 |
Segnala il documento |  |
Stampa | ||
| Evolutionary Psychology and Social Science |
| di Chris Haufe |
La teoria dell’evoluzione ha profondamente cambiato la nostra immagine dell’uomo e del suo posto nel mondo naturale. Con essa, infatti, si è costituito un nuovo programma di naturalizzazione della psiche umana e delle relazioni sociali. Questa sembra una banalità. Ma è proprio così? Quanto l’immagine che ciascuno di noi ha di sé stesso, dei suoi simili e della comunità di persone in cui vive, deriva dai principi fondamentali della teoria dell’evoluzione e quanto, invece, è il frutto di una tradizione indipendente da questa teoria? La grande popolarità che negli ultimi decenni hanno conosciuto la psicologia evolutiva e il darwinismo sociale sono forse indice del fatto che, finora, la teoria evolutiva non ha scalfito l’immagine restituitaci da una secolare tradizione religiosa e filosofica.
Ma in buona parte la situazione sta cambiando. Ne sono testimonianza le molteplici sponsorizzazioni di progetti legati allo studio della natura umana da un punto di vista evolutivo. Si va così costituendo una spaccatura sempre più netta. Da un lato vi è un nuovo, pionieristico metodo di indagine ispirato alla teoria dell’evoluzione, che si proclama portatore della verità in nome della scienza; dall’altro, vi sono le scienze sociali che proseguono la tradizione formatasi a partire almeno sin dal Settecento.
Una domanda quindi si pone per tutti coloro che oggi sono interessati al programma naturalista: in che misura l’immagine dell’uomo che sta alla base delle scienze sociali tiene conto dei risultati della teoria dell’evoluzione?
Nel saggio qui presentato, Chris Haufe, dottorando presso la Columbia University, discute la principale accusa sollevata da alcuni psicologi evolutivi contro agli scienziati sociali "tradizionali". Secondo l’accusa, gli assunti principali alla base di discipline quali la sociologia e la psicologia non sono consistenti con l’immagine della natura umana che discende dalla teoria dell’evoluzione. Ad esempio, tali discipline assumono che i processi mentali fondamentali sono independenti dal loro contenuto. Un assunto che – sostiene l’accusa – non è compatibile con l’idea che i processi mentali siano frutto dell’evoluzione, ovvero si siano formati sulla base di una necessità di meglio adattarsi all’ambiente circostante.
Haufe presenta i principali esempi di incompatibilità addotti dagli psicologi evolutivi, e discute il concetto di incompatibilità stesso, per poi scagionare le scienze sociali dall’accusa. (a.b.)
EVOLUTIONARY PSYCHOLOGY AND SOCIAL SCIENCE
Chris Haufe, Columbia University, New York
Introduction
Evolutionary psychologists have claimed that social science must "accept with grace the...tenet of mutual consistency among disciplines" in order to "move away from its present fragmented and insular form" (Cosmides et al. 1992: 12-13). In this paper I examine evolutionary psychologists' concept of consistency as well as the tenet of mutual consistency itself and argue that neither can bear the load required by the arguments against social science found in the evolutionary psychological literature.
Many evolutionary psychologists have criticized social scientists for paying insufficient attention to making their theories of human behavior "consistent with what is known in the natural sciences" (Cos-mides, et al. 1992: 4), in particular, evolutionary biology. Social science, they claim, has been con-ducted in an environment largely devoid of evolutionary considerations. By not taking into account the fact that our species evolved just like any other species, researchers have ignored natural selection, a causal process "known to govern all life" (Buss 1991: 461). A lack of proper concern for the forces of evolution would put any account of human psychology in jeopardy, for the simple reason that psycho-logical theories "inconsistent with evolutionary theory stand little chance of being correct" (Buss 1991: 461). In the succinct language of leading evolutionary psychologists Daly and Wilson, "[m]any...theories that are still debated by social scientists implicitly deny the action of natural selection, and are therefore surely wrong" (Daly and Wilson 1988: 7).
In stark contrast to the evolutionarily uninformed social sciences stands evolutionary psychology, a research program whose results are derived from, and thus assumed to be consistent with, the "strong deductive framework" of evolutionary theory (Tooby and DeVore 1987: 189). Because of its strong em-phasis on the centrality of evolution to any plausible theory of human psychology, evolutionary psychol-ogy avoids the risky business of theorizing about human nature which might turn out to be incompatible with what we've learned from evolutionary biology.
Examples of Evolutionary Inconsistency in Social Science
Domain-Generality
The most egregious violation of evolutionary theory perpetrated by social scientists is the assertion that the human mind consists primarily or solely of "general-purpose, content-independent, or content-free mechanisms" (Tooby and Cosmides 1992: 34). Mental mechanisms given free reign over what in-puts to focus on and what behaviors to produce would allow for an unwieldy variety of behavioral plas-ticity, in which the organism will rarely if ever provide the appropriate behavioral response to an envi-ronmental challenge.
Accordingly, to be endowed with broad behavioral plasticity uncon-nected to adaptive targets or environmental conditions is an evolution-ary death sentence, guaranteeing that the design that generates it will be removed from the population (Tooby and Cosmides 1992: 101).
The unerring eye of natural selection will weed such mechanisms out because "the property of freely varying behavior in all dimensions independent of conditions is not advantageous: It is evolutionarily and individually ruinous" (Tooby and Cosmides 1992: 101).
Passive Reception of Information
Another psychological property which social scientists have proposed is the "passive reception" of sociocultural norms, in which individuals learn appropriate behavior by imitating other members of their social group. Here, as in the case of domain-generality, social scientists have taken a "conceptual wrong turn" (Cosmides, et al. 1992: 13) because of their evolutionary ignorance.
"Evolution could not have produced a psyche that functioned as the passive receptacle of information transmitted from the social group, because (among other reasons) many members of the social group have antagonistic interests" (Tooby and Cosmides 1990: 44)
The idea here is simple: the interest of other members of one's social group will often be at odds with one's own interests. Anyone who merely imitated the behavior of others would either be exploited or, at a minimum, behave in a way that didn't necessarily cohere with his ultimate reproductive interests. Instead, evolution would had to have produced an "individually tailored adaptive system" (Tooby and Cosmides 1990: 44) which generated behavior appropriate to the individual's specific adaptive pre-dicament.
Unreliable "Status-Striving" Mechanisms
During a discussion of the evolutionary dynamics of psychological mechanisms, David Buss asserts that the evolution of psychological mechanisms which created a desire to climb the social ladder would be impossible "unless they reliably produced classes of acts that actually led to the increase or mainte-nance of positions within social hierarchies" (Buss 1991: 464). According to Buss, given that "[p]sychological mechanisms evolve because they have behavioral consequences," such mechanisms could only evolve in the instance where they faithfully caused a certain (presumably adaptive) behavioral output.
Family Violence and “Easy Striking Distance”
In their Lakatosian defense of evolutionary psychology, Ketelaar and Ellis assert that the psychologi-cal mechanism(s) for violence cannot be victim-general because this would undermine the “basic metatheoretical assumption of evolutionary theory...that natural selection favors nepotism.” If violence mechanisms did not discriminate between family members and non-family members, then genetic rela-tives would, ceteris paribus, be just as likely to be attacked as would non-relatives. This kind of mecha-nism would be less adaptive than one which discouraged violence against family members. According to Ketelaar and Ellis, for such a mechanism to have evolved would be inconsistent with evolutionary the-ory and “would require a reformulation of the basic assumptions” of that theory (Ketelaar and Ellis 2000: 5). Thus, we can rule out the possibility that violence mechanisms are victim-general, i.e., assuming that natural selection favors nepotism.
What's Supposed to be Inconsistent?
Uniting these claims of inconsistency is the notion of negative effects on fitness. In the case of do-main-general mechanisms, evolutionary psychologists argue that such mechanisms would reduce the pace of cognition to a debilitating rate, and that behavioral responses will more often than not be con-textually inappropriate, both of which are assumed to adversely affect the reproductive success of the organism. For passive receivers of information, their lack of consideration for their particular adaptive situation will result in behavior which goes against their evolutionary interests. Victim-general violence is assumed to lower inclusive fitness. And status-striving mechanisms which did not reliably produce an elevation in social status would not have been adaptive enough to evolve. The essence of the objection is best stated by Tooby and Cosmides themselves: "Adaptive tracking must, of course, have characterized the psychological mechanisms...during the Pleistocene, or such mechanisms could never have evolved", (Tooby and Cosmides 1989: 35) and elsewhere: "...a psyche that contained nothing but general-purpose information-processing procedures could not, in principle, generate adaptive behavior, and therefore is an evolutionary impossibility” (Tooby and Cosmides 1990: 27).
Now, this specific reference is to mechanisms "governing culture," but the message is general: that which is not adaptive cannot evolve. When a proposed behavior or morphological trait is perceived a priori to most likely have negative effects on the fitness of its bearer, it is determined that that behavior could not have evolved. To propose the existence of a trait that could not evolve is to be inconsistent with evolutionary theory. Therefore, to claim that we have some trait that is perceived to be the kind of thing that would have negative effects on fitness is to make a claim which contradicts evolutionary the-ory, a claim which has little chance of being correct.
A Priori Judgment of Negative Effects on Fitness
There are at least three serious problems with these claims of inconsistency. The first has to do with how we can know the effects of a given trait on an organism's fitness. The second has to do with do-mains of fitness. The third has to do with the notion of what could possibly evolve.
With the exception of a trait whose inability to evolve was logically necessary, there is no reason to think that we should be able to tell a priori that a particular trait would have negative effects on the fit-ness of any organism which possessed it. An a priori evaluation of a trait's evolutionary trajectory re-quires that the very concept of the trait entails either that it (a) cannot evolve, or (b) cannot but evolve. This is because, in the absence of a logically necessary trajectory, we cannot know in advance of the actual expression of the trait precisely how it is going to fit into the organism's overall phenotypic econ-omy. Moreover, we certainly cannot know whether a trait would have had negative effects on fitness without knowing the actual historical conditions in which it arose – i.e., without knowing which alternative morphs were available or what kinds of selection pressures the population encountered. But evolution-ary psychologists' charges of inconsistency attempt to do just that. The widespread pronouncement that domain-general cognitive structures could not possibly make up a significant portion of an organism's cognitive architecture because no such structures could have evolved implies that there are no possible phenotypic or ecological conditions under which domain-generality could have any effect on fitness other than an overwhelmingly negative, "ruinous" one. But whether this is the case is an entirely a posteriori matter which cannot be judged in any way other than experiment and observation - i.e., the way biologists usually determine what kinds of fitness effects a trait has. "Imaginations informed by evolutionary theory" (Daly and Wilson 1988: 13), no matter how fertile, cannot furnish us with answers to questions about fitness in advance of the actual performance data.
It is likely that there will be some resistance to the notion that, except for the case in which there is no possible world in which the trait could evolve, we cannot determine the evolutionary trajectory of a trait a priori, but examples can be adduced at will to allay any such fear. Here are just a few. Sterility immediately suggests itself, but quickly retreats owing to the preponderance of insect species where sterile males have been around for a long time. Traits which are extremely physiologically debilitating may also seem like attractive candidates. Again, however, we find these all over the place, from frog chucks to peacock trains. Theory suggests that these traits started off with relatively low costs but, owing to a female preference for them, quickly evolved into exaggerated versions. One explanation for the exaggeration is Ronald Fisher's (1930) "runaway" process, where the female preference for a trait causes it to reach increasingly extreme versions of itself until it becomes so physically debilitating that the fitness advantage gained through increased mating opportunities is counterbalanced by the sharp increase in mortality. Another explanation is Zahavi's (1975) "handicap mechanism," which suggests that males evolve physically taxing traits to show females how good their genes must be if they can withstand the burden of their handicap. These theories are virtual recipes for turning seemingly maladaptive traits into adaptive ones. Any a priori proposal for a trait's evolutionary trajectory would have to incorporate some way of circumventing this recipe. None of the "impossible" traits on offer in the field of evolutionary psychology provide a way of doing this.
Fitness Domains
This brings us to the second general problem. When we want to assess the fitness value of a trait, often what we must do is measure the performance of that trait with respect to a particular task. Good performance can provide some indirect evidence that the trait was selected to perform that task. Poor performance can provide indirect evidence that the trait was not selected to perform that task, but it cannot provide evidence - direct or indirect - that the trait was not selected to perform any task. Poor performance at a particular task may just indicate that we are assessing the value of the trait relative to the wrong task, a task for which it was not selected. This idea forms the foundation of methodological adaptationism, practitioners of which believe that moving on to another task analysis when the trait is shown to have failed at the original task is just how science gets done (Mayr 1983; Beatty 1987; Parker and Maynard Smith 1990).
For example, suppose we suspect a certain local species of bird to be a Batesian mimic based on the fact that it has colored spots similar to some sympatric unpalatable species of butterfly. If the bird is this sort of mimic, its viability should be increased by causing predatory species to avoid it as they avoid the local butterfly. To test our hypothesis, we paint the colored spots of half the members of this bird species so that their coloring is homogeneous. The other half we leave as they are. If our hypothesis is correct, the viability of the painted group should decrease as a result of having lost its ability to mimic the unpalatable butterflies. But lo, we find that viability in the homogeneously colored group actually increases. This could indicate that the colored spots were not selected for their positive effects on viability, contrary to our hypothesis. But it does not show that the spots are not an adaptation. In fact, a subsequent experiment shows that females prefer to mate with males with colored spots. Colored spots have positive effects on fitness through sexual selection and thus may still be an adaptation, even though they negatively impact survival.
What this example shows is that the domain in which we judge the value of a trait for reproductive success may just be the wrong domain in which to judge the value of that trait. So, even if evolutionary psychologists are capable of demonstrating that domain-general cognitive structures in any organism in any environment are, say, very slow, they still have a long way to go in showing that domain generality could not be an adaptation. Perhaps cognition actually slowed down in the transition from chimpanzees to humans in order to allow us to deal with a greater variety of information, or to expand the range of inferences of which we're capable. Perhaps some accidental property of the early female hominin brain produced a sensory bias which caused them to prefer slow thinkers. Of course, all of these alternatives are, like the charge of cognitive sloth, pure fantasy until the appropriate tests are carried out. The point is merely that even if we consider evolutionary psychologists to have had the last word on the issue of computational speed, there are many other domains in which the fitness value of domain generality needs to be assessed. A simple glance at our own anatomy should suffice to drive the argument home. Our relatively large brains made us, ceteris paribus, heavier than our smaller-brained ancestors. But it would be a serious error to suggest that larger brains couldn't have been selected for because of the negative effects on speed or balance created by the added brain mass.
Other Ways of Evolving
Let's suppose that evolutionary psychologists accept our challenge of testing the performance of a trait for each task for which it could reasonably be thought to have been selected and find that the per-formance is in every domain poor relative to other variants that probably existed when the trait was evolving. If there are no domains in which the trait could have outperformed contemporaneous variants, then we are justified in concluding that it is not an adaptation. It is quite another thing, however, to show that the trait could not exist . that, for example, domain-general cognition is "an impossible psychology" (Tooby and Cosmides 1992: 34).
One rather obvious way in which uniformly maladaptive traits can evolve is through genetic or func-tional correlation with other, adaptive traits (Price and Langen 1992: 309-310). Experiments with artificial selection have demonstrated a variety of correlated responses to selection for a specific trait. Selection for tameness in foxes, for instance, brought along with it several "juvenile-like characteristics which... appeared as a side effect of changing hormone concentrations" (ibid: 309). Similarly, selection for certain life-history traits can cause the persistence of traits that are severely maladaptive. A negative genetic correlation between traits that are positively functionally correlated (such as life-history traits) will cause tradeoffs between trait values when there is selection for one but not the other (Charlesworth 1990). Conversely, single genes may produce multiple traits (i.e., traits with a perfect genetic correla-tion), some of which are adaptive, some of which are maladaptive. Where the positive effects outweigh the negative, maladaptive traits can persist. Let it suffice to say that when there is the right kind of correlation between traits, maladaptive traits can and indeed do evolve. Thus there is nothing "impossible" about the evolution of a psychology that might have been slower than other variants, or a personality type that learned behavior through socialization rather than adaptive planning.
Living in a post-Spandrel world, evolutionary psychologists are presumably acutely aware of evolu-tionary avenues that do not follow the path of natural selection. Indeed, one can find exhortations to that effect across the literature. However, given the continual a priori evaluation of traits and references to things which "could not evolve" or are "impossible," I see no reason to think that the professed sensitivity to modes of evolutionary change other than natural selection is anything other than lip service. If evolutionary psychologists took seriously forces like drift and character correlation, there would be no reason for them to claim that, for example, domain-general cognitive architectures could not evolve, due to their possible negative effects on fitness.
Real Inconsistency
We have seen that claims of inconsistency on the part of evolutionary psychologists fail to show in-consistency, due to a mistaken conception of what it means to be "inconsistent with evolutionary theory." If evolutionary psychologists have painted an inaccurate picture of inconsistency, then what is the right picture?
In order to actually be inconsistent or incompatible with evolutionary theory (by which evolutionary psychologists typically mean "natural selection") an hypothesis would have to entail that the proposed trait was selected for because of its negative effects on reproductive success. This is something that Darwin's theory of natural selection cannot sustain logically. To make this condition clearer, consider what might (or might not) have caused the widespread existence of human noses. If we wanted to ex-plain the existence of human noses in a manner consistent with evolutionary theory, then we could not, for example, assert that noses caused the people who had them to have less children, and that those children (who acquired noses through genetic inheritance) also ended up having less children because of their noses. The reason that this explanation would be inconsistent with evolutionary theory is this: evolutionary theory entails that a property cannot be selected for because of its negative effects on fit-ness, and our explanation of the evolution of noses asserts that the nose was selected for because of its negative effects on fitness. Evolutionary theory says “p,” and our theory says “¬p.”
Of course, as I have already argued in detail, this does not mean that noses could not have evolved unless their fitness effects were non-negative. For noses to have negative effects on fitness and still evolve is perfectly consistent with Darwin's theory of evolution. Suppose the current version of the hu-man nose Nc was caused by the same mutation which gave us the opposable thumb, and suppose that Nc worked much more poorly than the previous version of the nose, Np. Even though the fitness effects of Nc relative to Np were negative, the overwhelmingly positive effects of the opposable thumb allowed the inferior Nc to spread through the population. Now, even though this example is strictly fiction, it nonetheless remains a fact that the ability of properties with negative fitness effects to become wide-spread is a phenomenon the possibility of which is not in serious doubt.
Social Science and Real Inconsistency
Although it may not be rare for social scientists to offer some casual remarks on why a particular mode of human behavior might have evolved, I know of no explanation proffered by a social scientist which takes the form of the human nose explanation outlined above––i.e., an account which attempts to explain the existence of a behavior by way of reference to its negative effects on fitness. The closest thing we see to such an explanation comes from evolutionarily-minded anthropologists who argue that certain behaviors evolved via the handicap mechanism, where the high viability costs of a behavior (e.g., hunting) signal that the person performing it must have "good genes" (Hawkes and Bird 2002). In this case, the negative fitness effects (in the domain of viability) of hunting are in a sense causally related to its evolution, but the reason hunting is adaptive is because of the preferential treatment received by men who hunt, not the viability-reducing effects. The fitness advantage caused by the preferential treatment outweighs the disadvantage incurred by the behavior itself.
The work of Hawkes and other evolution-savvy social scientists, while interesting and important, stands at the margin of social science. The vast majority of claims made by social scientists are com-pletely divorced from evolutionary considerations. But if we do consider it important for social scientists to make their theories consistent with evolutionary theory, we are presented with a puzzle: How do we determine whether a social scientific claim (SSC) is consistent with evolutionary theory when that claim does not mention evolution?
Above we saw the evolutionary psychological attempt to solve this puzzle––viz., to judge the poten-tial fitness consequences of a behavior and see whether, given those consequences, it could have evolved. This strategy turned out to be a chimera for two reasons: (1) we cannot judge a priori the fit-ness consequences of a behavior, and (2) fitness consequences alone cannot tell us whether a trait could have evolved. However, the strategy itself is suggestive of the kind of approach we must take. If indeed we are to check SSCs against evolutionary theory, we will need some way of rephrasing or trans-lating them into the predicates of evolutionary biology. Unfortunately, we have not yet been able to ac-complish this.
What can we say in general about the compatibility (by which I mean something like "mutual re-alizability of all statements") of two theories in advance of the requisite translation? If we have inde-pendent reasons to believe that each theory is true, we could infer that the theories are compatible from the fact that no two statements can be true and contradict each other. This kind of reasoning might be behind Cosmides' et al. assertion that "the theory of natural selection cannot, even in principle, be ex-pressed solely in terms of the laws of physics and chemistry, yet it is compatible with those laws" (Cos-mides, et al. 1992: 4). In the absence of shared predicates, this appears to be the only option we have available to us. Accordingly, we should apply the same rule when we find we are unable to translate the predicates of SSCs into the predicates of evolutionary biology. Where an SSC appears to be well-grounded independently of its relationship to evolutionary theory, we might properly conclude that it is compatible with evolutionary theory.
Contrastingly, Harman Holcomb III has suggested that the very fact that SSCs are formed in a se-mantics different from that of evolutionary biology is itself reason enough to suspect that SSCs are in-consistent with evolutionary theory.
"If social scientists are free to describe and explain behavior in ways that need not take into account biological evolution, then it is likely that their descriptions and explanations of human behavior will often be ir-relevant to or inconsistent with evolutionary theory" (Holcomb III 1993: 122).
I find the source of Holcomb's skepticism elusive. Why assume that social scientists are likely to make claims inconsistent with evolutionary theory if they do not use it as a guiding principle? A SSC that is internally consistent and well-supported should inspire us to give credence to it, irrespective of whether it is formulated in evolutionary semantics. Of course, given the lack of a common semantics between the social and evolutionary sciences, it's certainly possible that most SSCs could turn out to be inconsistent. But to predict in advance of the translation of one into the other (or both into a third lan-guage) that
it would be a miracle if a social science unguided by evolutionary con-siderations would on its own describe and explain human behavior in a way directly subsumable under evolutionary theory (Holcomb III 1993: 123)
strikes me as unreasonable. Given that we are admittedly incapable at present of judging the compati-bility of SSCs and claims made in evolutionary biology, it's unclear upon what Holcomb's pessimistic stance could be founded. If we have yet to determine the compatibility with evolutionary theory of even a single SSC, we could have no basis for thinking that their consistency with evolutionary theory would be unlikely or miraculous.
Is Consistency Important?
Quite apart from evolutionary psychology's misconceptions regarding what consistency with evolu-tionary theory is or whether we can determine whether some SSC achieves it, we should ask whether they're right to insist that social scientists' claims be consistent with evolutionary theory (along with the rest of the natural sciences). The fact that evolutionary psychologists are wrong about whether social scientists' claims and the claims of natural scientists are mutually consistent does not immediately let social scientists off the consistency hook. We still need to assess the requirement of consistency in and of itself.
There's a clear sense in which the demand of mutual consistency is, as Tooby and Cosmides (1992: 12) have emphasized, "absolute" and "incumbent upon all valid scientific knowledge". If statements in the natural sciences are true, social scientists' claims cannot both be true and contradict those state-ments. But the mutual consistency of true statements isn't a requirement restricted merely to the natural and social sciences. No two statements––be they within the same discipline, between disciplines, or not within any discipline––can both (1) be true and (2) contradict one another. Thus, the requirement of con-sistency extends not just from, for example, evolutionary biology to the social sciences, but even from evolutionary biology to the library sciences. Normally, though, evolutionary psychologists are not em-phatic that statements about, say, microfiche and statements in evolutionary theory be consistent with one another. Yet, the burden of consistency with evolutionary biology weighs no less heavily on the li-brary sciences than it does on the social sciences. Why, then, we might ask, are evolutionary psycholo-gists so insistent that the social sciences observe the consistency requirement, while the library sciences receive no mention at all?
The asymmetry with which evolutionary psychologists treat the social sciences, on the one hand, and the library (as well as all other) sciences suggests that their demands for consistency come not from the logical necessity that all truths must be mutually consistent, but rather from the methodological as-sumption that evolutionary theory––specifically, the theory of adaptation through natural selection––and theories of human psychology and behavior should have a lot do do with each other. If the real motiva-tion was the general requirement of mutual consistency then we should expect to see articles lambast-ing the library sciences for ignoring the principles of quantum mechanics and organic chemistry. But we have no such expectation, and with good reason.
As evidence for the existence of the assumption that the theory of adaptation through natural selec-tion and theories of human psychology and behavior must bear upon one another, consider the claim by Donald Symons that "there is no known scientific alternative to the theory that human nature is the product of natural selection" (Symons 1992: 147). Concerning the properties of organisms in general, even within evolutionary biology itself there are at least two "scientific alternatives" to the theory that those properties are the products of natural selection––viz., genetic drift and character correlation. For example, the notion that some mammalian tooth characters are the product of genetic drift rather than natural selection is an alternative to the theory that those characters are the product of natural selection (Lande 1976), and it's difficult to see precisely what about that alternative could be characterized as "un-scientific." What would undoubtedly be unscientific would be to rule out a priori the possibility that mammalian dentition is the product of genetic drift. Lande notes,
If the hypothesis of evolution by random genetic drift cannot be consis-tently ruled out from fossil evidence, there would be no basis for sup-posing that the phenotypic evolution is purely a result of natural selec-tion, and random genetic drift would emerge as a potentially significant motive force in evolution (1976: 321)
The same presumably is true for the aggregate property "human nature," which refers to the bundle of psychological and behavioral properties that are common to all humans in virtue of their shared genetic inheritance. For each property, there will be a variety of "scientific alternatives" to a selectionist account, many of which will come from outside evolutionary biology. As evolutionary psychological folk hero George C. Williams warned us many years ago,
This biological principle [i.e., adaptation through natural selection: CH] should be used only as a last resort. It should not be invoked when less onerous principles, such as those of physics and chemistry or that of unspecific cause and effect, are sufficient for a complete explanation (1966: 11).
Each of these alternatives will need to be entertained in a systematic, empirical way in order for our ex-planations of the focal properties to demand any sort of confidence.
Further evidence for the existence of an assumption that theories of human psychology and behav-ior need always to intersect with the theory of natural selection is provided by the assertion that there is an "irrefutable" logical progression from the premise that physiological properties are generally adaptive to the conclusion that "human nature, including human psychology, must have...an adaptive logic" (Daly et al. 1982: 23). Generally speaking, it does not follow that some biological property (e.g., our psychol-ogy) is an adaptation from the fact that another property (e.g., our physiology) is plausibly an adaptation. This inference procedure would effectively license us to conclude that all of our properties are adapta-tions in the case where we found that one of them was. Whether or not two organismal traits have simi-lar evolutionary histories (specifically in the respects relevant to determining whether they are adapta-tions) is going to depend on the particular historical facts of each trait considered separately, not on the supposed existence of a logical progression which dictates that a commitment to adaptation some-where is a commitment to adaptation everywhere.
The assumption that theories of human behavior and the theory of natural selection have a lot to do with each other is based on the view that it is true a priori that the theory of natural selection, in Jerry Fodor's words, "importantly constrains" theories of human psychology and behavior (Fodor 2000: 83). But the theory of natural selection can only constrain theories of human psychological and behavioral traits in the cases where those traits were actually selected for. And while it may be true in some cases––even every case––that a particular trait's evolutionary history suggests that it is the product of direct selection, it is not true a priori. For, as we have seen, even functional traits may arise through evolution-ary means other than natural selection. Thus, whether the theory of natural selection and theories of human psychology and behavior should have a lot to do with each other is not an a priori truth but will instead depend on the particular psychological or behavioral trait in which we are interested. So evolu-tionary psychologists are not justified in their use of the theory of natural selection as a logical constraint on the social sciences.
Now although it is not an a priori truth that evolutionary theory and theories of human behavior should have a lot to do with each other, that such a relationship exists between them may be a reason-able methodological assumption. There doesn't seem to be anything in principle wrong with adopting it as a methodological assumption and then going on to see whether work done under that assumption bears any fruit, in the same way there should not be anything in principle wrong with adopting adapta-tionism as a methodological assumption. But there is something in principle wrong with demanding that anyone who endeavors to study human behavior adopt that assumption, in the same way it would be wrong and scientifically ridiculous to demand that every evolutionary biologist adopt adaptationism as a methodological assumption. We know that many important properties of organisms are not adaptations, and to push on as though they were would certainly not help to expand the breadth of biological knowl-edge.
Conclusion
The arguments in this chapter have aimed at establishing three general claims. First, evolutionary psychologists are typically mistaken in their beliefs about what kinds of phenotypic properties are incon-sistent with evolutionary theory, a mistake which derives from a general misconception regarding what consistency with evolutionary theory actually is. Second, given the correct conception of consistency there is no clearly understood method for determining whether claims in the social sciences are consis-tent with evolutionary theory. Hence, there is no firm basis for the skepticism on the part of many evolu-tionary psychologists with respect to whether social scientific claims are consistent with evolutionary theory. And third, evolutionary psychologists' insistence that social scientific claims be consistent with evolutionary theory derives not from a principled commitment to the mutual consistency of scientific truths, as is usually claimed, but rather from an a priori assumption that the principles of evolutionary theory are necessary for understanding human psychology and behavior. While such an assumption may be methodologically justifiable, it is not a logically necessary constraint on our ability to understand human psychology and behavior.
1 This brings us back to the problem of a priori selectionism (Ahouse 1998). It's customary to measure performance relative to some variant which we know to exist or to have existed. But we have no clue whether the domain-specific cognitive architecture posited by evolutionary psychologists was extant at the time we were evolving. Obviously, if it wasn't around, it doesn't matter how poor domain-general mechanisms perform. They win by default.
2 This is accurate for the purposes of the discussion, but not as a general principle. Inconsistency with evolutionary theory is going to depend on what one takes to be the axioms (along with their implications) of that theory. It is not always clear what these axioms are, such there be. Natural selection is surely one. Genetic drift is probably another. Ironically, then, evolutionary psychologists claims about inconsis-tency are themselves inconsistent with evolutionary theory. For, they propose that a trait cannot evolve other than by direct selection, which is something which the axiom of genetic drift contradicts.
3 In fact, it is not clear why Cosmides et. al. think that the theory of natural selection is compatible with the laws of physics and chemistry. Nor is it clear how they were able to determine that the theory of natural selection cannot be expressed solely in terms of the laws of physics and chemistry.
4 I am grateful to Frank Thompson for helpful discussion on this point.
5 Thanks to Rachana Kamtekar for helpful discussion on this point.
6 See Fodor (2000: 80-84) for related comments on the consistency requirement.
References
Ahouse, J.C. (1998). "The Tragedy of a Priori Selectionism: Dennett and Gould on Adaptationism." Biol-ogy & Philosophy 13(3): 359-391.
Barkow, J. H., L. Cosmides, et al. (1992). The Adapted mind: evolutionary psychology and the generation of culture. New York, Oxford University Press.
Beatty, J. (1987). "Natural Selection and the Null Hypothesis" in Dupré (1987): 53-75.
Buss, D. M. (1991). "Evolutionary Personality Psychology." Annual Review of Psychology 42: 459-491.
Charlesworth, B. (1990). "Optimization Models, Quantitative Genetics, and Mutation." Evolution 44(3): 520-538.
Cosmides, L., J. Tooby, et al. (1992). "Introduction: Evolutionary Psychology and Conceptual Integra-tion" in J. H. Barkow, L. Cosmides and J. Tooby (1992): 3-15.
Daly, M. and M. Wilson (1988). Homicide. New York, A. de Gruyter.
Daly, M., M. Wilson, et al. (1982). "Male sexual jealousy." Ethology and Sociobiology 3(1): 11-27.
Dupré, J. (1987). The Latest on the Best: Essays on Evolution and Optimality. Cambridge, Mass.: MIT Press.
Fisher, R.A. (1930). The Genetical Theory of Natural Selection. Oxford: The Clarendon Press.
Fodor, J. A. (2000). The mind doesn't work that way: the scope and limits of computational psychology. Cambridge, Mass., MIT Press.
Hawkes, K. and R. B. Bird (2002). "Showing off, Handicap Signaling, and the Evolution of Men's Work." Evolutionary Anthropology 11: 58-67.
Holcomb III, H. R. (1993). Sociobiology, sex, and science. Albany, State University of New York Press.
Ketelaar, T. and B. J. Ellis (2000). "Are evolutionary explanations unfalsifiable? Evolutionary psychology and the Lakatosian philosophy of science." Psychological Inquiry 11(1): 1-21.
Kinzey, W.G., (ed) (1987). The Evolution of human behavior: primate models. Albany, State University of New York Press.
Lande, R. (1976). "Natural Selection and Random Genetic Drift in Phenotypic Evolution." Evolution 30(2): 314-334.
Mayr, E. (1983). "How to Carry out the Adaptationist Program?" American Naturalist 121(3): 324-334.
Parker, G. and Maynard Smith, J. (1990). "Optimization Theory in Evolutionary Biology." Nature 348(1): 27-33.
Price, T. and T. Langen (1992). "Evolution of correlated characters." Trends in Ecology & Evolution 7(9): 307-310.
Symons, D. (1992). "On the Use and Misuse of Darwinism in the Study of Human Behavior" in J. H. Barkow, L. Cosmides and J. Tooby (1992): 137-159.
Tooby, J. and L. Cosmides (1989). "Evolutionary psychology and the generation of culture, part I: Theo-retical considerations." Ethology and Sociobiology 10(1-3): 29.
Tooby, J. and L. Cosmides (1990). "On the universality of human nature and the uniqueness of the indi-vidual: The role of genetics and adaptation." Journal of Personality 58(1): 17-67.
Tooby, J. and L. Cosmides (1990). "The past explains the present: Emotional adaptations and the struc-ture of ancestral environments." Ethology & Sociobiology 11(4-5): 375-424.
Tooby, J. and L. Cosmides (1992). "The Psychological Foundations of Culture" in J. H. Barkow, L. Cos-mides and J. Tooby (1992): 19-136.
Tooby, J. and I. DeVore (1987). "The Reconstruction of Hominid Behavioral Evolution Through Strategic Modeling" in Kinzey (ed) (1987): 183-237.
Zahavi, A. (1975). "Mate Selection: A Selection for a Handicap." Journal of Theoretical Biology 53(1): 205-214.
Ma in buona parte la situazione sta cambiando. Ne sono testimonianza le molteplici sponsorizzazioni di progetti legati allo studio della natura umana da un punto di vista evolutivo. Si va così costituendo una spaccatura sempre più netta. Da un lato vi è un nuovo, pionieristico metodo di indagine ispirato alla teoria dell’evoluzione, che si proclama portatore della verità in nome della scienza; dall’altro, vi sono le scienze sociali che proseguono la tradizione formatasi a partire almeno sin dal Settecento.
Una domanda quindi si pone per tutti coloro che oggi sono interessati al programma naturalista: in che misura l’immagine dell’uomo che sta alla base delle scienze sociali tiene conto dei risultati della teoria dell’evoluzione?
Nel saggio qui presentato, Chris Haufe, dottorando presso la Columbia University, discute la principale accusa sollevata da alcuni psicologi evolutivi contro agli scienziati sociali "tradizionali". Secondo l’accusa, gli assunti principali alla base di discipline quali la sociologia e la psicologia non sono consistenti con l’immagine della natura umana che discende dalla teoria dell’evoluzione. Ad esempio, tali discipline assumono che i processi mentali fondamentali sono independenti dal loro contenuto. Un assunto che – sostiene l’accusa – non è compatibile con l’idea che i processi mentali siano frutto dell’evoluzione, ovvero si siano formati sulla base di una necessità di meglio adattarsi all’ambiente circostante.
Haufe presenta i principali esempi di incompatibilità addotti dagli psicologi evolutivi, e discute il concetto di incompatibilità stesso, per poi scagionare le scienze sociali dall’accusa. (a.b.)
EVOLUTIONARY PSYCHOLOGY AND SOCIAL SCIENCE
Chris Haufe, Columbia University, New York
Introduction
Evolutionary psychologists have claimed that social science must "accept with grace the...tenet of mutual consistency among disciplines" in order to "move away from its present fragmented and insular form" (Cosmides et al. 1992: 12-13). In this paper I examine evolutionary psychologists' concept of consistency as well as the tenet of mutual consistency itself and argue that neither can bear the load required by the arguments against social science found in the evolutionary psychological literature.
Many evolutionary psychologists have criticized social scientists for paying insufficient attention to making their theories of human behavior "consistent with what is known in the natural sciences" (Cos-mides, et al. 1992: 4), in particular, evolutionary biology. Social science, they claim, has been con-ducted in an environment largely devoid of evolutionary considerations. By not taking into account the fact that our species evolved just like any other species, researchers have ignored natural selection, a causal process "known to govern all life" (Buss 1991: 461). A lack of proper concern for the forces of evolution would put any account of human psychology in jeopardy, for the simple reason that psycho-logical theories "inconsistent with evolutionary theory stand little chance of being correct" (Buss 1991: 461). In the succinct language of leading evolutionary psychologists Daly and Wilson, "[m]any...theories that are still debated by social scientists implicitly deny the action of natural selection, and are therefore surely wrong" (Daly and Wilson 1988: 7).
In stark contrast to the evolutionarily uninformed social sciences stands evolutionary psychology, a research program whose results are derived from, and thus assumed to be consistent with, the "strong deductive framework" of evolutionary theory (Tooby and DeVore 1987: 189). Because of its strong em-phasis on the centrality of evolution to any plausible theory of human psychology, evolutionary psychol-ogy avoids the risky business of theorizing about human nature which might turn out to be incompatible with what we've learned from evolutionary biology.
Examples of Evolutionary Inconsistency in Social Science
Domain-Generality
The most egregious violation of evolutionary theory perpetrated by social scientists is the assertion that the human mind consists primarily or solely of "general-purpose, content-independent, or content-free mechanisms" (Tooby and Cosmides 1992: 34). Mental mechanisms given free reign over what in-puts to focus on and what behaviors to produce would allow for an unwieldy variety of behavioral plas-ticity, in which the organism will rarely if ever provide the appropriate behavioral response to an envi-ronmental challenge.
Accordingly, to be endowed with broad behavioral plasticity uncon-nected to adaptive targets or environmental conditions is an evolution-ary death sentence, guaranteeing that the design that generates it will be removed from the population (Tooby and Cosmides 1992: 101).
The unerring eye of natural selection will weed such mechanisms out because "the property of freely varying behavior in all dimensions independent of conditions is not advantageous: It is evolutionarily and individually ruinous" (Tooby and Cosmides 1992: 101).
Passive Reception of Information
Another psychological property which social scientists have proposed is the "passive reception" of sociocultural norms, in which individuals learn appropriate behavior by imitating other members of their social group. Here, as in the case of domain-generality, social scientists have taken a "conceptual wrong turn" (Cosmides, et al. 1992: 13) because of their evolutionary ignorance.
"Evolution could not have produced a psyche that functioned as the passive receptacle of information transmitted from the social group, because (among other reasons) many members of the social group have antagonistic interests" (Tooby and Cosmides 1990: 44)
The idea here is simple: the interest of other members of one's social group will often be at odds with one's own interests. Anyone who merely imitated the behavior of others would either be exploited or, at a minimum, behave in a way that didn't necessarily cohere with his ultimate reproductive interests. Instead, evolution would had to have produced an "individually tailored adaptive system" (Tooby and Cosmides 1990: 44) which generated behavior appropriate to the individual's specific adaptive pre-dicament.
Unreliable "Status-Striving" Mechanisms
During a discussion of the evolutionary dynamics of psychological mechanisms, David Buss asserts that the evolution of psychological mechanisms which created a desire to climb the social ladder would be impossible "unless they reliably produced classes of acts that actually led to the increase or mainte-nance of positions within social hierarchies" (Buss 1991: 464). According to Buss, given that "[p]sychological mechanisms evolve because they have behavioral consequences," such mechanisms could only evolve in the instance where they faithfully caused a certain (presumably adaptive) behavioral output.
Family Violence and “Easy Striking Distance”
In their Lakatosian defense of evolutionary psychology, Ketelaar and Ellis assert that the psychologi-cal mechanism(s) for violence cannot be victim-general because this would undermine the “basic metatheoretical assumption of evolutionary theory...that natural selection favors nepotism.” If violence mechanisms did not discriminate between family members and non-family members, then genetic rela-tives would, ceteris paribus, be just as likely to be attacked as would non-relatives. This kind of mecha-nism would be less adaptive than one which discouraged violence against family members. According to Ketelaar and Ellis, for such a mechanism to have evolved would be inconsistent with evolutionary the-ory and “would require a reformulation of the basic assumptions” of that theory (Ketelaar and Ellis 2000: 5). Thus, we can rule out the possibility that violence mechanisms are victim-general, i.e., assuming that natural selection favors nepotism.
What's Supposed to be Inconsistent?
Uniting these claims of inconsistency is the notion of negative effects on fitness. In the case of do-main-general mechanisms, evolutionary psychologists argue that such mechanisms would reduce the pace of cognition to a debilitating rate, and that behavioral responses will more often than not be con-textually inappropriate, both of which are assumed to adversely affect the reproductive success of the organism. For passive receivers of information, their lack of consideration for their particular adaptive situation will result in behavior which goes against their evolutionary interests. Victim-general violence is assumed to lower inclusive fitness. And status-striving mechanisms which did not reliably produce an elevation in social status would not have been adaptive enough to evolve. The essence of the objection is best stated by Tooby and Cosmides themselves: "Adaptive tracking must, of course, have characterized the psychological mechanisms...during the Pleistocene, or such mechanisms could never have evolved", (Tooby and Cosmides 1989: 35) and elsewhere: "...a psyche that contained nothing but general-purpose information-processing procedures could not, in principle, generate adaptive behavior, and therefore is an evolutionary impossibility” (Tooby and Cosmides 1990: 27).
Now, this specific reference is to mechanisms "governing culture," but the message is general: that which is not adaptive cannot evolve. When a proposed behavior or morphological trait is perceived a priori to most likely have negative effects on the fitness of its bearer, it is determined that that behavior could not have evolved. To propose the existence of a trait that could not evolve is to be inconsistent with evolutionary theory. Therefore, to claim that we have some trait that is perceived to be the kind of thing that would have negative effects on fitness is to make a claim which contradicts evolutionary the-ory, a claim which has little chance of being correct.
A Priori Judgment of Negative Effects on Fitness
There are at least three serious problems with these claims of inconsistency. The first has to do with how we can know the effects of a given trait on an organism's fitness. The second has to do with do-mains of fitness. The third has to do with the notion of what could possibly evolve.
With the exception of a trait whose inability to evolve was logically necessary, there is no reason to think that we should be able to tell a priori that a particular trait would have negative effects on the fit-ness of any organism which possessed it. An a priori evaluation of a trait's evolutionary trajectory re-quires that the very concept of the trait entails either that it (a) cannot evolve, or (b) cannot but evolve. This is because, in the absence of a logically necessary trajectory, we cannot know in advance of the actual expression of the trait precisely how it is going to fit into the organism's overall phenotypic econ-omy. Moreover, we certainly cannot know whether a trait would have had negative effects on fitness without knowing the actual historical conditions in which it arose – i.e., without knowing which alternative morphs were available or what kinds of selection pressures the population encountered. But evolution-ary psychologists' charges of inconsistency attempt to do just that. The widespread pronouncement that domain-general cognitive structures could not possibly make up a significant portion of an organism's cognitive architecture because no such structures could have evolved implies that there are no possible phenotypic or ecological conditions under which domain-generality could have any effect on fitness other than an overwhelmingly negative, "ruinous" one. But whether this is the case is an entirely a posteriori matter which cannot be judged in any way other than experiment and observation - i.e., the way biologists usually determine what kinds of fitness effects a trait has. "Imaginations informed by evolutionary theory" (Daly and Wilson 1988: 13), no matter how fertile, cannot furnish us with answers to questions about fitness in advance of the actual performance data.
It is likely that there will be some resistance to the notion that, except for the case in which there is no possible world in which the trait could evolve, we cannot determine the evolutionary trajectory of a trait a priori, but examples can be adduced at will to allay any such fear. Here are just a few. Sterility immediately suggests itself, but quickly retreats owing to the preponderance of insect species where sterile males have been around for a long time. Traits which are extremely physiologically debilitating may also seem like attractive candidates. Again, however, we find these all over the place, from frog chucks to peacock trains. Theory suggests that these traits started off with relatively low costs but, owing to a female preference for them, quickly evolved into exaggerated versions. One explanation for the exaggeration is Ronald Fisher's (1930) "runaway" process, where the female preference for a trait causes it to reach increasingly extreme versions of itself until it becomes so physically debilitating that the fitness advantage gained through increased mating opportunities is counterbalanced by the sharp increase in mortality. Another explanation is Zahavi's (1975) "handicap mechanism," which suggests that males evolve physically taxing traits to show females how good their genes must be if they can withstand the burden of their handicap. These theories are virtual recipes for turning seemingly maladaptive traits into adaptive ones. Any a priori proposal for a trait's evolutionary trajectory would have to incorporate some way of circumventing this recipe. None of the "impossible" traits on offer in the field of evolutionary psychology provide a way of doing this.
Fitness Domains
This brings us to the second general problem. When we want to assess the fitness value of a trait, often what we must do is measure the performance of that trait with respect to a particular task. Good performance can provide some indirect evidence that the trait was selected to perform that task. Poor performance can provide indirect evidence that the trait was not selected to perform that task, but it cannot provide evidence - direct or indirect - that the trait was not selected to perform any task. Poor performance at a particular task may just indicate that we are assessing the value of the trait relative to the wrong task, a task for which it was not selected. This idea forms the foundation of methodological adaptationism, practitioners of which believe that moving on to another task analysis when the trait is shown to have failed at the original task is just how science gets done (Mayr 1983; Beatty 1987; Parker and Maynard Smith 1990).
For example, suppose we suspect a certain local species of bird to be a Batesian mimic based on the fact that it has colored spots similar to some sympatric unpalatable species of butterfly. If the bird is this sort of mimic, its viability should be increased by causing predatory species to avoid it as they avoid the local butterfly. To test our hypothesis, we paint the colored spots of half the members of this bird species so that their coloring is homogeneous. The other half we leave as they are. If our hypothesis is correct, the viability of the painted group should decrease as a result of having lost its ability to mimic the unpalatable butterflies. But lo, we find that viability in the homogeneously colored group actually increases. This could indicate that the colored spots were not selected for their positive effects on viability, contrary to our hypothesis. But it does not show that the spots are not an adaptation. In fact, a subsequent experiment shows that females prefer to mate with males with colored spots. Colored spots have positive effects on fitness through sexual selection and thus may still be an adaptation, even though they negatively impact survival.
What this example shows is that the domain in which we judge the value of a trait for reproductive success may just be the wrong domain in which to judge the value of that trait. So, even if evolutionary psychologists are capable of demonstrating that domain-general cognitive structures in any organism in any environment are, say, very slow, they still have a long way to go in showing that domain generality could not be an adaptation. Perhaps cognition actually slowed down in the transition from chimpanzees to humans in order to allow us to deal with a greater variety of information, or to expand the range of inferences of which we're capable. Perhaps some accidental property of the early female hominin brain produced a sensory bias which caused them to prefer slow thinkers. Of course, all of these alternatives are, like the charge of cognitive sloth, pure fantasy until the appropriate tests are carried out. The point is merely that even if we consider evolutionary psychologists to have had the last word on the issue of computational speed, there are many other domains in which the fitness value of domain generality needs to be assessed. A simple glance at our own anatomy should suffice to drive the argument home. Our relatively large brains made us, ceteris paribus, heavier than our smaller-brained ancestors. But it would be a serious error to suggest that larger brains couldn't have been selected for because of the negative effects on speed or balance created by the added brain mass.
Other Ways of Evolving
Let's suppose that evolutionary psychologists accept our challenge of testing the performance of a trait for each task for which it could reasonably be thought to have been selected and find that the per-formance is in every domain poor relative to other variants that probably existed when the trait was evolving. If there are no domains in which the trait could have outperformed contemporaneous variants, then we are justified in concluding that it is not an adaptation. It is quite another thing, however, to show that the trait could not exist . that, for example, domain-general cognition is "an impossible psychology" (Tooby and Cosmides 1992: 34).
One rather obvious way in which uniformly maladaptive traits can evolve is through genetic or func-tional correlation with other, adaptive traits (Price and Langen 1992: 309-310). Experiments with artificial selection have demonstrated a variety of correlated responses to selection for a specific trait. Selection for tameness in foxes, for instance, brought along with it several "juvenile-like characteristics which... appeared as a side effect of changing hormone concentrations" (ibid: 309). Similarly, selection for certain life-history traits can cause the persistence of traits that are severely maladaptive. A negative genetic correlation between traits that are positively functionally correlated (such as life-history traits) will cause tradeoffs between trait values when there is selection for one but not the other (Charlesworth 1990). Conversely, single genes may produce multiple traits (i.e., traits with a perfect genetic correla-tion), some of which are adaptive, some of which are maladaptive. Where the positive effects outweigh the negative, maladaptive traits can persist. Let it suffice to say that when there is the right kind of correlation between traits, maladaptive traits can and indeed do evolve. Thus there is nothing "impossible" about the evolution of a psychology that might have been slower than other variants, or a personality type that learned behavior through socialization rather than adaptive planning.
Living in a post-Spandrel world, evolutionary psychologists are presumably acutely aware of evolu-tionary avenues that do not follow the path of natural selection. Indeed, one can find exhortations to that effect across the literature. However, given the continual a priori evaluation of traits and references to things which "could not evolve" or are "impossible," I see no reason to think that the professed sensitivity to modes of evolutionary change other than natural selection is anything other than lip service. If evolutionary psychologists took seriously forces like drift and character correlation, there would be no reason for them to claim that, for example, domain-general cognitive architectures could not evolve, due to their possible negative effects on fitness.
Real Inconsistency
We have seen that claims of inconsistency on the part of evolutionary psychologists fail to show in-consistency, due to a mistaken conception of what it means to be "inconsistent with evolutionary theory." If evolutionary psychologists have painted an inaccurate picture of inconsistency, then what is the right picture?
In order to actually be inconsistent or incompatible with evolutionary theory (by which evolutionary psychologists typically mean "natural selection") an hypothesis would have to entail that the proposed trait was selected for because of its negative effects on reproductive success. This is something that Darwin's theory of natural selection cannot sustain logically. To make this condition clearer, consider what might (or might not) have caused the widespread existence of human noses. If we wanted to ex-plain the existence of human noses in a manner consistent with evolutionary theory, then we could not, for example, assert that noses caused the people who had them to have less children, and that those children (who acquired noses through genetic inheritance) also ended up having less children because of their noses. The reason that this explanation would be inconsistent with evolutionary theory is this: evolutionary theory entails that a property cannot be selected for because of its negative effects on fit-ness, and our explanation of the evolution of noses asserts that the nose was selected for because of its negative effects on fitness. Evolutionary theory says “p,” and our theory says “¬p.”
Of course, as I have already argued in detail, this does not mean that noses could not have evolved unless their fitness effects were non-negative. For noses to have negative effects on fitness and still evolve is perfectly consistent with Darwin's theory of evolution. Suppose the current version of the hu-man nose Nc was caused by the same mutation which gave us the opposable thumb, and suppose that Nc worked much more poorly than the previous version of the nose, Np. Even though the fitness effects of Nc relative to Np were negative, the overwhelmingly positive effects of the opposable thumb allowed the inferior Nc to spread through the population. Now, even though this example is strictly fiction, it nonetheless remains a fact that the ability of properties with negative fitness effects to become wide-spread is a phenomenon the possibility of which is not in serious doubt.
Social Science and Real Inconsistency
Although it may not be rare for social scientists to offer some casual remarks on why a particular mode of human behavior might have evolved, I know of no explanation proffered by a social scientist which takes the form of the human nose explanation outlined above––i.e., an account which attempts to explain the existence of a behavior by way of reference to its negative effects on fitness. The closest thing we see to such an explanation comes from evolutionarily-minded anthropologists who argue that certain behaviors evolved via the handicap mechanism, where the high viability costs of a behavior (e.g., hunting) signal that the person performing it must have "good genes" (Hawkes and Bird 2002). In this case, the negative fitness effects (in the domain of viability) of hunting are in a sense causally related to its evolution, but the reason hunting is adaptive is because of the preferential treatment received by men who hunt, not the viability-reducing effects. The fitness advantage caused by the preferential treatment outweighs the disadvantage incurred by the behavior itself.
The work of Hawkes and other evolution-savvy social scientists, while interesting and important, stands at the margin of social science. The vast majority of claims made by social scientists are com-pletely divorced from evolutionary considerations. But if we do consider it important for social scientists to make their theories consistent with evolutionary theory, we are presented with a puzzle: How do we determine whether a social scientific claim (SSC) is consistent with evolutionary theory when that claim does not mention evolution?
Above we saw the evolutionary psychological attempt to solve this puzzle––viz., to judge the poten-tial fitness consequences of a behavior and see whether, given those consequences, it could have evolved. This strategy turned out to be a chimera for two reasons: (1) we cannot judge a priori the fit-ness consequences of a behavior, and (2) fitness consequences alone cannot tell us whether a trait could have evolved. However, the strategy itself is suggestive of the kind of approach we must take. If indeed we are to check SSCs against evolutionary theory, we will need some way of rephrasing or trans-lating them into the predicates of evolutionary biology. Unfortunately, we have not yet been able to ac-complish this.
What can we say in general about the compatibility (by which I mean something like "mutual re-alizability of all statements") of two theories in advance of the requisite translation? If we have inde-pendent reasons to believe that each theory is true, we could infer that the theories are compatible from the fact that no two statements can be true and contradict each other. This kind of reasoning might be behind Cosmides' et al. assertion that "the theory of natural selection cannot, even in principle, be ex-pressed solely in terms of the laws of physics and chemistry, yet it is compatible with those laws" (Cos-mides, et al. 1992: 4). In the absence of shared predicates, this appears to be the only option we have available to us. Accordingly, we should apply the same rule when we find we are unable to translate the predicates of SSCs into the predicates of evolutionary biology. Where an SSC appears to be well-grounded independently of its relationship to evolutionary theory, we might properly conclude that it is compatible with evolutionary theory.
Contrastingly, Harman Holcomb III has suggested that the very fact that SSCs are formed in a se-mantics different from that of evolutionary biology is itself reason enough to suspect that SSCs are in-consistent with evolutionary theory.
"If social scientists are free to describe and explain behavior in ways that need not take into account biological evolution, then it is likely that their descriptions and explanations of human behavior will often be ir-relevant to or inconsistent with evolutionary theory" (Holcomb III 1993: 122).
I find the source of Holcomb's skepticism elusive. Why assume that social scientists are likely to make claims inconsistent with evolutionary theory if they do not use it as a guiding principle? A SSC that is internally consistent and well-supported should inspire us to give credence to it, irrespective of whether it is formulated in evolutionary semantics. Of course, given the lack of a common semantics between the social and evolutionary sciences, it's certainly possible that most SSCs could turn out to be inconsistent. But to predict in advance of the translation of one into the other (or both into a third lan-guage) that
it would be a miracle if a social science unguided by evolutionary con-siderations would on its own describe and explain human behavior in a way directly subsumable under evolutionary theory (Holcomb III 1993: 123)
strikes me as unreasonable. Given that we are admittedly incapable at present of judging the compati-bility of SSCs and claims made in evolutionary biology, it's unclear upon what Holcomb's pessimistic stance could be founded. If we have yet to determine the compatibility with evolutionary theory of even a single SSC, we could have no basis for thinking that their consistency with evolutionary theory would be unlikely or miraculous.
Is Consistency Important?
Quite apart from evolutionary psychology's misconceptions regarding what consistency with evolu-tionary theory is or whether we can determine whether some SSC achieves it, we should ask whether they're right to insist that social scientists' claims be consistent with evolutionary theory (along with the rest of the natural sciences). The fact that evolutionary psychologists are wrong about whether social scientists' claims and the claims of natural scientists are mutually consistent does not immediately let social scientists off the consistency hook. We still need to assess the requirement of consistency in and of itself.
There's a clear sense in which the demand of mutual consistency is, as Tooby and Cosmides (1992: 12) have emphasized, "absolute" and "incumbent upon all valid scientific knowledge". If statements in the natural sciences are true, social scientists' claims cannot both be true and contradict those state-ments. But the mutual consistency of true statements isn't a requirement restricted merely to the natural and social sciences. No two statements––be they within the same discipline, between disciplines, or not within any discipline––can both (1) be true and (2) contradict one another. Thus, the requirement of con-sistency extends not just from, for example, evolutionary biology to the social sciences, but even from evolutionary biology to the library sciences. Normally, though, evolutionary psychologists are not em-phatic that statements about, say, microfiche and statements in evolutionary theory be consistent with one another. Yet, the burden of consistency with evolutionary biology weighs no less heavily on the li-brary sciences than it does on the social sciences. Why, then, we might ask, are evolutionary psycholo-gists so insistent that the social sciences observe the consistency requirement, while the library sciences receive no mention at all?
The asymmetry with which evolutionary psychologists treat the social sciences, on the one hand, and the library (as well as all other) sciences suggests that their demands for consistency come not from the logical necessity that all truths must be mutually consistent, but rather from the methodological as-sumption that evolutionary theory––specifically, the theory of adaptation through natural selection––and theories of human psychology and behavior should have a lot do do with each other. If the real motiva-tion was the general requirement of mutual consistency then we should expect to see articles lambast-ing the library sciences for ignoring the principles of quantum mechanics and organic chemistry. But we have no such expectation, and with good reason.
As evidence for the existence of the assumption that the theory of adaptation through natural selec-tion and theories of human psychology and behavior must bear upon one another, consider the claim by Donald Symons that "there is no known scientific alternative to the theory that human nature is the product of natural selection" (Symons 1992: 147). Concerning the properties of organisms in general, even within evolutionary biology itself there are at least two "scientific alternatives" to the theory that those properties are the products of natural selection––viz., genetic drift and character correlation. For example, the notion that some mammalian tooth characters are the product of genetic drift rather than natural selection is an alternative to the theory that those characters are the product of natural selection (Lande 1976), and it's difficult to see precisely what about that alternative could be characterized as "un-scientific." What would undoubtedly be unscientific would be to rule out a priori the possibility that mammalian dentition is the product of genetic drift. Lande notes,
If the hypothesis of evolution by random genetic drift cannot be consis-tently ruled out from fossil evidence, there would be no basis for sup-posing that the phenotypic evolution is purely a result of natural selec-tion, and random genetic drift would emerge as a potentially significant motive force in evolution (1976: 321)
The same presumably is true for the aggregate property "human nature," which refers to the bundle of psychological and behavioral properties that are common to all humans in virtue of their shared genetic inheritance. For each property, there will be a variety of "scientific alternatives" to a selectionist account, many of which will come from outside evolutionary biology. As evolutionary psychological folk hero George C. Williams warned us many years ago,
This biological principle [i.e., adaptation through natural selection: CH] should be used only as a last resort. It should not be invoked when less onerous principles, such as those of physics and chemistry or that of unspecific cause and effect, are sufficient for a complete explanation (1966: 11).
Each of these alternatives will need to be entertained in a systematic, empirical way in order for our ex-planations of the focal properties to demand any sort of confidence.
Further evidence for the existence of an assumption that theories of human psychology and behav-ior need always to intersect with the theory of natural selection is provided by the assertion that there is an "irrefutable" logical progression from the premise that physiological properties are generally adaptive to the conclusion that "human nature, including human psychology, must have...an adaptive logic" (Daly et al. 1982: 23). Generally speaking, it does not follow that some biological property (e.g., our psychol-ogy) is an adaptation from the fact that another property (e.g., our physiology) is plausibly an adaptation. This inference procedure would effectively license us to conclude that all of our properties are adapta-tions in the case where we found that one of them was. Whether or not two organismal traits have simi-lar evolutionary histories (specifically in the respects relevant to determining whether they are adapta-tions) is going to depend on the particular historical facts of each trait considered separately, not on the supposed existence of a logical progression which dictates that a commitment to adaptation some-where is a commitment to adaptation everywhere.
The assumption that theories of human behavior and the theory of natural selection have a lot to do with each other is based on the view that it is true a priori that the theory of natural selection, in Jerry Fodor's words, "importantly constrains" theories of human psychology and behavior (Fodor 2000: 83). But the theory of natural selection can only constrain theories of human psychological and behavioral traits in the cases where those traits were actually selected for. And while it may be true in some cases––even every case––that a particular trait's evolutionary history suggests that it is the product of direct selection, it is not true a priori. For, as we have seen, even functional traits may arise through evolution-ary means other than natural selection. Thus, whether the theory of natural selection and theories of human psychology and behavior should have a lot to do with each other is not an a priori truth but will instead depend on the particular psychological or behavioral trait in which we are interested. So evolu-tionary psychologists are not justified in their use of the theory of natural selection as a logical constraint on the social sciences.
Now although it is not an a priori truth that evolutionary theory and theories of human behavior should have a lot to do with each other, that such a relationship exists between them may be a reason-able methodological assumption. There doesn't seem to be anything in principle wrong with adopting it as a methodological assumption and then going on to see whether work done under that assumption bears any fruit, in the same way there should not be anything in principle wrong with adopting adapta-tionism as a methodological assumption. But there is something in principle wrong with demanding that anyone who endeavors to study human behavior adopt that assumption, in the same way it would be wrong and scientifically ridiculous to demand that every evolutionary biologist adopt adaptationism as a methodological assumption. We know that many important properties of organisms are not adaptations, and to push on as though they were would certainly not help to expand the breadth of biological knowl-edge.
Conclusion
The arguments in this chapter have aimed at establishing three general claims. First, evolutionary psychologists are typically mistaken in their beliefs about what kinds of phenotypic properties are incon-sistent with evolutionary theory, a mistake which derives from a general misconception regarding what consistency with evolutionary theory actually is. Second, given the correct conception of consistency there is no clearly understood method for determining whether claims in the social sciences are consis-tent with evolutionary theory. Hence, there is no firm basis for the skepticism on the part of many evolu-tionary psychologists with respect to whether social scientific claims are consistent with evolutionary theory. And third, evolutionary psychologists' insistence that social scientific claims be consistent with evolutionary theory derives not from a principled commitment to the mutual consistency of scientific truths, as is usually claimed, but rather from an a priori assumption that the principles of evolutionary theory are necessary for understanding human psychology and behavior. While such an assumption may be methodologically justifiable, it is not a logically necessary constraint on our ability to understand human psychology and behavior.
1 This brings us back to the problem of a priori selectionism (Ahouse 1998). It's customary to measure performance relative to some variant which we know to exist or to have existed. But we have no clue whether the domain-specific cognitive architecture posited by evolutionary psychologists was extant at the time we were evolving. Obviously, if it wasn't around, it doesn't matter how poor domain-general mechanisms perform. They win by default.
2 This is accurate for the purposes of the discussion, but not as a general principle. Inconsistency with evolutionary theory is going to depend on what one takes to be the axioms (along with their implications) of that theory. It is not always clear what these axioms are, such there be. Natural selection is surely one. Genetic drift is probably another. Ironically, then, evolutionary psychologists claims about inconsis-tency are themselves inconsistent with evolutionary theory. For, they propose that a trait cannot evolve other than by direct selection, which is something which the axiom of genetic drift contradicts.
3 In fact, it is not clear why Cosmides et. al. think that the theory of natural selection is compatible with the laws of physics and chemistry. Nor is it clear how they were able to determine that the theory of natural selection cannot be expressed solely in terms of the laws of physics and chemistry.
4 I am grateful to Frank Thompson for helpful discussion on this point.
5 Thanks to Rachana Kamtekar for helpful discussion on this point.
6 See Fodor (2000: 80-84) for related comments on the consistency requirement.
References
Ahouse, J.C. (1998). "The Tragedy of a Priori Selectionism: Dennett and Gould on Adaptationism." Biol-ogy & Philosophy 13(3): 359-391.
Barkow, J. H., L. Cosmides, et al. (1992). The Adapted mind: evolutionary psychology and the generation of culture. New York, Oxford University Press.
Beatty, J. (1987). "Natural Selection and the Null Hypothesis" in Dupré (1987): 53-75.
Buss, D. M. (1991). "Evolutionary Personality Psychology." Annual Review of Psychology 42: 459-491.
Charlesworth, B. (1990). "Optimization Models, Quantitative Genetics, and Mutation." Evolution 44(3): 520-538.
Cosmides, L., J. Tooby, et al. (1992). "Introduction: Evolutionary Psychology and Conceptual Integra-tion" in J. H. Barkow, L. Cosmides and J. Tooby (1992): 3-15.
Daly, M. and M. Wilson (1988). Homicide. New York, A. de Gruyter.
Daly, M., M. Wilson, et al. (1982). "Male sexual jealousy." Ethology and Sociobiology 3(1): 11-27.
Dupré, J. (1987). The Latest on the Best: Essays on Evolution and Optimality. Cambridge, Mass.: MIT Press.
Fisher, R.A. (1930). The Genetical Theory of Natural Selection. Oxford: The Clarendon Press.
Fodor, J. A. (2000). The mind doesn't work that way: the scope and limits of computational psychology. Cambridge, Mass., MIT Press.
Hawkes, K. and R. B. Bird (2002). "Showing off, Handicap Signaling, and the Evolution of Men's Work." Evolutionary Anthropology 11: 58-67.
Holcomb III, H. R. (1993). Sociobiology, sex, and science. Albany, State University of New York Press.
Ketelaar, T. and B. J. Ellis (2000). "Are evolutionary explanations unfalsifiable? Evolutionary psychology and the Lakatosian philosophy of science." Psychological Inquiry 11(1): 1-21.
Kinzey, W.G., (ed) (1987). The Evolution of human behavior: primate models. Albany, State University of New York Press.
Lande, R. (1976). "Natural Selection and Random Genetic Drift in Phenotypic Evolution." Evolution 30(2): 314-334.
Mayr, E. (1983). "How to Carry out the Adaptationist Program?" American Naturalist 121(3): 324-334.
Parker, G. and Maynard Smith, J. (1990). "Optimization Theory in Evolutionary Biology." Nature 348(1): 27-33.
Price, T. and T. Langen (1992). "Evolution of correlated characters." Trends in Ecology & Evolution 7(9): 307-310.
Symons, D. (1992). "On the Use and Misuse of Darwinism in the Study of Human Behavior" in J. H. Barkow, L. Cosmides and J. Tooby (1992): 137-159.
Tooby, J. and L. Cosmides (1989). "Evolutionary psychology and the generation of culture, part I: Theo-retical considerations." Ethology and Sociobiology 10(1-3): 29.
Tooby, J. and L. Cosmides (1990). "On the universality of human nature and the uniqueness of the indi-vidual: The role of genetics and adaptation." Journal of Personality 58(1): 17-67.
Tooby, J. and L. Cosmides (1990). "The past explains the present: Emotional adaptations and the struc-ture of ancestral environments." Ethology & Sociobiology 11(4-5): 375-424.
Tooby, J. and L. Cosmides (1992). "The Psychological Foundations of Culture" in J. H. Barkow, L. Cos-mides and J. Tooby (1992): 19-136.
Tooby, J. and I. DeVore (1987). "The Reconstruction of Hominid Behavioral Evolution Through Strategic Modeling" in Kinzey (ed) (1987): 183-237.
Zahavi, A. (1975). "Mate Selection: A Selection for a Handicap." Journal of Theoretical Biology 53(1): 205-214.