You are on page 1of 21

A Differential Association-Reinforcement Theory of Criminal Behavior

Author(s): Robert L. Burgess and Ronald L. Akers


Source: Social Problems, Vol. 14, No. 2 (Autumn, 1966), pp. 128-147
Published by: University of California Press on behalf of the Society for the Study of Social Problems
Stable URL: http://www.jstor.org/stable/798612 .
Accessed: 29/09/2013 15:16

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp

.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.

University of California Press and Society for the Study of Social Problems are collaborating with JSTOR to
digitize, preserve and extend access to Social Problems.

http://www.jstor.org

This content downloaded from 129.93.16.3 on Sun, 29 Sep 2013 15:16:04 PM


All use subject to JSTOR Terms and Conditions
128 SOCIAL PROBLEMS

surprising Theybecome dence on the factorsthat may in-


or unexpected.
moremeaningful wheninterpreted in fluencetheprocessof decisionmaking
thelightof studiesof therelationship bythehospitalpreadmission But
staff.
betweensocial class and psychiatricdiscussionswithvariousstaffmembers
treatmentand hospitalization. suggestthatan understandingof this
It was notpossiblewithinthescope processis crucialwhenaccountingfor
of thisstudyto gathersystematicevi- thedifferential
speedofhospitalization.

A DIFFERENTIAL ASSOCIATION-REINFORCEMENT
THEORY OF CRIMINAL BEHAVIOR

ROBERT L. BURGESS AND RONALD L. AKERS


University
of Washington

INTRODUCTION pirical test or thoroughrestatement


In spiteof thebodyof literature that beyondSutherland's own revisionin
has accumulated aroundthedifferential1947. Recognizingthatthe theoryis
associationtheoryof criminalbehav- essentially a learningtheory,Suther-
it has to receive crucial em- landrephrased it to stateexplicitly
that
ior,1 yet criminal behavioris learnedas anybe-
1 By 1960, Cressey had collected a 70- havioris learned.In Cressey's two re-
item bibliographyon the theory; see Ed- visionsof thetextbook, thetheory has
win H. Sutherlandand Donald R. Cressey, been deliberately leftunchangedfrom
of
Principles Criminology,6th ed., Chicago:
J. B. LippincottCo., 1960, p. vi. He has Sutherland's revision. Thus,thetheory
presentedan exhaustivereview of the mis- as it standsnow is postulatedupon
taken notions, criticisms, attempted re- theknowledge of thelearningprocess
formulations,and empirical tests of the extant20-25
theory containedin a sizable body of litera- years ago.2
ture. Donald R. Cressey, "Epidemiology Sutherland, himself, neverwas able
and Individual Conduct: A Case from to testdirectly or findspecificempiri-
Criminology,"PacificSociological Review, 3 cal supportforhis theory, buthe was
(Fall, 1960), pp. 47-58. For more recent convincedthat the
literature see Donald R. Cressey, "The two-edged theory
Theory of DifferentialAssociation: An In- -(1) genetic,differential association
troduction,"Social Problems, 8 (Summer, and (2) structural, differentialsocial
1960), pp. 2-5. James F. Short,Jr., "Dif-
ferentialAssociationas a Hypothesis:Prob- CriminalLaw, Criminologyand Police Sci-
lems of Empirical Testing," Social Prob- ence, 56 (September,1965), pp. 294-300.
lems, 8 (Summer,1960), pp. 14-25. Henry 2 The originalformalstatementappeared
D. McKay, "DifferentialAssociation and in Edwin H. Sutherland, Principles of
CrimePrevention:Problemsof Utilization," Criminology,3rd ed., Philadelphia: J. B.
Social Problems,8 (Summer,1960), pp. 25- LippincottCo., 1939, pp. 4-8. The terms,
37. Albert J. Reiss, Jr., and A. Lewis systematic"and "consistency"along with
Rhodes, "An EmpiricalTest of Differential some statementsreferringto social dis-
Association Theory," The Journal of Re- organizationand culture conflictwere de-
search in Crimeand Delinquency,1 (Janu- leted in the revised theory.Two sentences
ary,1964), pp. 5-18. Harwin L. Voss, "Dif- stating that criminal behavior is learned
ferentialAssociation and Reported Delin- were added and the terms "learned" and
quent Behavior: A Replication," Social "learning"were includedin othersentences.
Problems, 12 (Summer, 1964), pp. 78-85. The modalities of duration,priority,and
Siri Naess, "ComparingTheories of Crimi- intensitywere added. The revisedtheoryis
nogenesis," The Journal of Research in in Sutherland and Cressey, op. cit., pp.
Crimeand Delinquency,1 (July,1964), pp. 77-79. For Cressey's discussion of why he
171-180. C. R. Jeffery,"Criminal Behavior left the theoryin its 1947 form see ibid.,
and Learning Theory," The Journal of p. vi.

This content downloaded from 129.93.16.3 on Sun, 29 Sep 2013 15:16:04 PM


All use subject to JSTOR Terms and Conditions
CriminalBehavior 129

organization-accounted fortheknown Suggestedtheoreticalmodifications


data on the full rangeof crimes,in- have not been lacking,but the diffi-
cluding conventionalviolationsand cultywith these restatements is that
white-collar crimes.3The theoryhas theyare no morereadilyoperational-
receivedsomeotherempirical support,4 ized than Sutherland's.6 One recent
but negativecases have also been paper,however, byDeFleurand Quin-
found.5The attemptsto subjectthe ney,7 offersnew promise that the
theoryto empirical testare markedby theorycan be adequatelyoperational-
inconsistent findingsboth withinthe ized. They have presenteda detailed
same studyand betweenstudies,as strategy formakingspecific deductions
well as by highlycircumscribed and for empiricaltesting.But while they
qualified findingsand conclusions. haveclarified theproblemsin theder-
Whethertheparticular researcher con- ivation and generationof testable
cludesthathis findings do or do not hypothesesfrom differential associa-
seemto supportthe theory, nearlyall tion,theystillsee its empirical valida-
haveindicated difficultyin operational- tion as a verydifficult, thoughnot
izingtheconcepts and recommend that impossibletask.
the theorybe modifiedin sucha way Regardlessof the particularcriti-
thatit becomesmoreamenableto em- cisms,the exceptionstaken,and the
piricaltesting. difficulties
involvedin testingand re-
3 Ibid., pp. 77-80. Edwin H. Sutherland, formulating thetheory thathavebeen
White Collar Crime, New York: Holt, offered,few take exceptionto the
Rinehartand Winston, 1961, pp. 234-256 centrallearningassumptions in differ-
(originallypublished 1949). See also Cres- ential association.If we accept the
sey's Foreword,"ibid., p. x. basic assumption thatcriminalbehav-
4 John C. Ball, "Delinquent and Non-
Delinquent Attitudes Toward the Preva- ior is learnedby the same processes
lence of Stealing,"The Journalof Criminal and involvesthe same mechanisms as
Law, Crimnonology and Police Science, 48 conforming behavior, thenwe needto
(September-October, 1957), pp. 259-274. and makeuse of thecurrent
James F. Short, "DifferentialAssociation recognize
and Delinquency,"Social Problems,4, (Jan- knowledgeabout theseprocessesand
uary, 1957), pp. 233-239. Short, "Differ- mechanisms. Neitherthe extantstate-
ential Associationwith Delinquent Friends mentof thetheory northereformula-
and Delinquent Behavior," Pacific Sociolo- tionsof it make thenatureof
gical Review, 1 (Spring, 1958), pp. 20-25. explicit
Short, "DifferentialAssociation as a Hy- the underlying learning process in-
pothesis," op. cit. Voss, op. cit. Donald R. volved in differential association.In
Cressey,"Applicationand Verification of the short,no major revisionshave been
DifferentialAssociationTheory,"The Jour- made
nal of Criminal Law, Criminology and utilizing establishedlearning
Police Science, 43 (May-June,1952), pp. principles.
47-50. Cressey,OtherPeople's Money,Glen- That this typeof revisionof the
coe, Ill.: The Free Press,1953, pp. 147-149. theoryis neededhas been recognized
Glaser, op. cit., pp. 7-10.
5 Marshall Clinard, and somecriticism of differential
asso-
The Black Market,
New York: RinehartCo., 1952, pp. 285-
329. Marshall Clinard, "Rural Criminal 6 See Daniel Glaser, "CriminalityTheo-
Offenders,"AmericanJournalof Sociology, ries and Behavioral Images," American
50 (July, 1944), pp. 38-45. Edwin M. Journalof Sociology,61 (March, 1956), pp.
Lemert,"An Isolation and Closure Theory 433-444. Glaser, "DifferentialAssociation
of Naive Check Forgery,"The Journal of and Criminological Prediction," op. cit.,
Criminal Law, Criminology and Police pp. 10-13. Naess, op. cit., pp. 174-179.
Science,44, (September-October, 1953), pp. ? Melvin DeFleur and Richard Quinney,
293-307. Reiss and Rhodes, op. cit. Cressey, "A Reformulationof Sutherland'sDifferen-
"Application and Verificationof the Dif- tial Association Theory and a Strategyfor
ferentialAssociation Theory," op. cit., pp. EmpiricalVerification,"Journalof Research
51-52. Cressey,Other People's Money, op. in Crime and Delinquency, 3 (January,
cit., pp. 149-151. Glaser, op. cit.,pp. 12-13. 1966), p. 13.

This content downloaded from 129.93.16.3 on Sun, 29 Sep 2013 15:16:04 PM


All use subject to JSTOR Terms and Conditions
130 SOCIAL PROBLEMS

ciationhas revolvedaroundthe fact While sociologists knowa greatdeal


thatit does notadequatelyportraythe aboutthestructure of theenvironment
processby whichcriminalbehavioris fromwhichdeviantscome,we know
learned.But as Cresseyexplains: verylittleaboutthedetermining varia-
It is one thingto criticisethetheoryfor blesoperating withinthisenvironment.
failureto specify thelearning processac- The burdenof criminological theory
curatelyand anotherto specifywhich today is to combineknowledgeof
aspectsof thelearning processshouldbe structural pressureswithexplanations
includedand in whatway.8 of "whyonlysomeof thepersonson
Sutherland, of course,was as inter- whomthispressureis exertedbecome
estedin explaining the"epidemiology" non-conformists."1'
of crimeas in explaining how thein- It is forthisreasonthatthe recent
dividualcomesto engagein behavior effort by C. R. Jeffery to re-examine
in violationof thelaw and insisted that differential associationin light of
the two explanations mustbe consis- modernlearningtheorymarksa new
tent.9Differential social organization departure in theabundanceof thinking
(normative conflict)has beensuccess- and writingthathas characterized the
fulin "makingsense"of variations in intellectualhistoryof this theory,.2
crimerates.But differential association In spite of theirintricate axiomatiza-
has been less successful in explicating tion of the theory,DeFleur and
the processby whichthis differentialQuinney,for example,recognizethat
organization producesindividual crimi- even they have left the learning
nality. This seems to be due not to the processin differential associationun-
lack of importance of associationsfor specified. But, theynote,"modernre-
criminalbehaviorbut: inforcement learning theorywould
handle this problem . This is
... ratherto thefactthatthetheory out- '."1
ran the capacityof eitherpsychology or precisely what Jefferyproposedto do
socialpsychology to giveadequate,scien- and to theextentthatthisobjectiveis
tificanswers to thequestionof whythere servedby discussinglearningtheory
are such qualitative(selective) differ- and criminalbehaviortogether, he is
encesin humanassociation.io
at leastpartiallysuccessful. However,
It now appears,however, thatthere Jeffery does not in factmakeit clear
is a bodyof verified theorywhichis justhow Sutherland's differential
asso-
adequateto thetaskof accurately speci- ciationtheory maybe revised.His ex-
fyingthis process.Modern learning planationincorporates differentialre-
theoryseemscapableof providingin- inforcement:
sights into the problemof uniting -. [A] criminalactoccursin an en-
structuraland genetic formulations. vironment in whichin thepasttheactor
has beenreinforced
forbehavingin this
8 Cressey,"Epidemiologyand Individual manner, and the aversive consequences
Conduct,"op. cit., p. 54. attached to the behavior have been of
9 Sutherlandand Cressey,op. cit., p. 80. such a nature that they do not control
Albert K. Cohen, Alfred R. Lindesmith, or preventthe resoonse.14
and Karl F. Schuessler(eds.), The Suther-
land Papers, Bloomington: Indiana Uni- This statement, as it stands,bearsno
versityPublications,Social Science Series, obviousor directrelationto Suther-
No. 15, 1956, pp. 5-42. That Sutherland land's differential association,and no-
intended an explanation of the two-fold
problem of rates of crime and individual whereelse does Jeffery makeit clear
criminal behavior is, of course, the basic
point of Cressey's paper, "Epidemiology 11 Cressey,"The Theory of Differential
and Individual Conduct," op. cit. Association," op. cit., p. 5.
o10George B. Vold, Theoretical Crimi- 12 Jeffery, op. cit.
nology, New York: Oxford University 13 DeFleur and Quinney, op. cit., p. 3.
Press, 1958, p. 198. 14 Ibid., p. 295.

This content downloaded from 129.93.16.3 on Sun, 29 Sep 2013 15:16:04 PM


All use subject to JSTOR Terms and Conditions
CriminalBehavior 131

how differential reinforcement is a re- setof principles. Theyenablethehan-


formulation of differentialassociation. dlingof a greatvariety of observational
Jeffery does discussmodernlearning as well as experimental evidenceabout
principles, but he does not showhow humanbehavior.
theseprinciplesmay be incorporated It is thepurposeofthispapertotake
withinthe framework of Sutherland's thefirststepin thedirection to which
theory, nor how theseprinciplesmay Jeffery points.A restatement of the
lead to explanations of past empirical theory, not an alternative theory, will
findings. be presented,although,of necessity,
Jeffery's theoryand his discussion certain ideasnotintrinsic to differential
of criminal behaviorand learningthe- association will have to be introduced
oryremainsnot so muchincorrect as and additionswill be made to the
unconvincing.His presentationof original propositions.It should be
learningprinciples is supported wholly pointedoutthatDeFleurand Quinney
byreference to experiments withlower have been able to demonstrate that
organisms andhisextension to criminal Sutherland's propositions, whenstated
behavioris mainlythroughanecdotal in theformof settheory, appearto be
and illustrative material. The potential internally consistent.
By arranging the
value and impactof Jeffery's articleis propositions in axiomatic form,stating
diminished by not callingattention to them in logical ratherthan verbal
thealreadylargeand growingbodyof symbols, theyhave broughtthe theo-
literaturein experimental behavioral reticalgrammarup to date.'5 Such
science,especiallyevidenceusing hu- is not our intention in thispaper,at
man subjects,thathas directimplica- all. We recognizeand appreciatethe
tionsfordifferential associationtheory. importance of statingthepropositions
We are basicallyin agreement with in a formal,deductive fashion.We do
Jeffery that learningtheoryhas pro- feel,however,thatthistaskis, at the
gressedto the point whereit seems presenttime,subsidiary to the more
likelythatdifferential associationcan urgenttask of: (1) makingexplicit
be restated in a moresophisticated and the learningprocess,as it is now
testableformin thelanguageof mod- understood by modern behavioral
ern learningtheory.But thatrestate- science,fromwhichthe propositions
mentmustbe attempted in a thorough of differential association can be
fashionbeforewe can expectothersto derived;(2) fullyreformulating the
acceptit. Jeffery beginsto do thisand theory, statement bystatement, in light
his thoughtsare significant, but they of thecurrent knowledge of thislearn-
do nottakeintoaccountthetheoryas ing process;and (3) helpingcrimi-
a whole. nologists becomeawareof theadvances
The amountof empirical research in in learning theory andresearch thatare
the social psychologyof learning directly relevantto an explanation of
clearlyhas shownthattheconceptsin criminal behavior.16 No claimis made
learningtheory aresusceptible to oper- thatthisconstitutes a finalstatement.
ationalization.Therefore, applyingan If it has anyseminalvalueat all, that
integrated setof learningprinciples to is, if it provokesa seriousnew look at
differential associationtheoryshould 15 DeFleur and Quinney, op. cit.
adequately providetherevisionneeded 16 Our main concernhere, of course, is
for empiricaltesting.These learning with the nine statementsof the theoryas
principlesare based on literally thou- a genetic explanation of the process by
sands of experimental hourscovering which the individual comes to engage in
a wide rangeof thephylogenetic scale illegal behavior. We do not lose sight of
the fact, however, that this must be inte-
and morenearlyconstitute empiricallygrated with explanations of the variation
derivedlawsofbehavior thananyother and location of crime.

This content downloaded from 129.93.16.3 on Sun, 29 Sep 2013 15:16:04 PM


All use subject to JSTOR Terms and Conditions
132 SOCIAL PROBLEMS

thetheory andencourages furthereffort Accordingto thistheory, thereare


in this direction,our objectivewill two majorcategories of behavior.On
have beenserved. theone hand,thereis reflexive or re-
spondentbehaviorwhichis behavior
Differential Associationand Modern that is governedby the stimulithat
BehaviorTheory elicit it. Such behaviorsare largely
In thissectionthenineformalprop- associatedwiththe autonomicsystem.
ositionsin whichSutherland expressed The workofPavlovis of specialsignif-
his theorywill be analyzedin terms icancehere.On theotherhand,thereis
of behaviortheoryand researchand operantbehavior:behaviorwhichin-
will be reformulated as seven new volvesthecentralnervoussystem. Ex-
propositions. (See Table 1.) amples of operantbehaviorinclude
verbalbehavior,playingball, driving
I. "Criminal behavior is learned."VIII. a car, and buyinga new suit. It has
"The processof learningcriminalbe- been foundthatthisclassof behavior
haviorbyassociation withcriminal and is a functionof its past and present
anti-criminal patternsinvolvesall of environmentalconsequences.Thus,
the mechanisms thatare involvedin whena particular operantis followed
anyotherlearning." by certainkindsof stimuli,that be-
Sinceboththefirstand eighthsen- havior's frequency of occurrence
increasein the future.These stimuli
will
tencesin the theoryobviouslyforma
unitary idea,it seemsbestto statethem are called reinforcing stimulior re-
together.Sutherlandwas aware that inforcers20 and includefood, money,
thesestatements did notsufficientlyde- brain. However, most behavioral scientists
scribethelearningprocess,"7 butthese in this area are extremelycareful in hy-
twoitemsleaveno doubtthatdifferen-pothesizing interveningvariables or con-
tial associationtheory was meantto fit structs,whethertheyare egos, personalities,
responsesets,or some sort of internalcom-
into a generalexplanationof human puters. Generally they adopt the position
behaviorand, as muchis unambigu- thatthe only real value of a constructis its
ouslystatedin the prefatory remarks ability to improve one's predictions.If it
of thetheory: an "explanation of crim- does not, then it must be excluded in ac-
inal behaviorshouldbe a specific cordancewith the rule of parsimony.
part 20 It has been said by some that a tau-
of a generaltheoryof behavior."'s tologyis involvedhere.But thereis nothing
Modernbehaviortheoryas a general tautological about classifying events in
theory provides us with a good termsof their effects.As Skinner,op. cit.,
has noted, this criterionis both
idea of whatthemechanisms are that pp. 72-73,and
empirical objective.There is only one
are involvedin theprocessof acquir- sure way of tellingwhetheror not a given
ing behavior.'9 stimulus event is reinforcingto a given
individual under given conditionsand that
17 Cressey,1960, op. cit.,p. 54. is to make a direct test: observe the fre-
18 Sutherlandand Cressey,op. cit.,p. 75. quency of a selected behavior, then make
19 It should be mentionedat the outset a stimulus event contingentupon it and
thatthereis more than one learningtheory. observe any change in frequency.If there
The one we will employis called Behavior is a change in frequencythen we may
Theory. More specifically,it is that variety classifythe stimulus as reinforcingto the
of behavior theorylargely associated with individual under the statedconditions.Our
the name of B. F. Skinner. (Science and reasoningwould become circular,however,
Human Behavior, New York: Macmillan, if we went on to assertthata given stimulus
1953.) It differsfromother learningtheo- strengthensthe behavior because it is re-
ries in thatit restrictsitselfto the relations inforcing. Furthermore,not all stimuli,
between observable, measurable behavior when presented,will increasethe frequency
and observable, measurable conditions. of the behaviorwhichproducedthem.Some
There is nothingin this theorythat denies stimuli will increase the frequencyof the
the existence,or importance,or even the behavior which removes them, still others
inherentinterestof the nervous systemor will neither strengthennor weaken the

This content downloaded from 129.93.16.3 on Sun, 29 Sep 2013 15:16:04 PM


All use subject to JSTOR Terms and Conditions
CriminalBehavior 133

clothes,objectsof varioussorts,social ment. (3) A behaviormay produce


attention,approval,affection and social certainstimulus eventsand thereby de-
status.This entireprocessis called creasein frequency. Such stimuliare
positive reinforcement. One distin- called aversivestimulior, more re-
guishingcharacteristic of operantbe- cently,punishers.21The entirebehav-
havioras opposedto respondent behav- ioralprocessis calledpositivepunish-
ior,then,is thatthelatteris a function ment.(4) A behaviormayremoveor
of its antecedent stimuli,whereasthe terminate certainstimuluseventsand
formeris a function of its antecedent therebydecreasein frequency.Such
environmental consequences. stimuliarepositivereinforcers and the
Typically,operantand respondent processis termednegative punishment.
behaviorsoccurtogetherin an indi- (5) A behaviormayproduceor re-
vidual's everydaybehavior,and they movecertainstimulus eventswhichdo
interactin extremely intricateways. not changethe behavior'sfrequency
Consequently, to fullyunderstand any at all. Such stimuliare called neutral
set of patternedresponses, theinvesti- stimuli.(6) A behavior mayno longer
gatorshouldobservetheeffects of the producecustomary stimulus eventsand
operantson therespondents as well as therebydecreasein frequency.The
the effectsof the respondents on the stimuliwhichareproducedareneutral
operants.The connectionsbetween stimuli,and the process,extinction.
operantand respondent behaviorsare Whena reinforcing stimulus no longer
especiallycrucialto an analysisof atti- functions to increasethe futureprob-
tudes,emotional and conflict behaviors. abilityof thebehaviorwhichproduced
In everydaylife, different conse- it,we saytheindividualis satiated.To
quences are usuallycontingent upon restorethereinforcing property of the
differentclassesof behavior.This re- stimulus we need onlydeprivethein-
lationshipbetweenbehaviorand its dividualof it fora time.22
consequences functions to altertherate The increasein thefrequency of oc-
and formof behavioras well as itsre- currenceof a behaviorthat is rein-
lationshipto manyfeatures of the en- forcedis the verypropertyof rein-
vironment.The process of operant forcement thatpermitsthefascinating
reinforcement is the most importantvarietyand subtletythat occur in
processbywhichbehavioris generated operantas opposedto respondent be-
and maintained. Thereare,in fact,six havior.Anotherprocessproducing the
possible environmental consequences varietywe see in behavioris thatof
relativeto theLaw of OperantBehav- conditioning. When a primary or un-
ior. (1) A behaviormayproducecer- conditioned reinforcing stimulussuch
tain stimuluseventsand therebyin- as food is repeatedly paired with a
crease in frequency.As we have neutralstimulus, the latterwill even-
indicatedabove,suchstimuliarecalled tuallyfunction as a reinforcing stimu-
positivereinforcers and the processis lus as well. An illustration of this
called positivereinforcement. (2) A would be as follows. The milk a
behaviormay remove,avoid, or ter- motherfeedsto her infantis an un-
minate certainstimulusevents and conditioned reinforcer. If the food is
therebyincreasein frequency.Such
21 N. H. Azrin and D. F. Hake, "Con-
stimuliare termednegativereinforcersditioned
Punishment,"Journal of the Ex-
and the process,negativereinforce-perimentalAnalysis of Behavior, 8 (Sep-
tember,1965), pp. 279-293.
behavior them.See Rob-
whichproduced 22 See Jacob L. Gewirtz and Donald M.
RonaldL. Akers,"Are
ert L. Burgess, Baer, "Deprivation and Satiation of Social
?" The
OperantPrinciplesTautological Reinforcersas Drive Conditions," Journal
PsychologicalRecord, 16 (July, 1966), pp. of Abnormal and Social Psychology,57,
305-312. 1958, pp. 165-172.

This content downloaded from 129.93.16.3 on Sun, 29 Sep 2013 15:16:04 PM


All use subject to JSTOR Terms and Conditions
134 SocIAL PROBLEMS

repeatedly pairedwithsocialattention,through differentialreinforcement,the


affection, and approval,these latter child will eventually only speak the
will eventually becomereinforcing as word "daddy" when his fatheris
will the motherherselfas a stimulus presentor in other"appropriate" con-
object. Later these conditionedrein- ditions.We maysay thatthe father,
forcers can be usedto strengthen other as a stimulusobject,functionsas a
behaviorsby makingthesereinforcersdiscriminative stimulus(SD) setting
contingent upon thosenew behaviors. the occasionfor the operantverbal
Differential reinforcement mayalso response"daddy"becausein the past
altertheformof a response. This pro- suchbehaviorhas been reinforced un-
cess is calledshapingor responsedif- der suchconditions.
ferentiation. It can be exemplified by It has also been discovered thatthe
a childlearningto speak.At first, the patternor scheduleof reinforcement is
parentwill reinforce anyvocalization, as important as the amountof rein-
butas timewearson, and as thechild forcement. For example,a fixed-inter-
growsolder,the parentwill differen-val scheduleof reinforcement, where
tiallyreinforceonly those responses a responseis reinforced only aftera
whichsuccessfully approximate certain certainamountof time has passed,
criteria. The childwill be seento pro- producesa lowerrateof responsethan
ceed frommeregruntsto "baby-talk" thatobtainedwithreinforcement based
to articulate speech.23 on a fixed-ratio schedulewherea re-
Of course,organisms, whetherpi- sponseis reinforced onlyaftera certain
geons, monkeysor people, do not number of responses havealreadybeen
usually go around behavingin all emitted.Similarly a responserateob-
possiblewaysat all possibletimes.In tainedwith a fixed-ratio scheduleis
short,behaviordoes not occur in a lowerthanthatobtainedwitha vari-
vacuum; a given behavioris appro- able-ratioschedule,where reinforce-
priateto a givensituation. By appro- mentoccursfor a certainproportion
priatewe meanthatreinforcement has of responsesrandomlyvaried about
been forthcoming only undercertain somecentralvalue.A scheduleof rein-
conditionsand it is underthesecon- forcement, then,refersto theresponse
ditionsthatthe behaviorwill occur. contingencies upon which reinforce-
In otherwords,differential reinforce-ment depends. All of the various
mentnotonlyincreases theprobabilityschedulesof reinforcement, besides
of a response,it also makesthatre- producinglawful responsecharacter-
sponsemoreprobableupon the recur- istics,producelawfulextinction rates,
rence of conditionsthe same as or once reinforcement is discontinued.
similar to those that were present Briefly, behaviorreinforced on an in-
duringpreviousreinforcements. Such termittent schedule takes longer to
a processis calledstimuluscontrolor extinguish thanbehaviorreinforced on
stimulusdiscrimination. For example, a continuous schedule.
a childwhenhe is firsttaughtto say This concept, schedulesof reinforce-
"daddy"mayrepeatit whenanymale ment,is one theimplications of which
is present, or even,in theverybegin- are littleunderstood by manybehav-
ning,whenany adult is present.But ioral scientists, so a few additional
be the wordsare in order.Firstof all, social
23 This seems to processinvolved
in learning to become a marihuana user. reinforcements are for the mostpart
By successive approximations, the user intermittent. One obvious resultof
learns (from others) to close on the ap- thisfactis theresistance to extinction
propriate techniques and effectsof using and satiationof much social behav-
marihuana. See Howard S. Becker, Out-
siders, Glencoe, Ill.: The Free Press, 1963,
ior, desirableas well as undesirable.
pp. 41-58. This is not peculiarto humansocial

This content downloaded from 129.93.16.3 on Sun, 29 Sep 2013 15:16:04 PM


All use subject to JSTOR Terms and Conditions
CriminalBehavior 135

behavior,for even lower organisms have beenconditioned to it. These re-


seldom are faced with a continuous sponsesmay be competitive, thatis,
reinforcement schedule.Nevertheless,onlyone or theothercan occur.When
reinforcements mediatedby another thisis so,theparticular response which
organismare probablymuchless reli- doesoccurmayalso dependuponother
able than those produced by the discriminative stimulipresentin the
physicalenvironment. This is thecase situationthatcontrolonlyone or the
becausesocial reinforcement depends otherresponse.Finally,whilesomeof
upon behavioralprocessesin the rein- the stimulito whichan individualre-
forcerwhichare not undergood con- spondsemanatefromthe externalen-
trolby the reinforcee. A moresubtle, vironment, socialand otherwise, some
thoughessentially methodological, im- comefromhis ownbehavior.An indi-
plicationof thisis thatbecausemost vidual is, then,not onlya sourceof
social behaviorsare maintainedby responses, he is also a sourceof some
complexintermittent scheduleswhich stimuli-stimuli thatcaneffect hisown
havebeenshapedovera longperiodof behavior.
time,a socialobserver, newlyentering The mostgeneralbehavioralprin-
a situationmayhaveextreme difficultyciple is the Law of OperantBehavior
in immediatelydeterminingexactly whichsaysthatbehavioris a function
whatis maintaining a particular
behav- of its past and current environmental
ior or set of behaviors.Nor can the consequences. Therehavebeennumer-
individualhimselfbe expectedto be ous studieswithchildren26 as well as
able to identifyhis own contingenciesadults27 whichindicatethatindividual
of reinforcement.24 behaviorconformsto this law. Of
An important aspectof thistheory muchmoreinterestto sociologists is
is thepresentationof thegeneralways an experiment designedby Azrinand
that stimuli and responsescan be Lindsleyin 195628to investigate co-
formedintocomplexconstellations of operativesocial behavior.Theirstudy
stimulus-responseevents.Althoughthe demonstrated thatcooperative behavior
basicprinciplesaresimpleandmustbe couldbe developed,maintained, elimi-
separatedto distinguishand study natedand reinstated solelythrough the
them,in actuallifetheprinciples func- manipulation of the contingency be-
tionin concert,and consistof complex tweenreinforcing stimuliand the co-
arraysand constellations.25Such com- operativeresponse.This basic finding
plexitycan be seen in the fact that has receivedmuchsubsequent support.
single S-R eventsmay be combined It has also beendemonstrated thatnot
into sequenceson the basis of condi- only cooperativebehavior,but also
tioningprinciples.That is, responses competitive behaviorand leadingand
can be thought to havestimulus prop-
erties.In addition,morethanone re- 26 See, for example, S. W. Bijou and P.
sponsemaycomeunderthecontrolof T. Sturges, "Positive Reinforcersfor Ex-
a particularstimulus.Thus,whenthe perimental Studies with Children-Con-
stimulusoccurs,it will tendto set the sumables and Manipulatables," Child De-
30, 1959, pp. 151-170.
occasionforthevariousresponses that velopment,
27 J. G. Holland, "Human Vigilance,"
24 Cressey encounteredthis problem in Science, 128, 1959, pp. 61-67; Harold
tryingto get trustviolators to reconstruct Weiner, "ConditioningHistoryand Human
past associations. Cressey, Other People's Fixed-Interval Performance," Journal of
the Experimental Analysis of Behavior,
Money, op. cit., p. 149.
25 Arthur Staats, "An Integrated-Func- 7 (September,1964), pp. 383-385.
tional LearningApproach to Comolex Hu- 28 N. H. Azrin and O. R. Lindsley,"The
man Behavior," Technical Report 28, Con- Reinforcementof Cooperation Between
tract ONR and Arizona State University, Children," The Journal of Abnormal and
1965. Social Psychology,52 (January,1956).

This content downloaded from 129.93.16.3 on Sun, 29 Sep 2013 15:16:04 PM


All use subject to JSTOR Terms and Conditions
136 SOCIAL PROBLEMS

followingbehaviorare a functionof "normal" social behavior function


theirpastand presentconsequences. accordingto theLaw of OperantBe-
Anotherof thebehavioral principles havior.But whatabout"deviant"be-
we mentioned wasthatof stimulus dis- havior?Can we be sure thesesame
crimination. A discriminative stimulus principlesare operatinghere?Unfor-
is a stimulusin thepresenceof which tunatelytherehave been no studies
a particular operantresponseis rein- which attemptto test directlythe
forced.Muchof ourbehavior has come relevance of thesebehavioral principles
underthe controlof certainenviron- to criminalbehavior.But therehave
mental, including socialstimuli because been several experimental investiga-
in the past it has been reinforced in tionsof deviantbehaviorsemittedby
the presenceof thosestimuli.In an mentalpatients.For example,in a
experimentby Donald Cohen,29a studybyAyllonand Michael,30it was
normal13-year-old boy namedJustin, shown that the bizarrebehaviorsof
when placed under identicalexperi- psychoticsfunctionedaccordingto
mentalconditions emitted different be- theselearningprinciples.In thispar-
haviorsdependingupon whetherhis ticularstudyvariousbehavioralprob-
partner was his mother, brother, sister, lemsofpsychotic patientswere cured"
friend, ora stranger. The results of this through themanipulation of reinforce-
investigation demonstrated that Jus- mentcontingencies. Suchprinciples as
tin's social behaviorwas differentially extinction, negativeand positiverein-
controlled byreinforcement; butit also forcement, andsatiation wereeffectively
demonstrated thathisbehavior was dif- utilizedto eliminate theunwantedbe-
ferent depending upon the social haviors.31 This studywas one of the
stimulipresent,thus reaffirming the firstexperimental testsof the conten-
principleof stimulusdiscrimination.tionthatnotonlyconforming butalso
In otherwords,thedynamic properties manyunusual,inappropriate, or un-
of his social behavior,whetherco- desirablebehaviorsare shaped and
operative, competitive, leadingor fol- maintainedthroughsocial reinforce-
lowing,werecontrolled byhisprevious ment. In anotherexperiment Isaacs,
extra-experimental historywith his Thomas, and Goldiamond3"2demon-
teammates, althoughthe experimenterstratethatcomplexadjustivebehaviors
could changethosebehaviorsby ex- can be operantly conditioned in long-
perimentally alteringthecontingenciestermpsychotics bymanipulating avail-
of reinforcement. It is, of course,al- able reinforcers.
mosta truism to saythatan individual In yet anotherinvestigation,33" the
behavesdifferently in the presenceof 30 T. Ayllon and J. Michael, "The Psy-
different people. The significance of chiatricNurse as a Behavioral Engineer,"
this experiment, however,is thatthe Journal of the ExperimentalAnalysis of
investigator was able to isolate the Behavior, 2, 1959, pp. 323-334.
determining variablesand the prin- 31 There is, of course, no intentionon
our part to equate "mental" illness or simi-
ciplesby whichtheyoperatedto pro- larly severe behavior problemswith crimi-
duce thiscommonphenomenon. nal behavior.The only connectionthat we
While thisis by no meansa com- are making is that both may be seen to
plete survey of the relevant functionaccording to the same basic be-
experimental havioral principles and both may be in
testsof the behavioral opposition to establishednorms.
principles outlinedabove,it mayserve 32 W. Isaacs, J. Thomas, and I. Goldia-
to point out that many formsof mond, "Application of Operant Condition-
ing to Reinstate Verbal Behavior in Psy-
29 Donald J. Cohen, "Justin and His chotics,"Journalof Speech and Disorders,
Peers: an Experimental Analysis of a 25, 1960, pp. 8-12.
Child's Social World," Child Develop- 33 T. Ayllon and N. Azrin, "The Mea-
ment,33, 1962. surementand Reinforcement of Behavior of

This content downloaded from 129.93.16.3 on Sun, 29 Sep 2013 15:16:04 PM


All use subject to JSTOR Terms and Conditions
CriminalBehavior 137

personnel of a mental hospital ward dence we would amend Sutherland's


for schizophrenicsrecordedthe behav- firstand eighthpropositions to read:
ior of the patientsand providedconse- I. Criminal behavioris learnedaccord-
quences to it accordingto certainpre- ing to theprinciplesof operantcondi-
establishedprocedures.Without going tioning.
into the manyimportantdetailsof this
long investigation,we may note that II. "Criminalbehavioris learnedin
in each of the six experimentsthat interaction withotherpersonsin the
were carried out, the results demon- processof communication."
stratethat reinforcement was effective As DeFleur and Quinney have
in maintaining desired performances, noted,the major implication of this
even though these were "back-ward" proposition is thatsymbolic interaction
psychoticswho had resisted all pre- is a necessary conditionforthe learn-
vious therapy,includingpsychoanalysis, ing of criminalbehavior.36 Of direct
electroshocktherapy, lobotomies and relevanceto this is an experiment
so forth. designedtotesttherelativesignificance
In each experiment, the performanceof verbal instructions and reinforce-
fellto a nearzero levelwhentheestab- mentcontingencies in generating and
lished response-reinforcement relation maintaining
a certain class of behav-
was . The standard iors.37In brief,the resultsindicated
discontinued..
procedure forreinforcement had beento
providetokens... [exchanged]for a thatbehaviorcouldnot be maintained
varietyof reinforcers. Performance de- solely through verbal instructions.
creasedwhenthisresponse-reinforcement However,it was also discoveredthat
relationwas disrur~ted (1) by delivering it was an arduoustask to
tokensindependently of the response extremely
while still allowingexchangeof tokens shapea setof complexbehaviors with-
for the reinforcers (Exp II and III), out using verbal instructions as dis-
(2) by discontinuing the tokensystem criminativestimuli. Behavior was
entirely but providingcontinuing access
to the reinforcers (Exp IV), or (3) by quicklyand effectively developedand
discontinuing the deliveryof tokensfor maintained bya combination of verbal
a previouslyreinforced responsewhile instructionsand reinforcement conse-
simultaneously providingtokensfor a quences.Symbolic interaction is, then,
different, alternativeresponse(Exp I not of rein-
and VI). Further,the effectivenessof enough, contingencies
the reinforcement forcement
proceduredid not ap- mustalso be present.
From the perspectiveof modern
pear to be limitedto an all-or-nonebasis.
Patients selected and performedthe as-behaviortheory, two aspectsof social-
signmentthat provided the larger num- izationare usuallyconsideredto dis-
ber of tokens when reinforcement was
available for more than one assignmenttinguishit fromotherprocessesof
(Exp V).34 behavioralchange: (1) Only those
behavioralchangesoccurring through
Again, we cannot review all of the are consideredrelevant;(2)
relevant literature, yet perhaps the learning
the changesin behaviorhaving
three investigationscited will serve to only
theiroriginsin interaction
withother
emphasize that manyformsof deviant
behavior are shaped and maintained 1964. L. Krasner and L. UlIman, Research
by various contingenciesof reinforce- in Behavior Modification,New York: Holt,
ment.35Given this experimentalevi- Rinehart and Winston, 1965. L. Ullman
and L. Krasner, Case Studies in Behavior
Psychotics,"Journal of the Experimental Modification,New York: Holt, Rinehart
Analysisof Behavior,8 (November,1965), and Winston, 1964.
pp. 357-383. 36 DeFleur and Quinney, op. cit., p. 3.
34 Ibid., p. 381. 37 T. Ayllon and N. Azrin, "Reinforce-
35 See also J. J. Eysenck (ed.), Experi-
withMental Patients,"
mentand Instructions
ments in Behaviour Therapy, New York: Journal of the ExperimentalAnalysis of
PergamonPress, The Macmillan Company, Behavior,7, 1964, pp. 327-331.

This content downloaded from 129.93.16.3 on Sun, 29 Sep 2013 15:16:04 PM


All use subject to JSTOR Terms and Conditions
138 SOCIAL PROBLEMS

persons are consideredproductsof B, childA wouldlaterselectchildB


socialization.8 Sutherland's theory as a companion. The behavior of select-
may,then,be seen to be a theoryof ing childB was notthebehaviorthat
differential socializationsincehe, too, was reinforced. The experimental con-
restricted himselfto learninghaving ditionssimplypaired child B with
its origin in interaction with other positivereinforcement. In accordance
persons.While social learningis, in- with the principleof conditioning,
deed,important andevenpredominant,childB had becomea conditioned posi-
it certainly does notexhaustthelearn- tive reinforcer. As such any behavior
ing process.In short,we may learn whichproduced thepresence ofchildB
(and, thus, our behaviorwould be wouldbe strengthened, suchbehaviors,
modified)withoutany directcontact for example, as verbal responses
withanotherperson.As such,Suther- requestingchild B's company.Thus,
land's theorymay be seen to suffer as Staats42 hasnoted,theresults of this
froma significant lacuna in that it studyindicatethat the conceptsof
neglected thepossibility of deviantbe- reinforcingstimuliand groupcohesion
haviorbeinglearnedin nonsocialsitua- are relatedwhenanalyzedin termsof
tions.Consequently, to be an adequate an integratedsetof learning principles.
theoryof deviantbehavior,thetheory Glaser43has attempted to reformu-
mustbe amendedfurtherto include late Sutherland'sdifferential associa-
thoseformsof deviantbehaviorthat tiontheory in termsof socialidentifica-
are learnedin the absenceof social tion.It shouldbe recognized, however,
reinforcement. Other people are not thatidentification as well as modeling
the only sourceof reinforcement al- and imitativebehavior (which are
thoughtheyare the mostimportant.usuallyassociatedwithidentification)
As Jeffery39 has aptlynoted,stealing comprise justone feature of thesocial-
is reinforcing in and byitselfwhether izationprocess.Furthermore, suchbe-
otherpeople knowaboutit and rein- haviormay be analyzedquite parsi-
forceit sociallyor not.The samemay moniouslywith the principlesof
be said to apply to manyformsof modernbehaviortheory. For example,
aggressive behaviors.40 in a studyby Banduraand Ross,44a
There are manystudieswhichare child experienced the pairingof one
relevant to socialinteraction andsocial- adult with positivereinforcers. Pre-
izationon the one hand,and Suther- sumablythis adult would becomea
land'ssecondproposition on theother. conditionedreinforcer. And indeed,
For example,in a studyby Lott and later it was found that the child
Lott41it was foundthatwhenchildA imitatedthisadult morethanhe did
was reinforced in thepresence of child an adult who was not paired with
positivereinforcers.That is, the one
38 Paul E. Secord and Carl W. Backman, adult,as he becamea stronger rein-
Social Psychology,New York: McGraw- forcer, had also becomea stronger
Hill, 1964. SD
Jeffery,op. cit.
for imitatingor followingbehavior.
39
40 For some evidencethat aggressivebe-
havior may be of a respondentas well as 42 ArthurStaats,Human Learning,New
an operantnature,see N. Azrin,R. Hutchin- York: Holt, Rinehartand Winston, 1964,
son, and R. McLaughlin, "The Opportu- p. 333.
nity for Aggression as an Operant Rein- 43 Glaser, "Criminality Theories and
forcerduring Aversive Stimulation,"Jour- Behavioral Images," op. cit.
nal of the ExperimentalAnalysis of Be- 44 A. Bandura,D. Ross, and S. Ross, "A
havior,8 (May, 1965), pp. 171-180. ComparativeTest of the Status Envy, So-
41 B. E. Lott and A. J. Lott, "The For- cial Power and the Secondary Reinforce-
mationof PositiveAttitudesToward Group ment Theories of IdentificationLearning,"
Members," The Journal of Abnormal and Journalof Abnormaland Social Psychology,
Social Psychology,61, 1960, pp. 297-300. 67, 1963, pp. 527-534.

This content downloaded from 129.93.16.3 on Sun, 29 Sep 2013 15:16:04 PM


All use subject to JSTOR Terms and Conditions
CriminalBehavior 139

Thus, Bandura's and Ross's results criminativestimulusfunctionof a


demonstrate thatimitating or follow- group.For example,specificindivid-
ing behavioris at leastin parta func- uals as physicalstimulimay acquire
tionof thereinforcing valueof people discriminativecontroloveran individ-
as social stimuli. ual's behavior.The child in our ex-
On thebasisoftheseresultsit is sug- ample above is reinforced for certain
gestedthata changein the reinforcingkindsof behaviorsin the presenceof
value of an individual will change his his parent,thusthe parent'spresence
power as a stimuluscontrolling other may come to controlthis type of
people's behaviorin variousways. An behavior. He is reinforced fordifferent
increasein the reinforcing value of an
individualwill increaseverbal and motor behaviors in thepresenceof his peers,
approach, or companionable responses, who thencometo set theoccasionfor
respectfulresponses, affectionatebehav- this type of behavior.
ior, followingbehavior,smiling,pleasant
Consequently
conversation,sympatheticresponses and this proposition mustbe amendedto
the like.45 read: II. Criminalbehavioris learned
both in nonsocialsituationsthatare
The relevanceof these studiesis
or discriminative, and
that theyhave isolatedsome of the reinforcing
through that social interactionin
determining variableswhereby thebe- whichthe behavior other
haviorof one personis influenced or is of persons
reinforcing or discriminative for
changedby thebehavior of another as criminal behavior.
well as the principlesby whichthese
variablesoperate.We have,of course, III. "The principalpartof thelearn-
onlyscratched thesurface.Manyother ing of criminal behavioroccurswithin
variablesare involved.For instance, intimate personalgroups."
not all people are equallyeffective in In termsof ouranalysis, theprimary
controlling or influencing thebehavior wouldbe seen to be themajor
group
of others.The personwhocan mediate sourceof an individual'ssocial rein-
the mostreinforcers will exercisethe forcements. The bulk of behavioral
most power. Thus, the parent,who whichthechildreceives occurs
controls training
moreof hischild'sreinforcers,at a timewhen the
will exercisemore power than an the trainers, usually
older siblingor the temporary parents,possessa verypowerful
"baby system of reinforcers.In fact,we might
sitter."As the child becomesolder
characterizea primary groupas a gen-
and less dependentupon the parent eralizedreinforcer (one associated with
for many of his reinforcers, other manyreinforcers, conditioned as well
individualsor groups such as his as
unconditioned).And, as we sug-
peersmayexercisemorepower.Carry- gestedabove,as thechildgrowsolder,
ing the analysisone step further, the groups other than the familymay
personwhohas accessto a largerange cometo controla majority of an indi-
of aversivestimuliwill exertmore
vidual'sreinforcers,e.g., theadolescent
power than one who has not. Thus a
peergroup.
peergroupmaycometo exercisemore To say thatthe primarygroup is
power over a child's behaviorthan theprincipalmolderof an individual's
the parent even thoughthe parent behavioral is not to ignore
repertoire
maystillcontrola largeshareof the social learningwhich may occur in
child'spositivereinforcers. other contexts.As we noted above,
In additionto thereinforcing func- learningfromsocial models can
tionof an individualor group,there be
adequately explained in termsof these
is, as seen in the Cohen and the behavioralprinciples. The analysiswe
Bandura and Ross studies,the dis-
employedtherecan also be extended
45 Staats, 1964, op. cit., p. 333. to learningfromthe massmediaand

This content downloaded from 129.93.16.3 on Sun, 29 Sep 2013 15:16:04 PM


All use subject to JSTOR Terms and Conditions
140 SOCIAL PROBLEMS

from "reference"groups. In any case, motivation maybe seen to be a func-


we mayalterthisproposition to read: tionof theprocessesbywhichstimuli
III. The principalpartof thelearning acquireconditioned reinforcing value
of criminalbehavioroccursin those and becomediscriminative stimuli.Re-
groupswhichcomprise theindividual's inforcersand discriminative stimuli
majorsourceof reinforcements. here would become the dependent
the independentvariables
IV. "When criminal behavior is variables;
wouldbe the conditioning procedures
learned,thelearning includes(a) tech- previously mentioned and thelevelof
niquesof committing thecrime,which
deprivation.For example, when a
are sometimes verycomplicated, some- prisoneris deprived of contactwith
timesverysimple; (b) the specific members of theoppositesex,suchsex
directionof motives,drives,rational- reinforcers will become much more
izations,and attitudes." powerful. Thus, thosesexualreinforc-
A studyby Klaus and Glaser46as ers thatare available,such as homo-
well as manyotherstudies47 indicate sexualcontact, wouldcometo exerta
thatreinforcement contingencies areof greatdealofinfluence andwouldshape
primeimportance in learningvarious behaviorsthat would be unlikelyto
behavioraltechniques. And,of course, occurwithoutsuch deprivation. And,
many techniques,both simple and withoutgoing any furtherinto this
complicated, arespecific to a particular topic,somestimulimaybe morerein-
deviantact suchas jimmying, picking forcing,under similarconditionsof
locks of buildingsand cars,picking deprivation, forcertainindividualsor
pockets,short-and big-contechniques, groupsthanfor others.Furthermore,
counterfeiting andsafe-cracking. Other the satiationof one or moreof these
techniquesin criminalbehaviormay reinforcers wouldallowforan increase
be learnedin conforming or neutral in therelative strength of others.
contexts,e.g., drivinga car, signing Much, therefore,can be learned
checks,shootinga gun, etc. In any aboutthe distinctive characteristics of
event,we need not alterthe firstpart a groupbyknowingwhattheavailable
of thisproposition. and effective reinforcers are and the
The secondpartof thisproposition
does,however, deservesomeadditional analysis isandthat reinforcing stimuli, both
positive negative, elicit certain re-
comments.Sutherland'smajor focus spondents.Unconditionedreinforcerselicit
hereseemsto be motivation. Much of these responses without training, condi-
whatwe havealreadydiscussedin this tioned reinforcerselicit such responses
throughrespondentconditioning.Staats and
paper oftengoes under the general Staats (Complex Human Behavior, New
headingof motivation. The topic of York: Holt, Rinehartand Winston, 1964)
motivation is as important as it is com- have characterizedsuch respondentsas "at-
plex.This complexity is relatedto the titude" responses. Thus, a positive rein-
factthatthe same stimulusmayhave forcer elicits a positivehave attitude. Further-
two functions: it maybe botha rein- more, theserespondents stimuluschar-
acteristicswhich may become discriminative
forcing stimulus and a discriminative stimuli setting the occasion for a certain
stimuluscontrolling thebehavior which class of operantscalled "striving"responses
is followedby reinforcement.48 Thus, for positive reinforcersand escape and/or
avoidancebehaviorsfornegativereinforcers.
46 D. J. Klaus and R. Glaser, "Increas- These respondents and their attendant
ing Team ProficiencyThrough Training," stimuli may be generalized to other rein-
Pittsburg:American Instituteof Research, forcing stimuli. Thus, striving responses
1960. can be seen to generalize to new positive
47 See Robert L. Burgess, "Communica- reinforcerssince these also will elicit the
tion Networks and Behavioral Conse- respondentresponses and their characteris-
quences," forthcoming. tic stimuli which have become SD's for
48 A central principle underlyingthis such behavior.

This content downloaded from 129.93.16.3 on Sun, 29 Sep 2013 15:16:04 PM


All use subject to JSTOR Terms and Conditions
CriminalBehavior 141

behaviorsuponwhichtheyare contin- of a grouparebyobserving reactionto


gent.Basically,we are contending that behavior, i.e., thesanctions appliedto,
thenatureof thereinforcer system and or reinforcement and punishment con-
the reinforcement contingencies are sequencesof, such behavior.We may
crucialdeterminants of individualand also learnwhat a group'snormsare
group behavior.Consequently, a de- throughverbalor writtenstatements.
scription of an individual'sor group's The individualgroup memberalso
reinforcers, and an understanding of learnswhat is and is not acceptable
theprinciples bywhichreinforcers af- behavioron the basis of verbalstate-
fect behavior,would be expectedto ments made by others,as well as
yielda greatdeal of knowledgeabout throughthe sanctions(i.e., the rein-
individualandgroupdeviantbehavior. forcingor aversivestimuli)appliedto
Finally,the rationalizations which his behavior(and othernormviola-
Cresseyidentifies withregardto trust tors) byothers.
violatorsand the peculiarextensions Behaviortheoryspecifiesthe place
of "defensesto crimes"or "techniques of normative statements and sanctions
of neutralization" by which deviant in the dynamicsof acquiring"con-
behavioris justified, as identified by forming"or "normative"behavior.
Sykesand Matza,49maybe analyzedas Just as the behaviorand even the
operantbehaviorsof the escape or physicalcharacteristics of the individ-
avoidancetypewhichare maintained ual mayservediscriminative functions,
becausetheyhave the effect of avoid- verbal behavior,and this includes
ing or reducingthe punishment that normative statements, can be analyzed
comesfromsocialdisapproval by one- as SD's. A normative statement can be
self as well as by others.We may, analyzedas an SD indicating thatthe
therefore, rewritethis propositionto members of a groupoughttobehavein
read: IV. The learningof criminalbe- a certainwayin certaincircumstances.
havior,includingspecifictechniques, Such "normative" behaviorwould be
attitudes, and avoidanceprocedures, is developedand maintained bysocialre-
a function oftheeffective and available inforcement. As we observedin the
reinforcers, and the existingreinforce-Ayllon-Azrin study51of instructions
mentcontingencies. and reinforcement contingencies, such
verbal behaviorwould not maintain
V. "The specificdirection of motives anyparticular class of behaviorsif it
and drivesis learnedfromdefinitionswerenotat leastoccasionally backedby
of thelegal codesas favorableor un- reinforcement consequences. Extending
favorable." theiranalysis, an individualwouldnot
"conform"to a normif he did not
In thisproposition, Sutherland ap-
pearsto be referring, at leastin part, have a past historyof reinforcement
to the concept"norm"whichmaybe for such conforming behavior.This
definedas a statement madebya num- is important, forearlierwe statedthat
ber of the membersof a group,not we canlearna greatdealabouta group
all of them,prescribing or byknowing whattheeffective reinforc-
necessarily ers are and the behaviors which
proscribing certainbehaviors at certain upon
We often inferwhat thenorms they are contingent. We maynow say
times.5? thatwe can learna greatdeal about
49 Cressey, Other People's Money, op. an individual'sor a group'sbehavior
cir., pp. 93-138. G. M. Sykes and David whenwe are able to specify, only
not
Matza, "Techniques of Neutralization: A whatthe effective reinforcers are,but
Theory of Delinquency," American Socio-
logical Review, 22 (December, 1957), pp. Its ElementaryForms,New York: Harcourt,
664-670. Brace and World, 1961.
50 George C. Homans, Social Behavior: 51 Ayllon-Azrin,1964, op. cit.

This content downloaded from 129.93.16.3 on Sun, 29 Sep 2013 15:16:04 PM


All use subject to JSTOR Terms and Conditions
142 SOCIAL PROBLEMS

also what the rulesor normsare by The specificclass of behaviorswhich


whichthesereinforcers are applied.52 are learnedand theirfrequencyof
For thesetwotypesof knowledge will occurrence are a functionof the rein-
tellus muchaboutthetypesof behav- forcerswhichare effective and avail-
ior thattheindividualwill developor able, and therulesor normsbywhich
the typesof behaviorsthatare domi- thesereinforcers are applied.
nantin a group.
Forexample,ithas oftenbeennoted VI. "A person becomes delinquent
thatmostofficial criminal actsarecom- because of an excess of definitions
mittedbymembers of minority favorableto violationof law over
groups definitions unfavorable to violationof
who live in slums.One distinguishing
characteristicof a slumis thehighlevel law."
of deprivation ofmanyimportant social This propositionis generallycon-
reinforcers.Exacerbating thissituation sideredthe heartof Sutherland's the-
is thefactthatthesepeople,in contrast ory; it is the principleof differential
to othergroups,lack the behavioral association.It followsdirectlyfrom
repertoires necessary to producerein- proposition V, and we mustnow refer
forcement in theprescribed ways.They back to thatproposition. In proposi-
have not been and are not now ade- tion V, the use of the preposition
quatelyreinforced for lawfulor nor- "from"in the phrase,"learnedfrom
mativebehavior. Andas we knowfrom definitions of thelegal codesas favor-
theLaw of OperantReinforcement, a able or unfavorable," is somewhat mis-
reinforcer will increasethe rate of leading.The meaninghere is not so
occurrence of anyoperantwhichpro- muchthatlearningresultsfromthese
ducesit. Furthermore, we would pre- definitions as it is thattheyformpart
dictthatgivena largenumberof indi- of thecontent of one'slearning, deter-
vidualsundersimilarconditions, they miningwhichdirection one'sbehavior
are likelyto behavein similarways. willgo in relation to thelaw,i.e.,law-
Withinsuch groups,manyformsof abidingor lawbreaking.
social reinforcement maybecomecon- These definitions of the law make
tingent uponclassesofbehaviors which lawbreaking seemeitherappropriate or
are outsidethelargersociety'snorma- inappropriate. Those definitions which
tive requirements. Norms and legal place lawbreaking in a favorablelight
codes, as discriminative stimuli,will in a sense can be seen as essentially
onlycontrolthebehaviorof thosewho normsof evasionand/ornormsdi-
haveexperienced theappropriate learn- rectlyconflicting with conventional
ing history. If an individualhas been, norms.Theyare, as Sykesand Matza
andis,reinforced forsuch"normative" and Cresseynote,"techniques of neu-
behavior,thatbehaviorwill be main- tralization," "rationalizations," or "ver-
tainedin strength. If he has notbeen, balizations"whichmake criminalbe-
and is not now reinforced for such haviorseem"all right"or justified, or
behaviorstheywouldbe weak,if they which providedefensesagainstself-
existedin his repertoire at all. And, reproach and disapproval from
importantly, the reinforcement system others.53The principleof negative
mayshapeand maintainanotherclass
of behaviorswhichdo resultin rein- 53 Sykes and Matza, op. cit., Cressey,
OtherPeople's Money, op. cit., pp. 93-138;
forcement and suchbehaviorsmaybe Donald R. Cressey,"The DifferentialAsso-
considereddeviantor criminal byother ciation Theory and Compulsive Crimes,"
members of the group.Thus we may Journalof CriminalLaw, Criminologyand
formulate thisproposition to read: V. Police Science, 45 (May-June,1954), pp.
29-40; Donald R. Cressey,"Social Psycho-
logical Foundationsfor Using Criminalsin
52 Staats and Staats,op. cit. the Rehabilitationof Criminals," Journal

This content downloaded from 129.93.16.3 on Sun, 29 Sep 2013 15:16:04 PM


All use subject to JSTOR Terms and Conditions
CriminalBehavior 143

reinforcement would be of majorsig- of a groupbymakingvariousformsof


nificancein theacquisition and mainte- social reinforcement, such as social
nanceof such behaviors. approval,esteem, and statuscontingent
This analysissuggeststhat it may upon thatbehavior.
not be an "excess" of one kind of The concept"excess" in the state-
definition overanotherin thesenseof ment,"excessof definitions favorable
a cumulativeratio,but ratherin the to violationof law," has beenparticu-
sense of the relativeamountof dis- larlyresistant to operationalization. A
criminative stimulusvalue of one set translation of thisconceptin termsof
of verbalizations or normativestate- modernbehaviortheory wouldinvolve
mentsover another.As we suggested the "balance"of reinforcement conse-
in the last section,normativestate- quences,positiveand negative.The
mentsare, themselves, behaviorsthat Law of Differential Reinforcement is
are a function of reinforcement conse- crucialhere.That is, a personwould
quences.They,in turn,mayserveas engagein thosebehaviorsfor which
discriminative stimuliforotheroperant he had beenreinforced mosthighlyin
behaviors(verbaland nonverbal).But the past. (The readermayrecallthat
recallthatreinforcement mustbe forth- in theAyllon-Azrin studywithschizo-
coming,at leastoccasionally, beforea phrenics, it was foundthatthepatients
verbal statementcan continueas a selectedandperformed thosebehaviors
discriminative stimulus. Bear in mind, whichprovidedthe most reinforcers
also, thatbehaviormayproducerein- whenreinforcement was availablefor
forcingconsequences even in the ab- morethanone response.)Criminal be-
sence of any accompanying verbal haviorwould,then,occurunderthose
statements. conditionswhere an individualhas
In otherterms, a personwillbecome been mosthighlyreinforced for such
delinquentif the officialnorms or behavior, andtheaversive consequences
laws do not performa discriminativecontingentupon the behaviorhave
functionand thereby control"norma- beenof sucha naturethattheydo not
tive" or conformingbehavior.We performa "punishment function."54
know fromthe Law of DifferentialThis leadsus to a discussion of propo-
Reinforcement thatthatoperantwhich sitionVII. But,first, letus reformulate
producesthe mostreinforcement will the sixth propositionto read: VI.
becomedominantif it resultsin rein- Criminalbehavioris a functionof
forcement. Thus, if lawful behavior normswhich are discriminative for
did not resultin reinforcement, the criminalbehavior,the learning of
strengthof the behaviorwould be whichtakesplace whensuchbehavior
weakened,and a stateof deprivation
54 This, then, is essentiallydifferential
would result,whichwould, in turn, reinforcement as Jefferypresents it. We
increasethe probability thatotherbe- have attemptedto show how this is con-
haviorswould be emittedwhich are gruent with differentialassociation. Fur-
reinforced, and such behaviorswould of ther,while Jefferyignoresthe key concepts
'definitions"and "excess" we have in-
be strengthened. And,of course,these corporated them into the reformulation.
behaviors, thoughcommonto one or These definitions,viewed as verbalizations,
moregroups,maybe labelleddeviant become discriminativestimuli; and "ex-
bythelargersociety. Andsuchbehavior cess" operatesto produce criminalbehavior
two relatedways: (1) verbalizationscon-
patterns, themselves, mayacquirecon- inducive to law violation have greater dis-
ditionedreinforcing value and, subse- criminativestimulusvalue than other ver-
quently,be enforcedby the members balizations,and (2) criminalbehavior has
been more highly reinforcedand has pro-
of Research in Crime and Delinquency, 2 duced fewer aversive outcomes than has
(July,1965), pp. 45-59. See revisedpropo- law abiding behavior in the conditioning
sitionIV. historyof the individual.

This content downloaded from 129.93.16.3 on Sun, 29 Sep 2013 15:16:04 PM


All use subject to JSTOR Terms and Conditions
144 SOCIAL PROBLEMS

is more highlyreinforced than non- Priority,frequency, duration,and


criminalbehavior. intensity of associationwith criminal
and groupsare important to
VII. "Differentialassociationsmay persons
theextentthattheyinsurethatdeviant
varyin frequency, duration,priority,behaviorwill receivegreateramounts
and intensity." of reinforcement at morefrequentin-
In terms of ouranalysis,theconcepts tervalsor with a higherprobability
frequency, duration,and priorityare than conforming behavior.But the
straightforward enough.The concept frequency, probability, and amountof
intensity could be operationalized to reinforcement are thecrucialelements.
designatethe numberof the individ- This meansthatit is thecomingunder
ual's positiveand negativereinforcersthe controlof contingencies of rein-
anotherindividualor groupcontrols, forcement thatselectively producesthe
as well as the reinforcement value of criminal definitionsand behavior.
thatindividual or group.As previously Consequently, letus rewrite thispropo-
suggested thegroupwhichcanmediate sitionto read: VII. The strength of
themostpositivereinforcers andwhich criminalbehavioris a directfunction
has the mostreinforcement value, as of theamount,frequency, and proba-
well as access to a largerrange of bilityof its reinforcement.
aversivestimuli,will exertthe most
controlover an individual'sbehavior. IX. "While criminalbehavioris an
There is a good reasonto suspect, expression of generalneedsandvalues,
however,thatSutherland was not so it is not explainedby thosegeneral
muchreferring associa- needs and values since noncriminal
to differential
tionswithotherpersons, as differentialbehavioris an expression of thesame
associations withcriminalpatterns. If needsand values."
this suppositionis correct,then this In thisproposition, Sutherland may
proposition can be clarified
byrelating have beenreacting, at leastin part,to
it to differentialcontingenciesof rein- the controversy regarding the concept
forcement rather thandifferential
social "need."This controversy is nowessen-
associations. Fromthisperspective, the tiallyresolved.For, we have finally
experimental evidencewithregardto come to the realizationthat "needs"
thevariousschedulesof reinforcement areunobservable, hypothetical,fictional
is of majorimportance. Therearethree inner-causal agentswhichwereusually
aspectsof the schedulesof reinforce-inventedon thespotto providespuri-
mentwhichare of particularimpor- ous explanationsof some observable
tancehere: (1) the amountof rein- behavior. Futhermore,they were
forcement: the greaterthe amountof inferredfrompreciselythe same be-
reinforcement, thehighertheresponse haviortheyweresupposedto explain.
rate; (2) the frequency of reinforce- While we can ignorethe reference
mentwhichrefersto the numberof to needs, we must discuss values.
reinforcements per giventimeperiod: Values may be seen as reinforcers
the shorterthe timeperiod between whichhavesaliencefora number ofthe
reinforcements, thehighertheresponse membersof a group or society.We
rate;and (3) theprobability of rein- agreewithSutherland to theextent that
forcement whichis the reciprocalof he meansthatthenatureof thesegen-
responses perreinforcement: thelower
the ratioof responsesper reinforce-sis of Behavior (October, 1962), pp. 543-
ment,thehighertherateof response."55597. Because the emission of a fixed ratio
or variable ratio of responses requires a
55 R. T. Kelleher and L. R. Gollub, "A period of time, the rate of respondingwill
Review of Positive Conditioned Reinforce- indirectlydeterminethe frequencyof re-
ment,"Journalof the ExperimentalAnaly- inforcement.

This content downloaded from 129.93.16.3 on Sun, 29 Sep 2013 15:16:04 PM


All use subject to JSTOR Terms and Conditions
CriminalBehavior 145

eralreinforcers do notnecessarily deter- process, we haveseenthattheprinciple


mine which behavior they will of differential reinforcement is of cru-
strengthen. Money,orsomething elseof cial importance.But we must also
generalvalue in society, will reinforce attendto a thirdquestion,namely,
anybehaviorthatproducesit. This re- what sustainsthe patternor contin-
inforcement may depend upon non- gencyof reinforcement? We onlyhave
criminalbehavior,but it also maybe- hintedat someof thepossiblyimpor-
comecontingent upona setofbehaviors tant variables.We have mentioned
thatare labelledas criminal.Thus,if briefly, for example,structural factors
Sutherland can be interpretedas mean- such as the level of deprivation of a
ing thatcriminaland noncriminal be- particulargroup with regardto im-
havior cannotbe maintainedby the portantsocialreinforcers, and thelack
same set of reinforcers, we mustdis- of effective reinforcement of "lawful"
agree.However,it maybe thatthere behavior56 and theconcomitant failure
are certainreinforcing consequences to develop the appropriate behavioral
whichonlycriminal behaviorwill pro- repertoires to producereinforcement
duce, for the behaviorfinallyshaped legally.57 We have also suggested that
will dependupon the reinforcer that those behaviorswhich do resultin
is effective forthe individual.Never- reinforcement may, themselves, gain
theless,it is thereinforcement,notthe reinforcement value and be enforced
specific natureof thereinforcer, which by themembers of thegroupthrough
explainstherateand formof behavior. the manipulation of variousformsof
But since this issue revolvesaround socialreinforcement suchas socialap-
contingencies of reinforcement which provalandstatus, contingentuponsuch
are handledelsewhere, we will elimi- behaviors.58 In short,new normsmay
natethislastproposition. developandthesemaybe termed delin-
CONCLUDING REMARKS quent by the larger society.
Therearemanyothertopicsthatare
The purposeof thispaperhas been of directrelevanceto the problemof
the applicationof the principlesof deviantbehaviorwhichwe have not
modernbehavior theoryto Sutherland's beenable to discussgiventherequire-
differential associationtheory.While mentsof space.For instance, no men-
Sutherland'stheoryhas had an en- tionhas beenmadeof someoutstand-
during effectupon the thinkingof ingresearch in theareaofpunishment.
students ofcriminal behavior,ithas,till This topic is, of course,of prime
now, undergoneno major theoreticalimportance in theareaof crimepreven-
revisiondespitethefactthattherehas tion.To illustrate someof thisresearch
been a steadyand cumulative growth and its relevance,it has been found
in the experimental findingsof the experimentally thattheamountof be-
processes of learning. havior suppressionproducedby re-
There are threeaspectsof deviant
56 RobertK. Merton,Social Theory and
behaviorwhichwe have attempted to
Social Structure,Glencoe, Ill.: The Free
deal with simultaneously, but which Press, pp. 161-195. For a more complete
shouldbe separated.First,how does discussionof social structurein termsrele-
an individualbecomedelinquent,or vant to this paper, see Robert L. Burgess
how does he learndelinquentbehav- Parts and Don Bushell,Jr.,Behavioral Sociology,
ior? Second,whatsustainsthisdelin- IV and V, forthcoming, 1967.
57 Ibid., and Richard A. Cloward, "Il-
quentbehavior ? We haveattempted to legitimate Means, Anomie, and Deviant
describethe waysin whichthe prin- Behavior," American Sociological Review,
ciples of modernbehaviortheoryare 2458(April, 1959), pp. 164-177.
Albert K. Cohen, Delinquent Boys;
relevant to thedevelopment and main- The
tenanceof criminalbehavior.In the The Free Culture of the Gang, Glencoe, Ill.:
Press, 1955

This content downloaded from 129.93.16.3 on Sun, 29 Sep 2013 15:16:04 PM


All use subject to JSTOR Terms and Conditions
146 SOCIAL PROBLEMS

TABLE 1
A DIFFERENTIAL ASSOCIATION-REINFORCEMENT
THEORY OF CRIMINAL BEHAVIOR

Sutherland's
Statements Reformulated Statements
1. "Criminalbehavioris learned." 1. Criminalbehavioris learnedaccording
8. "The processof learningcriminalbe- to the principlesof operantcondition-
haviorby association withcriminal and ing.
anti-criminal patternsinvolvesall of
themechanisms thatareinvolvedin any
otherlearning."
2. "Criminalbehavioris learnedin inter- 2. Criminalbehavioris learnedboth in
actionwithotherpersonsin a process nonsocialsituations thatare reinforcing
of communication." or discriminative and throughthatso-
cial interactionin whichthe behavior
of otherpersonsis reinforcing or dis-
criminativeforcriminalbehavior.
3. "The principalpartof the learningof 3. The principalpart of the learningof
criminalbehavioroccurswithininti- criminalbehavior occursin thosegroups
matepersonalgroups." whichcomprisethe individual'smajor
sourceof reinforcements.
4. "When criminalbehavioris learned, 4. The learningof criminalbehavior,in-
thelearningincludes(a) techniques of cluding specifictechniques,attitudes,
committing the crime,whichare some- and avoidanceprocedures, is a function
timesverycomplicated, sometimes very of theeffectiveandavailablereinforcers,
simple; (b) the specificdirectionof and the existingreinforcement contin-
motives,drives, rationalizations, and gencies.
attitudes."
5. "The specificdirection of motivesand 5. The specificclass of behaviorswhich
drivesis learnedfromdefinitions of the are learnedand theirfrequency of oc-
legalcodesas favorable or unfavorable." currence area function ofthereinforcers
whichare effective and available,and
the rulesor normsby whichthesere-
inforcersare applied.
6. "A personbecomesdelinquent because 6. Criminal behavior is a functionofnorms
of an excessof definitions favorable to which are discriminative for criminal
violationof law over definitions un- behavior,the learningof whichtakes
favorable to violationof law." place when such behavioris more
highlyreinforced thannoncriminal be-
havior.
7. "Differential associationsmay varyin 7. The strength of criminalbehavioris a
frequency, duration,priority,and inten- directfunction oftheamount, frequency,
sity." and probability of its reinforcement.
9. "While criminal behavioris an expres- 9. (Omit fromtheory.)
sion of generalneedsand values,it is
not explainedby thosegeneralneeds
and values since noncriminal behavior
is an expression of thesameneedsand
values."

aversivestimuliis a more,it has beendiscovered


sponse-contingent thatif an
directfunction of the aversivestimulusis repeatedly
of the intensity paired
butthata mildaver- withpositivereinforcement,
aversivestimulus, and rein-
sive stimulusmayproducea dramatic forcement is not availableotherwise,
if it is paired the aversivestimulusmay becomea
behavior-suppression
withreinforcement for an alternative discriminative
stimulus(SD) forrein-
and incompatiblebehavior.Further- forcement and, consequently,not de-

This content downloaded from 129.93.16.3 on Sun, 29 Sep 2013 15:16:04 PM


All use subject to JSTOR Terms and Conditions
CriminalBehavior 147

crease the behavior'sfrequencyof withdifferential can be at-


association
occurrence. tributedto its scoringlow on these
Thereare, in conclusion, numerous criteria,especially(1) and (3). We
criteria
thathavebeenusedto evaluate submitthat the reformulated theory
theories.One suchsetis as follows: presentedhereanswerssomeof these
(1) The amount of empirical support problemsand bettermeets each of
for the theory'sbasic propositions. these criteria.It is our contention,
(2) The "power" of the theory,i.e., the moreover, thatthereformulatedtheory
amount of data that can be derived not only specifies the conditions
fromthe theory'shigher-order prop- under which criminal behavior is
ositions. learned,but also some of the condi-
(3) The controllingpossibilities of the tionsunderwhichdeviantbehavior in
theory,including (a) whether the generalis acquired.Finally,whilewe
theory's propositions are, in fact,
have not statedour propositionsin
causal principles,and (b) whether
the theory'spropositionsare stated strictlyaxiomaticform,a close exami-
in such a way thattheysuggestpos- nationwillrevealthateachof thelater
sible practical applications. propositionsfollowfrom,modify,or
What dissatisfaction therehas been clarifyearlierpropositions.

SANCTIONS*

JACK P. GIBBS
StateUniversity
Washington

Few conceptsin sociologyhave Despite its importance, the concept


wider applicationthan sanction,if sanctionremainsambiguousand its
onlybecauseit entersintothe notion definition presentsa formidable prob-
of norms and related distinctionslem. Thereare few formaltreatments
(e.g., laws versuscustoms).Giventhe of theconcept,and manywriters treat
conceptuallinkbetweensanctionsand sanctionas a primitive term.Leaving
norms,it followsthat the definitionthe conceptundefined impliesthatits
of crimeor deviantbehaviorand the meaningis generally understood, and
delimitation of relatedfieldsrequire the assumptionis by no means un-
reference to sanctions.The importancejustified. Evenwithout a formaldefini-
of sanctionsis no less conspicuous in tion,mostobservers would agreethat
substantive theory, on so- the executionof a felonor settinga
particularly
cialorder.As a casein point,verylittle finefor a traffic offenserepresents a
remainsif thiselementis deletedfrom sanction.Nonetheless, the practiceof
Hobbesiantheory;and,withreference leavingsanctions undefined has certain
to contemporary issues,the much as reciprocityand consensusas bases
theoretical
functionalist school is distinguishedof social order.This is particularlytrue for
fromthe conflict school in termsof, Malinowski. For criticismof Malinowski
interalia, emphasison sanctions.1 on this point, see E. Adamson Hoebel,
The Law Primitive of MAlan,Cambridge:
* Some parts of this paper were written Harvard UniversityPress, 1954, pp. 190-
in connectionwith a Russell Sage fellow- 210; and William Seagle, "PrimitiveLaw
ship for the studyof law. and ProfessorMalinowski," American An-
1 On the whole, functionalistsdo not thropologist,39 (April-June, 1937), pp.
emphasize formal punitive sanctions as 275-290.

This content downloaded from 129.93.16.3 on Sun, 29 Sep 2013 15:16:04 PM


All use subject to JSTOR Terms and Conditions

You might also like