Kroeber Anthropological Society Papers, Nos. 71-72, 1990

Is Sociobiology Methodologically Flawed? P. Thomas Schoenemann

The argument between human sociobiologists and their critics is a case ofcompeting paradigms. Two frequent criticisms ofsociobiology are: 1) it is inherently dangerous to assume that biology has an important influence on behavior; and 2) the assumption ofevolutionary adaptation is a methodological error. These criticisms are aimed at the underlying assumptions of sociobiology with the intent of demonstrating that sociobiology isfundamentallyflawed. It is argued that thefirst criticism disregards examples ofhwnan suffering that have occurred because ofan ignorance ofbiology and insteadfocuses exclusively on suffering that has occurred because ofa misuse ofbiology, while the second criticism represents a misunderstanding ofthe nature ofevolutionary biological research. Individual sociobiolo- gical argwnents may be incorrect, but the researchparadigm is not intrinsicallyflawed.

INTRODUCTION sociobiology suffers from serious methodologi- cal flaws because it emphasizes a search for Human sociobiology means different things adaptations. As we shall see, however, this cri- to different people. The classic definition (Wil- ticism betrays a profound lack of insight into son 1975:2) states that sociobiology is "...the how the field of evolutionary biology allows us systematic study of the biological basis of all to understand the world around us. social behavior." Because it is recognized that an organism's biology evolved (as opposed to being specially created), sociobiology necessarily THE PURSUIT OF KNOWLEDGE involves an application of evolutionary biological principles to the understanding of social be- My basic argument will hinge on the fact that havior. There remains, however, an extreme humans are severely limited in their ability to diversity of views on sociobiology's goals, perceive and understand the complexity of the methods and accomplishments (or lack thereof, world around them. As scientists, we are con- depending on one's perspective). This can be fronted with a bewildering array of empirical traced to the fact that sociobiology deals with observations which no human is able to compre- something that no one can claim to be agnostic hend as a single unit. In order to make sense of about. You may not, for example, have a vested this knowledge, we have no option but to break it interest in the truths of ceramic tile heat conduc- down into smaller, manageable pieces. We use tion, or of cloud formation, but you are likely to hypotheses, or theoretical constructs (which in have some understanding of human behavior. turn derive from our current scientific paradigm; Because humans are highly interactive social ani- Kuhn 1970) to guide our explorations. Whether mals, one simply cannot afford to be ignorant of the history of scientific understanding has pro- other people's behavior. To its severest critics, ceeded in a tortuous and twisted route, as Kuhn human sociobiology is seen as a dangerous and (1970) has argued, or simply by a series of suc- misguided attempt to speculate on the limits to cessive modifications of existing theories, it is human behavior; so dangerous and misguided, in clear that the process requires intermediate steps. fact, that they feel it should be rejected out of Paradigms are constructed precisely because our hand as a research paradigm. How valid is this search for understanding proceeds at a snail's claim? pace through a series ofintermediate steps. Para- In this article I will assess two of the most digms serve not only to organize and make sense common a priori reasons for rejecting the of the mass of empirical observations, but be- sociobiological paradigm. The first criticism cause they represent ways of perceiving the emphasizes the potential negative effects of faulty world, they also serve to guide future research. studies that attempt to detect/explain biological Paradigms define which questions "make sense" influences on behavior. This viewpoint ignores to ask; that is, which questions are likely to pro- the other side of the coin, that there are serious vide meaning when answered. Even "holists" negative consequences to the assumption that (those who emphasize that the whole is more biology is irrelevant to behavior, even as a than the sum of its parts) reject the "...romantic working hypothesis. The second argument, ad- notions that one can grasp the functional intri- vanced by some evolutionary biologists, is that cacies of complex systems without conducting 50 scientific and technological studies of individual potential for the abuse of biological ideas in components" (Soule 1985:728). We simply can- political arenas, then any theories purporting to not go from a position of complete ignorance to explain human behavior should operate under the one ofcomplete knowledge in a single step. assumption that biology is irrelevant. It is impor- Human sociobiology is an attempt to explain tant to keep in mind that there are two basic types behavior within an evolutionary biological oferrors that can be made when discussing beha- framework, using exactly the same theoretical vior. The first error is to assume that biology is foundation that evolutionists have used to explain irrelevant when it is in fact relevant. The second morphological variation. In this sense, it repre- error is to assume that biology is relevant when it sents nothing more than a simple extension of a is in fact irrelevant. Critics of sociobiology feel much larger research program. The diversity of that this second kind of error is inherently more opinions about the merits of sociobiology stem dangerous (and likely) than the first. They feel from an extremely basic dichotomy (a conflict of that it is safer to ignore biology in discussions of paradigms) in how people choose to study be- human behavior than to acknowledge it. havior. There are those who are fundamentally That this fear is central to criticisms of interested in differences between human groups, sociobiology is evidenced by the opening chapter and consequently feel that cultural variability in- of Kitcher's (1985) critique of sociobiology. In dicates human transcendence over biology; and it he describes a policy in England which placed there are others who see meaning only in com- children into either academic or trade schools on mon underlying behavioral patterns, and are the basis of psychometric tests given at age ele- consequently more receptive to biological expla- ven. His point is that children were made to feel nations. like failures if they did not pass these tests, and These two basic points of view define quite that this quite possibly scarred them for life. In a nicely (but obviously only on a very general sense, he believes that the tests created the differ- level) the fundamental difference between the ences between children, rather then pinpointing sociocultural anthropology paradigm and the existing differences. He feels that the risks of as- evolutionary biology paradigm. No graduate stu- suming a biological basis ofbehavior are greater dent in sociocultural anthropology, for instance, than the potential benefits that such a belief might could get away with claiming that the people she bring, and therefore we should always operate studied in her fieldwork in East Africa were for under the assumption that biology is irrelevant. all practical purposes essentially the same as the Furthermore, he argues that we should hold re- Balinese. It is the nature of this field to empha- search on such subjects to a higher level of size uniqueness, and because of this, there is no scrutiny than other fields of inquiry. As Kitcher unifying theme that ties the field together, save (1985) wrote: the omnipotence of culture. In contrast, it is the nature ofevolutionary biology to emphasize com- If a single scientist, or even the whole monalities, links and underlying threads. A community of scientists, comes to graduate student in zoology could not get away adopt an incorrect view ofthe origins of with studying desert pupfish without putting a distant galaxy, an inadequate model of them into an evolutionary framework. Evolu- foraging behavior in ants, or a crazy tionary biology, like the rest of the natural explanation of the extinction of the sciences, is a field that attempts to explain the dinosaurs, then the mistake will not greatest number of differences with the least prove tragic. By contrast, if we are number of principles. Human sociobiology is wrong about the bases of human social nothing more than the application ofevolutionary behavior, if we abandon the goal of a biological analyses to human behavior (Ruse fair distribution of the benefits and 1985; Barash 1977), and therefore shares all the burdens of society because we accept detriments as well as all the benefits ofthis larger faulty hypotheses about ourselves and paradigm. our evolutionary history, then the con- sequences of a scientific mistake may be grave indeed (1985:9). THE DANGER OF BIOLOGY There is nothing wrong with the argument Perhaps the most frequent criticism of that we should weigh the costs and benefits of sociobiology stems from the historical record of implementing policies and doing research. The abuse of biology to support racist, ethnocentrist question is whether the second kind of error and sexist ideologies. The argument is that (being wrong in assuming that biological because there is, and has historically been, such a differences influence a particular behavior) is 51 inherently more dangerous than the first kind of field of genetics was decreed by him to be devoid ewror (being wrong in assuming biology is irrele- of meaning (Lemer and Libby 1976). He was vant to behavior). While Kitcher (1985) and quoted as saying, "Just what is this gene? Who other critics point to examples of the negative has seen it? Who has felt it? Who has tasted it?" consequences when a biological influence on (Medvedev 1969:257). For Lysenko, evolution- behavior was mistakenly assumed, it is equally ary change occurred in an organism when the possible to document cases where harm has oc- environment changed, because the organism was curred because biology was implicitly, if not nothing more than the sum of the environmental explicitly, assumed to be irrelevant. One specific conditions (Brill 1975). example is the situation with regard to lactose The scientific basis for Lysenko's theories intolerance. Until recently, it was assumed (by was non-existent. "Scientific" arficles supporting scientists and policy makers of European ances- his claims consisted of various mixtures of try) that all humans were essentially the same in extreme polemics, guilt by association (e.g., their ability to digest the milk carbohydrate lac- Mendel was a priest and therefore his laws were tose as adults. However, it has become clear that invalid), argumentwn ad hominem (arguments Europeans, along with some populations in Af- that appealed to feelings and prejudices instead of rica, are abnormal with regard to most human intellect), a rejection of statistical analysis, and groups (and mammals in general) in their ability circular reasoning (certain results could be ob- to continue producing into adulthood the neces- tained only under particular conditions, which sary enzyme (lactase) which allows them to were defined as the conditions under which the digest lactose. "Lactose intolerant" adults suffer results were obtained) (Lerner and Libby 1976). a variety of ailments upon ingestion of milk, An incredible series of claims were made con- including diarrhea (often severe) and other symp- cerning the miraculous transformation of one toms of a disturbed intestinal mucosa (Leininger species into another through purely environmen- 1978). These facts are now common in intro- tal manipulation. ductory textbooks in physical anthropology, but This fiasco had disastrous repercussions on what is not stressed is that milk has been com- Soviet agriculture. The minimum probable loss monly sent as famine relief to needy countries sustained in the Soviet Union due to its 20 year where most of the population is lactose intoler- failure to adopt American hybridization tech- ant, such as Asia. Milk-induced severe diarrhea niques is estimated to have amounted to 30 to 50 in malnounrshed individuals can be directly or billion kilograms ofcorn alone (Medvedev 1969: indirectly fatal. Since it is unlikely that mal- 180). Unfortunately, the consequences were not nourished individuals will be as cautious about limited to agriculture. By the end of 1936 all re- the foods they ingest as healthy individuals, the search in human genetics was suspended because potential for serious complications is amplified. it was considered inherently dangerous, racist, This demonstrates that behavioral differences (the degrading, and based on fallacious "bourgeois lack of milk usage among most groups) may well genetics". Those scientists who disagreed with have biological bases (lactose intolerance) and the Lysenko and his followers lost their jobs, were ignorance of this possibility can and probably has reassigned to lower positions, or were sent to led to unnecessary suffering and death. work camps where an unknown number per- While much attention is given to the Nazi ished. The first book in Russian on human atrociies carried out under the guise of "genetics" genetics was not published until 1964, which (e.g., Stein 1988), little mention is made of the meant that twenty-five successive classes of devastation wrought in the Soviet Union because physicians graduated from medical schools of a rejection of genetics. From the 1930's until "...without the slightest notion of the laws of the fall of Krushchev in 1964, Soviet genetic heredity" (Medvedev 1969:94). While resear- theory and research was directed by the agron- chers in other countries became increasingly omist T. D. Lysenko, who rose to supreme aware that the severe behavioral effects of gene- academic authority solely because his theories tically based metabolic disorders could usually be were tailored to be desirable to the prevailing po- avoided with special diets, Soviet doctors were litical environment (Medvedev 1969). Lysenko's forced to toil under a proscribed ignorance. theories included a sort of bastrdized version of This ignorance led directly to unnecessary Lamarck's theory of the inheritance of acquired suffering. Research into phenylketonuria, for characteristics along with an outight rejection of example, which is known to cause early mental Mendelian genetics. He insisted on the indivisi- retardation because of an enzymatic deficiency, bility of the organism and maintained that it was began in 1934 and a special diet had been devised never possible to separate environmental from by 1953 (Plomin et al. 1990). Galactosemia, rleditary influences, to the extent that the whole another enzymadc deficiency that also causes 52 mental retardation, was discovered in 1956 things that can happen to a science (Cavalli-Sforza and Bodmer 1971). In all, the (Tinbergen, quoted in Clutton-Brock & specific enzyme deficiencies for some 24 differ- Harvey 1979). ent genetic conditions (11 of which are known to cause retardation or early death) were discovered The second a priori objection to sociobio- prior to 1964, the year of Lysenko's replacement logy arose from within the ranks ofevolutionary (Cavalli-Sforza and Bodmer 1971). If Soviet biologists. Sociobiologists believe that behavior scientists had been allowed to research human itself should be considered an adaptation, just as genetics, it is likely that Soviet researchers would morphological characters, like the wings of also have made significant discoveries. While birds, have been considered adaptations (Barash we will never know the exact cost in human life 1977). As a working hypothesis, an animal's of the rejection of genetics in the Soviet Union behavior is not assumed to be the result simply of during this period, we can be sure it was not learning, in a sense "independently created" in trivial. each generation, but rather it is assumed to have a Is it possible, then, to justify the position heritable basis and therefore to have been shaped that it is always more humane to operate from a by natural selection. This idea has been chal- non-biological paradigm in the absence of more lenged primarily because of this emphasis on specific knowledge? Unfortunately, if we con- behavior as an adaptation (Lewontin 1979). tinue to operate from a non-biological mindset, Gould and Lewontin (1979) have been the pri- the only way to obtain specific knowledge of mary critics of the way many biologists place biologically important influences on behavior will exclusive emphasis on the delineation of adapta- be to stumble blindly onto them. The Lysenko tions in species. They argue that not enough affair, along with the situation with regard to lac- attention is given to such questions as "to what tose intolerance, serves to remind us that severe extent are wejustified in assuming a character (or costs have been paid as a direct result of igno- behavior) is an adaptation?", "How should we rance. While neither situation involved a define adaptation?" and "What other evolutionary conscious conspiracy to cause harm, this did not forces besides natural selection are operating to make the costs any more palatable. It would produce the phenotypic and genotypic variation have been easy to avoid such costs if the right we see today?" In defense, adaptationists ac- questions had been asked. It will always be knowledge these problems but argue that some more prudent to investigate every possibility, starting assumptions are necessary for any evolu- including the possibility that biological differ- tionary investigation (we do not start with perfect ences influence behavior. This is not the same as knowledge), and that there is no reason to as- stating that we should assume all behavioral sume that adaptationism as a worldng hypothesis variation is genetic. It is an appeal for more (i.e., as a guide to future research) is invalid. In knowledge. If differences exist, it is essential fact, as I will argue, the "adaptationist pro- that we understand them. Tmly humane policies gramme" (Gould and Lewontin 1979) is the only can only be derived from an understanding of useful paradigm for evolutionary research. As a boti biological and environmental influences. result, sociobiology cannot be rejected out of hand simply because it is "adaptatonist". How do evolutionary biologists define adap- JUST-SO STORIES tation? Lewontin (1978) has defined it as a "process of evolutionary change by which the In the post-Darwinian era, a reaction organism provides a better and better 'solution' against uncritical acceptance of the se- to the 'problem.' The 'problem' is set by the lection theory set in, which reached its external world, and the agent ofchange is natural climax in the great days ofComparative selection" (1978:56). Function is therefore im- Anatomy, but which still affects many plicit in the concept of adaptation. Futuyma physiologically inclined biologists. It (1986) states that "the analysis of adaptations was a reaction against maling uncritical entails showing that the trait has been developed guesses about the survival value, the by natural selection, and specifying the nature of function, of life processes and struc- the selective agent or agents that have favored the tures. This reaction, of course healthy trait" (1986:251). Clutton-Brock and Harvey in itself, did not (as one might expect) (1979) have argued that the concept ofadaptation result in an attempt to improve methods should not be restricted to traits ofknown genetic of studying survival value; rather it origin, since ontogeny always involves the deteriorated into lack of interest in the interaction ofenvironment and genes. Any con- problem -- one of the most deplorable ceivable trait would therefore have to be divided 53 into adaptive (genetic) and nonadaptive (envi- which combinations conferring lower genetic ronmental) segments. By not restricting the fitness can even become fixed (ifpopulation and! definition in this way Clutton-Brock and Harvey or fitness differences are small enough) (Gould (1979) acknowledge, however, that adaptations & Lewontin 1979). could arise from forces other than natural selec- Given these confounding variables, under tion. Most workers reserve the term adaptation what circumstances are biologists everjustified in for those features specifically built for some advancing adaptationist arguments? Gould and function by natural selection (Williams 1966; Lewontin (1979) give no explicit guidance, but Gould & Vrba 1982). remark that they "would not object so strenuous- Gould and Lewontin (1979) argue that the ly to the adaptationist program if its invocation, common strategy among evolutionists is to: 1) in any particular case, could lead in principle to atomize the organism into traits; 2) argue that the its rejection for want of evidence" (1979:587). traits are optimally designed for some function by What they object to is the use of adaptationism as natural selection; and 3) argue that traits not a working hypothesis. They emphasize that bio- perfectly adapted to some function are so because logists need to treat the organism as a whole of the inability to optimize more than one trait entity, not a collection ofdiscrete adaptations: "If without imposing expenses on others, such that a selection can break any correlation and optimize balance between opposing traits is reached. They parts separately, then an organism's integration find fault in this strategy for two main reasons. counts for little" (1979:596). Lewontin (1979) First, not enough care is given to the problem of states outright that "If sociobiological theory is to how to atomize traits, or how to find what make a lasting contribution to our understanding Lewontin (1979) calls "the 'natural' suture lines of evolution, it must abandon the naive adapta- for evolutionary dynamics" (1979:7). Both tionist program..." (1979:14). Gould (1977) and Lewontin use the example of This is a fundamental methodological issue, the human chin: it is not an adaptation but simply and to properly analyze it we must not forget a result of two growth fields (alveolar and how science proceeds. Adaptationism is simply mandibular) that, in humans, have regressed at one way ofordering the mass of empirical obser- different rates (the former faster than the latter). vations. It does not follow that we are Second, they argue that the exclusive focus unjustified in using it to guide our investigations on the adaptive function of traits, the "adapta- simply because it might yield "incorrect" an- tionist programme," obscures the fact that some swers. In fact, nonadaptative arguments suffer traits may not be adaptations. They feel that one from exactly the same theoretical problem. should not assume a trait is an adaptation even if Adaptationalism and nonadaptationalism do not it is clearly advantageous for the organism: "The represent opposite ends of a spectrum (such as mere existence of a good fit between organism "high metabolism" versus "low metabolism") but and environment is insufficient for inferring the rather the presence or absence of a phenomena. action of natural selection" (Gould & Lewontin The opposite of "adaptive" is not "less adaptive", 1979:592). Some evolutionists try one "adaptive but "not adaptive". We just cannot know, a pri- story" after another until they find one they can ori, which hypothesis is operational. But if we live with. These explanations are dubbed "just- start with the assumption of nonadaptation, the so stories" after Kipling's children's tales. Other quest for understanding is ended. Mayr (1983) evolutionists simply assume an adaptive expla- makes this explicit: "As a consequence of the nation exists without considering alternative, adaptationist dilemma, when one selectionist ex- nonadaptive explanations. Possible alternatives planation of a feature has been discredited, the include: 1) pleiotropy, where the trait is simply evolutionist must test other possible adaptationist one nonadaptive expression of a gene with many solutions before he can resign and say: This phe- effects (these other effects may or may not be nomenon must be a product of chance" (1983: adaptive themselves); 2) multiple selective peaks 326). We have no choice but to start from an as- of equal adaptiveness, such that chance deter- sumption, as a working hypothesis, that a given mines which peak an organism happens upon trait is an adaptation if it confers some advantage first (e.g., one-horned Indian rhinoceros versus to the organism. There can never be direct two-horned African varieties); 3) allometry, in "proof' of nonadaptation, only a lack of adapta- which traits are unequally affected by develop- tionist explanations that fit our observations. mental processes, so that an increase in the size Nonadaptive arguments therefore suffer of, for instance, cervine deer antlers is more than from an intractable problem. Lewontin (1978) proportionately greater than the corresponding points out that "biologists are forced to the increase in body size; and 4) random gene fre- extreme adaptationist program because the alter- quency changes due to finite population size, in natives, although they are undoubtedly operative 54 in many cases, are untestable in particular cases" portance of natural selection in evolution. While (1978:169). This is not true of most adaptive there have obviously been random, or nonselec- arguments (Mayr 1983). As Clutton-Brock and tive, forces helping to shape organisms, such Harvey (1979) point out: forces could not have been dominant. No one has been able to demonstrate how such complex- Though many functional arguments are ity could arise without selection. In fact, Wright initiallypost-hoc, relatively few are ad- (1980) has argued that pure random drift (with hoc: almost all claim generality of some no accompanying selection) must lead to degen- kind and are consequently refutable by eration and extinction. While Wright's adaptive more detailed observation, by experi- landscape metaphor incorporates random factors, ment or by examination oftheir validity he is in fact arguing that adaptations superior to in other groups of animals... The fact those produced solely by selection could be ob- that they originated in post-hoc expla- tained from a mixture of selection and drift. nations is not an objection: this is an Experiments have demonstrated that tremen- inevitable stage in any observational dous changes can be induced in a population by science. In fact we rather doubt that artificial selection, given some genetic variation post-hoc explanations are more mis- (e.g., Yoo 1980; and see Falconer 1981). Daw- leading than many cases where kins (1982) warns us that human subjective predictions are formulated and subse- judgement can be quite unreliable in assessing quently tested: it is seldom difficult to just how unimportant weak selection may be to produce facts that are generally con- have any effect. Mathematical calculations by gruent with a theory, and the process Haldane (1932) demonstrate selection pressure as can appear falsely conclusive (1979: weak as 1 in 1000 on an advantageous allele 550). would lead to fixation in only a few thousand generations, given reasonable assumptions about It is instructive to note that in Gould's (1984) population size. Lande (1976) has shown that, attempt to demonstrate nonadaptive variation in assuming the heritability we see in extant species the shells of land snails in the genus Cerion, he has been constant throughout horse evolution, empirically invalidates the adaptive arguments he only about two selective deaths per million indi- had previously advanced (Gould 1969). He viduals would account for the observed changes acknowledges that "some unexamined selective in the paleontological record. Selection, there- agent might be clinally distributed throughout the fore, can at least theoretically exert a tremendous islands" (1984:235) that would account for the influence on biological evolution. observed variation. In order to argue for non- The utility of these calculations on extant adaptation, Gould himself implicitly recognizes organisms to help us understand the past is that he must demonstrate that the adaptive hy- limited because we do not fully understand the potheses are inoperable. Thus, whether one constraints imposed by developmental processes operates from the adaptationist program or es- on traits. Allometry, of course, is one such pouses nonadaptationist explanations, one is still developmental constraint. But allometric con- obligated to test possible adaptive arguments. stants are not constant for all time. Dawkins (1982) points out that "the allometric constant is a parameter of embryonic development. Like any THE CASE FOR ADAPTATIONISM other such parameter it may be subject to genetic variation and therefore it may change over evo- Aside from methodological issues, however, lutionary time" (1982:33). Clutton-Brock and there are other arguments that support the Harvey (1979) have noted that allometric con- adoption of adaptationism as an initial working stants vary between phylogenetic groups. It is hypothesis. Mayr (1978) points out that our certainly not inconceivable that these differences world view is dominated by the knowledge that are due to selection. But even this can be tested: both cosmic evolution and biological evolution if no other major ecological or behavioral factor have proceeded under "more or less directional is correlated with increased antler size in relation natural processes consistent with the laws of to body size, for example, then we may confi- physics" (1978:47). Yet he goes on to state that dently assert that the allometric constant, as far as biological evolution is fundamentally more we are able to judge, is nonadaptive. Clutton- complicated than cosmic evolution; that living Brock and Harvey (1979) found that large deer systems are "far more complex than any non- species tend to be strongly polygynous. Since living system" (1978:47). The tremendous males of polygynous species are likely to com- complexity of living organisms points to the im- pete for access to females, they may well invest 55 more heavily in large antlers than the males of 1964:37). This must be true, since mutations are small, non-polygynous deer species. The acid continually occurring at all loci and would accu- test would be to see how much ofthe variation in mulate unchecked in neutrl genes. antler size can be explained by these other vari- While it is true that adaptationists maintain ables. that natural selection is inherently an optimizing There are evolutionary biologists that have process, this does not mean that they ignore con- no faith in their ability to make decisions about straints. Dawkins (1982), Clutton-Brock and what constitutes an adaptation and what consti- Harvey (1979), Cain (1964) and Mayr (1983) all tutes a trivial character before actually studying list essentially the same possible constraints on an organism. Cain (1964) remarks that, before perfection. The difference is that adaptationists studies by Sheppard and himself demonstrating insist that we test for adaptations. This requires selective forces operating on the snail species that we construct plausible "just-so stories". Cepaea nemoralis and Cepaea hortensis, "it had Adaptationists operate from the assumption that been confidently asserted that it could not matter natural selection, given enough time, closely to a snail whether it had one band on its shell or approaches the optimal solutions given the possi- two" (1964:48). This remark comes frighten- ble constraints. In sum, there is no reason to ingly close to Gould's (1984) remark about vari- assume: 1) that nonadaptationism as a working ation in Cerion: "Are such small differences, hypothesis is most appropriate a priori; and 2) each involving a simple variation in basic onto- that adaptation and nonadaptation are mutually genetic pattern, necessarily adaptive? Must they exclusive explanations of a trait: in order to test make a difference to a snail?" (1984:235). for nonadaptation one is forced to exclude (test Dawkdns (1982) notes that history seems to be on for) adaptation. the side of the adaptationists, in that time after tdme aspects of organisms thought to be "trivial" have been found to be adaptive (1982:31). This CONCLUSION does not mean that every trait must be an adap- tation, but simply that "merely to fail on a casual Sociobiology is a research program aimed at inspection to see any selective significance in a understanding behavior as an evolutionary adap- particular variation does not license the observer tation. As such it carries all the underlying to proclaim that there is none" (Cain 1964:48). assumptions of the field of evolutionary biology. Every question must be tackled afresh, on its But because every paradigm has underlying as- own merits (Dawkins 1982). sumptions, this fact cannot be a criticism of Cain (1964) also calls into question the sociobiology, but instead is a criticism of human concept of neutral adaptation. What pattern of limitations. The main reasons that have been variation should we expect to find for truly proposed for why sociobiology should not be neutral traits? He argues that we would find used as a research program simply do not hold enormous variation. The example he cites is of water. Nonbiological explanations of behavior fingerprints. Every human being has dermal are not inherently less dangerous than biological ridges. This lack of variation in morphology ones. Human suffering has resulted from the suggests dermal ridges are functional (i.e., are uncritical acceptance of nonbiological expla- adaptive). However, provided there are enough nations. Critcisms of adaptationalism reflect a ridges to produce an adequate friction pad, and misunderstanding ofevolutionary biological rea- assuming they run in all directions so that the soning. Adaptationist hypotheses are an essential finger is not likely to slip in one direction more step in the process of gaining knowledge. Indi- tan another, "the exact pattern is imaterial and vidual sociobiological explanations of particular can be allowed to vary. The resulting variation is behaviors may well be incorrect, but the socio- certainly tremendous, and this, and not relative biological world view is not "wrong" apriori. constancy, is what we would expect of neutral characters" (1964:48). With this logic he main- tains that ancestral traits that characterize major ACKNOWLEDGMENTS clades (like the aquatic phase of amphibians) are not merely neutral holdovers (as per Gould This paper owes much of its genesis to dis- 1986), but represent adaptations to less special- cussions with Vincent Sarich. David Wake and izedways of life. If these characters were truly Reina Wong provided useful comments and aeutral and not soundly functional, they would criticisms on earlier drafts. varywildly, especially given their antiquity (Cain 56

REFERENCES CITED Barash, David P. (1977) Sociobiology and Be- Kuhn, Thomas S. (1970) The Structure of havior. New York: Elsevier. Scientific Revolutions, 2nd edition. Chicago: Brill, Henry (1975) Presidential address: Na- University of Chicago Press. ture and nurture as political issues. In R.R. Lande, Russel (1976) The maintenance of Fieve, D. Rosenthal, and H. Brill (eds.), genetic variability by mutation in a polygenic Genetic Research in Psychiatry. Baltimore: character with linked loci. Genetical Research Johns Hopkins University Press. Pp.283- 26:221-235. 288. Leininger, M. (1978) Some cross-cultural Cain, A.J. (1964) The perfection of animals. In universal and non-universal functions, beliefs, J.D. Carthy & C.L. Duddington (eds.), View- and practices of food. In M. Leininger (ed.), points in Biology, volume 3. London: Transcultural Nursing: Concepts, Theories, Butterworths. Pp.36-63. and Practices. New York: John Wiley and Cavalli-Sforza, L.L. and W.F. Bodmer (1971) Sons. Pp.203-219. The Genetics of Human Populations. San Lerner, I. Michael and W.J. Libby (1976) Francisco: Freeman. Heredity, Evolution, and Society. San Clutton-Brock, T.H. and Paul H. Harvey Francisco: W.H. Freeman and Company. (1979) Comparison and adaptation. Pro- Lewontin, Richard C. (1978) Adaptation. ceedings of the Royal Society of London B Scientific Anerican 239:156-169. 205:547-565. Lewontin, Richard C. (1979) Sociobiology as Dawkins, Richard (1982) The ExtendedPheno- an adaptationist program. Behavioral Science type: The Gene as the Unit ofSelection. New 24:5-14. York: Oxford University Press. Mayr, Ernst (1978) Evolution. Scientific Amer- Falconer, Douglas S. (1981) Introduction to ican 239:47-55. Quantitative Genetics. London: Longman. Mayr, Ernst (1983) How to carry out the Futuyma, Douglas. J. (1986) Evolutionary adaptationist program? American Naturalist Biology, 2nd edition. Massachusetts: Sinauer 121(3):324-334. Associates. Medvedev, Zhores A. (1969) The Rise and Fall Gould, Steven J. (1969) Character variation in of T.D. Lysenko. Translated by I. Michael two land snails from the Dutch Leeward Lerner. New York: Columbia University Islands: Geography, environment, and evolu- Press. tion. Systematic Zoology 18:185-200. Plomin, Robert, J.C. DeFries and G.E. Gould, Steven J. (1977) Ontogeny and phylo- McClearn (1990) Behavioral Genetics, 2nd geny. Cambridge, Massachusetts: Belknap edition. New York: W.H. Freeman and Press. Company. Gould, Steven J. (1984) Covariance sets and Ruse, Michael (1985) Sociobiology: Sense or ordered geographic variation in Cerion from Nonsense? 2nd edition. Boston: D. Reidel Aruba, Bonaire and Cura9ao: A way of Publishing Company. studying nonadaptation. Systematic Zoology Soule, M.E. (1985) What is conservation bio- 33:217-237. logy? BioScience 35:727-734. Gould, Steven J. (1986) Of kiwi eggs and the Stein, George J. (1988) Biological science and Liberty Bell. Natural History 95:21-29. the roots of Nazism. American Scientist 76: Gould, Steven J. and Richard C. Lewontin 50-58. (1979) The spandrels of San Marco and the Williams, George C. (1966) Adaptation and Panglossian paradigm: A critique of the Natural Selection. Pxinceton, New Jersey: adaptationist programme. Proceedings ofthe Princeton University Press. Royal Society ofLondon B 205:581-598. Wilson, Edward 0. (1975) Sociobiology: The Gould, Steven J. and Elizabeth S. Vrba (1982) New Synthesis. Cambridge, Massachusetts: Exaptation -- A missing term in the science of Harvard University Press. form. Paleobiology 8:4-15. Wright, Sewell (1980) Genic and organismic Haldane, J.B.S. (1932) The Causes of Evolu- selection. Evolution 34:825-843. tion. London: Logman's Green. Yoo, B.H. (1980) Long-term selection for a Kitcher, Philip (1985) Vaulting Ambition, quantitadve character in large replicate Sociobiology and the Quest for Human populadons of Drosophila melanogaster. I. Nature. Cambridge, Massachusetts: The MIT Respoe to selection. GnedicalResearch 35: Press. 1-17.