been one of the #reat isasters. In its search for internal
e)planation$ supporte by the false sense of cause associate "ith feelin#s an intro&
spective observations$ mentalism has obscure the environmental anteceents "hich
"oul have le to a much more effective analysis. *o ar#ue that 8mins kill$ not #uns8
may be simply to insist that "e shall not control assassins by makin# #uns unavailable$
but other means of control "ill be ne#lecte so lon# as "e accept the e)planation
that mins kill. *he ob'ection to the inner "orkin#s of the min is not that they are
not open to inspection but that they have stoo in the "ay of the inspection of more
important thin#s.

*he psyche$ like the min$ is a metaphor "hich is mae plausible by the
seemin# relevance of "hat a person feels or introspectively observes but "hich is es&
tine to remain forever in the epths. By contrast$ the environment is usually
accessible. We nee to kno" a #reat eal more about comple) contin#encies of rein&
forcement$ an it "ill al"ays be har to eal "ith that particular set to "hich any
one person is e)pose urin# his life$ but at least "e kno" ho" to #o about finin#
out "hat "e nee to kno".

*he ar#onauts of the psyche have for centuries saile
THE I::ER WOR59 O2 MOTIVATIO: A:9 EMOTIO: 1)0

the stormy seas of the min$ never in si#ht of their #oal$ revisin# their charts from
time to time in the li#ht of "hat seeme like ne" information$ less an less sure of
their "ay home$ hopelessly lost. *hey have faile to fin the Golen ?leece.

*heir pli#ht is su##este by the espair "ith "hich solutions are propose for
current problems. ! sin#le issue of a news#a#er re#orted commencement aresses
by three university presients$ "ho offere the follo"in# su##estions: @0E 8*he
confience an hopefulness an ea#erness to procee$ "hich historically have been
the fruits of faith$ are no" too frequently simply not there or too feebly there.8 @+E
8What this country esperately nees is a totality of outlook that "ill put a spiritual
face upon !merican society.8 @2E 8!merica has yet to release her moral po"er.8

*his kin of thin# has been #oin# on for centuries. It is surprisin# that so many
intelli#ent people refuse to ask "hat is "ron#.
)) he Self and *thers
'o Back

3t is often said that a science of beha/ior studies the human or.anism but ne.*ects the
person or se*f. 4hat it ne.*ects is a /esti.e of animism) a doctrine which in its crudest form he*d
that the bod- was mo/ed b- one or more indwe**in. spirits. 4hen the resu*tin. beha/ior was
disrupti/e) the spirit was probab*- a de/i*8 when it was creati/e) it was a .uidin. .enius or muse.
(races of the doctrine sur/i/e when we speak of a +ersonait-, of an e.o in e.o ps-cho*o.-) of
an 3 who sa-s he knows what he is .oin. to do and uses his bod- to do it) or of the ro*e a person
p*a-s as a persona in a drama) wearin. his bod- as a costume.

3n a beha/iora* ana*-sis a person 3s an or.anism) a member of the human species) which
has ac=uired a repertoire of beha/ior. 3t remains an or.anism to the anatomist and ph-sio*o.ist)
but it is a person to those to whom its beha/ior is important. Comp*e, contin

THE SE8# A42 OTHERS )8.

.encies of reinforcement create comp*e, repertoires) and) as we ha/e seen) different
contin.encies create different persons in the same skin) of which so6ca**ed mu*tip*e persona*ities
are on*- an e,treme manifestation. 4hat happens when a repertoire is ac=uired is the important
thin.. (he person who asserts his freedom b- sa-in.) 93 determine what 3 sha** do ne,t)9 is
speakin. of freedom in or from a current situation: the 3 who thus seems to ha/e an option is the
product of a histor- from which it is not free and which in fact determines) what it wi** now do.

A person is not an ori.inatin. a.ent8 he is a *ocus) a point at which man- .enetic and
en/ironmenta* conditions come to.ether in a 0oint effect. As such) he remains un=uestionab*-
uni=ue. Co one e*se %un*ess he has an identica* twin& has his .enetic endowment) and without
e,ception no one e*se has his persona* histor-. >ence no one e*se wi** beha/e in precise*- the
same wa-. 4e refer to the fact that there is no one *ike him as a person when we speak of his
identit-. %(he @atin idem means same) and when asked whether someone is rea**- so6and6so) we
ma- rep*- co**o=uia**-) 9(he sameG9 or) 9>imse*fG9 or we ma- sa- that a person who comp*ains
of bein. anno-ed b- his nei.hbors is 9the se*fsame person9 who anno-s others.&

A number of terms describin. a person and his re*ation to others need to be considered.

$nowing *neself

3n askin. what a person can know about himse*f) we are *ed at once to another =uestion: 5ho can
know about 9homA (he answer is to be found in the contin.encies which produce both a
knowin. se*f and a known. A distinction between two se*/es in the same
ABOUT BEHAVIORISM)8/

skin is made when we sa- that a tennis p*a-er 9.ets mad at himse*f9 because he misses an eas-
shot. >e is an.r- because somethin. has hurt him) and he has done the thin. that hurt8 hence be
is mad at himse*f. >e ma- e/en strike himse*f a..ressi/e*-. A simi*ar distinction is made in
se*f6know*ed.e.

A** species e,cept man beha/e without knowin. that the- do so) and presumab*- this was
true of man unti* a /erba* communit- arose to ask about beha/ior and thus to .enerate
se*f6descripti/e beha/ior. Se*f6know*ed.e is of socia* ori.in) and it is usefu* first to the com6
munit- which asks the =uestions. @ater) it becomes important to the person himse*f6for e,amp*e)
in mana.in. or contro**in. himse*f in wa-s to be) discussed short*-.

<ifferent communities .enerate different kinds and amounts of se*f6know*ed.e and
different wa-s in which peop*e e,p*ain themse*/es to themse*/es and others. Some produce the
deep*- introspecti/e intro/erted) or inner6directed) person) others the out.oin. e,tra/ert. Some
produce peop*e who act on*- after a carefu* consideration of the possib*e conse=uences) others
the thou.ht*ess and impu*si/e. Some communities produce peop*e particu*ar*- aware of their
reactions to art) music) or *iterature) others of their re*ations with the peop*e around them. (he
=uestions asked b- menta*istic ps-cho*o.ists and those asked b- beha/iorists natura**- produce
different kinds of se*f6know*ed.e. (he- first emphasi2e how a person fee*s about thin.s.

(here is *itt*e doubt of the historica* priorit- of the inner search. 3t was what Socrates
meant b- 9Onow th-se*f.9 %(hat in0unction appears on the wa** of a :oman bath beneath a
mosaic of a ske*eton6an anatomica* /ersion of the se*f.& ;ontai.ne spoke of 9sp-in. on himse*f9
and of 9disco/erin. the sprin.s which

THE SE8# A42 OTHERS )80

set him in motion.9 3t is the priorit- en0o-ed b- fee*in.s and introspecti/e*- obser/ed states o/er
past and present en/ironments.

Iuestions about fee*in.s tend to be c*ose*- associated with a sense of se*f or a se*f6ima.e.
(he- emphasi2e what a person is) his current state of bein.. +,istentia*ists) phenomeno*o.ists)
and humanistic ps-cho*o.ists ha/e encoura.ed se*f6obser/ation in this search for se*f. Qo.a has
been defined as a set of practices 9b- which the indi/idua* prepares for *iberation of the se*f.9
7n*- the *iberated se*f can assert) 93 do what 3 do because of what 3 am)9 or) 94hat 3 do not do or
wi** to do is not me.9 9Because 3 am what 3 am)9 said <iderot) 93 write the kind of p*a-s 3 do.9
Buffon put it in a we**6known phrase: %8e st-e, c:est ;homme$%

Bs-choana*-sis .i/es a person a c*earer ima.e of himse*f) main*- b- inducin. him to
e,p*ore his fee*in.s) and the se*f6know*ed.e it encoura.es is often ca**ed insi.ht) a term c*ose to
9introspection.9 (he patient is to *earn to fee* his own emotions) to acknow*ed.e fee*in.s as6
sociated with punished beha/ior) and so on.

Structure is natura**- emphasi2ed in an ana*-sis of bein.) and there is a re*ated /ersion of
de/e*opmenta*ism which emphasi2es becomin.. 1rom the present point of /iew an- chan.e is in
a repertoire) and it must be attributed to chan.in. contin.encies. 4hen a chan.e is disrupti/e) a
person ma- not fee* that he knows himse*f8 he is said to e,perience a crisis in his identit-. 3t is
difficu*t to maintain an identit- when conditions chan.e) but a person ma- concea* from himse*f
conf*ictin. se*/es) possib*- b- i.norin. or dis.uisin. one or more of them) or b- brandin. one) a
stran.er) as in e,p*ainin. uncharacteristic beha/ior b- sa-in.) 93 was riot m-se*f.9

(he /erba* communit- asks) 9>ow do -ou fee*A9
ABOUT BEHAVIORISM )88

rather than) 94h- do -ou fee* that wa-A9 because it is more *ike*- to .et an answer. 3t takes
ad/anta.e of the a/ai*ab*e information) but it has on*- itse*f to b*ame if other kinds of
information are not a/ai*ab*e. 3t has not unti* recent*-) induced peop*e to e,amine the e,terna*
conditions under which the- *i/e. As the re*e/ance of en/ironmenta* histor- has become c*earer)
howe/er) practica* =uestions ha/e be.un to be asked) not about fee*in.s and states of mind) but
about the en/ironment) and the answers are pro/in. increasin.*- usefu*.

The shift from introspecti/e to en/ironmenta* e/idence does not .uarantee that
se*f6know*ed.e wi** be accurate) howe/er. 4e do not a*wa-s obser/e the contin.encies to which
we are e,posed. 4e ma- keep records of what has happened) as in a diar-) but in .enera* our
information is sketch-. 4e are not a*wa-s watchin. what happens as we beha/e) and when asked
how we wou*d beha/e under .i/en circumstances) we often make a bad .uess) e/en thou.h we
ha/e been in simi*ar circumstances in the past. (hen) as usua*) we are *ike*- to e,p*ain the
ine,p*icab*e b- attributin. it to .enetic endowment6assertin.) 93 was born that wa-)9 or) 9(hat5s
the kind of person 3 am.9
3t is ne/erthe*ess important to e,amine the reasons for one5s own beha/ior as carefu**- as
possib*e because the- are essentia*) as 3 ha/e said) to .ood se*f6mana.ement. 5e shou*d not be
surprised that the more we know about the beha/ior of others) the better we understand
ourse*/es. 3t was a practica* interest in the beha/ior of 9the other one9 which *ed to this new kind
of se*f6know*ed.e) (he e,perimenta* ana*-sis of beha/ior) to.ether with a specia*
se*f6descripti/e /ocabu*ar- deri/ed from it) has made it possib*e to app*- to onese*f much of
what has been *earned about the beha/ior of others) inc*udin. other species.

THE SE8# A42 OTHER )8"

(hose who seek to know themse*/es throu.h an e,p*oration of their fee*in.s often c*aim
an e,c*usi/e kind of know*ed.e) 7n*- those who ha/e been ps-choana*-2ed) for e,amp*e) are.
said to know what ps-choana*-sis means) and the m-stic c*aims e,periences which cannot be
communicated or known to others e,cept throu.h simi*ar channe*s. But it ma- be ar.ued as we**
that on*- those who understand an e,perimenta* ana*-sis and its use in interpretin. human
beha/ior can understand themse*/es in a scientific or techno*o.ica* sense.

$nowing +nother Person

3n askin. wh- another person beha/es as be does) we ma- a*so distin.uish between what he fee*s
or introspecti/e*- obser/es and what has happened to him. <isco/erin. how he fee*s) or what he
thinks) is part of *earnin. what he is or is comin. to be or becomin.. A first step is to make
contact with him) possib*- in an 4encounter9 or 9confrontation.9 3n an- case) it re=uires .ood
9interpersona* re*ations9 and an abi*it- to share fee*in.s throu.h s-mpath-) a word which once
meant simp*- 9fee*in. with.9 Sensiti/it- trainin. is desi.ned to he*p. (he obser/er is to become
in/o*/ed and) *ike the mathematician who is said to think intuiti/e*- because he has not taken the
e,p*icit steps which *ead to a conc*usion) he is to intuit the fee*in.s of others that is) to know
them direct*- without necessari*- bein. ab*e to e,p*ain how he does so.

Ce/erthe*ess) one person does not make direct contact with the inner wor*d of another)
and so6ca**ed know*ed.e of another is often simp*- an abi*it- to predict what he wi** do. (hus)
how we** the members of a trainin. staff percei/e %and hence know& their train

ABOUT B#HAVIORISM )"'

ees has been said to be indicated b- how we** the- can predict how the trainees wi** answer a set
of =uestions. But we understand another person in part from his e,pression of fee*in.s. Actors
were once said to be ab*e to 9re.ister9 0o-) sorrow) and so on with facia* e,pressions) postures)
and mo/ements) and the audience read these e,pressions and hence understood the characters
and their moti/es) presumab*- because it had *earned to do so in rea* *ife with rea* peop*e.

4e can use an e,pression of fee*in.s b- askin. how we wou*d beha/e if we ourse*/es bad
the fee*in.s thus e,pressed. 7r we can ask what kinds of beha/ior a .i/en e,pression has tended
to accompan- in the past. (hus) we predict what a person who *ooks an.r- wi** do not b-
stoppin. to ask what we wou*d do if we *ooked an.r- but b- rememberin. what peop*e who *ook
an.r- .enera**- do. (he attribution of fee*in.s to others is ca**ed empath-. A person is said to
9pro0ect his fee*in.s9 into another. 4hen he pro0ects them into an inanimate thin.) he is
ob/ious*- makin. a mistake) and his beha/ior has been ca**ed the pathetic fa**ac-. (he 9an.r-
sea9 beha/es in an an.r- fashion) but we do not suppose that it fee*s an.r-. 4e mere*- infer that
for a time it wi** continue to beha/e in an an.r- wa-. 4e can a*so be wron. when we pro0ect
fee*in.s onto other peop*e. A person can 9act bra/e*- whi*e fee*in. afraid)9 but he does so with
different parts of his bod-) with different repertoires. 4e ma- be ab*e to disco/er how he
9rea**-9 fee*s b- a*terin. the contin.encies. 3f he is actin. bra/e*- because of pre/ai*in. socia*
contin.encies in which 9showin. fear9 is punished) we ma- be ab*e to chan.e the contin.encies
so that he wi** act as if afraid. 4hat he (et was in both cases) .enerated b- certain features of the
situation rather than b- the beha/ior which simu*ated bra/er-. A person who sa-s
THE SE8# A42 OTHERS )")

he (ees bra/e when he is rea**- fee*in. afraid is *ike a person who acts bra/e*- when fee*in.
afraid) and we can disco/er what *ie 9rea**-9 fee*s b- a*terin. the contin.encies. Bs-chotherap- is
particu*ar*- important when the contin.encies responsib*e for a /erba* report are so powerfu* that
the person himse*f does not 9know that he is afraid.9 (he therapist 9he*ps him to disco/er his
fear.9 4hen he acts bra/e*- whi*e fee*in. afraid) that is the kind of person he is at that moment.
4e do not need to assume that there is a fearfu* person *urkin. in the depths.

4e mistrust reports of fee*in.s) especia**- when the- conf*ict with other e/idence. A
curious e,amp*e was common in the ear*- da-s of anesthesia) when man- peop*e resisted a
ma0or operation on the) .rounds that the dama.e done to the bod- was c*ear*- associated with
pain and that it was possib*e that the anesthetic mere*- b*ocked the e,pression) to.ether with its
*ater reco**ection) rather than the pain itse*f.

4e find it easier to know what another person is fee*in. if he tries to communicate or
con/e- his fee*in.s /erba**-. Con/e- means to transport or transmit) and communicate means to
make common to both speaker and *istener) but what is rea**- con/e-ed or made commonA 3t is)
of course) =uite inade=uate to sa- that 9man trans*ates his e,perience into sound wa/es that
another person can understand6that is) so that the *istener can retrans*ate the sounds into a
comparab*e e,perience.9 (he meanin. of an e,pression is different for speaker and *istener8 the
meanin. for the speaker must be sou.ht in the circumstances under which he emits a /erba*
response and for the *istener in the response be makes to a /erba* stimu*us. At best the end
product of communication cou*d be said to be the fact that the *istener5s response is appropriate
to the speaker5s situa
ABOUT B#HAVIORISM )",

tion. A description of the bodi*- state fe*t b- the speaker does not b- itse*f produce a simi*ar state
to be fe*t b- the *istener. 3t does not make a fee*in. common to both.

Another techni=ue for 9communicatin. a fee*in.9 is to describe a situation which arouses
the same fee*in.. As we describe somethin. b- sa-in. what it *ooks ike, and thus enab*e the
*istener to respond to it as he has a*read- responded to somethin. e*se) so we can induce the
*istener to fee* as we fee* b- describin. a situation which creates a condition fe*t in the same
wa-. 4e saw an e,amp*e of this in Oeat5s report of how he fe*t on first *ookin. into Chapman5s
>omer. (he no/e*ist 4 communicates9 with the reader b- describin. situations which .enerate
fee*in.s. %(he same practice is usefu* in 9communicatin. ideas9: an ar.ument is de/e*oped from
which the reader comes to the same conc*usion as the writer.&

(erms describin. pri/ate e/ents are necessari*- ine,act. (his is true of the wor*d of ideas
%it does not he*p much to be to*d that 9a .ood *ecturer shou*d communicate bein.9&) and e/en
more specific references to 9what is in the speaker5s mind9 are fau*t-. Cot a** contin.encies can
be rep*aced with ru*es) and some contin.enc-6shaped beha/ior is be-ond the reach of /erba*
description. Simi*ar*-) the most precise description of a state of fee*in. cannot correspond
e,act*- to the state fe*t. (he fee*in.s of the m-stic or the aesthete are 9ineffab*e)9 and there are
other fee*in.s that can be known on*- b- passin. throu.h a re*e/ant histor-. 7n*- one who has
*i/ed in a concentration camp can rea**- know what 9it fee*s *ike)9 because there is nothin. *ike it
to .enerate comparab*e fee*in.s in others. 3f it is true that on*- those who ha/e been
ps-choana*-2ed can know
THE SE8# A42 OTHERS )"3

what it fee*s *ike) then presumab*- there is nothin. e*se that fee*s *ike it.

4e tr- to disco/er how another person fee*s for man- reasons. A .ood dea* of our
beha/ior is reinforced b- its effect on others) and it is presumab*- more reinforcin. if the effect
is c*ear. (hus) we act to reinforce those we *ike or *o/e and to a/oid harmin. them) in part
because of what the- do in return. %(he tendenc- cou*d be innate) since there is sur/i/a* /a*ue)
for e,amp*e) in the beha/ior of a mother who feeds and cares for her -oun. and protects them
from harm and who) in doin. so) pro/ides conditions which c*assif- as positi/e and ne.ati/e
reinforcers) but socia* contin.encies of reinforcement .enerate comparab*e beha/ior.& 3t is
important that the recipient show that we ha/e been successfu*) and he can do so b- reportin. his
fee*in.s. A person bein. massa.ed sa-s that it fee*s .ood8 a person for whom a particu*ar piece
of music is bein. p*a-ed sa-s that he *ikes it. 4hen these 9si.ns of fee*in.9 are absent) we ma-
ask or otherwise in/esti.ate how a person fee*s.

(here ma- seem to be a more compe**in. reason for probin. the fee*in.s of others. 3f it is
9not the beha/ior that counts but how a person fee*s about his beha/ior)9 the disco/er- of
fee*in.s shou*d be the first order of business. But how a person fee*s about his beha/ior depends
upon the beha/ior and upon the conditions of which it is a function) and we can dea* with these
without e,aminin. fee*in.s. 4hen we are he*pin. peop*e to act more effecti/e*-) our first task
ma- seem to be to chan.e how the- fee* and thus how the- wi** act) but a much more effecti/e
pro.ram is to chan.e how the- act and thus) incidenta**-) how the- fee*.

3n a beha/ioristic ana*-sis knowin. another person is

ABOUT BEHAVIORISM )"&

simp*- knowin. what he does) has done) or wi** do and the .enetic endowment and past and
present en/ironments which e,p*ain wh- he does it. 3t is <ot an eas- assi.nment) because man-
re*e/ant facts are out of reach) and each person is indubitab*- uni=ue. But our know*ed.e of
another person is *imited b- accessibi*it-) not b- the nature of the facts. 4e cannot know a**
there is to know) as we cannot know a** we shou*d *ike to know about the wor*ds of ph-sics and
bio*o.-) but that does not mean that what remains unknown is of a different nature. As in other
sciences) we often *ack the information necessar- for prediction and contro* and must be satisfied
with interpretation) but our interpretations wi** ha/e the support of the prediction and contro*
which ha/e been possib*e under other conditions.

4e can know another person in the other sense of knowin. discussed in Chapter #. 4e
understand other peop*e short of takin. action) and the mere perception of others must be
inc*uded amon. our responses to them. A** this depends upon what others do) much more than
upon what the- fee* or report the- fee*.

!anaging *neself

Se*f6mana.ement raises the same =uestion as se*f6know*ed.e: 4ho are the mana.in. and
mana.ed se*/esA And a.ain the answer is that the- are repertoires of beha/ior. (he inte**ectua*
se*f6mana.ement discussed in Chapter ! is a matter of chan.in. a situation unti* a response
appears which so*/es a prob*em) the prob*em6so*/in. repertoire makin. the repertoire containin.
the successfu* so*ution more effecti/e. (he two repertoires are more easi*- distin.uished in
ethica* se*f6mana.ement. (he mana.ed se*f is composed of what is si.nificant*- ca**ed se*fish
beha/ior6the prod

THE SE8# A42 OTHERS )".

uct of the bio*o.ica* reinforcers to which the species has been made sensiti/e throu.h natura*
se*ection. (he mana.in. se*f) on the other hand) is set up main*- b- the socia* en/ironment)
which has its se*fish reasons for teachin. a person to a*ter his beha/ior in such a wa- that it
becomes *ess a/ersi/e and possib*- more reinforcin. to others.

Se*f6mana.ement is often represented as the direct manipu*ation of fee*in.s and states of
mind. A person is to chan.e his mind) use his wi** power) stop fee*in. an,ious) and *o/e his
enemies. 4hat he actua**- does is chan.e the wor*d in which he *i/es. 3n both inte**ectua* and
ethica* se*f6mana.ement he ana*-2es contin.encies and ma- e,tract and app*- 6ru*es. But /er-
*itt*e se*f6mana.ement in this sense cou*d be *earned in one *ifetime. >ence the /a*ue of fo*k
wisdom) ru*es of thumb) pro/erbs) ma,ims) and other ru*es to be fo**owed to ad0ust more
e,pedient*- to the contin.encies the- describe. An i**uminatin. e,amp*e is the 'o*den :u*e. 3t
wou*d be impossib*e to construct a tab*e of commandments app*icab*e to a** the thin.s peop*e do
which affect others) but to disco/er whether a +articuar act is *ike*- to be punished because it
affects others a/ersi/e*- the indi/idua* is en0oined to e,amine the effect on himse*f. (his is the
ear*- and ne.ati/e form of the :u*e) but he ma- a*so *ook for reinforcin. effects. (he 0oint :u*e
te**s him to a/oid actin. if the effect wou*d be a/ersi/e to himse*f and to act if the effect wou*d
be reinforcin.. Cote that he is not asked to e,amine his putati/e (eein*s or to predict the fee*in.s
his beha/ior wou*d induce in others8 he is to see whether it is the kind of conse=uence *ie wou*d
act to achie/e. 3n e,aminin. such an effect on himse*f %as b- reca**in. his histor- or .enera*i2in.
from it&) he ma- we** respond to conditions of his own bod- rather than to the chan.es

ABOUT BEHAVIORISM )"/

induced in his beha/ior. (he conditions fe*t in association with reinforcers are sa*ient8 but
se*f6mana.ement is concerned with conse=uences) man- of them due to action taken b- others)
and the ru*e is more e,act*- app*ied if a person reco**ects not what he has fe*t but what he has
done when others ha/e treated him in a .i/en wa-.

Some we**6known techni=ues of se*f6mana.ement are desi.ned to brin. a person5s histor-
into p*a- in offsettin. an a/ersi/e effect) 1or e,amp*e) drinkin. a*coho*ic be/era.es often has
two opposin. conse=uences: an immediate reinforcement and a deferred punishment. After bein.
punished) a person ma- 9reso*/e9 not to drink a.ain. A reso*ution is a kind of se*f6made ru*e)
desi.ned to e,tend the effect of punishment into the future) but on a *ater occasion the immediate
reinforcin. effect ma- sti** take o/er. :eca**in. the reso*ution is a .esture of se*f6mana.ement)
thou.h possib*- ineffecti/e. A/oidin. situations in which one is *ike*- to drink %9a/oidin.
temptation9& is possib*- more effecti/e.

A common techni=ue of inte**ectua* se*f6mana.ement is to arran.e a situation6for
e,amp*e) a stud- or studio 6in which there is *itt*e to interfere with a .i/en kind of beha/ior) (he
c*oister and the hermita.e ha/e simi*ar effects in ethica* se*f6mana.ement. (he artist who paints
photo.raphica**- is under the powerfu* contro* of his mode*) but if he can brin. his persona*
histor- into p*a-) his work wi** show a kind of .enera*it-) because it wi** be *ess c*ose*- tied to
one situation. >e wi** ha/e 9e,tracted the essentia*s9 b- attenuatin. the contro* e,erted b- the
current settin.. (he same princip*e under*ies the practice of Ren) in which the archer) for
e,amp*e) *earns to minimi2e the particu*ar features of a sin.*e instance. Both the artist and the
archer are said

THE SE8# A42 OTHERS )"0

to 9transcend9 the immediate situation8 the- become 9detached9 from it.

Bersona* histor- asserts itse*f in se*f6contro* or se*f6mana.ement in other wa-s. (he
indi/idua* who refuses to 9.o under9 in a concentration camp) who is not 9broken9 b- efforts
made to demean or destro- his di.nit- or identit-) has transcended his current en/ironment. (o
sa- that he is ab*e to in0ect a different meanin. into that en/ironment is simp*- to sa- that he is
under the more Bowerfu* contro* of his histor-.

(he .oa* of se*f6mana.ement is often ca**ed se*f6fu*fi**ment or se*f6actua*i2ation.
1u*fi**ment seems to be concerned with achie/ement) with a/oidin. restraints and disco/erin.
positi/e reinforcers. Actua*i2ation seems to ha/e more to do with ma,imi2in. .enetic and
en/ironmenta* histories in order to free a person from immediate settin.s. 3n both cases the
emphasis is c*ear*- upon the here and now) oi* bein. or we**6bein. or momentar- becomin..

A .ood dea* of interest has recent*- been shown in the so6ca**ed se*f6contro* of autonomic
responses) such as chan.es in heart rate) or b*ood pressure) or b*ushin.) or sweatin.. (hese ref*e,
mechanisms ha/e been ca**ed in/o*untar-) and as we saw in Chapter 4) this wou*d Seem to set
them apart from operant beha/ior) but the conditions needed for operant conditionin. can be ar6
ran.ed. Autonomic beha/ior is usua**- concerned with the interna* econom-) and there ha/e
been few effects on the en/ironment which wou*d make operant conditionin. re*e/ant) but a
conspicuous indicator that a response is occurrin. can be set ti+, and operant contin.encies can
thus be estab*ished. A .i/en heart rate) for e,amp*e) can turn on a *i.ht) which is then fo**owed
b- a reinforcin. conse=uence. But s*owin. or speedin. the pu*se is no more se*f6contro* than
s*owin. or speed
ABOUT BEHAVIOMSM )"8

in. one5s stride when wa*kin.. (he on*- difference is that the pu*se is not norma**- fo**owed b-
reinforcin. conse=uences which brin. it under operant contro*. Conse=uences are sometimes
made more conspicuous in the operant conditionin. of ske*eta* musc*es) (hus) it is easier to *earn
to wi..*e one5s cars b- *ookin. in a mirror to impro/e feedback) and s*i.ht mo/ements of a
partia**- para*-2ed *imb are sometimes amp*ified for the same reason.

7ne can contro* one5s pu*se to some e,tent b- beha/in. in wa-s which affect it) speedin.
the heart rate b- e,ercisin. /io*ent*- and s*owin. it b- re*a,in.. (he direct operant contro* of
autonomic beha/ior can be demonstrated on*- when indirect contro* is e*iminated. ;an- -ears
a.o a co**ea.ue and 3 tried to reinforce chan.es in the /o*ume of the forearm) presumab*- re6
f*ectin. the re*a,ation of b*ood /esse*s. 7ne of us wou*d put his forearm in a water6fi**ed 0acket
%ca**ed a p*eth-smo.raph&) the /o*ume of which was indicated on a dia*. 4e found that we cou*d
mo/e the dia* in a direction which indicated that the /o*ume of our arm had increased) but we
*ater disco/ered that we were doin. so b- breathin. more and more deep*-. B- ho*din. a .reater
amount of residua* air in our *un.s) we were s=uee2in. b*ood into the arm. (here are wa-s in
which these mediatin. responses can be e*iminated) and the pure operant contro* of autonomic
beha/ior ma- be possib*e. 3t is not) howe/er) the se*f6mana.ement with which we are here
concerned.

4hen techni=ues of se*f6mana.ement ha/e been *earned) the instructiona* contin.encies
maintained b- the /erba* communit- ma- no *on.er be needed. Beha/ior resu*tin. from .ood
se*f6mana.ement is more effecti/e and hence .enerous*- reinforced in other wa-s. 3t is possib*e
that a much more precise kind of contro*
THE SE8# A42 OTHERS )""

ma- be.in to be e,erted b- pri/ate effects) in which case the prob*em of pri/ac- faced b- the
/erba* communit- is surmounted. Se*f6mana.ement then becomes as automatic in its dependence
on pri/ate stimu*i as the ski**ed mo/ements of an acrobat) but a*thou.h these contin.encies ma-
*ead to effecti/e pri/ate se*f6stimu*ation) the- do not *ead to se*f6know*ed.e. 4e ma- be as un6
conscious of the stimu*i we use in se*f6mana.ement as of those we use in e,ecutin. a
handsprin..

!anaging +nother Person

7ne person mana.es another in the sense in which he mana.es himse*f. >e does not do so
b- chan.in. fee*in.s or states of mind. (he 'reek .ods were said to chan.e beha/ior b- .i/in.
men and women menta* states) such as pride) menta* confusion) or coura.e) but no one has been
successfu* in doin. so since. 7ne person chan.es the beha/ior of another b- chan.in. the wor*d
in which he *i/es. 3n doin. so) he no doubt chan.es what the other person fee*s or introspecti/e*-
obser/es.

O+erant <onditionin*$ +/er-thin. we know about operant conditionin. is re*e/ant to
makin. beha/ior more or *ess *ike*- to occur upon a .i/en occasion. (his is the traditiona* fie*d
of rewards and punishments) but much sharper distinctions can be made in takin. ad/anta.e of
what we know about contin.encies of reinforcement. Hnfortunate*-) the reinforcers most often
used are ne.ati/e: .o/ernmenta* and re*i.ious contro* is based main*- on the threat of
punishment %9power9&) and noninstitutiona* practices are often of the same sort. Amon. positi/e
reinforcers are the .oods and mone- of economic contro* in a.ricu*ture) trade)

ABOUT BEHAVIORISM ,''

and industr- and) *ess forma**-) in dai*- *ife %9wea*th9 or 9pri/i*e.e9&. 3nterpersona* contact is
fre=uent*- a matter of appro/a* %9presti.e9& or censure) some forms of which are probab*-
effecti/e for .enetic reasons %9(he desire for approbation is perhaps the most deep*- seated
instinct of ci/i*i2ed man9&) but which usua**- deri/e their power from their e,chan.e with other
reinforcers.

3n traditiona* terms) one person arran.es positi/e or ne.ati/e contin.encies in order to
create interests) pro/ide encoura.ement) insti** incenti/es or purposes) or raise consciousness in
another person. 3n doin. so) he brin.s him under the contro* of /arious features of his
en/ironment. >e discontinues reinforcement in order to dissuade or discoura.e. >e uses
reinforcers deri/ed from deferred conse=uences to 9.i/e a person somethin. to *ook forward to.9
3n doin. so) he need not promote se*f6know*ed.e) but an increase in se*f6know*ed.e is re*e/ant
%94e must make the actua* pressure more pressin. b- addin. to it the consciousness of pres6
sure9&.

2escribin* <ontin*encies$ Arran.in. contin.encies of operant reinforcement is often
confused with describin. them. (he distinction is as important as that between
contin.enc-6shaped and ru*e6.o/erned beha/ior. 4hen we warn a person b- sa-in.) 9Come
inside. 3t is .oin. to rain)9 or b- puttin. up a si.n at an intersection readin.) 9Stop)9 we describe
beha/ior %comin. in or stoppin.& and identif- or imp*- re*e/ant conse=uences. 4e do not
necessari*- arran.e the contin.encies. A stop si.n ma- simp*- indicate the kind of intersection at
which dri/ers are *ike*- to ha/e troub*e) as the si.n 9(hin 3ce9 beside a pond deters the skater
without threatenin. puniti/e action b- the authorities.

(>+ S+@1 AC< 7(>+:S 2$1

But contri/ed a/ersi/e conse=uences are usua**- added. (he chi*d who sta-s out when to*d to
come in wi** not on*- .et wet) he wi** be punished for disobedience) (he dri/er who does not
stop at the intersection wi** not on*- run the risk of an accident) be wi** .et a ticket. %(he si.n
wi** be particu*ar*- effecti/e if a punisher6a po*iceman6is /isib*e.&

A warnin.) *ike the ru*es discussed in Chapter ") .i/es e,p*icit reasons in the form of a
%possib*- incomp*ete& description of contin.encies. A person who responds because of a warnin.
is beha/in. rationa**-) in the sense of app*-in. a ru*e) and this is particu*ar*- *ike*- to be said if)
thou.h he ma- ha/e *earned to respond because of past warnin.s) he does so now because he has
ana*-2ed the situation and) so to speak) warned himse*f. >e describes his own beha/ior and the
contin.encies responsib*e for it and as a resu*t is more *ike*- to beha/e in an appropriate wa- on
future occasions. (he *aw makes an important point of this8 the person who has wei.hed the
conse=uences of his action) who knows the effect his beha/ior wi** ha/e) is especia**- sub0ect to
punishment. 4e a*so ta*k about conse=uences66we supp*- reasons6when we e,hort a person to
act or ur.e or persuade him to act) (o ur.e is to make more ur.ent b- addin. conditioned
a/ersi/e stimu*i8 to persuade is to add stimu*i which form part of an occasion for positi/e
reinforcement. A more e,p*icit kind of ru*e is a contract. A *abor contract specifies amon. other
thin.s what a worker is to do and how much *ie is to be paid. A contract is put in force when
chi*dren are to*d that if the- beha/e we**) the- wi** .et a treat. (he worker and the chi*d ma- then
beha/e in order to be paid or treated) respecti/e*-) but the beha/ior ma- be weak. (he ru*e ma-
ha/e to be supp*emented b- additiona*

ABOUT BEHAVIORISM ,',

contin.encies) such as a super/isor5s threat of dischar.e or repeated si.ns of disappro/a* from a
parent.

Emotiona and Moti!ationa Measures$ 4hen we are in a position to do a person .ood6that
is) do somethin. he ca**s .ood6we can make that somethin. contin.ent on a .i/en topo.raph- of
beha/ior) which is then stren.thened) and we can brin. beha/ior under the contro* of a .i/en
stimu*us. 3f we 9do .ood9 without respectin. an- contin.ent re*ation.) we ma- satiate a person
and in doin. so reduce both the probabi*it- that he wi** en.a.e in beha/ior reinforced b- that
.ood and his susceptibi*it- to further reinforcement b- it. 4e ma- a*so create an emotiona*.
disposition to do .ood to us. Contrariwise) b- withho*din. the .ood) we. ma- e,tin.uish an-
beha/ior which has been reinforced b- it) but if we withho*d without respect to what is bein.
done) we create a state of depri/ation in which beha/ior reinforced b- that .ood is stron. and in
which the .ood is hi.h*- reinforcin.) and we create an emotiona* disposition to barm us. 4e
ourse*/es and the ob0ect of our attention ma- fee* or introspecti/e*- obser/e man- re*e/ant states
of our bodies) but the mana.ement of the contin.encies is the effecti/e step.

A number of fami*iar fie*ds of mana.ement ma- be brief*- discussed.

Teachin*$ +/er-one has suffered) and unfortunate*- is continuin. to suffer) from
menta*istic theories of *earnin. in education. 3t is a fie*d) in which the .oa* seems to be ob/ious*-
a matter of chan.in. minds) attitudes) fee*in.s) moti/es) and so on) and the +stab*ishment is
therefore particu*ar*- resistant to chan.e. Qet the point of education can be stated in beha/ioria*
terms: a teacher arran.es contin.encies under which the student ac=uires beha/ior which wi** be
usefu* to
THE SE8# A42 OTHERS ,'3

him under other contin.encies *ater on. (he instructiona* contin.encies must be contri/ed8 there
is no wa- out of this. (he teacher cannot brin. enou.h of the rea* *ife of the student into the
c*assroom to bui*d beha/ior appropriate to the contin.encies he wi** encounter *ater. (he
beha/iors to be constructed in ad/ance are as much a matter of producti/e thinkin. and creati/it-
as of p*ain facts and ski**s.

>ere is a samp*e of what is standin. in the wa- of effecti/e education: 3t is said that
9attitudes e,pressed in the structure of schoo* s-stems affect the co.niti/e and creati/e potentia*
of /irtua**- e/er- chi*d) as do the fee*in.s and persona*ities of teachers and their super/isors.9
(he 9attitudes e,pressed in the structure of schoo* s-stems9 presumab*- represent the beha/ior of
desi.nin. and constructin. schoo*s and instructiona* pro.rams8 the 9fee*in.s and persona*ities of
teachers and their super/isors9 are presumab*- inferred from their beha/iors8 and the thin.s
which 9affect the co.niti/e and creati/e potentia*9 of a chi*d are presumab*- the conditions under
which the chi*d ac=uires the kinds of beha/ior discussed in Chapter !. A trans*ation reads: 9(he
inte**ectua* and creati/e beha/ior of a chi*d is chan.ed b- the schoo* to which he .oes) its
instructiona* pro.rams) and the beha/ior of his teachers and super/isors.9 (his *acks the
profundit- of the ori.ina*) but profundit- here is certain*- obscurit-) and the trans*ation has the
merit of te**in. us where to be.in to do somethin. about teachin..

+ducation co/ers the beha/ior of a chi*d or person o/er a period of man- -ears) and the
princip*es of de/e*opmenta*ism are therefore particu*ar*- troub*esome. (he metaphor of .rowth
be.ins in the 9kinder.arten9 and continues into 9hi.her9 education) di/ertin. atten
ABOUT BEHAVIORISM ,'&

tion from the contin.encies responsib*e for chan.es in the students5 beha/ior.

He+in*$ Bs-chotherap- has been much more e,p*icit*- committed to menta*istic s-stems
than has education. (he i**ness which is the ob0ect of therap- is ca**ed menta*) and we ha/e
a*read- e,amined 1reud5s menta* apparatus and a few intraps-chic processes said to be disturbed
or deran.ed in the menta**- i**. 4hat is wron. is usua**- e,p*ored in the rea*m of fee*in.s. %At
one time it was su..ested that the ps-chiatrist shou*d take @S< in order to disco/er what it fee*s
*ike to be menta**- disturbed.&

;easures taken to chan.e fee*in.s66as in 9de/e*opin. the e.o9 or 9bui*din. a /ita* sense
of se*f96work b- constructin. contin.encies of reinforcement) b- ad/isin. a patient where
fa/orab*e contin.encies are to be found) or b- supp*-in. ru*es which .enerate beha/ior *ike*- to
be reinforced in his dai*- *ife. Beha/ior therap- is often supposed to be e,c*usi/e*- a matter of
contri/in. reinforcin. contin.encies) but it =uite proper*- inc*udes .i/in. a patient warnin.s)
ad/ice) instructions) and ru*es to be fo**owed.

4hen a prob*em ca**in. for therap- is due to a shorta.e of socia* or intimate*- persona*
reinforcers) a so*ution ma- be difficu*t. 3t ma- be ob/ious that a person wou*d profit from
reinforcement with attention) appro/a*) or affection) but if these are not the natura* conse=uences
of his beha/ior6if he does not merit attention) appro/a*) or affection6it ma- not be possib*e to
contri/e the needed contin.encies. Simu*ated attention) appro/a*) or affection wi** e/entua**-
cause more prob*ems than it so*/es) and e/en the deiberate use of deser/ed attention cheapens
the coina.e.

94hat is needed)9 sa-s Car* :o.ers) 9is a new concept of therap- as offerin. he*p) not
contro*.9 But these
THE SE88 A42 OTHERS ,'.

are not a*ternati/es. 7ne can he*p a person b- arran.in. an en/ironment which e,erts contro*)
and if 3 am ri.ht) one cannot he*p a person without doin. so. So ca**ed humanistic ps-cho*o.ists
contro* peop*e if the- ha/e an- effect at a**) but the- do not a**ow themse*/es to ana*-2e their
practices. 7ne unfortunate resu*t is that the- cannot teach them6and ma- e/en sa- that teachin. is
wron.. 9>e*p9 points to the interests of the person he*ped and 9contro*9 to the interests of the
contro**er) but before we decide that the first is .ood and the second bad) we shou*d ask whether
the contro**er is affected b- his own .ood or the .ood of others. 4e must *ook at wh- peop*e
he*p others) e,ertin. contro* as the- do so. (he cu*ture of the therapist shou*d *ead him to act in
wa-s which are .ood 6for the person he is he*pin.) and the prob*em of those who are concerned
for therap- is to .enerate such a cu*ture) not to find humane therapists. But this is to anticipate
the ar.ument of the fo**owin. chapter.

6o!ernin*$ 3n the broadest sense the term shou*d inc*ude a** mana.ement) but it is usua**-
confined to .o/ernmenta* and re*i.ious practices) particu*ar*- those which are puniti/e and
which are said to bui*d a sense of responsibi*it-. @ike dut- %what is due or owed to others& and
ob*i.ation %what one is obi*ed to pa-&) responsibi*it- su..ests a/ersi/e conse=uences) and we
sometimes sa- that a person is responsib*e simp*- in the sense that he responds to a/ersi/e
contin.encies. 4e ho*d him responsib*e b- maintainin. such contin.encies. %4e ho*d him
accountab*e in the more .enera* sense of keepin. an account of his beha/ior to see whether it
meets specifications upon which escape from punishment ma- be contin.ent.& 3t does not fo**ow
that he has a responsibi*it-8 the simp*e fact is that


ABOUT BEHAVIORISM ,'/

certain kinds of contin.encies ha/e affected him. 3f the- ha/e fai*ed to do so) it is because *ie is
uncontro**ab*e) not irresponsib*e) and the term contro**abi*it- has in some. =uarters tended to
rep*ace that of responsibi*it-.

Entertainin*$ 3t ma- be said that there is one fie*d in which what is done is not rea**-
mana.ement) a*thou.h an effect on other peop*e is e,treme*- important) (he. artist) the
composer) or the writer of poetr- or fiction produces somethin. 6which seems to be 0ustified
so*e*- b- the fact that it is reinforcin.) no attention bein. paid to the contin.encies. %:e*i.ious
art) ceremonia* music) and books with a messa.e are desi.ned to induce action) and simi*ar
reinforcin. effects are used for educationa*) therapeutic) and other purposes. :einforcin.
pictures) furnishin.s) and back.round music arc used to make stores) offices) and hote* *obbies
function as conditioned reinforcers) to increase the *ike*ihood that peop*e wi** come back a.ain.
But 3 am speakin. here of 9pure9 art.& At best the artist) composer) or writer acts to produce
somethin. which reinforces him) and he is most *ike*- to continue to be producti/e when that is
the case. But we must not o/er*ook the fact that the consumer of art) music) and *iterature is a*so
reinforced. 7ne *ooks at pictures) .oes to .a**eries to see them) bu-s them) or bu-s copies of
them in order to *ook at them because one is reinforced when one does so. 7ne p*a-s music
which is reinforcin.) .oes to concerts) or bu-s recordin.s. 7ne bu-s and reads books) (he fact is
probab*- not entire*- irre*e/ant to artist) composer) or writer) but e/en if it were) there is no
reason wh- a beha/ioristic account cou*d not *ist the reinforcin. effects of works of art) music)
and *iterature and dea* with them as such) rather in the manner of the contemp*ati/e knowin.
discussed in Chapter #.

THE SE88 A42 OTHERS ,'0

(he fact that conspicuous beha/ior) is *ackin. does not mean that menta* *ife has been
demonstrated.

he Self and *thers

Beop*e used to suppose that the- know themse*/es better than the- knew others %or than other
knew them&. 4hat the- meant is that the- knew their own fee*in.s and introspecti/e*- obser/ed
states better than those of others. Se*f6know*ed.e is then a matter of bein. in contact with
onese*f. 4hen peop*e be.an to disco/er wh- others beha/ed as the- did) a different kind of se*f66
know*ed.e arose) which took .enetic endowment) en/ironmenta* histor-) and current settin. into
account. (he historica* priorit- of se*f6know*ed.e based upon introspection .a/e wa- to
know*ed.e of en/ironmenta* contin.encies.

(he order of disco/er- was re/ersed in se*f6mana.ement. Beop*e *earn rather easi*- to
contro* others. A bab-) for e,amp*e) de/e*ops certain methods of contro**in. his parents when he
beha/es in wa-s *eadin. to certain kinds of action. Chi*dren ac=uire techni=ues of contro**in.
their peers) and the- become 6ski**fu* in this *on. before the- contro* themse*/es. (he ear*-
instruction the- recei/e in chan.in. their own fee*in.s or introspecti/e*- obser/ed states b-
e,ercisin. wi** power or a*terin. emotiona* and moti/ationa* states is not /er- effecti/e. (he
se*f6mana.ement which be.ins to be tau.ht in the form of pro/erbs) ma,ims) and ru*es of thumb
is a matter of chan.in. the en/ironment. (he contro* of others) *earned at an ear*- date) comes at
*ast to be used in se*f6contro*) and e/entua**- a fu**6f*ed.ed techno*o.- of beha/ior *eads to
ski**fu* se*f6mana.ement.

(hat it a*so *eads to the ski**fu* mana.ement of others raises serious prob*ems to which
we now turn.
*+ The ,!estion of )ontrol
'o Back

A scientific ana"7sis of behavior must8 I be"ieve8 assume that a personBs behavior
is contro""ed b7 his genetic and environmenta" histories rather than b7 the person
himse"f as an initiating8 creative agentA but no part of the behavioristic position has
raised more vio"ent ob;ections. !e cannot rove8 of course8 that human behavior as a
0ho"e is fu""7 determined8 but the proposition becomes more p"ausib"e as facts accumu-
"ate8 and I be"ieve that a point has been reached at 0hich its imp"ications must be
serious"7 considered.

!e often over"oo$ the fact that human behavior is a"so a form of contro". That an
organism shou"d act to contro" the 0or"d around it is as characteristic of "ife as breathing
or reproduction. A person acts upon the environment8 and 0hat he achieves is essentia"
to his surviva" and the surviva" of the species. #cience and techno"og7 are mere"7
manifestations of this essentia"

THE 8UESTION O4 7ONTRO9 2('

feature of human behavior. 5nderstanding8 prediction8 and e:p"anation8 as 0e"" as
techno"ogica" app"ications8 e:emp"if7 the contro" of nature. The7 do not e:press an
>attitude of domination> or a >phi"osoph7 of contro".> The7 are the inevitab"e resu"ts of
certain behaviora" processes.

!e have no doubt made mista$es. !e have discovered8 perhaps too rapid"78
more and. more effective 0a7s of contro""ing our 0or"d8 and 0e have not a"0a7s used
them 0ise"78 but 0e can no more stop contro""ing nature than 0e can stop breathing or
digesting food. Contro" is not a passing phase. Do m7stic or ascetic has ever ceased to
contro" the 0or"d around himA he contro"s it in order to contro" himse"f. !e cannot
choose a 0a7 of "ife in 0hich there is no contro". !e can on"7 change the contro""ing
conditions.

)o!ntercontrol

'rganiFed agencies or institutions8 such as governments8 re"igions8 and economic
s7stems8 and to a "esser e:tent educators and ps7chotherapists8 e:ert a po0erfu" and
often troub"esome contro". It is e:erted in 0a7s 0hich most effective"7 reinforce those
0ho e:ert it8 and unfortunate"7 this usua""7 means in 0a7s 0hich either are immediate"7
aversive to those contro""ed or e:p"oit them in the "ong run.

Those 0ho are so contro""ed then ta$e action. The7 escape from the
contro""er-moving out of range if he is an individua"8 or defecting from a government8 be-
coming an apostate from a re"igion8 resigning8 or p"a7ing truant-or the7 ma7 attac$ in
order to 0ea$en or destro7 the contro""ing po0er8 as in a revo"ution8 a reformation8 a
stri$e8 or a student protest. In other 0ords8 the7 oppose contro" 0ith countercontro".

ABOUT BEHAVIORISM 2/(

A condition ma7 be reached in 0hich these opposing forces are in e6ui"ibrium8 at
"east temporari"78 but the resu"t is se"dom an optima" so"ution. An incentive s7stem ma7
reconci"e a conf"ict bet0een management and "abor8 nations ma7 maintain a ba"ance of
po0er8 and governmenta"8 re"igious8 and educationa" practices ma7 be effective ;ust
short of defection8 apostas78 or truanc78 but the resu"ts are b7 no means 0e"" -designed
socia" environments.

-thics and )o$passion

!e spea$ of a benevo"ent ru"er8 a devoted teacher8 a compassionate therapist8 and a
pub"ic-spirited industria"ist8 as if their behavior 0ere s7mptomatic of inner traits of
character. !hen 0e as$ 0h7 a person is benevo"ent8 devoted8 compassionate8 or
Bpub"ic-spirited8 0e find ourse"ves e:amining the effect his behavior has on others. (The
5ti"itarians referred to effects of this sort in defining uti"it7 as >that princip"e that
approves or disapproves of ever7 action 0hatsoever8 according to the tendenc7 0hich it
appears to have to augment or diminish the happiness of the part7 0hose interest is in
6uestion8> but this 0as the approva" or disapprova" of a third part78 not of the part7
immediate"7 affected b7 the action.) The conse6uences responsib"e for benevo"ent8
devoted8 compassionate8 or pub"ic-spirited behavior are forms of countercontro"8 and
0hen the7 are "ac$ing8 these much-admired features of behavior are "ac$ing8

The point is i""ustrated b7 five fie"ds in 0hich contro" is not offset b7 countercontro"
and 0hich have therefore become c"assica" e:amp"es of mistreatment. The7 are the
care of the ver7 7oung8 of t1"e aged8 of prison
THE 8UESTION O4 7ONTRO9 2//

ers8 of ps7chotics8 and of the retarded. It is often said that those 0ho have these peop"e
in charge "ac$ compassion or a sense of ethics8 but the conspicuous fact is that the7 are
not sub;ect to strong countercontro". The 7oung and the aged are too 0ea$ to protest8
prisoners are contro""ed b7 po"ice po0er8 and ps7chotics and retardates cannot organiFe
or act successfu""7. ?itt"e or nothing is done about mistreatment un"ess countercontro"8
usua""7 negative8 is introduced from outside.

Countercontro" is no doubt not the on"7 reason 0h7 one person treats another
person 0e"". !e might act in such a 0a7 that another person is reinforced and re-
inforces us in turn. The human genetic endo0ment ma7 inc"ude some such tendenc78
as parenta". care of the 7oung8 for e:amp"e8 seems to i""ustrate. Ear0in pointed to the
surviva" va"ue of a"truistic behavior8 in a passage I sha"" 6uote "ater8 though on"7 ver7
specia" $inds of innate behavior seem to be invo"ved. In an7 case8 the 0a7 one person
treats another is determined b7 reciproca" action. !e gain nothing b7 turning to fee"ings.
It is often said that peop"e comfort the distressed8 hea" the sic$8 and faced the hungr7
because the7 s7mpathiFe 0ith them or share their fee"ings8 but it is the behavior 0ith
0hich suc* fee"ings are associated 0hich s*ould have had surviva" va"ue and 0hich is
modified b7 countercontro". !e refrain from hurting others8 not because 0e >$no0 ho0 it
fee"s to be hurt8> but (1) because hurting other members of the species reduces the
chances that the species 0i"" survive8 and () 0hen 0e have hurt others8 0e ourse"ves
have been hurt.

The c"assica" concept of *u2an!as 0as defined as a set of virtues8 but an7
fee"ing of virtue cou"d be thought of as a b7-product of conduct. A man. 0ho practiced
humanitas 0as confident in the sense of being usua""7
ABOUT BEHAVIORISM 2/2

successfu"A he treated others 0e"" and 0as as a resu"t 0e"" treated b7 themA he p"a7ed
an active part in governmentA and so on.

An >important determinant of mora" behavior and a ma;or component of character
deve"opment> is said to be >0i""ingness to fo""o0 ru"es8> but a person >0i""s> to fo""o0 a
ru"e because of the conse6uences arranged b7 those 0ho state the ru"e and enforce it.
The distinction bet0een ru"e-governed and contingenc7-shaped behavior is missed
0hen a test of >socia"iFation> is said to >assess the degree to 0hich a person has
interna"iFed the ru"es8 va"ues8 and conventions of his societ7.> )eop"e punished each
other "ong8 before behavior 0as ca""ed bad or 0rong and before ru"es 0ere formu"ated8
and a person ma7 have been >socia"iFed> b7 these@ )unitive contingencies 0ithout
benefit of ru"es.

)eop"e do begin to ca"" behavior good or bad or right or 0rong and to reinforce or
punish according"78 and ru"es are eventua""7 stated 0hich he"p a person conform to the
practices of his communit7 and he"p the communit7 maintain the practices. A person
0ho "earns these ru"es and behaves b7 e:p"icit"7 fo""o0ing them sti"" has not interna"iFed
them8 even 0hen he "earns to contro" himse"f and thus to ad;ust ever more effective"7 to
the contingencies maintained b7 the group. #ocia" behavior does not re6uire that the
contingencies 0hich generate it shou"d be formu"ated in ru"es or8 if the7 have been
formu"ated8 that a person shou"d $no0 the ru"es. . is e:traordinari"7 important8 ho0ever8
that socia" practices be formu"ated.

!e sometimes sa7 that 0e acted in a given 0a7 because 0e $ne0 it 0as right or
fe"t that it 0as right8 but 0hat 0e fee" 0hen 0e behave mora""7 or ethica""7 depends on
the contingencies responsib"e for our behavior. !hat 0e fee" about the behavior of
others depends
THE 8UESTION O4 7ONTRO9 2/3

on its effect on usA 0hat 0e fee" about our o0n behavior to0ard others depends on the
action others ta$e. ,ic bodi"7 conditions $no0n or fe"t ma7 be particu"ar"7 conspicuous
0hen the sanctions are strong8 A person 0ho has been e:posed to the promise of
heaven and the threat of he"" ma7 fee" stronger bodi"7 states than one 0hose behavior is
mere"7 approved or censured b7 his fe""o0 men. But neither one acts because he $no0s
or fee"s that his behavior is rightA "ie acts because of the contingencies 0hich have
shaped his behavior and created the conditions he fee"s.

A theo"ogica" 6uestion of some anti6uit7 is this@ Is man sinfu" because he sins8 or
doeQ he sin because he is sinfu"4 1ar: raised a simi"ar 6uestion and ans0ered it this
0a7@ >it is not the consciousness of man that determines his e:istenceA rather it is his
socia" e:istence that determincs his consciousness.> B!i""iam Mames fo""o0ed suit in the
fie"d of emotion@ >!e do not cr7 because 0e arc sadA 0e are sad because 0e cr7.> In a""
three formu"ations an important detai" is "ac$ing@ nothing is said about 0hat is
responsib"e for both the state and the behavior. And if 0e are as$ed8 >Is a person mora"
because "ie behaves mora""78 or does he behave mora""7 because "ie is mora"4> 0e must
ans0er8 >Deither.> Ge behaves mora""7 and 0e ca"" him mora" because he "ives in a
particu"ar $ind of environment.

Countercontro" is not too hard to e:p"ain 0hen contro" is immediate"7 aversive-for
e:amp"e8 0hen it is e:erted b7 punishment or the threat of punishment. There are
presumab"7 re"evant contingencies of surviva"@ 0hen unab"e to escape8 organisms
0hich attac$ a predator successfu""7 have a competitive advantage. But 0hen the
avcrsivc conse6ucnecs of contro" are deferred8 as in e:p"oitation8 countvrcontroff"ng
action is "ess "i$e"7. 1ost of thoso 0ho had great 0ea"th used it
ABOUT BEHAVIORISM 2/4

0ithout being sub;ect to ver7 much countercontro" unti" the nineteenth centur7. It has
been said of Gege" that be 0as one of the first to rea"iFe that a modern. s7stem of trade
and industr7 had >spontaneous"7 arisen from the 0or$ings of rationa" se"f-interest8> and
that "a0 and government 0ere no0 necessar78 not mere"7 to protect the societ7 and its
individua" members8 but to contro" the un"imited greed for persona" 0ea"th that ne0 pro-
ductive techni6ues had un"eashed. This cou"d on"7 be done8 be be"ieved8 if a genera"
sense of decenc7 pervaded societ7. A fe0 emendations are needed. To sa7 that trade
and industr7 >arose from the 0or$ings of rationa" se"f-interest> is simp"7 to sa7 that men
discovered ne0 0a7s of ac6uiring mone7 and goods. Their >greed> 0as un"imitA in the
sense that there 0as no countercontro". Gence the need for "a0s restricting trade and
industr78 but these re6uired "ega" action b7 in;ured peop"e rather than 8I >genera" sense
of decenc7.> It is not enough to cite the behavior from 0hich 0e infer a sense of
decenc78 as8 it 0as not enough to cite the behavior from 0hich 0e infer the compassion
of those 0ho have he"p"ess peop"e in their charge. !e must "oo$ at countercontro""ing
contingencies.

1an has been said to be superior to the other anima"s because he has evo"ved a
mora" or ethica" sense. >B7 far the most important characteristic of human beings is that
0e have and e:ercise mora" ;udgment.> But 0hat has evo"ved is a socia" environment in
0hich individua"s behave in 0a7s determined in part b7 their effects on others. Eifferent
peop"e sho0 different amounts and $inds of mora" and ethica" behavior8 depending
upon the e:tent of their e:posure to such contingencies. 1ora"s and ethics have been
said to invo"ve >attitudes to0ard "a0 and government 0hich have ta$en centuries in the
bui"ding8> but it is much more p"ausib"e

THE 8UESTION O4 7ONTRO9 2/5

to sa7 that the behavior said to e:press such attitudes is generated b7 contingencies
that have deve"oped over the centuries. An a"titude to0ard government as distinct from
behavior can scarce"7 have survived for centuriesA 0hat have survived are
governmenta" practices. ?ega" behavior depends on more than >an attitude of deference
to0ard government> as the ro"e of government depends on more than >an accomp"ished
fact of po0er8> and to sa7 that >"a0 is an achievement that needs to be rene0ed b7
understanding the sources of its strength> is to point direct"7 to the need to understand
and maintain governmenta" contingencies.

7ne of the most tra.ic conse=uences of menta*ism is dramatica**- i**ustrated b- those
who are earnest*- concerned about the p*i.ht of the wor*d toda- and who see no he*p e,cept in a
return to mora*it-) ethics) or a sense of decenc-) as persona* possessions. A recent book on
mora*s is said to show hope rather than despair because the author 9percei/es a .rowin.
awareness of each man for his fe**ows8 an increasin. respect for the ri.hts of others)9 and he sees
these as 9. . . steps toward a secure wor*d communit-) based on e/er6widenin. rea*ms of
re*atedness and empath-)9 and a pastora* *etter insists that our sa*/ation 9*ies in a return to
Christian mora*s.9 But what is needed is a restoration of socia* en/ironments in which peop*e
beha/e in wa-s ca**ed mora*.

B"aming peop"e in order to shape ethica""7 acceptab"e behavior has an
unfortunate resu"t. #amue" But"er made the point in E$e,*on3 0here peop"e 0ere
b"amed for ph7sica" but not mora" i""ness. Compare t0o peop"e8 one of 0hom has been
cripp"ed b7 an accident8 the other b7 an ear"7 environmenta" histor7 0hich ma$es him
"aF7 and8 0hen criticiFed8 mean. Both cause great inconvenience to others8 but one dies
a mart7r88 the other a scoundre". 'r compare t0o chi"dren--one cripp"ed b7
ABOUT BEHAVIORISM 2/"

po"io8 the other b7 a re;ecting fami"7. Both contribute "itt"e to othcrs and cause troub"e8
but on"7 one is b"amed. The inain difference is that on"7 one $ind of disabi"it7 is
correctab"e b7 punishment8 and even then on"7 occasiona""7. It is tempting to sa7 that
on"7 one person in each case could do something about his condition8 but shou"d 0e not
sa7 that 0e cou"d do something besides b"aming him4

To attribute mora" and ethica" behavior to environmenta" contingencies seems to
"eave no room for abso"utes. It suggests a $ind of re"ativism in 0hich 0hat is good is
0hatever is ca""ed good. 'ne ob;ection to this is that it refers to reinforcers but not to the
maintained contingencies in 0hich the7 appear. !e a"so tend to ob;ect 0hen 0hat
another group ca""s good differs 0ide"7 from 0hat 0e ca"" good8 if our practices conf"ict.
But an environmenta" account is not re"ativism in this sense. The >boo-hurrah theor7> of
ethica" emotivists 0as an appea" to fee"ings sharp"7 "oca"iFed in time and p"ace and
unre"ated to an7 apparent reasons for ethica" and mora" standards. 2thica" and mora"
contingencies of reinforcement have their o0n conse6uences8 to 0hich I sha"" turn in a
moment.

The #tr!ggle for Freedo$

1anBs success in freeing himse"f from the irritations and dangers of his ph7sica"
environment and from the punitive and e:p"oitative aspects of his socia" euvironment
has been perhaps his greatest achievement. it has "eft him free to deve"op other $inds
of behavior 0ith high"7 reinforcing conse6uences-in the sciences8 arts8 and socia"
re"ations. At the same time it has given him the fee"ing of freedom8 and perhaps noi
fee"ing has caused more troub"e.
THE 8UESTION O4 7ONTRO9 2/#

As I pointed out in Chapter %8 operant behavior under positive reinforcemcnt is
distinguished b7 the "ac$ of an7 immediate"7 antecedent event 0hich cou"d p"ausib"7
serve as a cause8 and as a resu"t it has been said to sho0 the inner origination ca""ed
free 0i"". .ef"e: behavior has its stimu"us and is therefore ca""ed invo" untar78 and
negative"7 reinforced operant behavior is emitted in the presence of the aversive
condition from 0hich the behavior brings escape. 5nder these conditions 0e do not
spea$ of 0hat 0e 0ant to do but of 0hat 0e have to do to avoid or escape from
punishment. !e ma78 through an >act of 0i""8> choose to submit to punishment8 but on"7
because other conse6uences of 0hich there is no immediate"7 antecedent cause ma$e
our submission >vo"untar7.>

The important fact is not that 0e fee" free 0hen 0e have been positive"7
reinforced but !*a! ,e do no !end !o esca0e o$ coun!e$a!!ac.- <ee"ing free is an
important ha""mar$ of a $ind of contro" distinguished b7 the fact that it does not breed
countercontro". The strugg"e for freedom has seemed to move to0ard a 0or"d in 0hich
peop"e do as the7 "i$e or 0hat the7 0ant to do8 in 0hich the7 en;o7 the right to be "eft
a"one8 in 0hich the7 have been >redeemed from the t7rann7 of gods and governments
b7 the gro0th of their free 0i"" into perfect strength and se"f-confidence.> It 0ou"d appear
to be a 0or"d in 0hich peop"e have fu"fi""ed themse"ves8 have actua"iFed themse"ves8
and have found themse"ves8 in the sense in 0hich these e:pressions are used in e:is-
tentia"ism8 phenomeno"og78 and 2astern m7sticism. It is a 0or"d in 0hich the contro" of
human behavior is 0rong8 in 0hich >the desire to change another person is essentia""7
hosti"e.> 5nfortunate"7 the fee"ing of being free is not a re"iab"e indication that 0e have
reached such a 0or"d.
ABOUT BEHAVIORISM 2/&

The fact that positive reinforcement does not breed countercontro" has not gone
unnoticed b7 0ou"d-be contro""ers8 0ho have simp"7 shifted to positive means. Gere is
an e:amp"e@ A government must raise mone7. If it does so through ta:ation8 its citiFens
must pa7 or be punished8 and the7 ma7 escape from this aversive contro" b7 putting
another part7 in po0er at the ne:t e"ection8 As an a"t9-rnativc8 the government
organiFes a "otter78 and instead of being fo$ced to pa7 ta:es8 the citiFen vo"untari"7 bu7s
tic$ets. The resu"t is the same@ the citiFens give the government mone78 but the7 fee"
free and do not protest in the second case. Deverthe"ess the7 are being contro""ed8 as
po0erfu""7 as b7 a threat of punishment8 b7 that particu"ar"7 pu0e"fu" (variab"eratio)
schedu"e of reinforcement discussed8 in Chapter %8 the effect of 0hich is a"" too c"ear"7
sho0n in the behavior of the compu"sive or patho"ogica" gamb"er.

Contro" is concea"ed 0hen it is represented as changing minds rather than
behavior. )ersuasion is not a"0a7s effective8 but 0hen it is8 it breeds "itt"e or no
countercontro". !e persuade in part b7 describing potentia""7 reinforcing conse6uences.
A 0e""-$no0n eco"ogist has discussed the possibi"it7 of ma$ing industries pa7 for the
right to po""ute air8 "and8 and 0ater. This re6uires either "egis"ation or vo"untar7
agreement b7 industr78 and >in our $ind of democrac7> either is possib"e on"7 >b7
persuasion8 b7 creating a favorab"e c"i2a!e of 0ublc opinion.> Mourna"ists and those 0ho
contro" the mass media must p"a7 an important ro"e. Another appea" to persuasion "ed
to the fo""o0ing comment in the ?ondon Times@

Do0 it is the ma;orit7 that never had it so good8 and it is democratica""7
determined to maintain that situation. >!e must persuade . . . persuade . 8 . persuade . .
.> sa7s 1r. Men$ins. >'ur on"7 hope is to appea" to the "atent

THE 8UESTION O4 7ONTRO9 2/'

idea"ism of a"" men and 0omen of good 0i"".> But that is evange"ism8 not po"itics. . . . It is
hoped that in his subse6uent speeches M$- Men$ins 0i"" discuss theJ po"itica" techni6ues
0hereb7 the ma;orit7 can be contro""ed.

The contro" of behavior is concea"ed or disguised in education8 ps7chotherap78
and re"igion8 0hen the ro"e of teacher8 therapist8 or priest is said to8 be to guide8 direct8
or counse"8 rather than to manage8 and 0here measures 0hich cannot be so disguised
a:e re;ected as intervention. #ocia" proposa"s often careftff"7 omit an7 reference to
means@ 0e need8 for e:amp"e8 to ma$e >better uti"iFation of human resources8> the
contro" invo"ved in >uti"iFation> not being specified.

The embarrassment of those 0ho find themse"ves in a position 0here the7 must
recommend contro" is e:emp"ified b7 the Eec"aration of )rincip"es issued b7 the
#toc$ho"m Conference on the 2nvironment he"d in 1-&1. The first princip"e begins8
>1an has the fundamenta" right to freedom8 e6ua"it78 and ade6uate conditions of "ife8 in
an environment of a 6ua"it7 that permits a "ife of dignit7 and 0e"" -being8 and he bears a
so"emn responsibi"it7 to protect and improve the environment for future generations.> Do
other species has rights and responsibi"ities in this sense8 and it is difficu"t to see ho0
the7 cou"d have evo"ved as fundamenta" human traits or possessions under natura"
se"ection un"ess 0e regard them as contro""ing and countercontro""ing practices. To
assert a right is to threaten action against those 0ho are said to infringe it. Thus8 0e act
to restrain those 0ho force us to act (and 0ho thereb7 reduce our fee"ing of freedom)8 or
0ho ta$e more than their share of avai"ab"e goods8 or 0ho fou" the 0or"d in 0hich 0e
"ive. !e ;ustif7 and e:p"ain our behavior 0hen 0e c"aim the right to restrain them. Those
0ho defend human rights point to measures to be ta$en against those 0ho in

ABOUT BEHAVIORISM 22(

fringe them. The Bi"" of .ights8 for e:amp"e8 protects the individua" against certain $inds
of "ega" action.

1an >bears a so"emn responsibi"it7> not to contro" others aversive"78 not to ta$e
more than a ;ust share of goods8 and not to fou" the environment8 in the sense that he
0i"" be criticiFed or punished b7 those 0ho suffer if he does so. The responsibi"it7 is not
a persona" possession but a propert7 of the (main"7 "ega") contingencies to 0hich
peop"e are e:posed. B7 turning from rights and responsibi"ities to the behaviors
attributed to them or said to bc ;ustified b7 them8 and in turning to the socia" (usua""7
govemmenta") contingencies 0hich shape and maintain those behaviors8 0e escape
from a centuries-o"d controvers7 and move to0ard possib"7 effective action.

The dec"aration of the #toc$ho"m conference contained t0ent7-si: princip"es. The
conference had no mi"itar7 or economic8 and ver7 "itt"e educationa"8 po0erA it cou"d on"7
ma$e recommendations. In the 2ng"ish version 0e find that e"even princip"es asserted
that states8 p"anners8 po"icies8 and so on must ta$e certain $inds of action. <ive asserted
that the7 s*ould3 and three that the7 s*all- <ive simp"7 pointed out that action is
essentia"8 and one ac$no0"edged a sovereign right. )erhaps it 0ou"d be unfair to as$
more of this particu"ar conference8 but it 0as ca""ed to meet possib"7 the greatest current
threat to the species8 and it is c"ear that it made "itt"e progress because it cou"d not
accept the fact that an essentia" step 0as the restriction of certain freedoms.

The )ontrolling #ocial -nviron$ent

)eop"e have suffered so "ong and so painfu""7 from the contro"s imposed upon them that
it is eas7 to under
TG2 352#TI'D '< C'DT.'? 1

stand 0h7 the7 so bitter"7 oppose an7 form of contro". A simp"e ana"7sis of contro""ing
practices8 such as that in the preceding chapter8 is "i$e"7 to be attac$ed simp"7 because
it cou"d be misused b7 contro""ers. But in the "ong run an7 effective countercontro"
"eading to the >"iberation> of the individua" can be achieved on"7 b7 e:p"icit design8 and
this must be based upon a scientific ana"7sis of human behavior. !e must sure"7 begin
0ith the fact that hurnan behavior is a"0a7s contro""ed. >1an is born free8> said
.ousseau8 >and is ever70here in chains8> but no one is "ess free than a ne0born chi"d8
nor 0i"" he become free as he gro0s o"der8 Gis on"7 hope is that he 0i"" come under the
contro" of a natura" and socia" environment in 0hich he 0i"" ma$e the most of his genetic
endo0ment and in doing so most successfu""7 pursue happiness. Gis fami"7 and his
pecrs are part of that environment8 and he 0i"" benefit if the7 behave in ethica" 0a7s.
2ducation is another part of that environment8 and he 0i"" ac6uire the most effective
repertoire if his teachers recogniFe their ro"e for 0hat it is rather than assume that it is to
"eave him free to deve"op himse"f. Gis government is part of that environment8 and it 0i""
>govern "east> if it minimiFes its punitive measures. Ge 0i"" produce 0hat he and others
need most effective"7 and "east aversive"7 if incentive conditions are such that he 0or$s
carefu""7 and industrious"7 and is reinforced b7 0hat "ic does. A"" this 0i"" be possib"e not
because those 0ith 0hom he associates possess mora"it7 and a sense of ethics or de-
cenc7 or compassion8 but because the7 in turn are contro""ed b7 a particu"ar $ind of
socia" environment.

The most important contribution of a socia" environment-a contribution 0ho""7
abandoned in the return to a thoroughgoing individua"isin-"ias to do 0ith the mediation of
the future. The bruta" prospect of over
ABOUT BEHAVIORISM 222

popu"ation8 po""ution8 and the e:haustion of resources has given the future a ne0 and
re"ative"7 immediate significance8 but some concern for the future has8 of course8 "ong
prevai"ed. It has been said that a hundred 7ears ago >there 0ere fe0 men a"ive8 0hether
5ti"itarians or re"igious peop"e8 0ho then thought of the goodness of an act as being in
the act itse"f or in the 0i"" that 0i""ed itA a"" 0as in the conse6uences8 for their happiness
tomorro0 or the B"ife hereafterBA both 0ere matters of future re0ard.> But goodness in the
"ight of 0hich an act ma7 be ;udged is one thingA inducing peop"e to be good or to act
0e"" >for the sa$e of a future conse6uence> is another. The important thing is that
institutions fast "onger than individua"s and arrange contingencies 0hich ta$e a
reasonab"7 remote future into account. The bohaviora" processes are i""ustrated b7 a
person 0ho 0or$s for a promised return8 0ho p"a7s a game in order to 0in8 or 0ho bu7s
a "otter7 tic$et. !ith their he"p8 re"igious institutions ma$e the prospect of art after"ife
reinforcing8 and governments induce peop"e to die patriotic deaths.

!e ob;ect to much of this8 but the interests of institutions sometimes coincide 0ith
the interests of individua"s@ governnients and re"igions sometimes induce peop"e to
behave 0e"" 0ith respect to each other and to act together for protection and support.
)roverbs and ma:inis8 as 0e"" as e:p"icit codes of "a08 strengthen behavior having
deferred conse6uences. B7 himse"f an individua" can ac6uire ver7 "itt"e behavior 0ith re-
spect to the future in. hisi o0n "ifetime8 but as a member of a group he prefits from the
socia" environment maintained b7 the group. This is a fact of the greatest importance
because it "eads to an ans0er to t0o basic 6uestions@ Go0 can 0e ca"" a particu"ar
instance of the
THE 8UESTION O4 7ONTRO9 223

contro" of human behavior good or bad8 and 0ho is to design and maintain contro""ing
practices4

The -vol!tion of a )!lt!re

The socia" environment I have been referring to is usua""7 ca""ed a cu"ture8 though a
cu"ture is often defined in other 0a7s--as a set of customs or manners8 as a s7stem of
va"ues and ideas8 as a net0or$ of communication8 and so on. As a set of contingencies
of reinforcement maintained b7 a group8 possib"7 formu"ated in ru"es or "a0s8 it has a
c"ear-cut ph7sica" status8 a continuing e:istence be7ond the "ives of members of the
group8 a changing pattern as practices are added8 discarded8 or modified8 and8 above
a""8 po0er. A cu"ture so defined contro"s the behavior of the members of the group that
practices it.

It is not a mono"ithic thing8 and 0e have no reason to e:p"ain it b7 appea"ing to a
group mind8 idea8 or 0i"". If there are indeed >sevent7-three e"ements of cu"ture common
to ever7 human societ7 sti"" e:isting or $no0n to histor78> then there must be
sevent7-three practices or $inds of practices in ever7 set of contingencies ca""ed a
cu"ture8 each of 0hich must be e:p"ained in terms of conditions prevai"ing before the
cu"ture emerged as such. !h7 do peop"e deve"op a "anguage4 !h7 do the7 practice
some $ind of marriage4 !h7 do the7 maintain mora" practices and formu"ate them in
codes4 #ome ans0ers to 6uestions of this sort are to be found in the bio"ogica"
characteristics of the species8 others in >universa" features> of the environments in
0hich peop"e "ive.

The important thing about a cu"ture so defined is that it evo"ves. A practice arises
as a mutation8 it affects


ABOUT BEHAVIORISM 224

the chances that the group 0i"" so"ve its prob"ems8 and if the group survives8 the practice
survives 0ith it. It has been se"ected b7 its contribution to the effectiveness of those 0ho
practice it. Gere is another e:amp"e of that subt"e process ca""ed se"ection8 and it has
the same fami"iar features. 1utations ma7 be random. A cu"ture need not have been
designed8 and its evo"ution does not sho0 a purpose.

The practices 0hich compose a cu"ture are a mi:ed bag8 and some parts ma7 be
inconsistent 0ith others or in open conf"ict. 'ur o0n cu"ture is sometimes ca""ed sic$8
and in a sic$ societ78 man 0i"" "ac$ a sense of identit78 and fee"ings of competenceA he
0i"" see the suspension of his o0n thought structures . . 8 to enter into a more fruitfu"
re"ationship 0ith those around him as betra7a"A he 0i"" approach the 0or"d of human
interaction 0ith a sense of rea" despairA and on"7 0hen he has been through that
despair and "earnt to $no0 himse"f 0i"" he attain as much of 0hat is se"f -fu"fi""ing as the
human condition a""o0s.

In trans"ation@ a sic$ societ7 is a set of contingencies 0hich generates disparate
or conf"icting behaviors suggesting more than one se"f8 0hich does not generate the
strong behavior 0ith 0hich a fee"ing of competence is associated8 0hich fa;Ms to
generate successfu" socia" behavior and hence "eads a person to ca"" the behavior of
others betra7a"8 and 0hich8 supp"7ing on"7 infre6ucnt reinforcement8 generates the
condition fe"t as despair. Another 0riter has said that our cu"ture is >in convu"sions
o0ing to its state of va"ue contradiction8 its incorporation of opposing8 and conf"icting
va"ues8> but 0e ma7 sa7 that the va"ues8 here as e"se0here8 refer to reinforcers8 and
that it is the contingencies of 0hich the7 are a part 0hich are opposing and conf"icting.
THE 8UESTION O4 7ONTRO9 225

The societ7 0i"" be >cured> if it can be changed in such a 0a7 that a person is
generous"7 and consistent"7 reinforced and therefore >fu"fi""s himse"f> b7 ac6uiring and
e:hibiting the most successfu" behavior of 0hich he is capab"e. Better 0a7s of teaching
(introduced for 0hatever reason8 possib"7 on"7 because of immediate conse6uences for
teacher or student) 0i"" ma$e a more effective use of the human genetic endo0ment.
Better incentive conditions (introduced for 0hatever reason8 possib"7 on"7 in the
interests of management or "abor) mean more and better goods and more en;o7ab"e
0or$ing conditions. Better 0a7s of governing (introduced for 0hatever reason8 possib"7
mere"7 in the interests of governed or governor) mean "ess time 0asted in persona"
defense and more time for other things. 1ore interesting forms of art8 music8 and
"iterature (created for 0hatever reason8 possib"7 simp"7 for the immediate reinforcement
of those creating or en;o7ing them) mean fe0er defections to other 0a7s of "ife.

In a 0e""-$no0n passage in ,: :escen! of Man3 Ear0in 0rote@
'bscure as s the prob"em of the advance of civi"iFation8 0e can at "east see that the
nation 0hich produced8 during a "engthened period8 the greatest number of high"7
inte""ectua"8 energetic8 brave8 patriotic8 and benevo"ent men8 0ou"d genera""7 prevai"
over "ess favored nations.
The point survives 0hen the appea" to character is corrected b7 spea$ing of >a nation
0hich maintains a socia" environment in 0hich its citiFens behave in 0a7s ca""ed
inte""igent8 energetic8 brave8 patriotic8 and benevo"ent.> Ear0in 0as spea$ing of the
surviva" va"ue of a cu"ture.

There are remar$ab"e simi"arities in natura" se"ection8 operant conditioning8 and
the evo"ution of socia" environments. Dot on"7 do a"" three dispense 0ith a


ABOUT BEHAVIORISM 22"

prior creative design. and a prior purpose8 the7 invo$e the notion of surviva" as a va"ue.
!hat is good for the species is 0hat ma$es for its surviva". !hat is good for the
individua" is 0hat promotes his 0e""-being. !hat is good for a cu"ture is 0hat permits it
to so"ve its prob"ems. There are8 as 0e have seen8 other $inds of va"ues8 but the7
eventua""7 ta$e second p"ace to surviva".

The notion of evo"ution is mis"eading-and it mis"ed both Gerbert #pencer and
Ear0in-0hen it suggests that the good represented b7 surviva" 0i"" natura""7 0or$ itse"f
out. Things go 0rong under a"" three contingencies of se"ection8 and the7 ma7 need to
be put right b7 e:p"icit design. Breeding practices have "ong represented a $ind of
intervention in the evo"ution of the species8 and geneticists are no0 ta"$ing about
changing genetic codes. The behavior of the individua" is easi"7 changed b7 designing
ne0 contingencies of reinforcement. De0 cu"tura" practices are e:p"icit"7 designed in
such fie"ds as education8 ps7chotherap78 peno"og78 and economic incentives.

The design of human behavior imp"ies8 of course8 contro"8 and possib"7 the
6uestion most often as$ed of the behaviorist is this@ !ho is to contro"4 The 6uestion
represents the age-o"d mista$e of "oo$ing to the individua" rather than to the 0or"d in
0hich he "ives. It 0i"" not be a benevo"ent dictator8 a compassionate therapist8 a devoted
teacher8 or a pub"ic-spirited industria"ist 0ho 0i"" design a 0a7 of "ife in the interests of
ever7one. !e must "oo$ instead at the conditions under 0hich peop"e govern8 give he"p8
teach8 and arrange incentive s7stems in particu"ar 0a7s. In other 0ords 0e must "oo$ to
the cu"ture as a socia" environment. !i"" a cu"ture evo"ve in 0hich no individua" 0i"" be
ab"e to accumu"ate vast po0er and use it for his o0n

THE 8UESTION O4 7ONTRO9 22#

aggrandiFement in 0a7s 0hich are harmfu" to others4 !i"" a cu"ture evo"ve in 0hich
individua"s are not so much concerned 0ith their o0n actua"iFation and fu" fi""ment that
the7 do not give serious attention to the future of the cu"ture4 These 6uestions8 and
man7 others "i$e thern8 are the 6uestions to be as$ed rather than 0ho 0i"" contro" and to
0hat end3 No one steps outside the causa" stream. Do one rea""7 intervenes. 1an$ind
has s"o0"7 but erratica""7 created environments in 0hich peop"e behave more effective"7
and no doubt en;o7 the fee"ings 0hich accompan7 successfu" behavior. It is a continuing
process.

04 What Is Inside the Skin/
Go Back

! behavioristic analysis rests on the follo"in# assumptions: ! person is first of all
an or#anism$ a member of a species an a Subspecies$ possessin# a #enetic
eno"ment of anatomical an physiolo#ical characteristics$ "hich are the prouct of
the contin#encies of survival to "hich the species has been e)pose in the process of
evolution. *he or#anism becomes a person as it acquires a repertoire of behavior
uner the contin#encies of reinforcement to "hich it is e)pose urin# its lifetime.
*he behavior it e)hibits at any moment is uner the control of a current settin#. It is
able to acquire such a repertoire uner such control because of processes of
conitionin# "hich are also part of its #enetic eno"ment.

In the traitional mentalistic vie"$ on the other han$ a person is a member of
the human species "ho behaves as he oes because of niany internal characteristics

WHAT IS I:SI9E THE S<I:& 22,

or possessions$ amon# them sensations$ habits$ intelli#ence$ opinions$ reams$
personalities$ moos$ ecisions$ fantasies$ skills$ percepts$ thou#hts$ virtues$ in&
tentions$ abilities$ instincts$ ayreams$ incentives$ acts of "ill$ 'oy$ compassion$
perceptual efenses$ beliefs$ comple)es$ e)pectancies$ ur#es$ choice$ rives$ ieas$
responsibilities$ elation$ memories$ nees$ "isom$ "ants$ a eath instinct$ a sense
of uty$ sublimation$ impulses$ capacities$ purposes$ "ishes$ an i$ represse fears$ a
sense of shame$ e)traversion$ ima#es$ kno"le#e$ interests$ information$ a supere#o$
propositions$ e)periences$ attitues$ conflicts$ meanin#s$ reaction formations$ a "ill
to live$ consciousness$ an)iety$ epression$ fear$ reason$ libio$ psychic ener#y$
reminiscences$ inhibitions$ an mental illnesses.
(o" are "e to ecie bet"een these t"o vie"s?

-rounds or "omparison

Si"#li*it3. We cannot say that one is simpler than the other$ since references to
mental states an activities make istinctions "hich must be recast in terms of
contin#encies of survival or reinforcement. It is possible$ inee$ that a behavioral
analysis "ill be more comple). !lthou#h some scheules of reinforcement$ for
e)ample$ prouce familiar effects "hich have been introspectively observe an
name$ many yiel entirely une)pecte results.

Use in -ontrol. !ccessibility is a ifferent matter. :o one has ever irectly
moifie any of the "ental activities or traits liste above. *here is no "ay in "hich
one can make contact "ith them. *he boily conitions felt as such can be chan#e
sur#ically$ electrically$ or "ith ru#s$ but for most practical purposes

AROUT BEHAVIORISM206

they are chan#e only throu#h the environment. When a evotee of mentalism
confesses that 8"e have not learne much about these problems in some"hat over
t"o thousan years of reflective thou#ht$8 "e may ask "hy reflective thou#ht has not
sooner come uner suspicion. Behavior moification$ althou#h still in its infancy$ has
been successful$ "hereas mentalistic approaches continue to fail$ an once the role
of the environment has been mae clear$ its accessibility is often surprisin#.

Use in .redi*tion. A ecision is perhaps more ifficult if "e simply "ant to
preict behavior. What a person feels is a prouct of the contin#encies of "hich his
future behavior "ill also be a function$ an there is therefore a useful connection
bet"een feelin#s an behavior. It "oul be foolish to rule out the kno"le#e a person
has of his current conition or the uses to "hich it may be put. (e may say that he
oes "hat he 8feels like oin#8 "ithout askin# "hy he feels that "ay$ an "e may ask
him to tell us "hat he feels like oin# an use his ans"er "ithout further inquiry$ as
"e prepare for his behavior. In casual iscourse the limits of accuracy note in
Chapter + are not necessarily serious$ but "e can nevertheless preict behavior more
accurately if "e have irect kno"le#e about the history to "hich feelin#s are to be
trace.

!ttitues$ opinions$ or intelli#ence$ as states inferre from behavior$ are also
useless in control$ but they permit us to preict one kin of behavior from another
kin kno"n to be associate "ith it$ presumably because of a common cause.

Use in Inter#retation. When human behavior is observe uner conitions "hich
cannot be e)actly
WHAT IS I:SI9E THE S<I:&201

escribe an "here histories are out of reach$ very little preiction or control is
possible$ but a behavioristic account is still more useful than a mentalistic one in
interpretin# "hat a person is oin# or "hy he behaves as he oes uner such
circumstances$ ! listener usually has no trouble in ientifyin# the ieas a speaker is
e)pressin#$ althou#h he has no inepenent evience$ but if "e are #oin# to #uess$ it
is more helpful to #uess about #enetic eno"ment an environmental history than
about the feelin#s "hich have resulte from them.

How 2ar Ba*k& When a person says that he acte 8because he felt like actin#$8
"e can put little faith in the 8because8 until "e have&. e)plaine "hy he ha the
feelin#$ but it has been ob'ecte that "e must stop some"here in follo"in# a causal
chain into the past an may as "ell stop at the psychic level. Clearly that is "hat is
one most of the time in mentalistic iscussions$ an that is "hy they block further
inquiry. It is true that "e coul trace human behavior not only to the physical
conitions "hich shape an maintain it but also to the causes of those conitions an
the causes of those causes$ almost a in8init+rv$ but there is no point in #oin# back
beyon the point at "hich effective action can be taken. *hat point is not to be foun
in the psyche$ an the e)planatory force of mental life has steaily ecline as the
promise of the environment has come to be more clearly unerstoo.

Relation to Other S*ien*es. !nother question is this: Which position more reaily
promotes a cooperative interchan#e "ith the social sciences on the one han an
physiolo#y on the other? (ere$ a#ain$ the behavioristic position seems to take first
place. *he social sciences "ere once heavily 8psycholo#istic.8

ABOUT BEHAVIORISM 202

4conomics ha its 4conomic 3an an political science its %olitical !nimal$ but "hen it
"as reco#ni-e that the psycholo#ical properties of these creatures "ere obviously
bein# invente precisely to e)plain the phenomena at issue$ psycholo#ism "as
re'ecte. ! behavioristic formulation may be sai to restore the role of the iniviual
in social science. *he result is not the 8behavioralism8 of political action @"hich$ as "e
have seen$ is a version of structuralismE but a ne" approach to the conitions of
"hich economic an political behavior are functions.

*he behavioral account is also close to physiolo#y: it sets the task for the
physiolo#ist$ 3entalism$ on the other han$ has one a #reat isservice by leain#
physiolo#ists on false trials in search of the neural correlates of ima#es$ memories$
consciousness$ an so on.

Is a -hoi*e :e*essar3& *here are those "ho "oul have it both "ays an "ho
continue to call psycholo#y the science of behavior an mental life. *o o so is to
return to .that three&sta#e sequence in "hich the physical environment acts upon the
or#anism to #enerate mental or psychic activities$ some of "hich ultimately fin
e)pression in physical action. *he pu--lin# question of ho" a physical event causes a
mental event$ "hich in turn causes a physical event$ remains to be ans"ere or
ismisse as unans"erable @a specialist in the physiolo#y of vision has sai that 8the
transition from the e)citations in the corte) to the sub'ective e)perience efies
e)planation8E.
*he problem coul be avoie if "e coul stay "ithin the mental or psychic sta#e.
In the 8intrapsychic life of the min8 mental causes have mental effects an amon#
them arc$ states of a"areness or consciousness$ an if this inner "orl coul be
observe in a

WHAT IS I:SI9E THE S<I:& 200

purely solipsistic "ay$ if the stuent of mental life ha no reason to appeal to
physical action$ even in communicatin# "ith others$ an if mental life playe no
isruptive role to be taken into account by the behaviorist$ everyone "oul be
satisfie. But psycholo#y as the stuy of sub'ective phenomena$ istinct from the
stuy of ob'ective behavior$ "oul then not be a science an "oul have no reason to
be.

! science of behavior must consier the place of private stimuli as physical
thin#s$ an in oin# so it provies an alternative account of mental life. *he question$
then is this: What is insie the skin$ an bo" o "e kno" about it? *he ans"er is$ I
believe$ the heart of raical behaviorism.

(h!siolog!

*he or#anism is$ of course$ not empty$ an it cannot be aequately treate simply as
a black bo)$ but "e must carefully istin#uish bet"een "hat is kno"n about "hat is
insie an "hat is merely inferre.

Aon# before %lato 8iscovere the min$8 the Greeks "ere e)plainin# behavior
"ith a curious mi)ture of anatomy$ physiolo#y$ an feelin#s. *hey mae a #reat eal
of the lun#s$ probably in part because the voice$ "hich seeme to e)press feelin#s
an ieas$ neee breath$ an because breathin# stoppe "hen life stoppe$ Both
8psyche8 @from the GreekE an 8spirit8 @from the AatinE ori#inally meant breath. Cther
precursors of behavior seeme locate in the heart$ "hich beat fast in emotion an
also stoppe beatin# "hen a person ie.

*races of this kin of physiolo#y have survive to the present ay. We say that
a person.s heart is not in his "ork or that it has been broken in a isastrous love

ABOUT B.HAVIORISM 204=

affair. We say that he has the #uts neee to stan up to his opponent or the #all to
call him a fool. When an#ry$ he vents his spleen. 3e may lack the brains neee for
his 'ob$ or his .brains may be ale or in turmoil. !t times his nerves may be
shattere$ fraye$ stretche to the breakin# point$ numb$ or @possibly be$ cause the
"or also once referre lo tenonsE taut or on e#e. We ia#nose. these inner states
as reaily as "e invent ieas an feelin#s$ an "e mi) them freely. ! ne"spaper
account of the Spassky&?ischer chess match rea as follo"s: 8Spassky.s bluner toay
coul very "ell be the result of the shakin# his nerves took in the earlier ,M&move
#ame. Cbviously Spassky has not recovere$ an his error toay may shake his con&
fience even more.8 It oes not matter "hether it "as his nerves or his confience
"hich "as shaken$ for the "riter has inverte both. In the same "ay co#nitive
psycholo#ists frequently use 8brain8 an 8min8 interchan#eably$ an all lan#ua#es are
sai to have certain invariant features because 8that is the "ay the brain is "ire.8 !
recent article in a scientific 'ournal reporte that 8the ri#ht hemisphere Fin
ri#hthanersH controls perceptual concepts; the left hemisphere #ets much of the
creit for the entire brain9s intelli#ence because it is the. brain9s mouthpiece
@lan#ua#e is store thereE.8 We are even tol that 8all normal hi#hly evelope brains
function in such a manner as to evelop natural moral ieas in response to their
e)periences of reality.8

*he science of physiolo#y be#an in much the same "ay. 4arly observations of
refle) action$ for e)ample$ "ere mae lon# before the activity of nerves coul be
etecte. =ifferent parts of the nervous system coul be isolate$ but "hat happene
in a #iven part coul only be inferre. *his "as true$ even in the first part of the
t"entieth century. *he synapse analy-e by Sir Charles

WHAT IS I:SI9E THE S<I:& 205

Sherrin#ton "as part of a conceptual nervous system$ an so "as the 8activity of the
cerebral corte)8 investi#ate by %avlov. ! conceptual nervous system cannot$ of
course$ be use to e)plain the behavior from "hich it is inferre.

%hysiolo#y an$ particularly "ith respect to behavior$ neurolo#y$ have of
course mae #reat pro#ress. 4lectrical an chemical properties of many neural activi&
ties are no" irectly observe an measure. *he nervous system is$ ho"ever$ much
less accessible than behavior an environment$ an the ifference takes its toll. We
kno" some of the processes "hich affect lar#e blocks of behavior&&sensory$ motor$
motivational$ an emotional&but "e are still far short of kno"in# precisely "hat is
happenin# "hen$ say$ a chil learns to rink from a cup$ to call an ob'ect by its name$
or to fin the ri#ht piece of a 'i#sa" pu--le$ as "e are still far short of makin#
chan#es in the nervous system as a result of "hich a chil "ill o these thin#s. It is
possible that "e shall never irectly observe "hat is happenin# in the nervous system
at the time a response occurs$ because somethin# like the (eisenber# principle may
apply: any means of observin# neural meiation of behavior may isturb the behavior.

! similar comparison may be mae of the technono#ical uses of behavioral an
physiolo#ical science. It has recently been$ sai that 8"e may be rapily acquirin# the
po"er to moify an control the capacities an activities of men by irect
intervention an manipulation of their boies an mins$8 an that the biolo#ical
human en#ineerin# that results "ill probably have profoun social consequences. It is
irect intervention an manipulation of the boy "hich is most often cite toay to
illustrate the an#ers of the control of behavior$ but a much more effective control is
alreay
ABOUT BEHAVIORISM 20

"ithin reach throu#h environmental manipulation. It is only the traitional fascination
"ith an inner life "hich a#ain leas to the ne#lect of the latter.

*o say that 8the only possible theoretical basis for the e)planation of human
behavior is to be foun in the physiolo#y of the brain an central nervous system$8
an that 8the aoption of this basis necessarily leas to the isappearance of
psycholo#y as an inepenent science$8 is also to overlook the possibility of a
behavioral science an of "hat it has to say about feelin#s an introspectively
observe states. ! similar entrapment in physiolo#y is illustrate by the statement 8If
man.s life is ever to be live alon# entirely rational lines$ free from such isturbances
as "ar$ crimes$ an economic booms an epressions$ he "ill have to fin a "ay of
increasin# the si-e of his brain.8 *here coul scarcely be a better e)ample of the am&
a#in# effect of the inner&irecteness of physiolo#ical$ as "ell as mentalistic$ inquiry.
If "e are to be free from "ar$ crimes$ an economic booms an epressions$ "e shall
have to fin a better social environment.

*he promise of physiolo#y is of a ifferent sort. :e" instruments an methos
"ill continue to be evise$ an "e shall eventually kno" much more about the kinds
of physiolo#ical processes$ chemical or electrical$ "hich take place "hen a person
behaves. *he physiolo#ist of the future "ill tell us all that can be kno"n about "hat
is happenin# insie the behavin# or#anism. (is account "ill be an important avance
over a behavioral analysis$ because the latter is necessarily 8historical8&that is to say$
it$ is confine to functional relations sho"in# temporal #aps. Somethin# is one toay
"hich affects the behavior of ail or#anism tomorro".

WHAT IS I:SI9E THE S<I:& 20%

:o matter ho" clearly that fact can be establishe$ a step is missin#$ an "e must
"ait for the physiolo#ist to supply it. (e "ill be able to sho" ho" an or#anism is
chan#e "hen e)pose to contin#encies of reinforcement an "hy the chan#e
or#anism then behaves in a ifferent "ay$ possibly at a much later ate. What he
iscovers cannot invaliate the la"s of a science of behavior$ but it "ill make the
picture of human action more nearly complete.

'ind or 5ervous S!stem/

But is this completion of a behavioral account not precisely the ob'ective of a
mentalistic analysis? =o "e not close the #ap bet"een behavior an the prior en&
vironmental history of "hich it is a function "hen "e feel or other"ise introspectively
observe the states of our boies arisin# from that history an responsible for that
behavior? Why shoul "e bother to ask about the nat+re of "hat is felt or
introspectively observe? Aet us take avanta#e of the position of the iniviual as an
observer of himself an allo" him to report on the meiatin# linka#e bet"een
behavior an its anteceent causes. I believe that this is the position of introspective
psycholo#y$ psychoanalysis$ an certain physicalistic. theories of kno"le#e that are
not committe to a pure solipsism.
*o a#ree. that "hat one feels or introspectively observes are conitions of one.s
o"n boy is a step in the ri#ht irection. It is a step to"ar an analysis both of seein#
an of seein# that one sees in purely physical terms. !fter substitutin# brain for min$
"e can then move on to substitutin# person for brain an recast the analysis in line
"ith the observe facts. But "hat is
ABOUT BEHAVIORISM 20)

8elt or intros#e*tivel3 o/served is not an important #art o8 the #h3siolo73 whi*h 8ills
the te"#oral 7a# in a historical analysis. ! severe limitation is to be seen in the
or#ans a person uses in observin# himself. !fter all$ "hat are the anatomy an
physiolo#y of the inner eye? So far as "e kno"$ self&observation must be confine to
the three nervous systems escribe in Chapter +&an interoceptive nervous system
#oin# to the viscera$ a proprioceptive nervous system #oin# to the skeletal frame$ an
an e)teroceptive system brin#in# a person mainly into contact "ith the "orl aroun
him. *hese three systems arose throu#h natural selection as the human species
evolve$ an they "ere selecte because of the role they playe in the internal an
e)ternal economy of the or#anism. But self&kno"le#e arose much later in the history
of the species$ as the prouct of social contin#encies arran#e by the verbal com&
munity$ an those contin#encies have not been active lon# enou#h to permit the
evolution of an appropriate nervous system.
Introspection has$ ha to use "hatever systems "ere available$ an they happene to
be systems "hich mae contact only "ith those parts of the boy that playe a role in
its internal an e)ternal economy. !ll that a person comes to kno" about himself "ith
their help are 'ust more stimuli an responses. (e oes not make contact "ith that
vast nervous system that meiates his behavior. (e oes not because be has no nerves
#oin# to the ri#ht places. *ryin# to observe much of "hat is #oin# on in one.s o"n
boy is like tryin# to hear supersonic souns or see electroma#netic raiation beyon
the visible ran#e$ *he brain is particularly lackin# in sense or#ans @its responses to
stimulation are not really sensin#E; it plays an e)traorinary role in behavior but not
as the ob'ect of that special behavior

WHAT IS I:SI9E THE S<I:& 20,

calle kno"in#. We can never kno" throu#h introspection "hat the physiolo#ist "ill
eventually iscover "ith his special instruments.
*he clue lies in the contin#encies of survival. /ust as "e cannot appeal to innate
eno"ment to e)plain #rammatical speech$ lo#ic$ or mathematics because #rammar$
lo#ic$ an mathematics have not been part of the human environment for a lon#
enou#h time$ so "e must question any effort to attribute introspective self&kno"le#e
to a nervous system especially aapte to that purpose. 5erbal behavior$ lo#ic$
mathematics$ an intros#e*tion have all been built on features of the human species
"hich ha alreay evolve for other reasons.

The "onceptual 5ervous S!stem

*he parts of the nervous system spoken of by early physiolo#ists "ere$ as "e have
seen$ lar#ely a matter of inference$ the classical e)ample bein# the synapse of
Sherrin#ton.s *he Inte7rative A*tion o8 the :ervo+s S3ste". Inference has yiele to
irect observation as instruments an methos have been improve$ an "ith #reat
#ains for physiolo#y. !nother "ay of ealin# "ith inference is to #ive it respectability
by convertin# it into an e)plicit moel or system. *here has arisen a kin of
thermoynamics of the nervous system$ in "hich #eneral la"s or principles are
establishe "ith little or no reference$ irect or inferre$ to the parts of the nervous
system involve. Information theory an cybernetics have both contribute to this
kin of speculation about "hat is #oin# on insie the hea. Such a moel or system
coul apply to either the mental or physical "orls or even to both$ an the problem
of ualism therefore$ seems to be avoie. Will a moel


ABOUT BEHA:IOIUSM 246

of the nervous system not serve until physiolo#y is more avance?

I believe the ans"er is no. *he stuy of the conceptual nervous system is
lar#ely concerne "ith those 8thou#ht processes8 iscusse in Chapter ,$ "hich$ as "e
have seen$ are real enou#h at the level of behavior but merely questionable
metaphors "hen move insie. !n like the stuy of consciousness or the real nervous
system$ a moel or system continues to turn attention in"ar$ a"ay from a #enetic
an personal history.

*4 #!$$ing .p
'o Back

The Introduction contains t0ent7 statements often made about behaviorism8 a"" of
them8 I be"ieve8 0rong. It is time to revie0 them in the "ight of the rest of this boo$.

1. 1ethodo"ogica" behaviorism and certain versions of "ogica" positivism cou"d be said
to ignore consciousness8 fee"ings8 and states of mind8 but radica" behaviorism does not
thus >behead the organism>A it does not >s0eep the prob"em of sub;ectivit7 under the
rug&A it does not >maintain a strict"7 behavioristic methodo"og7 b7 treating reports of
introspection mere"7 as verba" behavior>A and it 0as not designed to >permit con-
sciousness to atroph7.> !hat it has to sa7 about consciousness is this@ (a) #timu"ation
arising inside the bod7 p"a7s an important part in behavior. (b) The nervous s7stems
through 0hich it is effective evo"ved because of their ro"e in the interna" and e:terna"
econom7 of the organism. (c) In the sense in 0hich 0e sa7 that a person is conscious of
his surroundings8 he8 is con

ABOUT BEHAVIORISM242

scious of states or events in his bod7A he is under their contro" as stimu"i. A bo:er 0ho
has been >$noc$ed unconscious> is not responding to current stimu"i either 0ithin or
outside his s$in8 and a person ma7 continue to ta"$8 >unconscious of the effect he is
having on his "isteners> if that effect is not e:erting contro" over his behavior. <ar from
ignoring consciousness in this sense8 a science of behavior has deve"oped ne0 0a7s of
stud7ing it. (d) A person becomes conscious in a different sense 0hen a verba"
communit7 arranges contingencies under 0hich he not on"7 sees an ob;ect but sees
that he is seeing it. In this specia" sense8 consciousness or a0areness is a socia"
product. (e) Introspective $no0"edge of oneBs bod7-se"f-$no0"edge-is defective for t0o
reasons@ the verba" communit7 cannot bring se"f-descriptive behavior under the precise
contro" of private stimu"i8 and there has been no opportunit7 for the evo"ution of a
nervous s7stem 0hich 0ou"d bring some ver7 important parts of the bod7 under that
contro"8 (f) !ithin these "imits se"f-$no0"edge is usefu". The verba" communit7 as$s
6uestions about private events because the7 are the co""atera" products of
environmenta" causes8 about 0hich it can therefore ma$e usefu" inferences8 and
se"f-$no0"edge becomes usefu" to the individua" for simi"ar reasons. (g) Do specia" $ind
of mind stuff is assumed. A ph7sica" 0or"d generates both ph7sica" action and the
ph7sica" conditions 0ithin the bod7 to 0hich a person responds 0hen a verba" com-
munit7 arranges the necessar7 contingencies.

'ther species are a"so conscious in the sense of being under stimu"us contro".
The7 fee" pain in the sense of responding to painfu" stimu"i8 as the7 see a "ight or hear a
sound in the sense of responding appropriate"78 but no verba" contingencies ma$e them
conscious of pain in the sense of fee"ing that the7 are. fee"ing8 or of "ight

SUMMIN6 UP243

or sound in the sense of seeing that. the7 are seeing or hearing that the7 are hearing.

A comp"ete"7 independent science of sub;ective e:perience 0ou"d have no more
bearing on a science of behavior than a science of 0hat peop"e fee" about fire 0ou"d
have on the science of combustion. Dor cou"d e:perience be divorced from the ph7sica"
0or"d in the 0a7 needed to ma$e such a science possib"e. Eifferent verba" communities
generate different $inds and amounts of consciousness or a0areness. 2astern phi"os-
ophies8 ps7choana"7sis8 e:perimenta" ps7cho"og78 phenomeno"og78 and the8 0or"d of
practica" affairs "ead to the observation of ver7 different fee"ings and states of mind. An
independent science of the sub;ective 0ou"d be an independent science of verba"
communities.

1ust 0e conc"ude8 that a"" those 0ho have specu"ated about consciousness as a
form of se"f-$no0"edge-from. the Kree$s to the British empiricists to the phenome-
no"ogists-have 0asted their time4 )erhaps 0e must. The7 deserve credit for directing
attention to the re"ation bet0een a person and his environment (the scientific stud7 of
stimu"us contro" in the name of sensation and perception emerged from phi"osophica"
interests of that sort)8 but the7 have directed in6uir7 a0a7 from antecedent events in his
environmenta" histor7.

. It is hard to understand 0h7 it is so often said that behaviorism neg"ects innate
endo0ment. !atsonBs care"ess remar$ that he cou"d ta$e an7 hea"th7 infant and convert
him into a doctor8 "a07er8 artist8 merchant chief8 and8 7es8 even beggarman or thief can
scarce"7 be responsib"e8 because !atson himse"f repeated"7 referred to the >hereditar7
and habit e6uipment> of peop"e. A fe0 behaviorists8 particu"ar"7 1. .. /antor8 have
minimiFed if not denied a genetic contribution8 and in their enthusiasm for 0hat ma7 be
done through

ABOUT BEHAVIORISM 244

the environment8 others have no doubt acted as if a genetic endo0ment 0ere
unimportant8 but fe0 0ou"d contend that behavior is >end"ess"7 ma""eab"e.>
#ocia" and po"itica" issues have probab"7 p"a7ed a greater ro"e than has been
apparent8 and some have recent"7 conic into the open. The vie0 that "itt"e or nothing is
due to the environment has been inf"uentia" in education. #tudents are c"assified
essentia""7 as those 0ho do not need to be taught and those 0ho cannot be8 and the
doctrine of universa" education is cha""enged on the grounds that some chi"dren are
essentia""7 unteachab"e. But the ro"es of heredit7 and environment are to be discovered
through observation8 not assigned in conformit7 0ith po"itica" be"iefs. #pecies differ in
the speeds 0ith 0hich the7 can be conditioned and in the nature and siFe of the
repertoires the7 can maintain8 and it is probab"e that peop"e sho0 simi"ar inherited
differences. Deverthe"ess8 the topograph7 or form or behavior is on"7 rare"7 affected. To
sa7 that inte""igence or some other abi"it7 or trait is t0ent7 percent a matter of the
environment and eight7 percent a matter of genetics is not to sa7 that t0ent7 percent of
a personBs behavior is due to contingencies of reinforcement and eight7 percent to
genetic endo0ment. .aise one identica" t0in in China and the other in <rance and their
verba" behavior 0i"" be comp"ete"7 different. (The grammars of the t0o "anguages ma7
have certain features in common but not8 as 0e have seen8 because grammar has a
genetic basis.)

3. The e"iciting stimu"us is a particu"ar"7 c"ear e:amp"e of environmenta" action and
probab"7 for that reason 0as the first to be discovered and formu"ated. The notion of a
trigger 0hich re"eased stored behavior 0as suggested to Eescartes b7 certain h7drau"ic
devices that simu"ated human behavior8 and simp"e e:amp"es
SUMMIN6 UP245

0ere demonstrated during the nineteenth centur7 in segments of "iving organisms-in
decapitated sa"amanders8 for e:amp"e-b7 the ph7sio"ogist 1arsha"" Ga"". As 0e have
seen8 )av"ov sho0ed ho0 stimu"i might ac6uire the po0er to e"icit ref"e: responses
during the "ifetime of the individua"8 and a"" this "ed to the ambi tious program of
stimu"us-response ps7cho"og7.

The same appea"ingA simp"icit7 is perhaps responsib"e for the fact that the ref"e:
has retained its p"ace in nontechnica" "iterature as the stereot7pe of environmenta"
action8 as 0e"" as for the fact that behaviorism is so often said to treat behavior simp"7
as response to stimu"us. If that 0ere the case8 an organism 0ou"d have much of the
character of a puppet8 robot8 or machine. But stimu"i do not e"icit operant responsesA
the7 simp"7 modif7 the probabi"it7 that responses 0i"" be emitted. The7 do so because of
the contingencies of reinforcement in 0hich the7 have p"a7ed a part8 and the7 ma7 act
in combination 0ith other conditions8 possib"7 but not necessari"7 to the point at 0hich a
response8 occurs. This is a far different ro"e from that of the e"iciting stimu"us in a ref"e:.

%. Guman beings attend to or disregard the 0or"d in 0hich the7 "ive. The7 search
for things in that 0or"d. The7 genera"iFe from one thing to another. The7 discriminate.
The7 respond to sing"e features or specia" sets of features as >abstractions> or
>concepts.> The7 so"ve prob"ems b7 assemb"ing8 c"assif7ing8 arranging8 and rearranging
things. The7 describe things and respond to their descriptions8 as 0e"" as to descriptions
made. b7 others. The7 ana"7Fe the contingencies of reinforcement in their 0or"d and
e:tract p"ans and ru"es 0hich enab"e them to respond appropriate"7 0ithout direct
e:posure to the contingencies. The7 discover and use ru"es for deriving ne0 ru"es from
o"d. In a"" this8

ABOUT BEHAVIORISM 24"

and much more8 the7 are simp"7 behaving8 and that is true even 0hen the7 are
behaving covert"7. Dot on"7 does a behaviora" ana"7sis not re;ect an7 of these
>higher menta" processes>A it has ta$en the "ead in investigating the contingencies under
0hich the7 occur. !hat it re;ects is the assumption that comparab"e activities ta$e p"ace
in the m7sterious 0or"d of the mind. That assumption8 it argues8 is an un0arranted and
dangerous metaphor.

Do one can give an ade6uate account of much of human thin$ing. It is8 after a""8
probab"7 the most comp"e: sub;ect ever submitted to ana"7sis. The great achievements
of artists8 composers8 0riters8 mathematicians8 and scientists are no doubt sti"" be7ond
reach (in part8 as I have pointed out8 because "eaders in these fie"ds have been mis"ed
b7 menta"ism into giving use"ess reports of their activities). Do matter ho0 defective a
behaviora" account ma7 be8 0e must remember that menta"istic e:p"anations e:p"ain
nothing.

(. 2vo"utionar7 theor7 moved the purpose 0hich seemed to be disp"a7ed b7 the
human genetic endo0ment from antecedent design to subse6uent se"ection b7
contingencies of surviva". 'perant theor7 moved the purpose 0hich seemed to be
disp"a7ed b7 human action from antecedent intention or p"an to subse6uent se"ection b7
contingencies of reinforcement. A person disposed to act because "ie has been
reinforced for acting ma7 fee" the condition of his bod7 at such a time and ca"" it >fe"t
purpose8> but 0hat behaviorism re;ects is the causa" efficac7 of that fee"ing.

+. Contingencies of reinforcement a"so resemb"e contingencies of surviva" in the
production of nove"t7. The $e7 0ord in Ear0inBs tit"e 0as >origin.> Datura" sc"ection
e:p"ained the origination of mi""ions of different species on the surface of the earth8
0ithout appea"ing

SUMMIN6 UP24#

to a creative mind. In the fie"d of human behavior the possibi"it7 arises that
contingencies of reinforcement ma7 e:p"ain a 0or$ of art or the so"ution to a prob"em in
mathematics or science 0ithout appea"ing to a different $ind of creative mind or to a trait
of creativit7 or to the possibi"it7 that >men of genius have more creative nervous energ7
than "esser morta"s.>

In both natura" se"ection and operant conditioning the appearance of >mutations>
is crucia". 5nti" recent"78 species evo"ved because of random changes in genes or
chromosomes8 but the geneticist ma7 arrange conditions under 0hich mutations are
particu"ar"7 "i$e"7 to occur. !e can a"so discover some of the sources of ne0 forms of
behavior 0hich undergo se"ection b7 prevai"ing contingencies of reinforcement8 and for-
tunate"7 the creative artist or thin$er has other 0a7s of introducing nove"ties8 some of
0hich I revie0ed in Chapter &.

&. 'rigination is at the heart of the issue of a se"f or sense of se"f. A member of
the human species has identit78 in the sense that he is one member and no other. Ge
begins as an organism and becomes a person or se"f as he ac6uires a repertoire of
behavior. Ge ma7 become more than one person or se"f if he ac6uires more or "ess
incompatib"e repertoires appropriate8 to different occasions. In se"f-$no0"edge8 the
$no0ing se"f is different from the $no0n. In se"f-management the contro""ing se"f is
different from the contro""ed. But a"" se"ves are the products of genetic and
environmenta" histories. #e"f-$no0"edge and se"f-management are of socia" origin8 and
the se"ves $no0n and managed are the products of both contingencies of surviva" and
contingencies of reinforcement. Dothing about the position ta$en in this boo$ 6uestions
the uni6ueness of each member of the human species8 but the uni6ueness is
ABOUT BEHAVIORISM 24&

inherent in the sources. There is no p"ace in the scientific position for a se"f as a true
originator or initiator of action.

*. If 0e dismiss the pe;orative meaning of >superficia"> as "ac$ing in penetration
and the honorific meaning of >deep> as being profound8 then there is a grain of truth in
the contention that a behavioristic ana"7sis is superficia" and does not reach the depths
of the mind or persona"it7. The thrust of such an ana"7sis is to 6uestion the causa" ro"e
of 0hat is fe"t or introspective"7 observed 0ithin the s$in and to turn instead to genetic
histor78 environmenta" histor78 and the present setting -a"" of 0hich "ie outside. If
behaviorism too$ the "ine of a pure structura"ism8 abandoning the causa" ro"e of the
mind and putting nothing in its p"ace8 it 0ou"d be superficia" in an ob;ectionab"e sense8
but that is a ver7 superficia" vie0 of 0hat it actua""7 does.

-. 2:istentia"ists8 phenomeno"ogists8 and structura"ists8 fre6uent"7 contend that8 in
"imiting itse"f to prediction and contro"8 a science of behavior fai"s to grasp the essentia"
nature or being of man. #o-ca""ed humanistic ps7cho"og7 a"so deprecates prediction or
contro" in its concern for 0hat a person is8 here and no08 apart from his past or future8
and it attempts to ;ustif7 itse"f according"7@ >The theor7 of evo"ution is not a predictive
science8 7et it is high"7 respectab"e and important. Gence other nonpredictive
sciences-histor78 humanistic ps7cho"og7-ma7 be ;ustified.> But the theor7 of evo"ution is
not a science at a""A it is an interpretation of a ver7 "arge number of facts8 using severa"
re"evant sciences8 among them genetics and eco"og78 both of 0hich are or ma7 be
predictive and manipu"ative. A"though8 as 0e sa0 in Chapter -8 understanding8 "i$e
contemp"ation8 is a $ind of $no0ing 0hich often fa""s short of action8 it is derived from
conditions 0hich "ead
24' SUMMIN6 UP

to action. Both prediction and contro" are inherent in operant conditioning8 but the notion
is a"0a7s probabi"istic8 and 0e ma7 dea" 0ith a probabi"it7 0hen action is not ta$ing
p"ace. !e can substitute >understanding> for >$no0ing> onese"f or another person8 but
0hatever the condition at a given moment8 $no0"edge or understanding is >put to use>
on"7 0hen action is ta$en. The more thorough"7 0e understand the re"ation bet0een
human behavior and its genetic and environmenta" antecedents8 the more c"ear"7 0e
understand the nature or essence of the species.

10. A revie0 of a recent boo$ b7 an 2ng"ish behaviorist in a 0e""-B$no0n British
pub"ication contains the comment that >un"i$e behaviorists in America8 the author 0or$s
not 0ith rats but 0ith peop"e.> The reign of the 0hite rat in the ps7cho"ogica" "aborator7
ended at "east a 6uarter of a centur7 ago. An7thing "i$e an ade6uate samp"ing of the
species of the 0or"d is out of the 6uestion8 even for the fie"d etho"ogist8 but the e:-
perimenta" ana"7sis of behavior has been e:tended to a fair"7 "arge number of species8
among them Ho2o sa0ens.

There are e:ce""ent reasons for beginning 0ith simp"e cases and moving on on"7
as the po0er of the ana"7sis permits. If this means8 as it seems to mean8 that one
begins 0ith anima"s8 the emphasis is no doubt upon those features 0hich anima"s and
peop"e have in common. #omething is gained8 ho0ever8 since on"7 in this 0a7 can 0e
be sure of 0hat is uni6ue"7 human. It is not8 as one 0riter contends8 an error to >tr7 to
app"7 a s7stem deve"oped from "imited 0or$ 0ith anima"s to human societ7 and8 indeed8
to the 0ho"e rea"m of human e:perience.>B That is the direction-from simp"e to
comp"e:-in 0hich science moves. But one app"ies the s7stem b7 removing the "imits as
rapid"7 as pos
ABOUT BEHAVIORISM 25(

sib"e and 0or$ing direct"7 0ith human behavior. !e are admitted"7 far short8 as is
ever7one e"se8 of an ade6uate account of human societ7 or the 0ho"e rea"m of human
e:perience8 and it is true that >the arbitrar7 choice to use rats and pigeons to obtain data
necessari"7 ru"es out human freedom and dignit7>A it does so b7 definition8 since rats
and pigeons are not human. But if 0e are to ana"7Fe -the issues raised b7 freedom and
dignit78 it is 0e"" to use ever7 conceivab"e source of information about the under"7ing
processes.

2ven 0ith human sub;ects8 most of the ear"7 e:periments 0ere done 0here the
environment cou"d be most easi"7 contro""ed8 as 0ith ps7chotics and retardates8 (Its
success there has been rather grudging"7 admitted@ >Based on research 0ith anima"s8 it
0or$s gratif7ing"7 0ith those 0ho are a"read7 menta""7 "imited.>) But norma" chi"dren
soon began to be studied8 and then norma" adu"ts. #pecia" prob"ems natura""7 arise from
the re"ation of the sub;ect to the e:perimenter and from the "ong and comp"e: histor7 of
the sub;ect prior to the e:periment8 but enough has been done to suggest that the same
basic processes occur in both anima"s and men8 ;ust as the same $ind of nervous s7s-
tem is to be found in both. There are8 of course8 enormous differences in the comp"e:it7
of their repertoires.

Curious"7 enough8 the re"ation bet0een anima" and human behavior is sometimes
said to point in the other direction. >The stud7 of anima" behavior is uni6ue among the
sciences because it begins historica""7 and methodo"ogica""7 0ith human behavior8
prescinds from human e:perience8 and pro;ects this e:perience into other anima"s.> If
this 0ere the case8 0e shou"d have to sa7 that 0e are "imited in anima" research to
those topics 0hich can be prescinded from human e:perience and pro;ected into
anima"s. The American behaviorist
SUMMIN6 UP25/

2d0ard C. To"man said8 in fact8 that he designed his e:periments 0ith anima"s b7
e:amining 0hat he himse"f 0ou"d do in a given situation8 but the e:perimenta" ana"7sis
of behavior is far be7ond the point at 0hich introspection thro0s an7 "ight on the
processes being studied. It 0ou"d be a remar$ab"e person 0ho predicted his behavior
under the contingencies arranged in a modem "aborator7.

11. The ob;ection to arguing from anima"s to men and 0omen is in part an
ob;ection to e:trapo"ating from "aborator7 to dai"7 "ife8 and the point app"ies as 0e"" 0hen
the organism in the "aborator7 is human. The setting in the "aborator7 is designed to
contro" conditions. #ome are he"d as constant as possib"eA others are changed in an
order"7 fashion. The e6uipment used for these purposes comes bet0een e:perimenter
and organism. It obvious"7 does not promote -it a"most necessari"7 interferes 0ith-the
$ind of persona" $no0"edge 0hich is said to arise from the interpersona" re"ations dis-
cussed in Chapter 11. The e6uipment is not used because it does so8 even though a
c"inica" ps7cho"ogist has c"aimed that >e:perimenta" ps7cho"ogists use their gadgets and
machines to defend themse"ves against rea" invo"vement 0ith their sub;ects.> The7 use
them for the sa$e of a different $ind of >$no0ing another person.>

'bvious"7 0e cannot predict or contro" human behavior in dai"7 "ife 0ith the
precision obtained in the "aborator78 but 0e can neverthe"ess use resu"ts from the
"aborator7 to interpret behavior e"se0here. #uch an interpretation of human behavior in
dai"7 "ife has been criticiFed as metascience8 but a"" the sciences resort to something
much "i$e it. As 0e have ;ust seen8 the princip"es of genetics are. used to interpret the
facts of evo"ution8 as the behavior of substances under high pressures. and
temperatures are used to interpret geo"ogica"

ABOUT BEHAVIOMSM 252

events in the histor7 of the earth. !hat is happening in interste""ar space8 0here contro"
is out of the 6uestion8 is "arge"7 a matter of interpretation in this sense. 1an7
techno"ogica" app"ications pass through a stage of interpretation. !e cannot stud7 the
behavior of a ne0 t7pe of aircraft unti" it has been bui"t and f"o0n8 but it is designed and
bui"t according to princip"es estab"ished in the "aborator7. In much the same 0a7
princip"es emerging from an e:perimenta" ana"7sis of behavior have been app"ied in the
design of education8 ps7chotherap78 incentive s7stems in industr78 peno"og78 and in
man7 other fie"ds.

Those 0ho argue that "aborator7 resu"ts cannot account for human behavior in the
0or"d at "arge presumab"7 be"ieve that the7 $no0 0hat is happening in that 0or"d8 or at
"east that it can be $no0n. The7 are often spea$ing of casua" impressions. But if a
statement about behavior is "ess to be trusted in dai"7 "ife than in a "aborator7 setting8
0e must certain"7 as$ 0hether the impression against 0hich it is compared is an7 more
re"iab"e. Those 0ho fee" that the7 understand 0hat is happening in the 0or"d at "arge
ma7 be tested in a ver7 simp"e 0a7@ "et them "oo$ at the organism as it behaves in a
modem e:periment and te"" us 0hat the7 see. The contingencies current"7 under
investigation8 though e:treme"7 comp"e:8 are far "ess comp"e: than those in dai"7 "ife8
7et it is a"most impossib"e to discover 0hat is going on. Those fami"iar 0ith "aborator7
research 0i"" be more "i$e"7 to "oo$ for the important things and 0i"" $no0 0hat other
things to as$ aboutA the7 0i"" have a better understanding of 0hat the7 see. That is 0h7
the7 can more accurate"7 interpret dai"7 "ife. The "aborator7 ana"7sis ma$es it possib"e to
identif7 re"evant variab"es and to disregard others 0hich8 though possib"7 more
fascinating8 neverthe"ess have "itt"e or no bearing

SUMMIN6 UP253

on the behavior under observation. 1an7 of the techno"ogica" advances derived from
the stud7 of operant behavior have had the benefit of that $ind of interpretation.

1. Those 0ho sa7 that a science of behavior is oversimp"ified and naive usua""7
sho0 an oversimp"ified and naive $no0"edge of the science8 and those 0ho c"aim that
0hat it has to sa7 is either trivia" or a"read7 0e"" $no0n are usua""7 unfami"iar 0ith its
actua" accomp"ishments. To sa7 that behavior is nothing but a response to stimu"i is
oversimp"ified. To sa7 that peop"e are ;ust "i$e rats and pigeons is naive. To sa7 that the
science of behavior is a matter of ho0 rats can "earn to find their 0a7 in maFes or ho0
dogs come to sa"ivate 0hen the7 hear the dinner be"" is to spea$ of 0hat is trivia" or
a"read7 fami"iar. A distinguished phi"osopher of science recent"7 asserted that >even a
behaviorist can8 at the ver7 bes!3 predict that8 under the given conditions a rat 0i"" ta$e
t0ent7 to t0ent7-five seconds to run a maFe@ he 0i"" have no idea ho08 b7 specif7ing
more and more precise e:perimenta" conditions8 he cou"d ma$e predictions 0hich
become more and more precise-and8 n 0$nc0le3 0$ecse ,!*ou! l2!-; T*s s an
observation that might have been time"7 perhaps fift7 7ears ago. It is not surprising that
a 0riter 0ho dismisses behavior modification as a matter of >go"d stars and bac$ pats>
or that the editor of a ne0spaper 0ho assures his readers that a demonstration of
superstition in a pigeon 0i"" not e:p"ain the T0ent7-third )sa"m be"ieves that a science
of behavior is oversimp"ified.

!hen <reud ca""ed behavorism naive8 he 0as spea$ing of an ear"7 version and
comparing it 0ith his o0n e:treme"7 invo"ved account of the menta" apparatus an
account 0hich some of his fo""o0ers have fe"t to be in need of simp"ification. An7one
0ho tries to ma$e

ABOUT B4HAVIOMSM 254

some $ind of s7stematic order of 0hat is said to be ta$ing p"ace inside the s$in 0i""
necessari"7 come up 0ith an account 0hich is far from simp"e. But if 0e are to ca""
an7thing oversimp"ified8 it must be those menta"istic e:p"anations8 so readi"7 invented on
the spot8 0hich are appea"ing because the7 seem so much simp"er than the facts the7
are said to e:p"ain. It is easier to sa7 that a person is suffering from an:iet7 than from a
histor7 of sib"ing riva"r7 to 0hich the an:iet7 must eventua""7 be attributed8 as it is
simp"er to sa7 that >menta" retardates sho0 reactive inhibition> than to e:amine the
defective re"ation bet0een their behavior and the environments to 0hich the7 are
e:posed. The >comp"e:it7 of menta" organiFation> that behaviorism is said to
underestimate is the comp"e:it7 0hich arises from the effort to s7stematiFe formu"ations
that might better be abandoned.

A science of behavior is especia""7 vu"nerab"e to the charge of simp"ification
because it is hard to be"ieve that a fair"7 simp"e princip"e can have vast conse6uences in
our "ives. !e have8 "earned to accept simi"ar apparent discrepancies in other fie"ds. !e
no "onger find it hard to be"ieve that a bacterium or virus can e:p"ain the devastation of
a p"ague. or that the s"ipping of parts8 of the earthBs crust can e:p"ain the traged7 of a
cit7 "eve"ed b7 an earth6ua$e. But 0e find it much more difficu"t to be"ieve that
contingencies of reinforcement can rea""7 be the roots of 0ars8 sa78 or-at the. other
e:treme-of art8 music8 and "iterature.

A"" sciences simp"if7 the conditions the7 stud7 as far8 as possib"e8 but this does not
mean that the7 refuse to e:amine more comp"e: instances as soon as the7 can do so
profitab"7. The reader 0ho has reached this point 0i"" not be inc"ined to ca"" a
behavioristic ana"7sis of human behavior simp"e (un"ess he b"ames the comp"e:
SUMMIN6 UP255

ities he has encountered on m7 e:position)8 and I ma7 remind him that I have $ept facts
and princip"es to a bare minimum. The e:perimenta" ana"7sis of behavior is a rigorous8
e:tensive8 and rapid"7 advancing branch of bio"og78 and on"7 those 0ho are una0are of
its scope can ca"" it oversimp"ified.

As to fami"iarit78 it is true that a scientific ana"7sis has to some e:tent been
anticipated in phi"osoph78 theo"og78 "inguistics8 po"itica" science8 and man7 other fie"ds.
The importance of operant reinforcement8 for e:amp"e8 has "ong been recogniFed in
discussions of re0ards8 se"f-interest8 hedonism8 and 5ti"itarianism. 1ar: and Bentham
have8 been ca""ed behaviorists because the7 dre0 attention to the environment8 but the7
both be"ieved that the environment acted upon consciousness8 0hich in turn affected
human action. The great essa7ists have e:tracted ru"es of thumb and insights ver7
c"ose to some imp"ications of a scientific account. #omething of the same sort has been
said of a"most ever7 scienceA the Kree$s ta"$ed about the atom and the dimension of
time. The mista$e is an e:amp"e of structura"ism@ it is not 0hat 0as said but 0h7 it 0as
said that must be ta$en into account. <acts are not invented b7 the scientist8 and facts
about behavior have a"0a7s been conspicuous features of the 0or"d in 0hich peop"e
"ived8 but scientists 6uic$"7 pass be7ond the stage of fo"$ 0isdom and persona"
e:perience8 and this has a"read7 happened in the e:perimenta" ana"7sis of behavior.
,er7 "itt"e of 0hat is discovered in current research cou"d be said to be fami"iar to
an7one.

)erhaps the best evidence that a science of behavior has something ne0 to offer
is the success of its techno"ogica" app"ications8 but 0e shou"d not over"oo$ the evidence
to be found in the vio"ence 0ith 0hich the position is current"7 attac$ed. I doubt 0hether
so much
ABOUT BEHAVIORISM 25"

gunpo0der 0ou"d be spent on 0hat 0as obvious"7 a naive and outmoded trivia"it7.

13. Behaviorists are sometimes accused of ido"atr7A the7 are said to 0orship
science and to borro0 the trappings of science simp"7 in order to "oo$ scientific. This is
a fair"7 common criticism of a"" socia" or behaviora" sciencesA mere"7 to count or
measure is said to be to ape the natura" sciences. But it is hard to find an7 sign of this in
the histor7 of the e:perimenta" ana"7sis of behavior. 2ar"7 studies used simp"e
e6uipment8 and the data 0ere reported as simp"7 as possib"e. The under"7ing
assumption that behavior 0as order"7 rather than capricious cou"d scarce"7 be said to
have been adopted for honorific purposes. To estab"ish the dimensions of behavior and
re"ated variab"es8 to insist upon prediction and contro"8 to use mathematics 0here
6uantification permitted-these 0ere essentia" steps rather than 0indo0 dressing. The
armamentarium of science is much more conspicuous in information theor78 cognitive
ps7cho"og78 c7bernetics8 and s7stems ana"7ses8 0hich are rep"ete 0ith terms 2$e
>interface8> >gating8> >reverberating circuits8> >parameters of comp"e:it78> >over"oaded
channe"s8> and >c"osed feedbac$ "oops> (>resurrecting purpose and freedomJ>)8 and
0here. mathematics becomes an end in itse"f in spite of the "ac$ of an ade6uate
dimensiona" ana"7sis of the data.

If the critics 0ho spea$ of the >dead hand of scientism> mean it in the "itera" sense
of mortmain8 the7 are presumab"7 suggesting that there 0ere ear"7 hopes of a science
0hich have no0 been abandoned b7 a"" but a desperate fe08 but there is nothing in the
histor7 of the8 ana"7sis of behavior to suggest a setbac$. In fact8 compared 0ith man7
other sciences8 both the ana"7sis and its techno"ogica" app"ications have deve"oped 0ith
un-8 usua" speed. It is true that the death of behaviorism has
SUMMIN6 UP 25#

often been predicted and occasiona""7 announced. A good <reudian might spea$ of a
death 0ish8 and a paranoid behaviorist might vie0 the contentions I am e:amining as
murderous in intent. But the historian8 "i$e ever7one e"se8 is 0e"" advised to turn from
0ishes and intentions to environmenta" contingencies8 and a boo$ such as this is
designed to be part of the environment of those 0ho 0i"" continue to ta"$ about
behaviorism as a "iving phi"osoph7.

1%. There have been man7 dramatic app"ications of operant conditioning8 but
ver7 often 0hat is done seems in retrospect to be "itt"e more than the app"ication of
common sense. Deverthe"ess8 0e have to as$ 0h7 simi"ar changes 0ere not made
before the advent of an e:perimenta". ana"7sis. It is sometimes said that the7 0ere8 and
iso"ated instances in 0hich something ver7 much "i$e a modem behaviora" techno"og7
can be cited. But 0e ma7 sti"" as$ 0h7 these occasiona" instances8 scattered throughout
the centuries8 have not become standard practice. 2ven the fact that a practice 0or$s or
that it ma$es sense is often not enough to ensure its continued use8 and one reason is
that8 temporari"7 at "east8 the 0rong practices a"so seem to 0or$ and ma$e sense.

The important difference is in the time at 0hich the desirab"e resu"ts occur. In
spite of the fact that man78 peop"e find them ob;ectionab"e8 punishment and aversive
contro" are sti"" common8 and for a sing"e reason@ those 0ho use them are usua""7
immediate"7 reinforcedA on"7 the deferred resu"ts are ob;ectionab"e. )ositive
reinforcement is at "east e6ua""7 po0erfu" as a contro""ing measure8 but its effects are
usua""7 at "east s"ight"7 de-. ferred. It is on"7 0hen "aborator7 research demonstrates8
that positive reinforcement has 0orth0hi"e conse6uences that one "earnsA to 0ait for
them.
ABOUR BEHAVIORISM 25&

The disastrous resu"ts 'f common sense in the management of human behavior
are evident in ever7 0a"$ of "ife8 from internationa" affairs to the care of a bab78 and 0e
sha"" continue to be inept in a"" these fie"ds unti" a scientific ana"7sis c"arifies the
advantages of a more effective techno"og7.. It 0i"" then be obvious that the resu"ts are
due to more than common sense.

1(. In An In<u$) n!o Meann+ and T$u!* Bertrand .usse"" 0rote as fo""o0s@
>!hen the behaviorist observes the doings of anima"s8 and decides 0hether these sho0
$no0"edge or error8 he is not thin$ing of himse"f as an anima"8 but as an at "east
h7pothetica""7 inerrant recorder of 0hat actua""7 happens.> Ge 0as spea$ing 'f an ear"7
version of behaviorism8 and even so he 0as not right. It 0ou"d be absurd for the
behaviorist to contend that he is in an7 0a7 e:empt from his ana"7sis8 Ge cannot step
out of the causa" stream and observe behavior from some specia" point of vantage8
>perched on the epic7c"e of 1ercur7.> In the ver7 act of ana"7Fing human behavior he is
behaving-as8 in the ver7 act of ana"7Fing thin$ing8 the phi"osopher is thin$ing.

The behavior of "ogician8 mathematician8 and scientist is the most difficu"t part of
the fie"d of human behavior and possib"7 the most subt"e and comp"e: phenomenon
ever submitted to a "ogica" mathematica"8 or scientific ana"7sis8 but because it has not
7et been 0e"" ana"7Fed8 0e shou"d not conc"ude that it is a different $ind of fie"d8 to be
approached on"7 0ith a different $ind of ana"7sis. There is no reason 0h7 0e cannot as$
0hat a "ogician or mathematician does as he discovers ho0 or 0h7 ne0 ru"es can be
derived from o"d or 0h78 if the o"d can be said to be true8 the ne0 must be true too. It is
possib"e8 in fact8 that a behaviora" ana"7sis ma7 7ie"d a ne0 $ind of attac$ on fami"iar
prob"ems8 such as the parado:es or K4de"Bs theorem.

SUMMIN6 UP25'

#cientific $no0"edge is verba" behavior8 though not necessari"7 "inguistic. it is a
corpus of ru"es for effective action8 and there is a specia" sense in 0hich it cou"d be
>true> if it 7ie"ds the most effec!%e ac!on possib"e. But ru"es are never the contingencies
the7 describeA the7 remain descriptions and suffer the "imitations in-8 herent in verba"
behavior. As I pointed out in Chapter *8 a proposition is >true> to the e:tent that 0ith its
he"p the "istener responds effective"7 to the situation it describes. The account given b7
the spea$er functions in "ieu of the direct contro" b7 the environment 0hich has
generated it8 and the "istenerBs behavior can never e:ceed the behavior contro""ed b7
the situation de8 scribed. The tauto"ogica" truth of the "ogician or mathematician can be
provedA it is abso"ute. The canons of scientific methods are8 designed to ma:imiFe the
contro" e:erted b7 the stimu"us and to suppress other conditions8 such as incidenta"
effects upon the "istener 0hich "ead the spea$er to e:aggerate or Ge.

Traditiona" theories of $no0"edge run into troub"e because the7 assume that one
must thin$ before behaving (not to mention thin$ing before e:isting8 as in Cogito8 ergo
sum). Do one thin$s before he acts e:cept in the sense of acting covert"7 before acting
overt"7.

It is sometimes said that to spea$ the truth one must be freeA contro""ed behavior
is too restricted to be ;udged or accepted as true8 and the behaviorist 0ho insists that a""
behavior is determined evident"7 denies that it can therefore be true at a"". But one is
>free to spea$ the truthC on"7 0ith respect to a present setting. !hat he does in that
setting is a function of antecedent conditions in 0hich simi"ar settings have p"a7ed a
part. Certain"7 the procedure according to 0hich one arrives at the truth b7 deduction is
not freeA one is bound to the >"a0s of thought> and to other ru"es for deriving
ABOUT BEHAVIORISM 2"(

ne0 ru"es from o"d. The intuitions 0hich seem to antici pate deduction but are "ater
deductive"7 proved suggest a $ind of freedom8 but not if intuitive behavior is con-
tingenc7-shaped rather than ru"e-governed. Induction8 "i$e intuition or insight8 a"so
suggests freedom8 but it is the freedom from the constraints of ru"e-governed behavior
and can be ca""ed freedom on"7 unti" the contro" of the environment in evo$ing intuitive
behavior is fu""7 recogniFed.

The behavior of -the scientist is often reconstructed b7 scientific methodo"ogists
0ithin a "ogica" frame0or$ of h7pothesis8 deduction8 and the testing of theorems8 but the
reconstruction se"dom represents the behavior of the scientist at 0or$. Certain prob"ems
concerning the "imitations of human $no0"edge might be c"oser to so"ution if the
behavior of $no0ing 0ere ana"7Fed further. It has been said8 for >amp"e8 that science
has reached a "imit be7ond 0hich it cannot estab"ish the determinac7 of ph7sica"
phenomena8 and it has been argued that this ma7 be the point at 0hich freedom
emerges in human behavior. Behaviora" scientists 0ou"d probab"7 be 6uite content 0ith
the degree of rigor sho0n b7 ph7sics in spite of this apparent "imitation8 but there ma7
be something about the human organism 0hich ma$es indeterminac7 re"ative"7
important. !e can discover 0hether or not that is an important "imitation on"7 b7
deve"oping a science of human behavior to the point at 0hich indeterminac7 becomes
apparent.

There 0ou"d remain the possibi"it7 of investigating the behavior of the scientist
and the nature of scientific $no0"edge to see 0hether some abso"ute "imit has in fact
been reached. #imi"ar arguments have proved 0rong in the past. <or e:amp"e8 the
"imits of 0hat cou"d be seen through the microscope 0ere once c"ear"7 estab"ished from
a consideration of the 0ave "engths of
SUMMIN6 UP2"/

visib"e "ight. #ince then8 of course8 the e"ectron microscope has proved that the ear"ier
determination8 though right on the evidence8 0as 0rong 0ith respect to the "imits of
microscop7.

1+. It is not a"0a7s c"ear 0hat is meant 0hen it is said that a behaviora" ana"7sis
dehumaniFes man or destro7s man 6ua man. #ometimes the imp"ication seems to be
that its picture of man is incomp"ete@ >Behaviorism tried to bui"d a ps7cho"og7 0ithout
inc"uding man in his fu"" comp"e:it78> or8 >Behaviorism has omitted human phenomena
0hich do not fit a ph7sica"istic mode".> (Gumanistic ps7cho"og78 on the other hand8 is9
said to be a science >appropriate to man as a sub;ect matter8> >committed to dea"ing
0ith humanness in its o0n right8> and >comprehensive"7 human.>) But phrases "i$e >man
6ua manCB or >man in his humanit7>B te"" us ver7 "itt"e about 0hat has been "eft out.

#ometimes the imp"ication is that behaviorism neg"ects something a person does
because he is a member of the human species8 or $eeps him from doing something he
0ou"d other0ise do as such a member. The position of the <rench phi"osopher Keorges
#ore" has been paraphrased in this 0a7@

;an) at his best) that is) at his most human) seeks to fu*fi** himse*f) indi/idua**- and with
those c*ose to him) in spontaneous) unended) creati/e acti/it-) in work that consists of the
imposition of his persona*it- on a reca*citrant en/ironment.... >e acts and is not acted upon) he
chooses and is not chosen for.... >e resists e/er- force that seeks to reduce his ener.-) to rob him
of his independence mid his di.nit-) to ki** the wi**) to crush e/er-thin. in him that stru..*es for
uni=ue se*f6e,pression and reduce him to uniformit-) impersona*it-) monoton-) and) u*timate*-)
e,tinction.
This characteriFation of the species is "i$e"7 to be subscribed to b7 a"" those members of
the species 0ho can

ABOUT BEHAVIORISM 2"2

understand it8 but it does not identif7 an7thing essentia""7 human8 as can be sho0n b7
app"7ing it to another species. !e can readi"7 agree that a "ion ;umping through a hoop
in a circus is not behaving 6ua "ion8 and 0e might e"aborate in this 0a7@

*he lion at his best$ that is$ at his most leonie$ seeks to fulfill himself$
iniviually an "ith those close to him$ in spontaneous$ unene$ creative activity$
in "ork that consists of the imposition of his leoninity on a recalcitrant environment. .
. . (e acts an is not acte upon. (e chooses an is not chosen for.... (e resists every
force that seeks to reuce his ener#y$ &to rob him of his inepenence an his i#nity$
to kill the "ill$ to crush everythin# in him that stru##les for unique self&e)pression
an reuces him to uniformity$ unleoninity$ monotony$ an$ ultimately$ e)tinction.

I suspect that most "ions 0ou"d subscribe to this reassuring picture if the7 cou"d.

It is often said that a behavioristic account someho0 neg"ects something of 0hat a
person can be or do because it treats him as a machine. >1an comes to thin$ of
himse"f8> as 1artin Buber put it8 >as if he 0ere determined b7 the same mechanica" "a0s
that govern his refrigerator.> But to assert that human behavior is "a0fu" is not to sa7
that the "a0s 0hich govern it are as simp"e or as >mechanica"> as those that app"7 to the
operation of a refrigerator. Dor is the choice bet0een (a) an entire"7 techno"ogica"
societ7 in 0hich persons are run b7 machines and (b) >an era of humanit7 0ith man at
peace 0ith himse"f b7 comporting 0ith his natura" environment.> And 0e can scarce"7
den7 that man is an anima"8 though a remar$ab"e one. The comp"aint that )av"ov
converted Gam"etBs >Go0 "i$e a god> into >Go0 "i$e a dogJ> 0as ans0ered b7 Gam"et
himse"f@ >In action ho0 "i$e an ange"J In apprehension ho0 "i$e
SUMMIN6 UP2"3

a godJ The beaut7 of the 0or"dJ The paragon of ani ma"sJ> 1an s the paragon of
anima"s.

!hat is usua""7 meant in sa7ing that behaviorism dohumaniFes man is that it
neg"ects important capacities 0hich are not to be found in machines or anima"s8 such as
the capacit7 to choose8 have purposes8 and behave creative"7. But the behavior from
0hich 0e infer choice8 intention8 and origina"it7 is 0ithin reach of a behaviora" ana7"sis8
and it is not c"ear that it is 0ho""7 out of reach of other species. 1an is perhaps uni6ue in
being a mora" anima"8 but not in the sense that he possesses mora"it7A he has
constructed a socia" environment in 0hich he behaves 0ith respect to himse"f and
others in mora" 0a7s.

1an7 of these issues 0ere no doubt neg"ected in ear"7 versions of behaviorism8
and methodo"ogica" behaviorism s7stematica""7 ru"ed some of them out of account8 but I
$no0 of no essentia""7 human feature that has been sho0n to be be7ond the reach of a
scientific ana"7sis8 and I doubt 0hether those 0ho charge dehumaniFation 0ou"d 0ish to
rest their case on the inade6uac7 of a behaviora" account8 since the future might turn
too heavi"7 against them.

Behavior is the achievement of a person8 and 0e seem to deprive the human
organism of something 0hich is his natura" due 0hen 0e point instead to the
environmenta" sources of his behavior. !e do not dehumaniFe himA 0e dehumauncu"iFe
him. The essentia" issue is autonom7. Is man in contro" of his o0n destin7 or is he not4
The point is often made b7 arguing that a scientific ana"7sis changes man from victor to
victim. But man remains 0hat he has a"0a7s been8 and his most conspicuous
achievement has been the design and construction of a 0or"d 0hich has freed him from
constraints and vast"7 e:tended his tango.
ABOUT BEHAVIORISM2"4

Do doubt he has been care"ess. In the nineteenth centur7 manBs overriding
inhumanit7 to man came from the Industria" .evo"ution-the pa7ment of 0ages to a
hungr7 "abor force8 for e:amp"e8 neg"ected serious side effects. 1ar: is said to have
described this under the inf"uence of ear"ier romantic 0riters. #chi""er8 for e:amp"e8 had
0ritten@ >2n;o7ment 0as separated from "abor8 the means from the end8 e:ertion from
recompense. 2terna""7 fettered on"7 to a sing"e "itt"e fragment of the 0ho"e8 man
fashions himse"f on"7 as a fragmentA ever hearing on"7 the monotonous 0hir" of the
0hee" 0hich he turns8 he never disp"a7s the fu"" harmon7 of his being.> In other 0ords8
"abor no "onger had the reinforcing conse6uences 0hich generate the condition fe"t as
;o7A the contingencies sustained a ver7 narro0 repertoireA a person had no chance to
ac6uire most of the behavior of 0hich he 0as capab"e.

Toda7 other side effects are attracting more attention. 1an continues to bui"d
machines 0hich dehumaniFe him b7 dispensing 0ith behaviors that contribute to his
status as a person8 but he is a"so breeding at a dangerous rate8 e:haustirig the 0or"dBs
resources8 poI"uting the environment8 and doing "itt"e to re"ieve the threat of a nuc"ear
ho"ocaust. Deverthe"ess8 if the position I have presented here is correct8 he can remed7
these mista$es and at the same time bui"d a 0or"d in 0hich he 0i"" fee" freer than ever
before and achieve greater things.

Ge can do this on"7 if he recogniFes himse"f for 0hat he is. Ge has fai"ed to so"ve
his prob"ems because he has "oo$ed in the 0rong p"ace for so"utions. The e:traordinar7
ro"e of the environment opens the prospect of a much more successfu" future8 in 0hich
he 0i"" be most human and humane8 and in 0bich he 0i"" manage himse"f s$i""fu""7
because he 0i"" $no0 himse"f accurate"7.


SUMMIN6 UP2"5

A science of behavior has been said to dehumaniFe man because it is
reductionistic. It is said to dea" 0ith one $ind of fact as if it 0ere a different $ind-as is
done for e:amp"e8 b7 ph7sio"ogica" ps7cho"og7. But bebaviorism does not move from
one dimensiona" s7stem to another. It simp"7 provides an a"ternative account of the
same facts. It does not $educe fee"ings to bodi"7 statesA it simp"7 aTgue83 that bodi"7
states are and a"0a7s have been 0hat are -fe"t. It does not $educe thought processes to
behaviorA it simp"7 ana"7Fes the behavior previous"7 e:p"ained b7 the invention of
thought processes. It does not $educe mora"it7 to certain features of the socia"
environmentA it simp"7 insists that those features have a"0a7s beim responsib"e for
mora" behavior.
2ven so8 something in traditiona" formu"ations ma7 seem to be missing. The
prob"em is not pecu"iar to the behavioria" sciences. Go"d a s"ip of paper ;ust above a
cand"e f"ame and it 0i"" >catch fire.> !e spea$ of fire as something things catch and then
possess. !hen 0e stop the burning8 0e are said to >put the fire out>A 0e >6uench> it in
the sense of causing it to vanish. The metaphor is probab"7 harm"ess enough in casua"
discourse8 but it is not particu"ar"7 usefu" to the ph7sicist8 and a person 0ho is
accustomed to 6uenching fires b7 thro0ing 0ater on them 0i"" be unhapp7 0hen be is
to"d that 0ater simp"7 coo"s or s2o!*e$s a fire. Coo"ing and smothering do not seem "i$e
6uenching. But it 0ou"d be 0rong to sa7 that the process of 6uenching had been
>reduced> to coo"ing and smothering. The term has simp"7 been trans"ated into terms
having a broader reference.88 and8 as in simi"ar instances in a behaviora" ana"7sis8 0hat
seems to be missing is nothing to be ta$en serious"78 nor does it ;ustif7 the charge of
reductionism.

Behaviorism cou"d perhaps be ca""ed reductionistic

ABOUT BEHAVIORISM 2""

0hen it discusses the socia" sciences. It reduces socia" processes to the behavior of
individua"s8 But that cou"d scarce"7 be ca""ed a reduction of the ro"e of the indi vidua". It
is8 rather8 a $ind of aggrandiFement.

1&. 2ver7 science has probab"7 been accused at one time or other of neg"ecting
the uni6ueness of the individua" case in its search for genera" princip"es. There is more
in a sunset8 a storm at sea8 a b"ade of grass8 or a piece of music than is dreamt of in
phi"osophies or accounted for in science. 2:posure to a uni6ue contingenc7 of
reinforcement generates a specia" $ind of $no0ing8 and the fee"ings or introspective"7
observed states of mind associated 0ith it differ 0ide"7 from those produced 0hen a
person fo""o0s a ru"e or obe7s a "a0. The specia"ist ma7 come c"ose to the uni6ue
event8 but he never encompasses a"" of it.

)s7chotherap7 has natura""7 been concerned 0ith the individua? Its stoc$ in
trade is the case histor78 0ith its end"ess fascination8 and the intensive $no0"edge of an-
other person ac6uired in therapeutic sessions is no doubt be7ond the reach of a science
0hich emphasiFes genera"ities. )s7chotherap7 is "arge"7 responsib"e for the humanistic
ps7cho"og7 0hich comp"ains that behaviorism ignores the individua". As one ma7 $no0
the 0or"d in a sense 6uite different from $no0ing ph7sica" and bio"ogica" science8 so one
ma7 $no0 0eo0le in a sense 6uite different from $no0ing behaviora" science. It 0ou"d
be foo"ish to 6uestion the interest a person ta$es in others----in gossip8 in
autobiographies8 in nove"s8 in dramas8 in ne0s reports8 and so on. Individua"s are
important parts of the environment8 and a"though a science of behavior permits a
person to interpret 0hat he sees more effective"78 it 0i"" never te"" him the 0ho"e stor7
about the individua" case.
It is on"7 0hen 0e as$ 0hat is to be done 0ith $no0"edge
SUMMIN6 UP2"#

that 0e begin to e:amine the different forms of $no0ing more c"ose"7 and appreciate the
va"ue of the genera"it7. #cience must ba"ance costs and gains8 and though it ma7 bear
do0n hard on a uni6ue event8 especia""7 in a techno"ogica" app"ication8 it reaps a
greater harvest from genera" princip"es.

1*. Basic research in the science of behavior is essentia""7 manipu"ativeA the
e:perimenter arranges conditions under 0hich a sub;ect behaves in a given 0a78 and in
doing so he contro"s behavior. Because operant conditioning is conspicuous8 it is often
referred to as if it 0ere nothing more than a techni6ue to be used in the contro" of
others. There is no doubt that it can be so used for nonscientific purposes and 0i"" be so
used if the resu"ts are reinforcing. Among those 0ho have the po0er needed to contro"
others in this 0a7 are governmenta" and re"igious authorities and men possessing a
good dea" of mone78 !e escape from them or attac$ their po0er 0hen the7 resort to
aversive methods or methods 0hich have the deferred aversive conse6uences ca""ed
e:p"oitation. As 0e have seen8 those 0ho do not use their po0er in aversive or
e:p"oitative 0a7s do not refrain because the7 possess compassion8 or a sense of
ethics8 or a concern for the 0e"fare of others8 but because the7 have been sub;ected to
countereontro". Eemocrac7 is a version of countercontro" designed to so"ve the prob"em
of manipu"ation.

The difference bet0een conspicuous and inconspicuous contro" has "ed to man7
misunderstandings. It is sometimes said that chi"dren 0ho are being taught verba"
behavior are ac6uiring it through operant conditioning but that 0hat the7 "earn from a
noninstructiona" verba" communit7 is "earned through some other process. #imi"ar"78 it
has been said that government and economics are peihaps proper fie"ds for a
behavioristic
ABOUT BEHAVIORISM2"&

ana"7sis8 but that art8 "iterature8 music8 re"igion8 and dai"7 "ife are not. Chi"dren in a
c"assroom organiFed on the princip"es of a to$en econom7 are fe"t to be contro""ed8 but
0hen it proves possib"e to abandon e:p"icit reinforcement and to move on to such a
natura" reinforcer as successfu" accomp"ishment8 the chi"dren are said to be free.

!hat is at issue is not the behavioria" process but the contingencies.
Contingencies designed for e:p"icit purposes can be ca""ed manipu"ative8 though it does
not fo""o0 that the7 are e:p"oitativeA unarranged contingencies must be recogniFed as
having e6ua" po0er8 and a"so possib"7 unhapp7 conse6uences. It must not be forgotten
that e:hortation8 demagoguer78 evange"ism8 and so on are a"so behaviora" practices8 as
are simi"ar practices on a sma""er sca"e in dai"7 "ife. !e are a"" so used to being
contro""ed to our disadvantage that to ca"" a person harm"ess is to imp"7 that he is tota""7
ineffective or feeb"e-minded.

To sa7 that a"" contro" is manipu"ative and hence 0rong is to over"oo$ important
uses in education8 ps7chotherap78 government8 and e"se0here. A proposa" to terminate
behaviora" research or to se6uester its resu"ts on the grounds that the7 can be used b7
despots and t7rants 0ou"d be a disastrous mista$e8 because it 0ou"d undermine a"" the
important contributions of the cu"ture and interfere 0ith the counter-contro""ing measures
0hich $eep avcrsive and e:p"oitative contro" 0ithin bounds.

1-. In an operant ana"7sis of the stimu"us contro" of verba" behavior8 0e can
identif7 the referent of abstract terms but terms "i$e >mora"it7> and >;ustice> raise an
additiona" prob"em. It can be so"ved b7 recogniFing that the behavior 0e ca"" mora" or
;ust is a product of

SUMMIN6 UP2"'

specia" $inds of socia" contingencies arranged b7 governments8 re"igions8 economic
s7stems8 and ethica" groups. !e need to ana"7Fe those contingencies if 0e are to bui"d
a 0or"d in 0hich peop"e behave mora""7 and ;ust"78 and a first step in that direction is to
dismiss mora"it7 and ;ustice as persona" possessions.

0. A common reaction to behaviorism runs as fo""o0s@ >!hat 7ou have said ma7
a"" be true8 but I am not interested in m7 behavior. I am interested in m7 fee"ings-and in
the fee"ings of others. I read boo$s because the7 intrigue or e:cite meA I "isten to music
because it e:hi"arates meA I "oo$ at pictures because I find them beautifu"A I associate
0ith peop"e I "ove or 0ith 0hom I en;o7 ta"$ing about ever7da7 things 0ith ever7da7
0ords.> The same cou"d be said8 of course8 about an7 science@ >I do not care about
immuno"og78 I simp"7 0ant to avoid disease>A >I do not care about genetics8 I simp"7
0ant hea"th7 chi"dren>A >I do not care 0here energ7 comes from8 I simp"7 0ant comfort
and convenience.> A $no0"edge of medicine8 genetics8 and techno"og7 does not
interfere 0ith fee"ing 0e""8 having hea"th7 chi"dren8 or being comfortab"e8 and no one is
"i$e"7 to suppose that it does8 but simi"ar statements about behavior are debated. Pet
there is nothing in a science of behavior or its phi"osoph7 0hich need a"ter fee"ings or
introspective observations. The bodi"7 states 0hich are fe"t or observed are
ac$no0"edged8 but there is an emphasis on the environmenta" conditions 0ith 0hich
the7 are associated and an insistence that it is the conditions rather than the fee"ings
0hich enab"e us to e:p"ain behavior.

Those 0ho understand the theor7 or histor7 of music do not find music therefore
an7 "ess en;o7ab"e8 nor are those 0ho understand the techni6ues of the artist or

AROUT BEHAVIORISM 2#(

the histor7 of art "ess "i$e"7 to en;o7 paintings. It is true that a touch of m7ster7 ma7 be
reinforcing8 and 0e ma7 be particu"ar"7 moved b7 the apparent"7 ine:p"icab"e8 but if
there 0ere no offsetting gains8 educators 0ou"d have a good dea" to e:p"ain.

A distinguished critic of science has e:pressed an opposing vie0 in tD9 fo""o0ing
0a7@ >according to Hctho"og7I /eats is a"" 0rong@ the bird is not pouring forth its sou" in
ecstas78 for no0 0e $no0 that a"" it is don+ s serving notice on its fe""o0s that it c"aims
a certain territor7 for 0orm grubbing.> And he insists that 0e sha"" never $no0 0h7 birds
sing8 >but as poets 0e $no0-none better-ho0 their singing affects us and as to this
datum science has not a 0ord to sa7A it can on"7 "isten too.>

It seems to be imp"ied that to understand 0h7 birds sing 0ou"d interfere 0ith the
effect of their singing upon us8 and upon the poet8 and upon us 0hen 0e read 0hat he
has to sa7. The etho"ogist 0ou"d be 0rong to ta$e these effects into account in tr7ing to
discover 0h7 birds sing8 but he can neverthe"ess en;o7 bird song and a"so 0hat a poet
sa7s about it. The bird sings not because of ho0 it fee"s but because of certain contin-
gencies of surviva". Go0 the poet fee"s upon hearing it is doub"7 irre"evant to 0h7 it
sings8 but there is no reason 0h7 the poet cannot te"" us ho0 he fee"s or8 if he is a good
poet8 induce in us a condition fe"t in the same 0a7.

If 0e stop to "isten to a bird8 it is because 0e are reinforced for doing so8 and
science can "isten for other reasons. It can surve7 the e:tent to 0hich sound patterns
are or become reinforcing8 and in doing so it ma7 contribute to an e:p"anation of 0h7
peop"e compose and "isten to music. The conditions generated 0ithin the bod7 of the
"istener remain forever private8 but the

SUMMIN6 UP2#/

behaviora" scientist ma7 sti"" investigate the reinforcing effects the7 are associated 0ith
and possib"7 discover ho0 more reinforcing effects ma7 be achieved.

Dot on"7 has the most ardent behaviorist fee"ings "i$e ever7one e"seA on ba"ance
he has 6uite possib"7 more en;o7ab"e ones8 because there are states of the bod7-
associated8 for e:amp"e8 0ith fai"ure8 frustration8 or "oss -0hich are far from en;o7ab"e or
reinforcing8 and the7 are "ess "i$e"7 to be e:perienced b7 those 0ho practice scientific
se"f-$no0"edge and se"f-management. And it is hard to see ho0 a he"p! interest in8 or
affection for8 another person cou"d be endangered b7 improved understanding.

The Behaviorists Own Behavior

#o much for the misunderstandings and criticisms "isted in the Introduction. To them
shou"d perhaps be added the charge that the behaviorist constant"7 vio"ates his o0n
princip"es8 most obvious"7 b7 his continued use of menta"istic terms. Ge sa7s8 >I thin$>A
he as$s his readers to $eep something >in mind>A he summariFes the >purport> or
>purpose> of a passageA and so on. In the samp"e 0hich the reader has no0 had a
chance to e:amine8 I be"ieve I have been consistent in the fo""o0ing respects@

I have used tee"mica" terms in ma$ing a technica" point. I have preferred a
technica" term e"se0here 0hen it cou"d be used at no great cost. .ather than sa7 that
our prob"ern is >to create a concern for the future8> I have preferred to sa7 that it is >to
induce peop"e to act 0ith respect to a future.> I prefer the e:pression >It occurred to me .
. .> to >The thought occurred to me.> But e"se0here I have free"7 used the "a7 vocabu"ar7
0hi"e accepting the responsibi"it7 of providing a tech
ABOUT BEHAVIORISM 2#2

nica" trans"ation upon demand. There is no other 0a7 if a boo$ of this $ind is to be brief
and readab"e. The reader 0ho ob;ects must a"so ob;ect to the ph7sician 0ho te""s him
that he has caught a >co"d> (rather than a virus) or the a"manac 0hich te""s him 0hen the
sun 0i"" rise rather than 0hen it 0i"" become visib"e over the horiFon as the earth turns.
The convenience of the "a7 vocabu"ar7 does not 0arrant its use 0here a technica"
a"ternative 0ou"d be more he"pfu". 2ducation8 for e:amp"e8 has "ong suffered from
efforts to ana"7Fe teaching and "earning in "a7 terms.

The ob;ection is not a"0a7s a matter of vocabu"ar7. Those 0ho approacha
behavioristic formu"ation for the first time ma7 be surprised b7 the mention of se"f-con-
tro". Eoes this not su8-gest some $ind of inner determination4 'r happiness---does this
not mean that fee"ings are important4 The behavioristBs o0n behavior a"so seems to
vio"ate his princip"es. Gas he not decded to 0rite a boo$4 Is he not c"ear"7 $es0onsble
for it8 since it 0ou"d not e:ist if he had not 0ritten it4 Eoes he not urge his readers to
adopt a behavioristic point of vie04 According to traditiona" definitions of se"f -contro"8
happiness8 decision8 responsibi"it78 and urging8 the behaviorist is indeed inconsistent8
but according to his o0n definitions he is notA and 0hen the "atter are understood8
ob;ections of this sort "ose their force.

Another version ta$es this form@ >If human behavior is as fu""7 determined as the
behaviorist sa7s it is8 0h7 does he bother to 0rite a boo$4 Eoes he be"ieve that
an7thing matters4> To ans0er that 6uestion 0e shou"d have to go into the histor7 of the
behaviorist. Dothing he sa7s about human behavior serious"7 changes the effect of that
histor7. Gis research has not a"tered his concern for his fe""o0 men or his be"ief in the
re"evance

SUMMIN6 UP2#3

of a science or techno"og7 of behavior. #imi"ar 6uestions might as 0e"" be as$ed of the
author of a boo$ on respiration@ >If that is respiration8 0h7 do 7ou go on breathing4>

On the ositive #ide

Behaviorism has so often been defined in terms of its supposed shortcomings-of 0hat it
is said to ignore or neg"ect-that setting the record straight often appears to destro7 0hat
0as meant to be saved. In ans0ering these charges I ma7 seem to have >abandoned
the ver7 basis of behaviorism8> but 0hat I have abandoned are the vestiges of ear"78
statements of the position8 sub;ected to various e"aborations and criticisms over a
period of some si:t7 7ears. !hat survives can be put in a positive form@

1. The position I have ta$en is based8 as the read> 0as 0arned8 on a particu"ar
$ind of behavioria" science. I have chosen it in part no doubt because of m7 fami" iarit7
0ith it but main"7 because it has certain features especia""7 re"evant to the behavioristic
argument. It offers8 I be"ieve8 the c"earest possib"e statement of the causa" re"ations
bet0een behavior and environment. It ana"7Fes individua" data rather than group
averages. The comp"e:it7 of the e:perimenta" environment has gradua""7 increased unti"
it no0 approaches the comp"e:it7 of dai"7 "ife--in 0hich8 therefore8 e:trapo"ations from
the "aborator7 become increasing"7 usefu".

. !hat 0e have "earned from the e:perimenta" ana"7sis of behavior suggests
that the environment performs the functions previous"7 assigned to fee"ings and
introspective"7 obsorved inner states of the organism. This fact has been on"7 s"o0"7
recogniFed. 'n"7 ver7

ABOUT BEHAVIORISM 2#4

strong evidence of the ro"e of the environment cou"d offset the effects of menta"ism in
directing attention to supposed inner causes.

3. A behaviora" ana"7sis ac$no0"edges the importance of ph7sio"ogica" research.
!hat an organism does 0i"" eventua""7 be seen to be due to 0hat it is8 at the moment it
behaves8 and the ph7sio"ogist 0i"" someda7 give us a"" the detai"s. Ge 0i"" a"so te"" us
ho0 it has arrived at that condition as a resu"t of its previous e:posure to the
environment as a member of the species and as an individua".

%. A crucia" step in the argument can then be ta$en@ 0hat is fe"t or seen through
introspection is on"7 a sma"" and re"ative"7 unimportant part of 0hat the ph7si3"ogist 0i""
eventua""7 discover. In particu"ar it is not the s7stem 0hich mediates the re"ation
bet0een behavior and the environment revea"ed b7 an e:perimenta" ana"7sis.

As the phi"osoph7 of a science of behavior8 behaviorism ca""s for probab"7 the
most drastic change ever proposed in our 0a7 of thin$ing about man. It is a"most
"itera""7 a matter of turning the e:p"anation of behavior inside out.

The F!t!re of Behavioris$

A good dea" of 0hat is ca""ed behaviora" science is not behavioristic in the present
sense. #ome of it8 as 0e have seen8 avoids theoretica" issues b7 confining itse"f to the
form8 topograph78 or structure of behavior. #ome of it appea"s to the >conceptua"
nervous s7stems> of mathematica" mode"s and s7stems theories. 1uch of it remains
fran$"7 menta"istic. )erhaps this diversit7 is hea"thfu"@ different approaches cou"d be
regarded as mutations8 from 0hich a tru"7 effective behaviora" science 0i"" eventua""7 be
se"ected. Deverthe"ess8 the
SUMMIN6 UP 2#5

present condition is not promising. 2ven in a sing"e part of the fie"d it is unusua" to find
t0o authorities ta"$ing about precise"7 the same things8 and a"though nothing cou"d be
more re"evant to the prob"ems of the 0or"d toda78 the actua" accomp"ishments of
behaviora" science do not seem -to be e:tensive. (it has been suggested that the
science is >too 7oung> to so"ve our prob"ems. This is a curious e:amp"e of
deve"opmenta"ism8 in 0hich immaturit7 offers a $ind of e:oneration. !e forgive the bab7
for not 0a"$ing because he is not 7et o"d enough8 and b7 ana"og7 0e forgive the asocia"
or disturbed adu"t because he has not 6uite gro0n up8 but must 0e then 0ait unti" the
behaviora" sciences gro0 more effective4)

I contend that behaviora" science has not made a greater contribution ;ust
because it is not ver7 behavioristic. It has recent"7 been pointed out that an Internationa"
Congress on )eace 0as composed of statesmen8 po"itica" scientists8 historians8
economists8 ph7sicists8 bio"ogists--and not a sing"e. behaviorist in the strict sense.
2vident"7 behaviorism 0as regarded as use"ess. But ,e 2us! as$ 0hat the conference
achieved. It 0as composed of specia"ists from man7 different fie"ds8 0ho probab"7
spo$e the conmmonsense "ingua franca of the "a7man8 0ith its heav7 "oad of references
to inner causation. !hat might !*e conference have achieved if it cou"d have
abandoned this fa"se scent4 The currenc7 of menta"ism in discussions of human af fairs
ma7 e:p"ain 0h7 conferences on peace are he"d 0ith such monotonous regu"arit7 7ear
after 7ear.

To assert that a thoroughgoing behaviorism cou"d ma$e a great difference is
a"most inevitab"7 to be as$edA >!e""8 then8 0hat do 7ou suggest4 !hat 0ou"d 7ou do
about 0ar8 or popu"ation8 or po""ution8 or racia" discrimination8 or tho revo"t of the
7oung4> 5nfortu
ABOUT BEHAVIORISM 2#"

nate"78 to understand the princip"es invo"ved in so"ving a prob"em is not to have the
so"ution. To $no0 aerod7namics is not at once to $no0 ho0 to design a p"ane8 to $no0
p"ate tectonics is not at once to $no0 ho0 to predict earth6ua$es8 to understand the
doub"e he"i: is not at once to be ab"e to create a ne0 species. The detai"s of a prob"em
must be studied. /no0ing the basicA princip"es 0ithout $no0ing the detai"s of a practica"
prob"em is no c"oser to a so"ution than $no0ing the detai"s 0ithout $no0ing the basic
princip"es. But prob"ems can be so"ved8 even the big ones8 if those 0ho are fami"iar 0ith
the detai"s 0i"" a"so adopt a 0or$ab"e conception of human behavior.

!hen 0e sa7 that science and techno"og7 have created more prob"ems than
the7 have so"ved8 0e mean ph7sica" and bio"ogica" science and techno"og7. It does not
fo""o0 that a teehno"og7 of behavior 0i"" mean further troub"e. 'n the contrar78 it ma7 be
;ust 0hat is needed to sa"vage the other contributions. !e cannot sa7 that a science of
behavior has fai"ed8 for it has scarce"7 been tried. Amd it 0i"" not be given a fair tria" unti"
its phi"osoph7 has been c"ear"7 understood. A distinguishcd socia" phi"osopher has said8
>It is on"7 through a change of consciousness that the 0or"d 0i"" be saved. 2ver7one
must begin 0ith himse"f.> But no one can be+n 0ith himse"fA and if he cou"d8 it 0ou"d
certain"7 not be b7 changing his consciousness.

If it 0ere true that >an ever greater danger than nuc"ear 0ar arises from 0ithin
man himse"f in the form of smou"dering fears8 contagious panics8 primitive needs for
crue" vio"ence8 and raging suicida" destructiveness8> then 0e shou"d be "ost. <ortunate"78
the point of attac$ is more readi"7 accessib"e. It is the environment 0hich must be
changed. A 0a7 of "ife 0hich furthers the stud7 of human behavior in its re"ation to that
environment

SUMMIN6 UP 2##

shou"d be in the best possib"e position to so"ve its ma;or prob"ems. This is not behaviorism8
because the great prob"ems are no0 g"oba". In the behavioristic vie08 man can no0 contro" his
o0n destin7 because he $no0s 0hat must be done and ho0 to do it.