JOURNAL OF APPLIED BEHAVIOR ANALYSIS

BEHAVIORISM: PART OF THE PROBLEMOR PART OF THE SOLUTION?'

JAMES G. HOLLAND

UNIVERSITY OF PITTSBURGH

The form frequently taken by behavior-modification programs is analyzed in terms ofthe parent science, Behaviorism. Whereas Behaviorism assumes that behavior is the re-sult of contingencies, and that lasting behavior change involves changing the contin-gencies that give rise to and support the behavior, most behavior-modification programsmerely arrange special contingencies in a special environment to eliminate the "problem"behavior. Even when the problem behavior is as widespread as alcoholism and crime,behavior modifiers focus on "fixing" the alcoholic and the criminal, not on changingthe societal contingencies that prevail outside the therapeutic environment and continueto produce alcoholics and criminals. The contingencies that shape this method of dealingwith behavioral problems are also analyzed, and this analysis leads to a criticism of thecurrent social structure as a behavior control system. Although applied behaviorists havefrequently focused on fixing individuals, the science of Behaviorism provides the meansto analyze the structures, the system, and the forms of societal control that produce the"problems".DESCRIPTORS: behavior principles, behavior analysis, social control, political psy-

chology, community psychology, ethics

EDITOR's NOTE

The manuscript by Holland is clearly acontroversial one. Nonetheless, the manu-script raises interesting issues that deserve at-tention. In accord with JABA policy for dis-cussion articles, the manuscript is largelyunedited, as are the reviewers' comments.Each manuscript or critique represents theauthor's opinion and should not be inter-preted as representing JABA policy or theopinion of the editors or editorial board.

-Editor

During the struggles of the 1960s, EldridgeCleaver proclaimed: "If you are not a part of thesolution, you are part of the problem." And now

1This paper was presented as an invited address tothe 1975 convention of the American PsychologicalAssociation under the joint sponsorship of the Divi-sion for the Experimental Analysis of Behavior andthe APA Commission on Behavior Modification. Thepaper served also as the basis for a presentation to theVI International Seminar on Behavior Control inPanama, January, 1976. Reprints may be obtainedfrom James G. Holland, Psychology Department,University of Pittsburgh, Pittsburgh, Pennsylvania15260.

the question is past due for us: "Is Behaviorismpart of the solution or part of the problem?"There are many people ready with answers tothis question; many who object to what they seeas an oppressive political system and who seestratified control by corporate leaders. For thesepeople, Behaviorism is often seen as the problemand the behaviorist as another instrument intheir manipulation and in their exploitation. Butthe charge arises from too limited a view of be-haviorism. I will suggest to you that the analysisof behavior provides the means to analyze thestructures, the system, the forms of societal con-trol against which these very critics are rebelling.I will suggest that the view of humanity held bythe social reformer is supported, not refuted, bythe analysis of behavior. It is true that appliedbehavior analysts have, in overwhelming num-bers, been hired to do jobs in the service of thosein power and, even though the science, Behavior-ism, stands ready to be part of the solution, theapplied behaviorist has too often been part of theproblem. But tomorrow, if we are true to our sci-

163

1978, 11, 163-174 NUMBER 1 (SPRING 1978)

JAMES G. HOLLAND

ence, we can be part of the solution. For the sci-ence of the analysis of behavior is based on aview of people compatible with social reform.

BEHAVIOR-MODIFICATION SYSTEMSAND ESTABLISHEDPOWER STRUCTURES

The cases against the present systems of be-havior modification are certainly by now wellknown. In an earlier paper (Holland, 1975), Isuggested that the fears about a behaviorallycontrolled society could be evaluated by examin-ing the miniature planned societies that exist to-day-namely the token reinforcement systemsfound in prisons, mental hospitals, schools, thearmy, and industry. When the prospects of aplanned behavioral society are so evaluated, thedire predictions of the civil libertarian and thesocial reformer are amply confirmed. The con-cerns of "who controls the controller" expressa plea of manipulative or exploitive control inwhich people from one stratum (the controllers)set the goals, define contingencies, and dispensereinforcers to people on a lower, subjugated stra-tum (the controllees). I concluded my earlieranalysis as follows:

Guards reinforce prisoners; nurses reinforcepatients, and teachers reinforce students.The fear of manipulative control is wellfounded when a professional-client rela-tionship is lacking. Subjects of these behav-ior control systems are not clients. Behaviormodifiers in prisons are fundamentally andinescapably responsible to the warden orBureau of Correction, not the prisoner; inthe classroom, they are responsible to theprincipal or Board of Education, not to thestudents. Put simply, today's token econo-mies support established power structures.(Holland, 1975, p. 90)

Those in authority who hire the behavior an-alysts and the analysts themselves may see theircontrol as benevolent. Surely rulers usually do.

But the surprising thing about the role of behav-ior modifiers is the frequency with which theyaccept the inner traits seen by their employers asthe causes of the to-be-corrected behavior. Wil-liam Ryan refers to this as "blaming the victim"(Ryan, 1971). The people whose behavior is tobe modified are called "sociopaths" or "unmoti-vated" workers or "defective delinquents" or, ifvery young, "predelinquents" or "hyperactive"school kids. The behavior modifier then proceedsto arrange special reinforcement contingencies inspecial institutional environments. And the sub-jects whose behavior is so corrected are expectedto continue this new behavior regardless of thecontingencies that prevail outside the institution.However, our basic science, the ExperimentalAnalysis of Behavior, tells us that behavioradapts to contingencies-any contingency-notjust those arranged by the psychologist. The be-havior that bothers the employer of the behavioranalyst is itself the result of contingencies. Ifreal, lasting changes are to be made, it is thesecontingencies that must be changed. Lasting be-havior change requires the modification of thecontingencies that produce and maintain theoriginal behavior.

In this context I would like briefly to considerthree victims: the alcoholic, the criminal, andthe behavioral analyst, and discuss how behavior-ists have contributed to their problems, whilebehaviorism points toward a solution.

The AlcoholicFirst the drunk-often the object of behavior

therapy-surely an unfortunate in need of help.For the jailed or hospitalized alcoholic, aversiontherapy is common (Miller and Barlow, 1973).In classical conditioning procedures, the variousstimuli associated with drinking, such as thetaste of alcoholic beverages, visual stimuli, andeven "imagining" drinking, are paired with astrong aversive stimulus, such as electric shockor drug-induced nausea. This conditioning pro-cedure is expected to establish the various stimulithat accompany drinking as conditioned aversivestimuli.

164

BEHAVIORISM: PART OF THE PROBLEM OR PART OF THE SOLUTION?

Advocates of aversion therapy base confidencein their techniques on their understanding of lab-oratory findings. However, the facts of discrim-ination learning are neglected. The home, thebar, or streets in which the problem drinkingnormally takes place are easily discriminatedfrom the contrived conditions of the clinic, oreven the simulated bars used to promote transferof training. Differences between these contrivedsettings and the real-life situations are still fargreater than those found in laboratory discrim-ination learning studies where responding regu-larly comes under stimulus control. Behavior isadaptable. It adjusts to the contingencies of re-inforcement or punishment. Oddly, the behaviortherapist very often seems to miss this point. Thetherapist attempts a technical fix for special de-fects in the drunk by applying aversive stimuliin the clinic, but outside the clinic all those con-ditions prevail that maintain the behavior in thefirst place, and behavior adjusts to these condi-tions. One of the adjustments is drinking itself.And so the problem, as defined by the employerof the behavior modifier, returns-hence recidi-vism or the lack of transfer.

Perhaps given the way the problem is defined,the punishing aspects of the technique might ac-tually contribute to its use. Although the thera-pist may be well-intentioned, a society thatviews the "skid row" alcoholic as worthy of scornand retribution may encourage such abusivetreatment. It is interesting that aversive therapyis used almost exclusively for behaviors that innontherapeutic settings are the object of severesanction and retaliation. Many condone, or de-mand, the punishment of the homosexual, thechild molester, the violent person, the drug ad-dict, and the drunk. These victims are the objectof scorn and retribution; they are also the objectof aversion therapy. At the same time, aversiontherapy is notably absent in the treament of anumber of problems that do not generally re-pulse others. The phobic is not shocked when heshows fear; nor is the under-assertive personshocked for passivity, unlike the over-assertive,aggressive prisoner. A homosexual may be

shocked for having an erection when shown aphotograph, but the patient in therapy for im-potence is not shocked for failure to attain anerection in similar circumstances. If aversivetherapy is at all effective, why is it reserved foracts viewed as despicable? Is there an element ofsocial retaliation in its use?

I assume that therapists are motivated by hu-manitarian concerns to help the patient. But theproblems of the individual have been defined byothers in such a way that aversive means of be-havior change are justified. This can be illus-trated in a study by Gallant (1970), who evalu-ated a compulsory treatment program for "re-volving door" alcoholics. The "revolving door"alcoholic is one who is regularly arrested fordrunkenness (at least in New Orleans, where thestudy was conducted). Over 52% of the 210 pa-tients in the study had been arrested more than50 times and had averaged 14 convictions withinthe year of their forced participation in the study.A cooperative judge sentenced these people ei-ther to serve a 90-day jail term or to "volunteer"to participate in Gallant's experiment. Antabusewas used as a component of therapy. Antabuseis a drug that induces severe nausea when alco-hol is ingested. Hence, it is a form of aversiontherapy. You may judge whether the therapist'sscorn for revolving door alcoholics might haveeased his decision to use a tortuous procedure.In support of coerced participation, Gallant saysof these patients:

It was decided to undertake the presentcompulsory treatment project, which at-tempts to provide the "revolving door" al-coholic with "something to lose" if he re-turns to his former habits. To have "some-thing to lose" is essential to the apathetic,depressed, mistrusting and dependent alco-holic patient with a severe lack of essentialanxiety or concern. (Gallant, 1970, p. 734)

The alcoholics' problem is defined as drinkingand it is assumed that subjecting them to sometemporary aversive conditioning might "cure"their drinking habits.

165

JAMES G. HOLLAND

However, Gallant's description of the demo-graphic characteristics of the patient populationmight suggest some other definitions of the prob-lem. Of the 210 patients, only four lived witha wife, 24 with relatives, and three with friends.Most lived alone and over half had no perma-nent residence. They were poor, unemployed,and the medical work-up revealed a host of se-vere physical problems. In general, the patientswould seem to be an unhappy and lonely lotwith more than enough problems. But the thera-pist views alcohol to be the problem and treats itdirectly. Even when police harassment is recog-nized, the concern is with the problem it posesfor the evaluation data, not the problem it pre-sents to the alcoholic. According to Gallant:

It was not unusual for some of these "re-volving-door" alcoholics, who loitered inthe "skid-row" areas, to be arrested whenthey were not creating a disturbance . . .

the officers were cognizant of the past his-tory of the loiterer and would have littlehesitation in arresting the person if he ap-peared to be intoxicated. This unhealthy cy-cle had to be broken. Otherwise, two of theimportant criteria of efficacy, which are ar-rest rate and conviction rate, would presentconfusing and unreliable results. (Gallant,1970, p. 737)

So cards were issued to subjects in the experi-ment, which they could show to a police officerattempting to pick them up. Police officers wereinstructed not to arrest a cardholder "unless ofcourse, he has committed an illegal offense".

If the very theory on which behavior therapyis based is correct, then the solution to a behav-ioral problem cannot rest in the specially ar-ranged contingencies in the special environmentof the clinic. The contingencies of the naturalenvironment must be modified if the problem isto be corrected. The abject misery and lonelinessof Gallant's skid-row alcoholics could provide asolid operant basis for drinking as a means of es-cape into unconsciousness. That the long-termeffect adds to the poverty and loneliness simply

makes moderation more difficult by deepeningthe problems and establishing a vicious cycle.The short-term escape leads to wretched long-term effects, which in turn lead to another short-term escape as the cycle continues.The very arrest record that so concerns Gal-

lant raises questions about the nature of these re-inforcement contingencies. When behavior oc-curs in high frequency, it is likely that thebehavior is being reinforced. Gallant fails torecognize the significance of an anecdote he re-ports showing the functioning of natural contin-gencies in maintaining alcoholism. One of his"revolving door" alcoholics called the police.toreport that a drunk was lying unconscious by thephone booth. After hanging up, this alcoholicleft the booth and lay down beside it to await hisown arrest. For the poor, the homeless, and thelonely, the jail may contain many important re-inforcers. Borrowing jargon from the animalpsychologist, the jail may be the alcoholic's goalbox.

Other behavior analysts have shown thatdrinking is modifiable by direct reinforcementprocedures. Cohen, Liebson, and Faillace (1970)used a token reinforcement system to manipulatefrequency of imbibing alcohol by hospitalizedalcoholics, and thereby demonstrated the operantnature of drinking. It is, they showed, governedby its consequences. Quite properly, they sug-gest that drinking problems seen by the profes-sional are a result of contingencies in the drink-er's environment. They particularly point out therole of the professional in actually providingcontingencies that may maintain drinking-intheir words:

Some of the reinforcers which may be rele-vant to alcoholism in the outside world in-clude hospitals' and physicians' care, andwelfare, and rehabilitative programs. Theyoften provide powerful reinforcers, such asmoney, attention, drugs, medical and psy-chiatric care, guidance, and counseling tothe alcoholic. These services . . . are oftendispensed during or following excessive

166

BEHAVIORISM: PART OF THE PROBLEM OR PART OF THE SOLUTION?

drinking. Although these reinforcers maynot have initiated excessive drinking, theymay make substantial contribution towardsits maintenance. Perhaps some of these re-inforcers might be dispensed systematicallycontingent on moderation. (Cohen et al.,1970, pp. 764-765)

Gallant failed to find any advantage for hiscompulsory aversive-therapy program. I hope hewill cease to use coercive procedures and I hopehe has learned that special clinic procedures can-not solve behavioral problems that are main-tained by conditions in the natural environment.Alcoholism is not like appendicitis, in which re-moval of the malfunctioning part is the cure.Rather, alcoholism is adaptive behavior. Gallantmight instead try in his New Orleans skid-row aspecial community center with a free kitchen, atelevision, games, and furniture arranged forconversation. Then, his ex-patients should begranted admission contingent on being sober.What of the common environmental contin-

gencies that establish and maintain drinking?It would seem obvious that the taste and imme-diate pharmacological effects are reinforcing.Also, the withdrawal symptoms of the addict areimportant in maintaining drinking. But in oursociety, drinking plays such a prominent socialrole that contingent reinforcement for drinkingis widespread. To arrange for social contacts ina light, friendly spirit we hold a cocktail party.When a difficult problem is to be discussed in aless threatening or more leisurely manner, wearrange to get together over a beer. The cama-raderie of the neighborhood bar at the end ofthe work day is full of reinforcing social con-tacts over a drink. A boy approaches a girl withan offer to buy her a drink. In fact, the path tothe bedroom may include cocktails, wine withdinner, and after-dinner drinks. The nondrinkermisses many of these reinforcers; in fact, thenondrinking businessman who forgoes the mid-day cocktail and official business cocktail partiesmight be occupationally handicapped. In allof these instances, alcohol serves as an SD for

reinforcement of social behavior which one'speers would otherwise treat as frivolous, inap-propriate, sentimental, or otherwise punishable.

Drinking receives additional impetus from alarge industry. "Response cost" is minimized bygood merchandising with ample, convenientlylocated bars and liquor stores. The industryspends 200 million dollars a year advertisinghardy, masculine men "going around only once"to "grab all the gusto" they can get in the formof a beer.A serious effort to solve the social problems

involved in drinking is not made by focusing onthe so-called "flawed personality" of the un-sightly drunk. It is not to be solved by victimblaming. It requires, as every behaviorist shouldknow, a change in the environmental contingen-cies that constitute daily business and culturalpractices. Such change is not likely in the normalfunctioning of the grand corporate syndicate thatis America. Can you imagine the consequences ofa sudden 50% reduction in alcohol consump-tion? We had a small taste of a comparableproblem when the Arab oil embargo pressed usto solve our many problems associated with theoveruse of the automobile. But normalcy pre-vailed as techniques to keep profits up at any so-cial cost were developed.

The CriminalThe criminal, like the alcoholic, suffers from

victim blaming. The prison system has as oneaim the correction of alleged personality flawsviewed as the cause of criminal behavior. Theprofessional is hired to correct these traits and torehabilitate the prisoner. The vocabulary of theprison establishment reflects this approach. Pris-ons are called "correctional" institutions, guardsare called "correctional officers", and the "hole"or solitary confinement is called the "behavioraladjustment unit". When the behavior analystis hired he goes along with the system's defini-tion of the problem and sets up token economiesand behavior therapy programs to "correct" theindividual, even though the behaviorists' knowl-edge of the principles of behavior control should

167

JAMES G. HOLLAND

inspire a search for the controlling variables inthe world of the criminal. It should then be ofno great surprise that a thorough survey of fol-lowup research on prison rehabilitation pro-grams (including some behavioral ones) failedto find credible evidence for the effectiveness ofany in reducing recidivism (Martinson, 1974).

It is time, then, that the behaviorist stop tin-kering with what he or she knows are fictionalmental causes and look to the contingencies thatproduce the "criminal" behavior. This is a largeorder, but I would like to suggest a few tentativerelationships in the hope of getting the analysisstarted. First, it is necessary to recognize thescope and variety of the problem called crime.Although a special subset-the lower class-isdisproportionately represented in the prison pop-ulation, studies of anonymous self-reports (Dol-eschal and Klapmuts, 1973) show that over90% of the adult male population has commit-ted criminal acts.As behaviorists, we might correctly suppose

that legal and illegal behavior are similarly con-trolled. Most crime is for economic gain and sodirect reinforcement is at its base. For example,according to the 1970's FBI's Uniform CrimeReports, 87% of the crimes were classified asproperty crimes (which includes larceny over$50.00, automobile theft, burglary, and so on).Of the remaining "crimes against persons", halfwere robbery which, while involving personalconfrontation, are at base also economic. Thus,939% were crimes resulting in economic gain.The sensational ritual killer and child molesterso prominent in law-and-order discussions andyellow journalism reports are a very small por-tion of the imprisoned, and even there they areusually outcasts among the prisoners.

But the vast bulk of crime is not included inthe FBI statistics, which exclude much of whitecollar crime. Estimates made by the Presi-dent's Commission on Law Enforcement (1967)suggested that white-collar crimes of embezzle-ment, fraud, tax cheating, and forgery amount to$1.73 billion, while crimes of the poor-rob-bery, burglary, auto theft, and larceny-amount

to a little more than one-third of that amount(or $608 million). Poor and rich alike commitillegal acts for economic gain. Our society ishighly stratified, and at any level there is a strug-gle to increase status and happiness through in-dividualistic competitive efforts (legal or ille-gal). Status is defined by wealth and levels ofconsumption. The competitive system praisesthose who acquire even at the expense of others.

There are differences in deprivation and accessthat explain the class differences in crimes. Thepoor and unemployed have no opportunity tocheat on taxes or embezzle funds. Their level ofbasic needs forces them into the dangerous, butavailable street crimes that are public, and hencemore frequently result in arrest. The better-offneed not be involved in burglary. Their crimesare safer and can occur in the privacy of theirhomes or offices. In fact, their public stance maybe to defend truth and righteousness. To borrowa thought from Kurt Vonnegut, you may es-tablish a reputation for truthfulness so as to bebelieved when you lie.

Economic wealth, nature of deprivation, andresponse opportunity explain only part of thepattern of crime. Criminal justice is itself de-signed to deter crime, and as such involves avoid-ance contingencies. While this deterent effectdoes not stop the crimes from occurring, it playsan important role in shaping the nature and themagnitude of the crime. The act that occurs,given the opportunity, depends on its probabil-ity and probable magnitude of punishment.Therefore, street crimes occur in safe areas,which, ironically, also tend to be among thepoor. Law enforcement is better in better neigh-borhoods and the poor, black prowler more con-spicuous. Similarly, the tax cheat avoids outra-geous deductions that invite audit, and we haveall been victimized by small consumer fraudsthat leave us defenseless, because the cost of re-covering the loss is larger than the loss itself.

The Behavior AnalystHow is it that behavior analysts fail to focus

on the maintaining contingencies in the behavior

168

BEHAVIORISM: PART OF THE PROBLEM OR PART OF THE SOLUTION?

of the criminal or the drunk? Surely our behav-ior is the result of natural contingencies also.Hence, my third victim-ourselves and our aca-

demic colleagues. In the layered stratified hier-archies of control, we are high enough that it isparticularly important to look at sources of con-

trol and contingencies designed to shape our be-havior. Over the past decade, social scientistWilliam Domhoff has carried out research thathas given us a number of insights. I will present

a broad outline of his conclusions and encourage

you to read his books for their rather impressivearray of supporting data. Domhoff (1967) hasoperationally defined membership in a socialupper-class (which comprises less than one-halfof 1% of the population). He finds evidence thatthis hereditary social upper-class is also a gov-

erning class by examining the social backgroundsof policymakers in business and government.Members of the social upper-class, along withsome nonhereditary wealthy school chums andclub-mates who hold top executive positions intheir major corporations, govern first through thedirect control of major corporations both byownership (they hold 75% of corporate stock)and by direct representation in sizable numberson corporate boards. These boards interlock ex-

tensively, with some persons holding seats across

several corporations. In addition, the ruling classcontrol the major foundations which are so im-portant in defining public policy and problems.These foundations, of course, are themselvesproducts of the wealthy and serve as mattresses

under which the wealthy hide their money. Theuniversities that hire us are similarly dominatedby this ruling class, through individual and cor-

porate contributions and through heavy repre-

sentation on university board of trustees. Themedia, television, large-circulation magazines,and major metropolitan newspapers are almostall owned by ruling-class members and furtherinfluenced by the wishes of their major advertis-ers who (need I say it?) are the major corpora-

tions. The executive branch of the federal gov-

ernment is likewise controlled by this power

elite. The presidential nominee of either party

must, by the time of the convention, commandsufficient campaign funds to secure the nomina-tion. I trust recent events made clear where thesefunds come from. After the election, the richare amply rewarded with cabinet positions andspecial advisory roles. Much public policyadopted by the executive branch is not formedwithin the government at all, but by specialgroups and commissions that are dominated bythe ruling class. For example, the Council onForeign Affairs has been the important formu-lator of foreign policy (Domhoff, 1970).From these lofty heights the objectives of our

society are formulated, the aims and values areset in conformity with the perspective of theupper-class who, let me hasten to add, are benev-olent rulers. But what of our reinforcement onmeeting these designated objectives? In general,our lot in life is good. We are paid very well in-deed, relative to those whose behavior we targetfor change. Once we are within the system, wehave a high degree of security and can painlesslypass through periods during which newspapersreport 10% unemployment. We have a high de-gree of prestige and could easily believe we de-serve this because of superior intelligence, merit,motivation, and the rights of passage sufferedduring graduate comps, orals, and defense. It istempting for even the behaviorist to ignore hisscience and accept victim-praising inner causes.Wrapped in our security and benefits-low-

cost home mortgages, group insurance rates, tui-tion benefits for our families, retirement funds-we lead a worry-free life as long as we can avoidlosing it. Technically, you will recognize thiscontingency as an avoidance schedule, but unlikeaversion therapy programs this is avoidance ofwithdrawal of positive reinforcement. If yousometimes feel hassled by this control regimen,you no doubt appreciate the line from a songsung by Janis Joplin: "Freedom's just anotherword for nothing left to lose". But as long as wecan avoid losing our reinforcement system, weclimb a neatly arranged set of job classificationsin the university from instructor to professor.The various review processes tend to reinforce

169

JAMES G. HOLLAND

orthodoxy-political, professional, and social.Our young colleagues should have no fear, sinceif they manage to prove for six years that theiracademic freedom needs no defense, they will begranted tenure to assure their academic freedom.

But our aspiring young colleagues must alsopublish, and so must we. We generally proceedas though we believe that a productive line ofresearch requires funds. To obtain these funds,we go to organizations that define problems theywish to see solved. We have seen already that themajor source of funds is dominated by a smallruling class. The problems, then, are defined andpolicies set by those who gain the most from ourcurrent economic, social, and political systems.

The myth of inner causes is fostered becauseof the reinforcement provided to the elite andbecause of the myth's role in maintaining thecurrent system. Those high in the hierarchy ofpower are said to have ascended through greatpersonal merit. The rich were free to use theirinner resources, their will, their determination,their motivation, their intelligence to reach theirhigh level. Inner causes serve as justification forthose who profit from inequality.A special set of inner causes is reserved for

the poor. They are said to be lazy, to lack ambi-tion, to be untalented. Those who get the mostfrom the social system might find it punishingto view their own good fortune as the result of asystem that exploits those below and createspoverty and unhappiness. If so, verbal state-ments that attribute each person's position in so-ciety to personal traits, either innate or the resultof a "disadvantaged" subculture, would be rein-forced.And it is especially important for those on top

to convince those on the bottom that their plightis their own fault. So Time and Atlantic Monthlyprovide instant fame for the latest version ofSocial Darwinism, even when these views areconsidered faulty by many scientists. Sadly, eventhose who sell their talents as behavior modifierssometimes accept the victim-blaming definitionson which funding policies rest; and they attemptto fix, not environments, but the inner nature of

the individuals. Though behavior modifiers knowbetter, they often treat prisoners as though theyare maladapted and can be "fixed" in a behavior-modification system, returned to the reinforce-ment system from which they reached prison,and somehow expect that those contingenciesthat shaped the criminal behavior originally willnot now exert control.

Analyzing the System, Not the Victim

But clearly, behaviorists (and radical reform-ers) know that credit or blame rests with the sys-tem. In operant conditioning laboratories, behav-ior of individual organisms is shaped in preciseand replicable ways, using simple and complexcontingencies of reinforcement, avoidance, orpunishment. When the apparatus is changedfrom one schedule to another, the behaviorchanges from one stable pattern through a transi-tion phase to the new pattern appropriate for thenew set of contingencies. Automated equipmentcontrols these experiments and defines the con-trolling contingencies. To the pigeon, the contin-gencies programmed by the equipment are itssocietal system, and the pigeon behaves appropri-ately to that system. Our contingencies arelargely programmed in our social institutions andit is these systems of contingencies that determineour behavior. If the people of a society are un-happy, if they are poor, if they are deprived, thenit is the contingencies embodied in institutions, inthe economic system, and in the governmentwhich must change. It takes changed contingen-cies to change behavior. If social equality is agoal, then all the institutional forms that main-tain stratification must be replaced with formsthat assure equality of power and equality ofstatus. If exploitation is to cease, institutionalforms that assure cooperation must be developed.Thus, experimental analysis provides a support-ing rationale for the reformer who sets out tochange systems.

Behavior analysis also provides some basis foroptimism. A contingency analysis shows inherentflaws in a stratified control system which should,in time, bring about its change, and a wider un-

170

BEHAVIORISM: PART OF THE PROBLEM OR PART OF THE SOLUTION?

derstanding of Behaviorism should acceleratethis process. Behaviorism will be part of thesolution.

While meeting the stratified system's goals,our behavior-change programs also follow theform of power relationships characteristic of oursociety (Holland, 1975). Thus, ironically, evenas we serve power, our behavior-modificationsystems are beginning to give the social reformeran advantage. Behavior modification models thenormal societal control process and makes theprocess explicit and clearer. Reinforcement con-tingencies are clearly stated. Objectives arespelled out. Responsibility for dispensing rein-forcement is made explicit. As a model of theparent society, it gives us a chance for a behavioranalysis of control in that society.

There are several problems intrinsic to astratified control system. First, when the interestof the controller and the controllee are different,controllers should find it exceedingly difficult todesign any contingency management system sowell that the controllee will not find alternativeroutes to reinforcement which are subversive tothe aim of the controller. To a prisoner, successin a behavior-modification system can be ameans of early parole. The prisoner can fake at-titude change to obtain his tokens. Prisonerscall this practice shamming, and written instruc-tions for shamming circulate in an undergroundcommunication network in some prisons. Otherprisoners resist behavior modification by hiringlawyers and fighting in the courts; others refuseto accept reinforcement even when it includespromotion from solitary confinement.With stratification of power between control-

ler and controllee, there is a reinforcement basisfor struggle and resistance by the controlled.When goals are set from above to adapt the sub-jugated to standards imposed from above, thereis a clear behavioral basis for counter control. Itis not in the prisoners' interest to be shaped intoobedience to authority; there have been, and con-tinue to be, good reasons for their resistance toauthority. Authorities have repeatedly exploitedthem, giving privilege and wealth to those

with authority over them. Prisoners may gainmore in reinforcement through devising unpro-grammed alternatives than through carefullyprogrammed contingencies leading to small re-inforcers.

Second, exploitative systems using only posi-tive reinforcement are exceedingly difficult todesign. If wealth is to accumulate at the top, itmust be sparingly distributed among the con-trollees at the bottom. When a behavior-modifi-cation program seems entirely positive, usuallycoercion or restraint is required at the boundariesto keep the controllee in the system. A prisontoken economy still requires prison walls andarmed guards to keep the prisoners from leavingbefore they earn graduation from the program.In the token-economy program for Army basictraining (Datel and Legters, Note 1) it is un-likely that the recruit would diligently work forpoints toward a weekend pass if it were not forthe well-known aversive consequences of deser-tion.

Similarly, the effective use of minimal rein-forcement requires prior deprivation states. Theworking poor can be made to work for limitedgain only if they are kept poor. It is not likelythat people could be kept hungry in a wealthycountry without armed forces and a prison sys-tem. And, indeed, in the prison token economy,the prisoner begins in solitary confinement, anextreme state of deprivation, and works to earnpositive reinforcement in the form of points. Abehavior-modification system itself may use onlypositive reinforcement, but if the higher strataare to keep the greater share of the wealth andprivileges, restraint and coercion must be used tokeep the controllee within the rules of the sys-tem.

Aversive control generates counter control;although the carrot and the stick may generateless counter control than the stick alone, in timethe aversive conditions required to keep the con-trollee in contact with the limited reinforcementsystem should generate counter control. Unequalwealth and power require protection by coerciveforces. This generates counter control, class

171

JAMES G. HOLLAND

struggle, and the eventual replacement of thesystem. Here is a basis of natural selection forcultural practices which could favor an egalitar-ian system.

BEHAVIORISM AND AFUTURE SOCIETY

How does our science, Behaviorism, and itsapplication move on to furthering the solution?It would seem we need to move toward collectiveforms-forms based on cooperation, forms thatmaximize reinforcement for helping others, in-stead of reinforcement at the expense of others.Interesting efforts in doing such work are basedon what I call the "Frazier" model. The Fraziermodel draws from Walden II (Skinner, 1948),itself an extremely interesting example of be-havioral community that I think practical, egali-tarian, and socially radical. It came into beingbecause Frazier benevolently created it and was

himself somehow removed from the specific con-

tingencies of Walden II. There are numerous

real-life attempts at egalitarian behavioral socie-ties in the form of communes, notably TwinOaks in Virginia (Kinkade, 1973), and a pro-

foundly impressive one in which Keith Miller(1976) at the University of Kansas plays Frazierto a communal student-housing project. Miller'sefforts are impressive because they are completewith data collection, specification of job descrip-tions and procedures that evolved from the proj-ect. I have taken to playing Frazier in an under-graduate course I teach in behavioral control insociety. The class is formed into collectives ofabout 10 to 15 students. Members of each col-lective work together and prepare each other be-haviorally to analyze social systems for a col-lective project. They are tested individually, butthe assigned grade is the average for all individu-als in a collective. I have simply stopped givingcompetitive, individual grades and the resultsare most exciting. Attempts in total institutionshave been made as well. Fairweather (1964) ex-

perimented with group reinforcement systems on

a ward in a mental hospital. Here, the patients

take part in defining criterion behavior and as-sessment of performances for both themselvesand their peers. A group at Mendocino State inCalifornia (Rozynko, Swift, Swift, and Boggs,1973) has created communities of hospitalizedalcoholics who work together in behaviorallyanalyzing their problems. All of these examplesare limited in two important ways: they exist inspecial isolated environments and they havewithin them the contradiction of a Frazier-anelite designer.What we need today is an extensive and vig-

orous analysis of the controlling conditions inour society. What are the contingencies in thework place? What are the various means bywhich people are kept segmented and compart-mentalized and exploited by such pigeon-hold-ing? Now, from all I have said, it should beclear that we cannot expect funding for this ac-tivity. Promotion even within the academicworld will be difficult, since to be part of the so-lution the audience for our findings must includepeople other than academicians. I can hardlyexpect that the third of my trio of victims, thebehavior analyst, should suddenly be free of con-trols and be propelled toward this work in sys-tem-blaming. But unexpected help may layahead. With the liberal attack on behavior modi-fication, the establishment may simply throw usaway. The liberal reformer will have won a greatvictory and the ruling elite will laugh over cock-tails in their clubs because they know behaviorcontrol will continue in much the same form.Then, with our freedom, with nothing left tolose, the analysis may proceed toward solutions.Our old practices, useful when we embraced

victim-blaming and victim-praising must bemodified. Some new methodology is called for.I suggest that we need to work with people inall our social institutions in analyzing the contin-gencies that oppress them. We bring to this en-terprise certain specialized knowledge and skillsin the use of data; they bring direct specializedexperience with those day-to-day contingencies.They are also the audience for the results of theanalysis; we and they, as collaborators, can

172

BEHAVIORISM: PART OF THE PROBLEM OR PART OF THE SOLUTION?

evolve experimental solutions to society's prob-

lems.There is a precedent for the effort to listen

carefully to what people say and analyze thecontrolling contingencies of this behavior. Skin-ner's Verbal Behavior (Skinner, 1957) is rich injust such examples. It has long been a laboratoryaxiom that the organism knows best. We must

adopt this with a new seriousness. The person isa mirror for his or her response contingencies.Behavior is adaptive. It has a close integrity withthe controlling contingencies. The behavioristshould listen carefully as partner and colleaguein social reform. Our role in the change process

will be as a catalyst-to assist in the design ofsolutions.

I cannot overly emphasize the importance ofalways understanding the adaptive nature of be-havior, since the problem-definers traditionallyconsider so much behavior maladaptive. We us-

ually marvel at the intricate biological formsthat evolve through natural selection. These as-

tounding variations reflect generations of shift-ing environmental conditions. Similarly, a per-

son's behavior reflects response contingencies.We should be as respectful and awed by theseadaptations as we are of evolutionary adapta-tions.

What direction might this collective explora-tion of controlling contingencies take? I realizeit is presumptuous to hazard a guess in advanceof a series of necessary experiments. However,I think it probable that the analysis of control-ling contingencies will reveal deep contradic-tions in our stratified system, and will reveal pro-

found problems in our ideology of "doing your

own thing" and maximizing personal gain. Ithink the analysis will naturally draw us towarda community of equality, with service to othersand responsibilities for others as guiding princi-ples. One reason for my expectation is that supe-

rior behavior control is possible in a communityof peers. When all the day-to-day acts of eachindividual are evaluated communally in terms ofthe criteria of "service to others", small begin-nings of elitism or personal gain at the expense

of others can be immediately assessed. Shouldgroup members display elitist behavior, such asgaining special individual advantage at the ex-pense of others, the group would criticize thebehavior. The smallest beginning signs of such a"social disease" would be dealt with in thegroup. Unlike today's society, behaviors that canget someone into trouble would not go througha natural shaping process in which small exam-ples of unacceptable behaviors are reinforcedand allowed to grow in size until suddenly an in-dividual is in trouble with the legal system.

Monitoring of contingencies in peer groupscould be precise and detailed. A group of peerswho work and live together and discuss the sig-nificance of their work and life together, can de-tect the smallest progress toward fulfilling a goalor the earliest sign of incorrect thought andprovide the appropriate feedback promptly.There is no need for a separate, distant elite at-tempting to evaluate performance. Peers evalu-ate each other's movement toward shared goals.Because goals are shared, there is little to begained by deliberate deception or subversion.Because of the closeness among group membersand coworkers, fine subtleties can be analyzedfor positive or negative comment. Moreover,this peer approval is likely to be much morepowerful than the tokens dispensed by a disaf-fected elite.

In summary, fundamental truths adducedfrom an empirical laboratory and politically neu-tral science have laid the basis for this analysisof social systems. I have tried to show that in astratified system there are built-in contingenciesfor struggle by the oppressed. The standard useof victim-blaming and inner causes itself has abehavioral basis, in that it reinforces the statusquo; this is so even though the science of behav-ior has made such inner causes unsatisfactory asexplanations of behavior. A wider disseminationof the methods for analyzing controlling contin-gencies can accelerate the creation of a nonop-pressive society and the passing away of the so-cial problems for which victims are themselvesso often blamed.

173

174 JAMES G. HOLLAND

REFERENCE NOTE

1. Datel, W. E. and Legters, L. J. The psychologyof the Army recruit. Paper presented at the meet-ing of the American Medical Association, Chicago,Illinois, June, 1970.

REFERENCES

Cohen, M., Liebson, I. A., and Faillace, L. A. Themodification of drinking in chronic alcoholics.In N. K. Mello and J. H. Mendelson (Eds), Re-cent advances in studies of alcoholism. Rockville,Maryland: National Institute of Mental Health,1970. Pp. 745-766.

Doleschal, E. and Klapmuts, N. Toward a new crim-inology. Crime and Delinquency Literature, 1973,5, 607-626.

Domhoff, G. W. The higher circles: the governingclass in America. New York: Random House,1970.

Domhoff, G. W. Who rules America? EnglewoodCliffs: Prentice-Hall, 1967.

Fairweather, G. W. (Ed) Social psychology in treat-ing mental illness: An experimental approach.New York: John Wiley & Sons, 1964.

FBI Uniform Crime Reports (1970).Gallant, D. M. Evaluation of compulsory treatment

of the alcoholic municipal court offender. In N. K.Mello and J. H. Mendelson (Eds), Recent ad-vances in studies of alcoholism. Rockville, Mary-land: National Institute of Mental Health, 1970.Pp. 730-744.

Holland, J. G. Behavior modification for prisoners,patients, and other people as a prescription for

the planned society. Mexican Journal of BehaviorAnalysis, 1975, 1, 81-95.

Kinkade, K. A Walden Two experiment. NewYork: William Morrow, 1973.

Martinson, R. What works? Questions and answersabout prison reform. The Public Interest, No. 35,Spring 1974.

Miller, L. K. The design of better communitiesthrough the application of behavioral principles.In W. E. Craighead, A. E. Kazdin, and M. J. Ma-honey (Eds), Behavior modification: principles,issues, and applications. Boston: Houghton Mif-flin, 1976. Pp. 68-101.

Miller, P. M. and Barlow, D. H. Behavioral ap-proaches to the treatment of alcoholism. Journalof Nervous and Mental Disease, 1973, 157, 10-20.

President's Commission on Law Enforcement. Wash-ington, D.C.: U.S. Government Printing Office,1967.

Rozynko, V., Swift, K., Swift, J., and Boggs, L. J.Controlled environments for social change. In H.Wheeler (Ed), Beyond the punitive society. SanFrancisco: W. H. Freeman and Company, 1973.Pp. 71-100.

Ryan, W. Blaming the victim. New York: VintageBooks, 1971.

Skinner, B. F. Verbal behavior. New York: Apple-ton-Century-Crofts, 1957.

Skinner, B. F. Walden Two. New York: Macmillan,1948.

Received 6 August 1976.(Final Acceptance 23 August 1977.)