sign in sign up
<<
Home , Cultural anthropology, Cultural system, Culture, Enculturation, Marvin Harris

The Behavior Analyst 2007, 30, 37–47 No. 1 (Spring)

Cultural and Behavior Analysis: An Introduction to Harris Brian D. Kangas of

The year 2007 marks the 80th anniversary of the birth of (1927–2001). Although relations between Harris’ cultural materialism and Skinner’s radical have been promulgated by several in the behavior-analytic (e.g., Glenn, 1988; Malagodi & Jackson, 1989; Vargas, 1985), Harris himself never published an exclusive and comprehensive on the relations between the two . However, on May 23rd, 1986, he gave an invited address on this topic at the 12th annual conference of the Association for Behavior Analysis in Milwaukee, Wisconsin, entitled Cultural Materialism and Behavior Analysis: Common Problems and Radical Solutions. What follows is the publication of a transcribed audio recording of the invited address that Harris gave to Sigrid Glenn shortly after the conference. The identity of the scribe is unknown, but it has been printed as it was written, with the addendum of embedded references where appropriate. It is offered both as what should prove to be a useful asset for the students of behavior who are interested in the studyofcultural contingencies, practices, and epistemologies, and in commemoration of this 80th anniversary. Key words: cultural materialism, radical behaviorism, behavior analysis

Cultural Materialism and Behavior Analysis: Common Problems and Radical Solutions Marvin Harris

Cultural materialism is a in rejection of mind as a cause of which shares many episte- individual behavior, radical mological and theoretical principles behaviorism is not radically behav- with radical behaviorism. Like radi- iorist enough for a confrontation cal behaviorism, it stresses environ- with idealists’ anthropological para- ment over heredity, in opposition to digm, in spite of the fact that cultural biological reductionist . idealists identify mind as the locus of And like radical behaviorism, it sociocultural and causality. stands in opposition to psychological The cause of this disparity lies in paradigms such as cognitivism, which the radical behaviorist’s neglect of the stress mind as the cause of behavior. systemic of cultural phe- It is not clear, however, to what nomena and a consequent failure to extent cultural materialism and rad- identify the selective core of the ical behaviorism have a comparable sociocultural and how that position vis-a`-vis the anthropological selection takes place. analogue of cognitivism, which I call In the of the development cultural idealism, for reasons that will of paradigmatic attempts to solve the become clear later on. question of sociocultural causality, In fact, it is my contention that cultural materialism has confronted while cultural materialism and radi- and continues to confront a of cal behaviorism have a common basis long-established, common-sense but erroneous and unproductive beliefs Reprints may be obtained from Brian D. about human that are similar Kangas, Department of , Univer- sity of Florida, P.O. Box 112250, Gainesville, to those which confronted and con- Florida 32611-2250 (e-mail: [email protected]). tinue to confront the development of

37 38 MARVIN HARRIS radical behaviorism. Behaviorists gous to differential reproductive suc- have had to struggle against the cess on the biological level, common- that organisms, especially , ly known as natural selection, and possess faculties called mind and will to operant conditioning on the psy- whose exercise makes their behavior chological level, one must first have largely unpredictable. While cultural a clear formulation of what culture idealists generally also have recourse is and how we know what it is. to will, intention, goals, and other From a cultural materialist perspec- mentalist notions about the causes of tive, culture is the aggregate of behavior, it must have been recog- classes, with classes underlined, of so- nized that at least in the short run, cially conditioned, operant responses, is largely predict- which are associated with a particular able. This follows from their com- human group and which tend to be mitment to the study of a cultural replicated within and across genera- level of phenomena. For example, tions. The reason for laying stress on most cultural idealists would have no the word classes is twofold: On the trouble with the idea of predicting one hand, many individuals may that children born in a certain cate- exhibit the same class of behavior, gory of South India will and the class itself often continues to grow up to speak Malayalam, eat exist even after many of the individ- with their fingers, and take their uals who exhibit the characteristic meals seated cross-legged on the behavior have died. Human floor, just as they would have no are a distinctive, emergent - difficulty with the prediction that ary phenomenon. However, there are children in U.S. households would many rudimentary examples of cul- grow up to speak English, eat with tural inventory among primates and metal or plastic utensils, and take among primitive protohumans which their meals while seated on a chair. enable us to reconstruct the pathway Indeed, it is the predictable replica- that led to the fully evolved cultural tion of such behavior through the systems that have characterized hu- practice known as groups for at least the last which has given rise to the cultural 4,000 years. field of inquiry. But cultural idealists Classes of culturally condition- have typical psychological, cogniti- ed operants are conventionally iden- vist views concerning how encultura- tified by terms such as , tion takes place—that is, they have collecting, warfare, feuding, polygy- attributed enculturation to intention- nous , bride wealth, uxor- al, willful instruction or instinctual ilocality, reciprocal exchange, feast- imitation of one person or one ing, headhunting, cannibalism, chief- generation by another. doms, feudal states, capitalist firms, The fulcrum of cultural idealists’ and multinational corporations—and resistance to the development of so forth. These components, conven- a of culture lies elsewhere. tionally called traits, customs, tradi- Specifically, it lies in the problem tions, , and so forth, do of why the acculturated behavioral not exist in isolation from each other. repertoires of different human groups Rather, they present preconditions are different in some respects and for each others’ characteristic activi- similar in others. Or more precisely, ties and consequences. For example, it lies in the problem of how to no feasting without hunting, no explain the divergent, convergent, hunting without hunters, no hunters and parallel evolution of sociocultur- without spears, and so forth. More- al systems. over, much empirical data indicate In order to specify the cultural that most if not all of the classes of selection principle that is analo- culturally conditioned operant re- CULTURAL MATERIALISM AND BEHAVIOR ANALYSIS 39 sponse chains are related to each of selection by consequences. In the other in the sense that changes in most general terms, new behavioral one are frequently followed by repertoires arise as operants in the changes in others. Thus, intrinsic to behavior of individuals. Some of the emergence of a cultural level of these get propagated within and phenomena is the arrangement of transmitted across generations, and cultural components in subsystems they are selected for. Other operants characteristic of particular , do not get propagated, or transmit- that is, sociocultural systems. ted. They are extinguished or selected Let me emphasize that the necessi- against. ty for dealing with these classes of Note that the existence of selection individual responses and sociocultur- for or against particular classes of al systems as a level of phenomena operants implies that not all reinfor- with a distinct set of selection pro- cing contingencies are equally de- cesses and lawful regularities arises terminative of cultural repertoires. not from any special emergent es- In other words, the issue is not simply sences peculiar to human cultural life, between an operant and uncondi- such as mind, will, intention, con- tioned behavior, but between one sciousness, symbolic thought, or even operant and another operant, each verbal behavior, as the majority of of which is controlled by reinforcing my anthropological colleagues con- contingencies. The only way the tend, but rather from the quantitative selection of one operant over another enormity of the numbers of individ- can be accounted for in cultural uals and responses that underwrite evolution without invoking mentalist that abstraction and classification of and teleological hypotheses is if their sociocultural components and all relative efficacy in achieving a partic- systems. When we speak, for exam- ular effect is different. Thus we are ple, of the nuclear as a compo- led, step by step, to the depiction of nent of U.S. culture, we are grouping cultural selection as a process that together, averaging, or abstracting selects alternative operants with re- from behavioral repertoires of mil- spect to particular effects or func- lions of specific individuals, dead as tions. well as alive, who characteristically What kind of function? An infinite emit or emitted millions of specific number of functions can be fulfilled responses concerned with nurturance by operants. Therefore, further prog- of children, coresidence, sexual ac- ress for specifying principles of cul- cess, and division of labor. tural selection requires delimitation In rejecting mentalist and essen- and classification of functions that tialist definitions at the cultural level, are served by cultural repertoires. cultural materialism is bound inextri- Several behavior analysts have cably to the radical behaviorist pro- suggested recently that the principal gram, for despite the enormous function governing cultural selec- quantitative escalation and complex- tion is cultural survival. And they ity, Skinner [1984] is absolutely cor- have specified three derivatives: (1) rect when he states that no new group cohesion, (2) effective action behavioral principles are needed to vis-a`-vis the physical environment, deal with phenomena on the cultural and (3) effective action vis-a`-vis other level. But this merely takes us to the groups. I believe that this formula- threshold of the problem, the prob- tion is inadequate. It is inadequate lem being how to account for the because the vast majority of , not of individuals alone, have been selected for by but of the behavior of the cultural consequences that did not evidence . As Skinner has suggested, any greater or lesser degree of pro- is an instance moting survival of the culture at the 40 MARVIN HARRIS time they were selected for. The test that behavior. To conform to the of whether it is the survival-promot- pattern of scientific parsimony, the ing effect of an that has number of such innate susceptibilities been selected for cannot be the should be restricted to susceptibilities survival of the culture—for then all or needs that have been operationally the components of any surviving defined through their power to con- culture would by definition have this dition responses. Hunger, sex, body characteristic, and the proposition contact, including protection from would be unacceptable. This is exact- disease—may not constitute an ex- ly the same tautology which plagued haustive list, but it is sufficient to earlier generations of Darwinists with define an array of basic functions regard to the principle called survival whose fulfillment by operants are of the fittest. Moreover, while capi- subject to cultural selection on the talization and political-military strat- basis of relative effectiveness. egy might be selected for their Admittedly, the identification, clas- immediate group survival effects, it sification, and measurement of innate is hard to see how to relate the same human needs is a weak link in the consequences to innovation in such logical and empirical scaffolding of matters as , the , and a behavioristically consistent science . of culture. I ask my radical behav- Finally, innovations of all sorts iorist friends present here, however, could go on accumulating for thou- to reflect on why this is the case. sands of years before their effects on Hopefully, as the interest in a consis- group survival would be subject to tent radical behavior approach to selection. In the meantime, it would cultural phenomena increases, new require a prescience to track the experimental data concerning Skin- effective consequences for survival, ner’s innate susceptibilities to rein- a prescience which no culture has at forcement will fill the gaps in our the present and has never possessed. current . It is a set of If we are to specify sociocultural basic, natural susceptibilities or functions as the measure of the needs that underwrites the selection differential effectiveness to the group of all the components and subcom- that are fulfilled by particular cultur- ponents of a particular culture. But al innovations, we must first insist this does not mean that for each that human beings are organisms need, there is a discrete component which have certain physical, chemi- or subsystem that is exclusively cal, biological, sociological, and psy- conditioned by the contingencies chological needs. Regardless of what corresponding to that need. During else cultures may do, they must the enculturation of, say, the Hindu service these needs. Indeed, the selec- aversion to the slaughter of cattle tion of the human capacity for the and the consumption of beef, appro- elaboration of cultural repertoires by priate behaviors are not shaped natural selection is explicable only as exclusively by reinforcement with selection consequent on the greater , nor during the enculturation efficiency of cultural over noncultural of polygynous marital customs are means of satisfying the requirements the appropriate behaviors shaped of individuals, not cultural survival exclusively by sexual reinforcement. and reproduction. This lack of direct relationship be- The importance of needs for ex- tween needs and contingencies of plaining sociocultural selection by reinforcement is compensated for consequences is suggested by Skin- by the interacting and interdepen- ner’s [1966] reference to the acquisi- dent relationship of the major com- tion of behavior because of innate ponents of sociocultural systems and susceptibilities to reinforcement for subsistence. CULTURAL MATERIALISM AND BEHAVIOR ANALYSIS 41

Before I describe these compo- ior. I agree with Sigrid Glenn that nents, the problem of defining behav- human speech functions to establish ior in a manner suitable for sociocul- and maintain relations between the tural analysis and distinct from receptors of some members and the behavior analysis must be broached. effectors of other members of the Needless to say, it has been of verbal community, thereby increasing considerable importance in the histo- the range of effective action for all ry of behavior analysis to establish members. But we must not lose sight the point that verbal behavior is of the considerable in psy- conditioned by the same processes chology, , and anthropolo- that condition other operants. And gy, which I cite here, which shows I’ve also been informed to emphasize that mands more often than not are the fact that the operational defini- followed by noncompliant responses, tion of thought is silent verbal and that tacts more often than not behavior, and that therefore thought present an inaccurate picture of what is also subject to the same behavior they purport to describe. analysis as are other forms of behav- Contra prevailing opinion, the ior. of verbal behavior in the evolution of From a cultural materialist per- human cultures can easily be over- spective, however, the most impor- emphasized. Indeed, such overesti- tant characteristic of thought is not mation is an intrinsic defect of both that it is verbal behavior but that it is psychological, cognitive, and cultural behavior which has no measurable idealism, for whom (the cultural effectivity apart from the external idealists) quite literally, in the begin- behavior that it may or may not ning there was the word—whereas, accompany or initiate. On the other quite literally, for cultural material- hand, external verbal behavior, pro- ists, in the beginning, there was the ducing mands and tacts, as well as foot—a proposition empirically con- most nonverbal behavior, move with, firmed by the last fifty years of embody, and move or transform human paleontology and the archae- aspects of the environment and ology of the lower Paleolithic. therefore have measurable effectivity Let me mention one more item in the form of costs and benefits before I link the components of independent of internal behavior. sociocultural systems with the inven- If the task of sociocultural analysis tory of the naturally determined involves the identification of the human needs. The challenge before cost–benefit consequences of alterna- us is to measure the costs and benefits tive behavioral repertoires, it makes of alternative ways of satisfying such little sense to treat the internal verbal needs. To do this we can employ behavior, which has no independent various currencies, such as protein, environmental consequences, as if it time, morbidity, or the failure rate, were as significant for cultural selec- life expectancy, sexual events, and in tion as environmentally consequen- some instances dollars and cents. tial forms of behavior. To do so, as I There are many methodological shall emphasize later on, is to slip problems associated with these mea- inadvertently toward the form of surements—what is especially baf- mentalism which is characteristic of fling is how to combine different cultural idealism. currencies in arriving at a final bal- Classes of mands and tacts, on the ance. An equally baffling methodo- other hand, are present throughout logical problem is whether one all major components and subsys- should rank or weight the cost– tems of sociocultural systems. They benefits balances that pertain to have an instrumental status in initi- different needs. Too clearly, this is ating and organizing effective behav- another area in which must 42 MARVIN HARRIS wait upon experimental evidence that tion of painting, sculpture, literature, radical behaviorists might well con- and of course religion, scientific and sider proper to their own expertise. educational edifices and apparatus, Returning now to the task of together with the associated behav- specifying the major metacompo- ioral performances. The characteris- nents of sociocultural systems, from tic which superstructural products a cultural materialist perspective, it is and activities share in common is clear that if basic naturally deter- that they are more remotely related mined needs of human beings are to to the satisfaction of basic naturally be served, it will be a set of cultural determined needs than the structural components concerned with the pro- components, and still more remotely duction of indispensable material related to them than the infrastruc- goods and services, such as food, tural components. shelter, , , and med- Now it is time to return to the ical care, in brief, a subsistence econ- internal, verbal behavior which ac- omy. For example, hunting, collect- companies most human responses. In ing, , irrigation agricul- each of the external behavioral sub- ture, , etc. systems are components that are sets The quantities and qualities of of elicitable instructions, rules, gram- goods produced by these classes of mars, guides, maps, goals, intentions, behavior, however, will always be explanations, rationalizations, values, a function of a set of technological , and other internal ver- components that are available for the bal classes of responses. extraction, transformation, and trans- Thus there is a cognitive infra- portation of natural resources—that structure, a cognitive structure, and is, a culture’s —which in a cognitive superstructure. For ex- turn is a function of the particular ample, rules for constructing har- resources and other conditions of the poons, for planting maize, or pro- specific environment available for gramming a computer are examples exploitation by that culture. of cognitive infrastructural compo- And finally, human sexual and nents. The rules for divorcing your economic needs almost predispose wife in Saudi Arabia or protocol for sociocultural systems to encourage kow-towing before a Chinese emper- the production of children … [gap or, or the philosophical justification between Side 1 and Side 2 of the for free enterprise are cognitive struc- audio recording] … power that are tural components, while the religious socially available. justification for building the Gothic Two great classes of institutions cathedral, or, for a Hopi priest, rules satisfy this function—domestic econ- for praying for rain, are examples of omy and . Domes- superstructural components. tic economic behavior centers on The sum of all these cognitive domiciles where feeding, nurturing, components constitutes a sociocultur- and reproduction take place. Political al system’s ideology. Bear in mind economic behavior takes place out- that all of the behaviors and sub- side of domiciles and regulates rela- systems inside as well as outside the tions between domiciles, and between skin are viewed here as classes of nondomestic groups, and between operants. If you can accept, pro- one population and another. Togeth- visionally at least, the taxonomic er domestic and political economy decisions that I’ve been making here, constitute what the cultural materi- I can now state the issue in cultural alists call the structure of sociocul- that is analogous to tural systems. the opposition between behaviorism There is a third subsystem, the and cognitivism on the psychological superstructure, involving the produc- level. (I’m sorry to have put you CULTURAL MATERIALISM AND BEHAVIOR ANALYSIS 43 through all that , but it’s er viewpoints and its relativist entail- the only way I know of to be clear ment by using the linguistically based concerning the fact that what we’re concepts of as modifiers talking about here are not homolo- of mental and behavioral events, but gous oppositions but analogous I shall not burden you with the ones.) details of the tribal wars now being The issue is not the operant status fought over this distinction. of mental behavior or thought— Suffice it to say that cultural rather, it is the relative importance materialists posit objective knowl- of behavioral versus ideological in- edge as a totally approachable form novations in , and of behavior potentially available to that is why the analogue of psycho- human beings regardless of cultural logical cognitivism is best described differences. But to come to the most as cultural idealism, and not as relevant point for behavior analysts, cultural cognitivism. The cultural cultural materialism holds that the idealist position is that sociocultural direction of sociocultural evolution is evolution is directed by ideological probabilistically determined by the innovation—I refrain from saying consequences of behavioral innova- ‘‘determined’’ for reasons to be given tions with a cost–benefit for pro- in a moment. duction and reproduction—the prob- Cultural idealists posit new attrac- ability that an innovation which tive thoughts arising for some reason arises from the infrastructure or the within the minds of certain individu- structure or superstructure or ideo- als, which are propagated and trans- logical subsystem will be propagated mitted. These thoughts become so- or transmitted is determined by called templates for behavior, which whether or not it results in a more is therefore allegedly modified to favorable or less favorable balance of correspond with the revised ideology. productive and reproductive costs as Indeed, so strong is this commit- measured by the varying currencies ment to the priority of thought over previously mentioned. behavior that culture itself, in the Inventions which have adverse prevailing idealists’ paradigm associ- infrastructural cost–benefit conse- ated with anthropological eminences quences will be selected against re- such as Levi-Strauss, Marshall Sah- gardless of where they arise in the lins, and , is regarded as system. A corollary of this principle is nothing but what I am here calling that changes in the external behav- ideology. But I only scratch the ioral infrastructure probabilistically surface of mentalism as it now reigns determine both the internal and in contemporary anthropology. Not external components of structure content with spontaneous ideas as and superstructure. directives of cultural evolution, cul- Let me make it clear that the cost– tural idealists subscribe to the further benefits which determine selection for complication of extreme relativism, or against a particular innovation of which I do not think there is need not have the same balance for a parallel for among psychological all the participants in a given socio- cognitivists. This converts external . Where sexual, age, behavior itself to nothing but what class, , and ethnic hierarchies a culture’s participants think it to be, exist, innovations may be selected for thereby canceling Western civiliza- on the grounds of their favorable tion’s 3,000-year-old quest for objec- cost–benefit consequences for the tive knowledge about human behav- superordinate individual regardless ior and sociocultural systems. of the consequences for subordinate I’ve attempted to deal with the individuals. This does not mean, as problem of participant versus observ- some dialectical materialists insist, 44 MARVIN HARRIS that every innovation selected for in reproductive and productive activi- class-stratified necessarily has ties of the Iranian people, the infra- adverse consequences for the sub- structural conditions for the over- ordinate. Optimization does not throw of the Shah were created mean optimal, even for ruling classes. during the prior period of Colonialist I should also hasten to assure you struggle over control for Iran’s oil that the principle of infrastructural resources. determinism allows for, indeed de- If and religion were really mands, the recognition of causally in command in evolutionary perspec- powerful but not dominant feedback tive, we should never be able to between the other components of explain why band organizations per- sociocultural systems and infrastruc- sisted for hundreds of thousands of ture. years, why villages were a rarity until All aspects of sociocultural systems 8000 B.C., and why the first states are causally linked to each other, but did not make their appearance until the links are not symmetrical. This a mere 6,000 years ago. One might qualification emerges from the dis- argue that these dates correspond to tinctive comparative and natural his- the amount of time needed for the torical evolutionary perspective of maturation of the ideas which were and anthropo- appropriate for the transformations logical , and it stands in question, but the fact is that the opposed to paradigms within anthro- transformations did not occur every- pology and in other disciplines which where; bands, villages, and agrarian ignore the long view. When sociocul- states linger on in some locales till tural differences and similarities are this very day; and it is both their examined synchronically (single time presence and their absence in specific span), causal relationships rapidly places as well as specific times that dissolve into an incoherent corpus needs to be explained. This cannot be of middle-range eclectic correlations done nomothetically without specify- linking infrastructural, structural, su- ing recurrent conditions that are perstructural, and ideologic compo- external to structure and superstruc- nents in infinite arrays. It then ture, conditions that lie nowhere else appears empirically demonstrable but along the constants and variables that there are no overall asymmetries of infrastructure, for reasons that I in the causal nexus. This error cannot shall expand on in a moment. be resolved simply by adding short- This present occasion is too brief term time depth—indeed, depen- to allow me a full-scale defense of the dence on short time frames com- principles of infrastructural determin- pounds the problem and leads to ism. Such a defense would rest on the impression that at certain histor- logical and empirical grounds and on ical moments of society structure or the corpus of testable theories pro- superstructure or ideology dominate duced under the auspices of cultural infrastructure. For example, in the materialism as compared with ideal- shortest term, it often appears that ists and other alternative paradigms. politics is in command—to use the The first point to be noted with phrase ‘‘applied’’ for the ill-fated respect to the idealists’ paradigm is attempt to change China from the that it is essentially indeterminate, top to the bottom during the epoch since it lacks any selection principle of the Great Leap Forward. to account for why certain ideas Expansion of time frame also rather than others are materialized exposes the fallacy of religion in and incorporated into a particular command, as in the case of contem- cultural repertoire at a particular porary Iran. While Shiite fundamen- time. There is an exception, and that talism has profoundly changed the is Hegelian, which asserts that there is CULTURAL MATERIALISM AND BEHAVIOR ANALYSIS 45 selection for ideas which increase the cause they make war. They make war sway of reason and freedom, a prin- because of population pressure and ciple, however, that clearly results in resource depletion. Whites are preju- counterfactual theories when applied diced against blacks because blacks to the contrasts between cultural are discriminated against. The reason primitives and civilized society. why they are discriminated against Implicit in most idealist ap- has to do with the economic role of proaches is the conviction that cul- unemployed blacks as buffers against tural evolution is dominated by white unemployment. Africans like stochastic processes, such as the children because they have a high random appearance of great leaders fertility rate. They have a high fertil- or the arbitrary appeal to certain ity rate because of the cost–benefits , rules, and . At best, of rearing children. Americans find idealist theories are historical partic- disgusting because they don’t ularist, that is, ideographic theories eat them. The reason they don’t eat which account for the selection and them is because relative to domesti- extinction of cultural innovations by cated ruminants and swine, insects appealing to specific sequences of are a highly inefficient source of changes in infinite regress. Historical animal flesh. , however, cannot account One might very well wonder how for why particular components are cultural idealism, which is devoid of present or absent in other sociocul- retrodictive or predictive principles, tural systems except by providing has been selected for and become a separate for each in- dominant in anthropology and other stance, thus, no predictions or retro- social . The answer may be dictions can be made, because there quite simple: The majority of Amer- are no principles for making such ican social scientists are paid to prove predictions or retrodictions. that human behavior at both the The hopelessness of this fact is of psychological and cultural level is course multiplied many times over primarily a result of will or chance. when idealists reject the possibility of Convinced that there are no nomo- achieving objective knowledge about thetic principles to be found, they what actually happens in other cul- don’t bother to look for them—and tures, according to the testimony of hence are never in any great danger a community of trained observers. of finding any. Also, idealists’ explanations of This behavior—to continue to cultural phenomena frequently take speculate—has been selected for be- a form that is analogous to the cause in our own particular form of explanation of operant behavior, by hierarchical state society, the hungry, appeals to preexisting mind-sets. unemployed, and otherwise frustrat- Why do Amazonian villages make ed and unfulfilled majority are ex- war? Because their males are aggres- pected to blame their losses on wrong sive. Why are blacks discriminated attitudes, bad values, weak wills, and against? Because whites are preju- lousy luck, rather than on the Alice- diced. Why do Africans have a high in-Wonderland design of the socio- fertility rate? Because they like chil- cultural system which governs their dren—or sex. Why do Americans lives. spurn insects as food? Because insects Indeed, has not the President of the are disgusting. Republic, this very week, confirmed For each such case, cultural mate- the importance of cultural idealism as rialist explanations reverse the causal a national paradigm by proposing and raise additional questions that hunger and homelessness in the which open the way to nomothetic world’s richest country results from solutions. Males are aggressive be- nothing but a failure of knowledge 46 MARVIN HARRIS about the rules which provide access psychology get imposed upon human to food and shelter? I hasten to add behavior in the most direct and that similar obfuscations are as ram- powerful . If the behavior of pant or even more so in the rest of the infrastructural subsystems is not sub- world’s civilized society. ject to these lawful restraints, then Regardless of their social func- there would be no reason to suppose tions, the claims of ideal- that cultural evolution is subjected to ist paradigms rest on whether or not selection by consequences. Any oper- it is possible to render testable and ant would then have an equal prob- parsimonious retrodictions or predic- ability of being selected for. And the tive explanations of sociocultural evolutionary trajectories of particular differences and similarities under the cultures would simultaneously exhib- auspices of nomothetic principles, it every combination of components such as the principles of empirical that human beings are capable of determinism. reinforcing in themselves and in The fact is that many cultural others. But such is not the case materialist theories have been ad- empirically. vanced with respect to a broad vari- On the contrary, cultural evolution ety of infrastructural, structural, su- is demonstrably more orderly than perstructural, and ideological com- biological evolution. The incidence of ponents, ranging from the origins of convergent and parallel evolution and the evolution of the produced under the lawful con- state, to the Hindu for straints of the cultural realm are cows and the American rejection of greater per unit of time and unit of dog meat. These theories have been culture than that which has been contested by every known form of produced by natural selection. trickery, deceit, distortion, and mis- It is now my painful duty to point representation in the out that when radical behaviorists journals, and I personally have been turn their attention to the analysis of charged with being everything from sociocultural phenomena, there is no a rabid Marxist, vulgar simplistic guarantee that they will support mechanical materialist, and teller of cultural materialism over the reigning just-so stories, to cryptic idealist and cultural idealist paradigm, other than CIA agent. The one thing that the extreme relativism of phenomen- cultural materialism has not elicited ologists. The problem is that by so far are alternative theories that are failing to identify the causal priorities more parsimonious, of wider scope, that are inherent in sociocultural and constitute a broader and better subsystems, they are as prone as integrated corpus. cultural idealists to accept the com- Since I am not here able to review mon-sense notion that fundamental the expanse of evidence for and changes in a can against the corpus of theory generat- be induced with equal probability of ed by the principles of infrastructural success by controlling reinforcement determinism, let me invoke a more schedules in any subsystem or any abstract form of argument. One component of any subsystem. But expects a priori that infrastructural behavior modification is not the same subsystems will exercise a greater de- as cultural modification. In behavior terminative influence over the other modification, the behavior analyst subsystems simply on the grounds controls the resources and the re- that the infrastructure is the interface inforcement schedule. In cultural between culture and nature, in the modification, the problem is that sense that it is through the infrastruc- others control the resources which ture that the constraints of the need to be controlled if the culture is of physics, chemistry, biology, and to be changed. CULTURAL MATERIALISM AND BEHAVIOR ANALYSIS 47

Since verbal behavior is cheap, the But I do not wish to end on path of least resistance seems to lie in a contentious note. I hope I have the control over the contingencies of made it clear that cultural material- verbal behavior as stated by Skinner ism and radical behaviorism are [1976], ‘‘what is needed is not a new natural allies, and that both strategies kind of government, but further rest upon a common epistemological knowledge about human behavior foundation. It was my good fortune and new ways of applying that to have been exposed to the world of knowledge to the design of cultural operant psychology while still in practices’’ [p. xvi]. college, and I have long awaited an I for one need to be reassured that opportunity such as this to express this means something different from my thanks. If you get anything out of the proposal that the word was in the this, it’s only because you and your beginning, and that it will make us mentors have put so much into this. free. The point is not that knowl- edge—that is, external and internal REFERENCES verbal behavior—is irrelevant to so- ciocultural change, but that infra- Glenn, S. S. (1988). Contingencies and meta- structural–structural conditions se- contingencies: Toward a synthesis of behav- verely restrict the effectiveness of ior analysis and cultural materialism. The Behavior Analyst, 11, 161–179. attempts to change Malagodi, E. F., & Jackson, K. (1989). simply by controlling educational Behavior analysts and : contingencies. We are not free to Troubles and issues. The Behavior Analyst, teach everything we would like to 12, 17–33. teach, nor are we free to learn Skinner, B. F. (1966). Contingencies of re- everything that would benefit social inforcement in the design of a culture. Behavioral Science, 11, 159–166. life. At least not on a scale that would Skinner, B. F. (1976). Walden two. Englewood be adequate for fundamental change. Cliffs, NJ: Prentice Hall. The priority task, therefore, for Skinner, B. F. (1984). The evolution of behavior analysts who wish to mod- behavior. Journal of the Experimental Anal- ify culture is to find ways to contrib- ysis of Behavior, 41, 217–221. Vargas, E. A. (1985). Cultural contingencies: ute to the modification of the existing A review of Marvin Harris’s Cannibals and systems of political-economic control Kings. Journal of the Experimental Analysis and its infrastructural base. of Behavior, 43, 419–428.