The Behavior Analyst 2003, 26, 181-193 No. 2 (Fall)

Behavior Analysis, Mentalism, and thePath to Social Justice

J. MooreUniversity of Wisconsin-Milwaukee

Traditional psychology is mentalistic in the sense that it appeals to inner causes in the explanationof behavior. Two examples of mentalism in traditional psychology are (a) dispositional attributionsand (b) conventional treatments of intelligence. These examples may be linked to such pernicioussocial -isms as racism and sexism by noting that some individuals justify engaging in discriminatoryconduct toward others by appealing to some deficient inner quality of those being discriminatedagainst. This sort of mentalistic appeal ultimately prevents some members of our society from beingintegrated into society and from progressing down the path of social justice. Behavior analysisoffers a constructional alternative to the mentalistic views of traditional psychology and allows oursociety as a whole to move down the path.Key words: mentalism, dispositional attributions, intelligence, racism, sexism, social justice

Day (1976) outlined three proposi-tions in terms of which he defined theradical behaviorist outlook. The first isa focal interest in the contingencies in-volved in the control of behavior. Thesecond is an opposition to mentalism.The third is social activism. This thirdproposition

involves at heart a particular conviction with re-spect to social planning, namely, that if we areto survive as a species we should begin at onceto restructure our social environment, in a piece-meal fashion, so that it acts to produce peoplewho have the behavioral equipment necessaryfor us all to survive. (p. 535)

Certainly the first concern-the analy-sis of contingencies-is vitally impor-tant to all behavior analysts. However,the second and third concerns-the op-position to mentalism and the socialactivism aimed at investing citizenswith the behavioral skills necessary foreveryone's survival-are just as im-portant, and they are the particular fo-cus of this article.On occasion some have said they do

not see what is wrong with mentalism,

This article is based on a presidential addressdelivered at the convention of the Associationfor Behavior Analysis, San Francisco, in May,2003.

Correspondence concerning this article maybe addressed to J. Moore, Department of Psy-chology, University of Wisconsin-Milwaukee,Milwaukee, Wisconsin 53201.

as long as the terms of a mental or cog-nitive orientation can be operationallydefined and thereby regarded as theo-retical and objective. They may evenargue that mentalisms are useful be-cause they have a heuristic contribu-tion, such as by suggesting novel waysof conceptualizing problems.The perspective of this article is that

there are many things wrong with men-talism, notwithstanding exculpatoryappeals to operational definitions, the-oretical dimensions, or heuristics. Theproblems created by mentalism ulti-mately have to do with Day's thirdconcern about social activism and sav-ing the world through behavior analy-sis. Let us begin an analysis of theseproblems by defining our terms.

MENTALISM

Defining Mentalism

In general terms, mentalism may bedefined as an approach to the study ofbehavior which assumes that a mentalor "inner" dimension exists that dif-fers from a behavioral dimension. Thisdimension is ordinarily referred to interms of its neural, psychic, spiritual,subjective, conceptual, or hypotheticalproperties. Mentalism further assumesthat phenomena in this dimension ei-ther directly cause or at least mediatesome forms of behavior, if not all.

181

182 J. MOORE

These phenomena are typically desig-nated as some sort of act, state, mech-anism, process, or entity that is causalin the sense of initiating or originating.Mentalism regards concerns about theorigin of these phenomena as inciden-tal at best. Finally, mentalism holdsthat an adequate causal explanation ofbehavior must appeal directly to the ef-ficacy of these mental phenomena. Acausal explanation that does not appealto these phenomena, and appeals toonly present or past behavioral, physi-ological, and environmental variables,is necessarily limited in scope and ad-equacy. It therefore cannot be seriouslyentertained by anyone interested in acomplete explanation of behavior.

In some instances mentalism takesthe form of the formal, explicit bifur-cation of the world into the two ex-haustive and mutually exclusive onto-logical dimensions of the mental andthe physical, or the mind and body.Classical psychophysical dualism ofthe sort proposed by traditional philos-ophers is an example of this sort ofmentalism.

In other instances mentalism takesthe form of an "epistemological dual-ism." Epistemological dualism is theassumption that two dimensions are in-herent in the knower rather than theknown. Epistemological dualism startswith the view that words or terms aresymbols that are created in the imme-diate experience of the knower. Thementalism consists in the stances that(a) the immediate experience is a men-tal (or equivalently, subjective) dimen-sion that differs from the physical and(b) knowledge is an outgrowth of pro-cesses in which knowers create andmanipulate symbols for physical ob-jects in their immediate experience.Epistemological dualism is particularlyconspicuous when scientists seek toexplain scientific activity in terms ofthe supposedly unique epistemologicalpowers of theories, logic, and con-structs. Sometimes psychologists whoappeal to the unique epistemologicalpowers of theories, logic, and con-structs believe they are not being men-

talistic because they are not bifurcatingobjects from the world into physicaland extraphysical components. How-ever, this belief is in error, because theway that these psychologists conceiveof the processes by which scientists be-come knowledgeable makes these psy-chologists epistemological dualists andtherefore mentalistic (for further dis-cussion of mentalism and epistemolog-ical dualism in scientific behavior, seealso Moore, 1999, 2000; Smith, 1986,pp. 116 ff.). Readers may note thatSkinner (1945) addressed this problemmany years ago when he stated thatconventional operationismhas not developed a satisfactory formulation ofthe effective verbal behavior of the scientist....Modem logic, as a formalization of "real" lan-guages, retains and extends this dualistic theoryof meaning and can scarcely be appealed to bythe psychologist who recognizes his own re-sponsibility in giving an account of verbal be-havior. ... If it turns out that our final view ofverbal behavior invalidates our scientific struc-ture from the point of view of logic and truth-value, then so much the worse for logic, whichwill also have been embraced by our analysis.(pp. 270, 271, 277)

By this definition, the positions of mostof the distinguished figures in Westernintellectual history who have hadsomething to say about psychology-Descartes, Kant, Freud, Piaget, andothers-are mentalistic in one way oranother. In addition, a large portion ofcontemporary psychology is mentalis-tic: cognitive psychology certainly andun-self-consciously, but also much ofpersonality theory, social psychology,developmental psychology, sensation-perception, neuroscience, cognitiveneuroscience, the medical model ofpsychopathology, and ironically evenmediational S-O-R neobehaviorism.Indeed, a reasonable argument is thatone gets famous in contemporary psy-chology by thinking up new mental-isms and calling them theories, giventhe commitment to epistemological du-alism if not classical psychophysicaldualism.

Sources of MentalismAs we have defined mentalism

above, it is encountered as a particular

BEHAVIOR ANALYSIS AND SOCIAL JUSTICE 183

way of explaining behavior, and an ex-planation is verbal behavior. Verbal be-havior is operant behavior, and as be-havior analysts we analyze it in termsof the contingencies that promote it.Even mentalistic verbal behavior maybe so analyzed. Mentalistic verbal be-havior is often thought to be of concernbecause it purports to refer to entitiesthat exist in another dimension and thatcannot directly be part of a respectablescientific explanation because they arenot publicly observable. However,there is no such other dimension, andthere are no such entities to which theverbal behavior in question supposedlyrefers. Therefore, mentalistic verbalbehavior cannot be of concern becauseit literally refers to another dimensionor supposed entities in another dimen-sion. On the present view, all verbalbehavior, even that which is calledmentalistic, is a function of naturalisticfactors and processes that exist inspace and time, in the physical and ma-terial dimension. Thus, mentalistic ver-bal behavior is of concern because ofthe factors that cause it and its conse-quences. The task is to identify thosefactors and their consequences.

Consider the following two passagesfrom Skinner's writing, in which heoutlined those factors and consequenc-es. Here is the first passage:

Tuming from observed behavior to a fanciful in-ner world continues unabated. Sometimes it islittle more than a linguistic practice. We tend tomake nouns of adjectives and verbs and mustthen find a place for the things the nouns aresaid to represent. We say that a rope is strong,and before long we are speaking of its strength.We call a particular kind of strength tensile, andthen explain that the rope is strong because itpossesses tensile strength. The mistake is lessobvious but more troublesome when matters aremore complex. There is no harm in saying thata fluid possesses viscosity, or in measuring andcomparing different fluids or the same fluid atdifferent temperatures on some convenient scale.But what does viscosity mean? A sticky stuffprepared to trap birds was once made from vis-cus, Latin for mistletoe. The term came to mean"having a ropy or glutinous consistency," andviscosity "the state or quality of being ropy orglutinous." The term is useful in referring to acharacteristic of a fluid, but it is nevertheless amistake to say that a fluid flows slowly because

it is viscous or possesses a high viscosity. Astate or quality inferred from the behavior of afluid begins to be taken as a cause.

Consider now a behavioral parallel. When aperson has been subjected to mildly punishingconsequences in walking on a slippery surface,he may walk in a manner we describe as cau-tious. It is then easy to say that he walks withcaution or that he shows caution. There is noharm in this until we begin to say that the walkscarefully because of his caution. ...The extraordinary appeal of inner causes and

the accompanying neglect of environmental his-tories and current setting must be due to morethan a linguistic practice. I suggest that it hasthe appeal of the arcane, the occult, the hermitic,the magical-those mysteries which have heldso important a position in the history of humanthought. It is the appeal of an apparently inex-plicable power, in a world which seems to liebeyond the senses and the reach of reason.(Skinner, 1974, pp. 165-166, 169)

Here is the second passage:

We almost instinctively look inside a system tosee how it works. We do this with clocks, aswith living systems. Some early efforts to un-derstand and explain behavior in this way havebeen described by Onians in his classic Originsof European Thought. It must have been theslaughterhouse and the battlefield that gave manhis first knowledge of anatomy and physiology.The various functions assigned to parts of theorganism were not usually those that had beenobserved introspectively. ... Observation of abehaving system from within began in earnestwith the discovery of reflexes. ...We have not advanced more rapidly to the

methods and instruments needed in the study ofbehavior precisely because of the diverting pre-occupation with a supposed or real inner life....

It is easier to make the point in the field ofmedicine. Until the present century very littlewas known about bodily practices in health anddisease from which useful therapeutic practicescould be derived. Yet it should have been worth-while to call in a physician. Physicians sawmany ill people and should have acquired a kindof wisdom, unanalyzed perhaps but still of valuein prescribing simple treatments. The history ofmedicine, however, is largely the history of bar-baric practices-bloodlettings, cuppings, poul-tices, purgations, violent emetics-which muchof the time must have been harmful. My pointis that these measures were not suggested by theintuitive wisdom acquired from familiarity withillness; they were suggested by theories, theoriesabout what was going on inside the ill person.Theories of mind have had a similar effect, lessdramatic, perhaps, but quite possibly far moredamaging. ... But philosophy and psychologyhave had their bleedings, cuppings, and purga-tions too, and they have obscured simple wis-dom. They have diverted wise people from apath that would have led more directly to an

184 J. MOORE

eventual science of behavior. ... We have beenmisled by the almost instinctive tendency tolook inside any system to see how it works, atendency doubly powerful in the case of behav-ior because of the apparent inside informationsupplied by feelings and introspectively ob-served states. Our only recourse is to leave thatsubject to the physiologist, who has, or willhave, the only appropriate instruments or meth-ods. (Skinner, 1978, pp. 73-74, 77, 81)

On the basis of such treatments, there-fore, mentalism may be understood asa function of our conventional linguis-tic practices, embedded in a matrix ofunderlying cultural assumptions. Ulti-mately, mentalistic explanations aresupported by the social reinforcementinherent in conceiving of the causes ofbehavior in culturally approved ways.Mentalism may therefore be regardedas the dominant orientation in our so-ciety, as evidenced in our general cul-tural outlook, our system of jurispru-dence, most organized religions, andother social institutions (see alsoMoore, 1981, 1990, for further treat-ment of this entire matter).

AnalysisTo be sure, in some instances men-

talistic positions may appear to explainevents successfully. The extent to whichthey appear to do so is because theytake contingencies into account, if onlyindirectly and implicitly, even thoughmentalistic positions generally mini-mize any interest in environmental fac-tors and relations as explanatory. Thepoint is simply that mentalistic posi-tions would be even more scientificallysuccessful if they took contingenciesinto account directly and explicitly.From another perspective, mental-

ism entails many negative consequenc-es, which more than offset any osten-sible contributions. In brief, mentalisticstatements are troublesome becausethey (a) are incomplete and vague, (b)obscure important details, (c) allay cu-riosity by getting us to accept fictitiousway stations as explanatory, (d) im-pede the search for relevant environ-mental variables, (e) misrepresent thefacts to be accounted for, (f) misrep-resent the processes involved when a

scientist becomes knowledgeable, (g)falsely assure us about the state of ourknowledge, and (h) lead to the contin-ued use of scientific techniques thatshould be abandoned (e.g., hypotheti-co-deductive theory testing), for ex-ample, because they are wasteful.

In the final analysis, mentalism isobjectionable because it is distinctlyunpragmatic. That is, people's mental-istic statements about the causes of be-havior are incorrect because their ver-bal behavior is under the control offactors that are cherished for irrelevantand extraneous reasons. These incor-rect statements then interfere with theexplanation of behavior in terms ofcontingencies at the level of phyloge-ny, ontogeny, and the culture. Ulti-mately, as Skinner (1974) put it, "wemust remember that mentalistic expla-nations explain nothing" (p. 230).A further unfortunate consequence

of a pervasive mentalism is that thefull power of a science of behavior isnot brought to bear on the human con-dition. In particular, the power of a sci-ence of behavior is not brought to bearon social problems, and our culturewill not be able to move down the pathof social justice and ensure that all cit-izens are well educated, integrated intothe mainstream, and provided with thebehavioral equipment that is necessaryfor us all to survive, precisely becauseof mentalism. Let us now look at twospecific examples of mentalism to seehow and why it is so problematic.

ATTRIBUTION THEORYThe Traditional View

The first example is attribution the-ory. As rendered in traditional psy-chology, attribution theory is an aspectof "person perception" and "socialcognition." It is related to the work ofHeider (1958), who proposed that peo-ple attribute observed behavior to ei-ther an internal (i.e., a personal factor)or an external (i.e., a situational factor)cause, but not both. Examples of inter-nal causes are such entities as traits,attitudes, intelligence, expectancies,

BEHAVIOR ANALYSIS AND SOCIAL JUSTICE 185

and cognitions. When one appeals tosuch internal causes in an explanationof behavior, one is said to make a dis-positional attribution. Examples of ex-ternal causes are environmental or so-cial demands. When one appeals tosuch external causes, one is said tomake a situational attribution.A currently influential version of at-

tribution theory is derived from thework of Kelley (1967). Although Kel-ley did not frame the matter in exactlythe following terms, his approach isthat individuals implicitly seek answersto certain questions as they attempt tomake dispositional or situational attri-butions. The first question concernsconsistency: Does the observed personact in the same way in the same situ-ation? The second question concernsdistinctiveness: Does the observed per-son act in a similar way in a similarsituation? The third question concernsconsensus: Are others acting in thesame way in the observed situation?Given one set of answers to these ques-tions, Kelley argued that we tend tomake a dispositional attribution. Givenanother set of answers, we tend tomake a situational attribution. Once wehave started to favor one kind of attri-bution, we then discount the other pos-sibility. Traditional researchers havefurther noted that there is a strong biasin our culture toward making disposi-tional attributions, which on a behav-ior-analytic view is not surprising giv-en the prevalence of mentalism. Tra-ditional researchers call this bias the"fundamental attribution error."

Related to the fundamental attribu-tion error is the "consistency paradox."That is, if the notion of a dispositionalattribution is at all valid, it implies thatthe disposition is a more or less per-manent part of a person's psychologicalmake-up and is presumably alwayspresent to cause the person to behave inmore or less the same way across timeand place. Therefore, to make a dispo-sitional attribution is to imply that aperson's behavior should be reasonablyconsistent across time and place. At is-sue is whether behavior is actually con-

sistent enough across time and place tojustify the great frequency with whichpeople tend to make dispositional attri-butions. In other words, given that peo-ple make N dispositional attributions(where N is very large), is behavior ac-tually consistent enough to justify N at-tributions, or should people be makingfewer than N? It turns out that the evi-dence suggests that behavior is not re-ally consistent enough, yet people par-adoxically persist in making N disposi-tional attributions, despite the negativeevidence. Many traditional social psy-chologists and personality theoristsspend a great deal of time seeking toresolve this paradox and understandwhy people persist in making a largenumber of dispositional attributions (seeBem & Allen, 1974; Mischel, 1999, pp.428-429). Unfortunately, it is not clearthat traditional researchers spend acomparable period of time discussingwhether any inner cause in the sense ofa dispositional attribution in fact exists,and should ever be invoked as an ex-planation of behavior. In other words,from a behavior-analytic view N shouldalways be zero. That N is greater thanzero is understandable as an illustrationof the power of mentalism in our cul-ture. When N is greater than zero, be-havior analysts argue that the verbal be-havior purportedly explaining the ob-served behavior of another person isunder the control of mentalistic precon-ceptions that interfere with an effectivenaturalistic explanation of the behaviorin question.

Implications of the Traditional View

There is a further point about dis-positional attributions that is important.This point is that disparaging disposi-tional attributions about certain classesof people tend to become a license orjustification for treating those peopledifferently based on the supposed innerentities that are taken to cause their be-havior. In short, disparaging disposi-tional attributions become a way of ex-cluding people from society rather thanintegrating them into society, and tra-

186 J. MOORE

ditional psychology contributes to thisproblem when it legitimizes such men-talisms as dispositional attributions. In-deed, from a behavior-analytic per-spective this process is at the heart ofsuch pernicious social -isms as racismand sexism.To see how dispositional attribu-

tions are related to the pernicious social-isms, let us now consider the follow-ing scenario:

1. The members of Group A regardas normative their own characteristicsand behavior (e.g., skin color, hairstyle, gender, age, place of residence,ethnic heritage, social customs andconventions, sexual orientation, facialfeatures, language, religion).

2. The members of Group A then re-gard themselves to be superior to, notjust different from, another group(Group B) whose characteristics andbehavior differ.

3. The members of Group A thenmake disparaging dispositional attri-butions about the behavior of membersof Group B.

4. The members of Group A inferthat the disparaging attributions accu-rately reflect an inferior underlyingpsychological reality of Group B.

5. The members of Group A assumethat this inferred and inferior underly-ing reality is immutable because it ismental, and that its expression in thebehavior of Group B cannot be pre-vented or modified, so why even try.

6. Finally, the members of Group Ainvoke the disparaging dispositional at-tributions as justification for engagingin discriminatory actions with respectto members of Group B, such as de-nying reinforcers or delivering punish-ing or aversive stimuli.

In sum, the discriminatory actionsmay be seen as a function of the socialreinforcement within Group A thatmaintains the behavior that controls thedelivery of consequences to Group B.If one wants to use the term power inthis sort of analysis, the term is pre-sumably occasioned by the controlover the contingencies.

CountercontrolFurther analysis suggests that groups

being disparaged sometimes try to ex-ert countercontrol in either or both oftwo ways. One is that they may arguethat their mental causes really are goodenough, and that they should not bedisparaged. A second way is that theymake equally uncomplimentary dispo-sitional attributions about the dispar-agers. Both arguments are regrettablebecause they concede the premise.Question is never raised as to whymake the dispositional attributionsabout the causes of anyone's behavior,disparaging or not. Ironically, althoughtraditional psychologists may cautionagainst the fundamental attribution er-ror in our culture, Hineline (1990, p.311) has noted that the error is itselfan example of the fundamental attri-bution error.

In any case, society does not seemlikely to follow the path to social jus-tice if it follows the practices of tradi-tional psychology and uncritically le-gitimizes dispositional attributionsabout broad classes of people. Indeed,the whole history of social oppressionfollows the making of disparaging dis-positional attributions about othergroups. Surely the Holocaust is suffi-cient illustration.

Behavior Analysis As aConstructional Alternative

Not surprisingly, behavior analystsapproach this problem entirely differ-ently. Indeed, as Skinner (1974) put itwhen talking about the uselessness ofinner causes, "This kind of thing hasbeen going on for centuries. It is sur-prising that so many intelligent peoplerefuse to ask what is wrong" (p. 170).On a behavior-analytic view, disposi-tional attributions certainly do notidentify genuine inner or mental causesof behavior because there are literallyno inner or mental causes to be iden-tified. Rather, dispositional attributionsare another type of fanciful explana-tory fiction, serving as a contiguouscause. Those who make dispositional

BEHAVIOR ANALYSIS AND SOCIAL JUSTICE 187

attributions are conforming to tradi-tional modes of explanation by virtueof the social reinforcement for actingin accord with mentalistic cultural ex-pectations. People whose behavior isbeing described in terms of a disposi-tional attribution are simply respondingto their environment. However, attri-bution theory does a very poor job ofspecifying the contingencies that gen-erate the responding with which the at-tribution is concerned. Repertoires maywell be organized, and some stimuli orclasses of stimuli may be related toothers in a way called "structural," butthere are no inner causes in the senseof dispositional attributions. Similarly,some individuals may have been ex-posed to contingencies that have pro-moted inadequate repertoires, but this'does not mean their inadequate behav-ior is caused by an entity from an innerdimension. As Hineline (1990) dis-cussed, an analysis of contingencieswould put the matter in good order. In-deed, Goldiamond (1975) pointed outthat constructional therapeutic or re-habilitative interventions are most ef-fectively based on explicit knowledgeof underlying contingencies.

INTELLIGENCELet us now consider a second ex-

ample of mentalism that is problematicin traditional psychology: the conven-tional conception of intelligence.

The Traditional View

On the traditional view, following inimportant respects from certain notionsrelating to evolution, individual differ-ences, and population genetics in thelate 1800s, intelligence is taken to beone of the naturally occurring charac-teristics of a population. It is regardedas a mental characteristic rather than aphysical or morphological characteris-tic like height or weight. As anothernaturally occurring characteristic of apopulation, it is presumed to be nor-mally distributed across the membersof the population.

This conception is reflected in many

of society's practices. For example,consider common teaching practices.Many hold that exposure to a subjectmatter builds general intelligence bygiving particular kinds of experiencesonly indirectly and imperfectly reflect-ed in repertoires. Grades should be as-signed according to a normal curve be-cause intelligence is thought to be dis-tributed according to a normal curve,and grades should follow to some ap-proximation from the amount of "in-telligence" that students possess(Moore, 2001).

Readers familiar with the history ofintelligence testing may recall that ear-ly in the first decade of the 20th cen-tury, the French government commis-sioned Alfred Binet and Theophile Si-mon to develop an objective diagnostictest that would identify children whowere unable to profit by the standardform of instruction in the publicschools, and who should be taught inspecial schools where they would re-ceive remedial instruction. The testwas reasonably successful, and the ba-sic form of Binet's test was importedto the U.S. and modified in the seconddecade of the 20th century by variousfigures of the time. One of the mostinfluential was Lewis Terman of Stan-ford University, who developed theversion we now refer to as the Stan-ford-Binet test. In the U.S., the prin-cipal use of such mental tests was inthe school system. However, two otherinterrelated uses emerged. One was totest Army recruits during the mobili-zation brought about by World War I.The stated aim was to make the Armymore efficient by assigning those pos-sessing a high amount of intelligenceto technically demanding jobs or as of-ficers and those possessing a lowamount of intelligence to menial jobs.Lewis Terman, Carl Brigham, andRobert Means Yerkes were three psy-chologists who worked on this project.A second use was to screen immigrantsto determine whether they were "fee-ble minded." The stated aim was to en-sure that the mental stock of the U.S.population was not reduced by allow-

188 J. MOORE

ing mentally defective individuals toenter the country. This aim resulted inthe Immigration Restriction Act of1924 and revitalization of the eugenicsmovement.

Group Differences in Intelligence

Analysis of the literature of the timereveals a fair amount of blatant preju-dice and racism based on the prevail-ing mentalistic views of intelligence.For example, Brigham (1923) soughtto evaluate the scores of various ethnicgroups making up the U.S. populationas revealed in the data of the Armytesting program, so that immigrationpolicy could be readjusted:

We may consider that the population of the Unit-ed States is made up of four racial elements, theNordic, Alpine, and Mediterranean races of Eu-rope, and the negro. If these four types blend inthe future into one general American type, thenit is a foregone conclusion that this future blend-ed American will be less intelligent than thepresent native born American, for the generalresults of the admixture of higher and lower or-ders of intelligence must inevitably be a meanbetween the two. ... Our own data from thearmy tests indicate clearly the intellectual su-periority of the Nordic race group. This superi-ority is confirmed by observation of this race inhistory. The Alpine race, according to our fig-ures, which are supported by historical evidence,seems to be considerably below the Nordic raceintellectually. ... Our data on the Alpine Slavshow that he is intellectually inferior to the Nor-dic, and every indication would point to a low-ering of the average intelligence of the Nordicif crossed with the Alpine Slav. ... We mustnow frankly admit the undesirable results whichwould ensue from a cross between the Nordic inthis country with the Alpine Slav, with the de-generated hybrid Mediterranean, or with the ne-gro, or from the promiscuous intermingling ofall four types. ... We must face a possibility ofracial admixture here that is infinitely worse thanthat faced by any European country today, forwe are incorporating the negro into our racialstock, while all of Europe is comparatively freefrom this taint. ... According to all evidenceavailable, then, American intelligence is declin-ing, and will proceed with an accelerating rateas the racial admixture becomes more and moreextensive. The decline of American intelligencewill be more rapid than the decline of the intel-ligence of European national groups, owing tothe presence here of the negro.... The steps thatshould be taken here to preserve or increase ourpresent intellectual capacity must of course bedictated by science and not by political expedi-

ency. Immigration should not only be restrictivebut highly selective. (pp. 205-210)

In another passage, Brigham (1923)cites approvingly the previous work ofMadison Grant, a New York lawyerwith no particular expertise in psycho-logical testing:In regard to the Irish, Mr. Madison Grant says:... [Ireland] "has more than its share of gen-eralized and low types surviving in the livingpopulations, and these types ... have imparteda distinct and very undesirable aspect to a largeportion of the inhabitants of the west and southand have greatly lowered the intellectual statusof the population as a whole. The cross betweenthese elements and the Nordic appears to be abad one, and the mental and cultural traits of theaborigines have proved to be exceedingly per-sistent and appear especially in the unstable tem-perament and the lack of coordinating and rea-soning power, so often found among the Irish."(pp. 184-185)

We previously noted that one of theleading figures in the testing movementwas Robert Means Yerkes, a specialistin comparative animal learning andperhaps the godfather of the cognitiveanimal learning movement. Considernow the following remarks from Yer-kes (1923), who also sought to pro-mote the practical success of the Armytesting program in identifying ethnicdifferences:The intellectual status of the negro is greatly in-ferior to that of the white, and the figures alreadypresented as typical are supported by measure-ments of the practical value of the negro soldierand by opinions of commanding officers, whoagree that he lacks initiative, displays little lead-ership, and cannot safely accept responsibility.

Almost as great as the intellectual differencebetween negro and white in the army are thedifferences between white racial groups.Of natives of England serving in the United

States Army only 8.7 per cent graded D or lowerin intelligence; of natives of Poland, 69.9 percent. In the English group, 19.7 per cent gradedA or B, and in the Polish group, one half of oneper cent. The race differences are so pro-nounced, and of such obvious practical signifi-cance in connection with immigration, that isseems excusable to present the rank order of theseveral racial groups. (p. 364)

To be sure, eventually Brigham(1930) distanced himself from his ear-lier analysis:That study, with its entire superstructure of ra-cial differences, collapses entirely. ... This re-

BEHAVIOR ANALYSIS AND SOCIAL JUSTICE 189

view has summarized some of the more recenttest findings which show that comparative stud-ies of various national and racial groups may notbe made with existing tests, and which show, inparticular, that one of the most pretentious ofthese comparative racial studies-the writer'sown-was without foundation. (pp. 164, 165)

However, Brigham did so more on thebasis of what he decided was an inad-equate methodology than on the rejec-tion of an ideology derived from men-talistic assumptions about intelligence.Unfortunately, the damage had beendone. Again, a stereotyped mentalisticconcern with inner causes had taken atragic toll on citizens.

Gender Differences in Intelligence

An equally common practice is toapply certain concepts of intelligenceto supposed gender differences be-tween men and women. Of course, thisargument is just one aspect of the larg-er issue, namely, the supposed geneticor biological basis of intelligence. Inthis regard, let us consider the early ap-proach of Hippocrates to personalitytheory. Hippocrates talked in terms ofa balance among the four "humors":blood, phlegm, black bile, and yellowbile. Are we to assume by virtue of thisapproach that women, because of theirmenstrual periods, must be routinelyout of intellectual balance throughblood loss? Are we to assume thatwomen must therefore be emotionallyunstable and constitutionally incapableof being intellectually rigorous? Arewe to distrust all women to perform anintellectually demanding job, insteadalways selecting any man and payinghim more because he is not so peri-odically "disadvantaged" and is al-ways on an even emotional keel? To soview men and women is the height ofmentalism, and to be a behavior ana-lyst is to be opposed to mentalism.

Shields (1982) describes anothercase of a pernicious mentalism, also re-lated to gender. In the late 19th centu-ry, many scientists adopted a set of be-liefs known as the "variability hypoth-esis." According to this hypothesis,males were more variable in their men-

tal as well as physical characteristicsthan were females. Thus, the distribu-tion of characteristics for males wouldbe flatter, with a greater number at thehigh and low extremes but fewer in thecenter of the distribution, than that forfemales. The implication was that thepercentage and number of smart maleswas higher than that of smart females,and so it should not be surprising thatmore males than females were in po-sitions that demand intelligence. Thedifferences between males and femaleswere presumed to be inherent, which isto say biologically based. Consequent-ly, a biological model of intellectualheritage rose to prominence, and manyluminaries of the time subscribed tothis model. Darwin himself had regard-ed males as intellectually superior tofemales-more courageous, pugna-cious, and energetic, with a more in-ventive genius, and possessing morequalities that were necessary in thestruggle for life (Shields, 1982, p. 772).Thorndike suggested that the educa-tional system should channel womeninto fields in which only modest levelsof ability were necessary:

Not only the probability and the desirability ofmarriage and the training of children as an es-sential feature of a woman's career, but also therestriction of women to the mediocre grades ofability and achievement should be reckoned withby our educational systems. The education ofwomen for ... professions ... where a very fewgifted individuals are what society requires, isfar less needed than for such professions as nurs-ing, teaching, medicine, or architecture, where[only] the average level is essential. (as cited inShields, 1982, p. 782)

Despite the pioneering work of HelenBradford Thompson (Wooley) andLeta Stetter Hollingworth, these sortsof myths persisted, and have been thefoundation of many mentalistic treat-ments of supposed inherent gender dif-ferences.The examples cited to this point

have largely been historical. Yet, wehave only to look at contemporary psy-chology to see treatments of brainfunction that seek to localize specificmodular "cognitive abilities" in cer-tain regions of the brain. Cognitive

190 J. MOORE

neuroscience specializes in such treat-ments, or "chasing ghosts with Geigercounters" as Faux (2002) has wonder-fully put it. Executive functions are inthe prefrontal cortex; speech centersare over here, in this hemisphere orlobe; face recognition centers are overthere; spatial imagery centers are uphere; creative thought is down there;logic and analytic ability are in be-tween and slightly to the left; poetry isup and to the right; memories belowthem in the amygdala or hippocampus;and so on. Suffice it to say that the fac-tual basis for such claims is clearlysuspect (Faux, 2002).

Implications of the Traditional View

As seen in the passages from Brig-ham (1923) and Yerkes (1923) citedabove, attributions about intelligenceas a mental power or ability parallelthe various dispositional attributionsabout behavior. An implication of amentalistic view of intelligence is thatit is acceptable to treat others prejudi-cially because they are not worthy ofany other kind of treatment. If theyaren't smart enough, it must be becausethey lack the genetic or other disposi-tional characteristics to ever be smartenough. Accordingly, why bother toinvest in them by giving them a decenteducation, job, housing, or medicalcare? Why cast pearls before swine?

For present purposes, are we obligedto assume that if some women do notperform certain tasks as well as somemen, it must be because the neurons ofthese women are defective, and allwomen should therefore be kept bare-foot and pregnant in the kitchen? Arewe obliged to assume that if somemembers of Group B do not performcertain tasks as well as some membersof Group A, it must be because theneurons of all members of Group B aredefective, and they should be deniedreinforcers? To so approach these ques-tions is the height of mentalism, and tobe a behavior analyst is to be opposedto mentalism.

Behavior Analysis As aConstructional Alternative

On a behavior-analytic view, intel-ligence is yet another mentalistic con-tiguous cause, a fanciful explanatoryfiction that is cherished for irrelevantand extraneous reasons. Our verbalpractices and cultural assumptions leadus to say a person does something in-telligently, then does something thatshows intelligence, and finally that theperson has intelligence. What started asan adverb becomes a noun, and peoplego off looking in another dimensionfor the thing the noun is said to rep-resent.

Clearly, questions about intelligenceare questions about (a) the presenceand absence of certain forms of stim-ulus control and (b) organizations ofrepertoires. People do differ, just asother organisms differ. For example,they have different genetic endow-ments that underlie different sensitivi-ties to environmental stimulation. Thepoint is that contingencies give rise tostructure within and between aspects ofrepertoires, and on a behavior-analyticview there is no justification for infer-ring that any resulting structure withinand between repertoires is the result ofa monolithic causal entity in anotherdimension. Certainly some life experi-ences add or detract in specific or gen-eralized ways to the development ofstimulus control and the organizationof repertoires, but culturally basedmentalistic assumptions only interferewith constructional therapeutic or re-habilitative intervention (see Goldia-mond, 1975). Behavior analysis doesunderstand that repertoires obviouslydiffer among individuals and that therepertoires of some individuals arewell less than optimal. No doubt thaton average the repertoires of those inthe higher socioeconomic classes orwho earn higher annual incomes aremore developed in a descriptive sensethan those in lower classes or who earnlower incomes. However, behavioranalysis advocates socially construc-tional intervention to prevent deficient

BEHAVIOR ANALYSIS AND SOCIAL JUSTICE 191

repertoires in the first place or rehabil-itate them if they already exist. It re-jects statements that give license to dis-criminatory treatment of the individualon the basis of some presumed dispo-sitional quality or entity that is thoughtto cause the deficient repertoires.Again, it seems unlikely that the ap-propriate interventions will be as ro-bust as they need to be, and that wewill progress down the path of socialjustice, if we continue to appeal to tra-ditional, mentalistic definitions of in-telligence (cf. Herrnstein & Murray,1994).To be sure, certain bodily activities

are accomplished by some anatomicalstructures and not others. Persons whoexperience strokes or other forms ofbrain injuries may not be able to per-form certain classes of responses, suchas those on intelligence tests, by virtueof the localized damage they sustain tothose structures. The point is that we(our nervous systems) are changed bythe contingencies we experience, andwe (our nervous systems) "store"these changes. Well, we (our nervoussystems) must store them somewhereand somehow. If these stored changesare selectively altered or otherwise dis-rupted, the behavior mediated by thisunderlying physiology may well be se-lectively altered or disrupted. Re-searchers can knock out genes in lab-oratory animals and inhibit proteinsynthesis, so that the experiences donot actually "change" the nervous sys-tem of the animals in a way that thechanges are stored. However, all of thisis a long way from the mentalistic stor-age and retrieval metaphors so uncrit-ically taken from folk psychology andinstitutionalized in contemporary tra-ditional psychology.

In addition, the 23rd pair of chro-mosomes in the cells of males presum-ably has an XY configuration, and thisconfiguration differs from the XX con-figuration of females. The resultingbalance of male and female sexual hor-mones in the bodies of these malesmeans their hypothalamus and nodoubt other regions of their brains dif-

fer from those females. Presumably,those males will also react in differentways to some stimuli than do females.

Again, none of this implies thatthere is an underlying "cognitive abil-ity" that resides in one or another brainregion, operated by one or another ho-munculus; that this mental or cognitiveability is genetically determined; orthat this approach of traditional, men-talistic folk psychology to the questionof human behavior is in any way rea-sonable. As Skinner (1974) put it,"The behavioral account ... sets thetask for the physiologist. Mentalism,on the other hand, has done a great dis-service by leading physiologists onfalse trails in search of the neural cor-relates of images, memories, con-sciousness, and so on" (p. 217).

The Pygmalion Effect

Relevant to the present discussionare studies on the "Pygmalion effect"by Rosenthal and Jacobson (1968).These studies examined how teachers'"expectations" influenced their behav-ior toward children. In these studies,children in an elementary school weregiven a nonverbal intelligence test sup-posedly designed to predict academicblooming or intellectual gain. For arandomly selected 20% of the students,something like a deception was in ef-fect, such that their teachers were toldthat the test indicated the studentswould show unusual intellectual gainsduring the academic year. Eightmonths later, the children were retest-ed. The 20% of the children who wereselected at random did show a signifi-cantly greater gain on the test than didthe children in a control group. Pre-sumably, the teachers translated theirexpectations into behavior that led thestudents to do better. Indeed, follow-upstudies by Meichenbaum, Bowers, andRoss (1968) and Rubovits and Maehr(1973) showed that students benefitwhen their teachers increase their pos-itive interactions and praise studentsmore than another group of studentsserving as controls. The point is, why

192 J. MOORE

can't all students be treated like thosein the experimental group so that allstudents can show gains? Yet, a legit-imate question is whether these con-structional interventions will even becontemplated if one is committed tothe view that intelligence is largelysome kind of an innate mental char-acteristic that is normally distributedwithin a population and one can dovery little about who ends up with howmuch in the distribution.

SUMMARY ANDCONCLUSIONS

In conclusion, we note that towardthe end of the 19th century, John StuartMill suggested that "Of all the vulgarmodes of escaping from the consider-ation of the effect of social and moralinfluences upon the human mind, themost vulgar is that of attributing thediversities of conduct and character toinherent natural differences" (as citedin Gould, 1977, p. 247). Now, morethan 100 years later, a reasonable ques-tion is whether much has changed. Un-fortunately, perhaps even tragically, itappears not. Even the most casual in-spection reveals traditional psychologyhas uncritically institutionalized thementalism of inner causes that comesfrom folk psychology. Such mental-isms as found in dispositional attribu-tions and conventional conceptions ofintelligence have led us to the perni-cious social -isms of racism and sex-ism. These views interfere with bring-ing all members of society into contactwith effective, constructional interven-tions on the path to social justice. In-deed, there is no assurance that our cul-ture as a whole can progress far or ef-fectively down the path of social jus-tice without ridding itself ofmentalism, and it is not clear that tra-ditional psychology is up to the task.

However, for our society to progressmost effectively down the path of so-cial justice, behavior analysts must alsounderstand clearly what mentalism is,what problems it creates, and how be-havior analysis represents an alterna-

tive, so that we ourselves do not con-tribute to those problems. This require-ment applies to all aspects of our ac-tivity, but especially to the scientificbehavior we call theorizing and ex-plaining. For example, if we cannot an-alyze our own behavior as scientists-our theorizing and explaining-with-out an implicit commitment to mental-ism and epistemological dualism, whatchance do we have of getting others toanalyze in a nonmentalistic way anykind of behavior? Day (1969) raisedthis matter some years ago:

Strange blends of Skinner and conventional be-haviorism abound.... Mentalism among Skin-nerians is rampant.... I have taken the libertyof speaking here directly to some of those whopreach most loudly a supposedly Skinnerianline. One hardly knows where to begin to ana-lyze the grossly uninformed verbal material thatis generated concerning Skinner's work by thetypical psychologist. (pp. 326-327)

Let us be clear: Our uncritical talkabout scientific epistemology (and weshould not forget that Skinner came tobehaviorism because of its bearing onepistemology; Skinner, 1978, p. 124) isa factor that contributes directly to themaintenance of mentalism, and if it iscontributing to mentalism, it is contrib-uting indirectly to racism and sexism.Presumably, much of the activitycalled mentalistic is supported in oneway or another by social reinforce-ment, rather than because it yieldsmore effective prediction and controlof behavior. The good news is that ifmentalism is just a product of socialreinforcement, then presumably we canprovide social reinforcement for non-mentalism, and mentalism will be ex-tinguished. That state of affairs will ul-timately leave us in a better position togo out and actually do the better sci-ence that enhances the welfare of hu-manity.

Consider once again some wordsfrom Skinner:Do I mean to say that Plato never discovered themind? Or that Acquinas, Descartes, Locke, andKant were preoccupied with incidental, often ir-relevant by-products of human behavior? Or thatthe mental laws of physiological psychologistslike Wundt, or the stream of consciousness of

BEHAVIOR ANALYSIS AND SOCIAL JUSTICE 193

William James, or the mental apparatus of Sig-mund Freud have no useful place in the under-standing of human behavior? Yes, I do. And Iput the matter strongly because, if we are tosolve the problems that face us in the world to-day, this concern for mental life must no longerdivert our attention from the environmental con-ditions of which human behavior is a function.(1978, p. 51)

Clearly, an authentically scientific viewof human nature offers exciting possi-bilities for humans to achieve their fullpotential. We have not yet seen whatwe can make of ourselves, perhaps be-cause behavioral science is not yet be-havioristic enough (Skinner, 1971, p.215; 1974, p. 257). In principle, as be-havioral science does become more be-havioristic, we can learn more aboutmaking ourselves better readers, writ-ers, citizens, and parents. We can learnmore about preventing and rehabilitat-ing inadequate repertoires. We canlearn more about conserving our nat-ural resources, maintaining the integ-rity of the environment, managing pop-ulation growth, developing our econo-mies in sustainable ways, preservingthe dignity of human capital, and al-locating resources to human needs in-stead of weapons of mass destruction.Given such a rich view of human na-ture, the prospects that all members ofour society can walk together on thepath to social justice are promising in-deed.

REFERENCES

Bem, D., & Allen, A. (1974). On predictingsome of the people some of the time: Thesearch for cross-situational consistencies. Psy-chological Review, 81, 506-520.

Brigham, C. (1923). A study ofAmerican intel-ligence. Princeton, NJ: Princeton UniversityPress.

Brigham, C. (1930). Intelligence testing of im-migrant groups. Psychological Review, 37,158-165.

Day, W. F, Jr. (1969). Radical behaviorism inreconciliation with phenomenology. Journalof the Experimental Analysis of Behavior, 12,315-328.

Day, W. F, Jr. (1976). The case for behaviorism.In M. H. Marx & F E. Goodson (Eds.), The-ories in contemporary psychology (2nd ed.,pp. 534-545). New York: Macmillan.

Faux, S. F (2002). Cognitive neuroscience from

a behavioral perspective: A critique of chasingghosts with Geiger counters. The BehaviorAnalyst, 25, 161-173.

Goldiamond, I. (1975). Insider-outsider prob-lems: A constructional approach. Rehabilita-tion Psychology, 22, 103-116.

Gould, S. J. (1977). Ever since Darwin. NewYork: Norton.

Heider, F (1958). The psychology of interper-sonal relations. New York: Wiley.

Herrnstein, R. J., & Murray, C. (1994). The bellcurve. New York: Free Press.

Hineline, P N. (1990). The origins of environ-ment-based psychological theory. Journal ofthe Experimental Analysis of Behavior, 53,305-320.

Kelley, H. H. (1967). Attribution theory in so-cial psychology. In D. Levine (Ed.), NebraskaSymposium on Motivation (Vol. 15, pp. 192-238). Lincoln: University of Nebraska Press.

Meichenbaum, D. H., Bowers, K. S., & Ross, R.R. (1968). Modification of classroom behav-ior of institutionalized female adolescent of-fenders. Behaviour Research and Therapy, 6,343-353.

Mischel, W. (1999). Introduction to personality(6th ed.). Fort Worth, TX: Harcourt Brace.

Moore, J. (1981). On mentalism, methodologi-cal behaviorism, and radical behaviorism. Be-haviorism, 9, 55-77.

Moore, J. (1990). On mentalism, privacy, andbehaviorism. Journal of Mind and Behavior,11, 19-36.

Moore, J. (1999). The basic principles of be-haviorism. In B. Thyer (Ed.), The philosoph-ical legacy of behaviorism (pp. 41-68). Lon-don: Kluwer.

Moore, J. (2000). Varieties of scientific expla-nation. The Behavior Analyst, 23, 173-190.

Moore, J. (2001). On certain assumptions un-derlying contemporary educational practices.Behavior and Social Issues, 11, 49-64.

Rosenthal, R., & Jacobson, L. (1968). Pygma-lion in the classroom. New York: Holt, Rine-hart, and Winston.

Rubovits, P C., & Maehr, M. L. (1973). Pyg-malion black and white. Journal of Person-ality and Social Psychology, 25, 210-218.

Shields, S. A. (1982). The variability hypothe-sis: The history of a biological model of sexdifferences in intelligence. Signs, 7, 769-797.

Skinner, B. F (1945). The operational analysisof psychological terms. Psychological Review,52, 270-277, 291-294.

Skinner, B. F (1971). Beyondfreedom and dig-nity. New York: Knopf.

Skinner, B. F (1974). About behaviorism. NewYork: Knopf.

Skinner, B. F (1978). Reflections on behavior-ism and society. Englewood Cliffs, NJ: Pren-tice Hall.

Smith, L. D. (1986). Behaviorism and logicalpositivism. Stanford, CA: Stanford UniversityPress.

Yerkes, R. M. (1923). Testing the human mind.Atlantic Monthly, 131, 358-370.