Definitions and Levels of Although understandable given their strictly enlargement resulted in the capacity to Analysis biological interpretation of exaptations, this practice religion, and (c) that religious approach is problematic because the authors practices function to assuage the fear of death

sought to apply their standards of evidence to (which would itself be required to meet the Theodore Beauchaine the list of exaptations proposed by Gould evidentiary standards of a spandrel). Al- State University of New York (1991), namely language, religion, principles though the first two of these conjectures may at Stony Brook of commerce, warfare, reading, writing, and be true, they can at best be confirmed only at

the fine arts. All of the items on this list are the pseudoempirical level. Gould (1991) rec- Although Buss, Haselton Shackelford, psychological phenomena borne of a ognized this and suggested that the term Bleske, and Wakefield (1998) presented a structural complexity (i.e.. the human brain) aptation be applied to cases where a lack of topic that deserves mainstream attention in that is not well understood and that was not historical evidence precludes the determina- psychology, I am concerned about their arti- sufficiently considered by the authors. Here tion of whether a characteristic is an adapta- cle on several grounds. My main concern is again. Buss et al. may have misinterpreted tion or an exaptation. Furthermore, confirm- with their interpretations of Gould’s (1987 Gould, who did not use human brain size ing or disconfirming the third conjecture is 1991, 1997) writings. Buss et al. suggested merely as an "example of an exaptation" (p. far more difficult than establishing the utility that Gould (1991) was inconsistent in his 539), as the authors stated. Rather, Gould of female orgasm as a sperm transport usage of the term exaptation, and that his suggested that the human brain, by virtue of mechanism, because the functional level of “stated definitions seem to require that these both complexity and flexibility, is “the best analysis is psychological, not structural. effects and culturally useful features available case for predominant exaptation-- All of the proposed exaptations listed (exaptations) must contribute to fitness and in other words, for a near certainty that by Gould (1991), because they are specified have specifiable biological functions [italics exaptations must greatly exceed adaptations at the psychological level, are similarly pre- added] to qualify as exaptations” (p.541). Yet in number and importance” (p. 55). Gould cluded from meeting the strict evidentiary nowhere in Gould’s (1987 1991, 1997) offered the practice of religion as an example standards set forth by Buss et al. (1998). Yet, writings did he suggest that the function and suggested that our enlarged brains force it would be unfortunate indeed if psycholo- performed by an exaptation needs to be us to confront our own mortality. Because it gists, in an effort to meet such standards, biological. As Buss et al. reported, Gould is quite unlikely that brain enlargement were to reject Gould's distinctions and con- (1991) offered two definitions of exaptation: evolved to serve this end, the practice of tinue in adaptationist practices. We should "a feature, now useful to an organism, that religion performs an exapted function by recognize the error in logic of inferring evo- did not arise as an adaptation for its present moderating, through a variety of themes, our lutionary cause from current consequence, role, but was subsequently coopted for its evolutionarily functional fear of death. Thus, whether or not we can empirically demon- current function" (p.43) and "features that recognition of our own mortality is a strate the existence of psychological exapta- now enhance fitness, but were not built by spandrel, and the practice of religion is an tions. natural selection for their current role" (p.47). exaptation. Thus, although exaptations always provide a Moreover, it is a priori unlikely that REFERENCES current function, that function need not be such complex psychological phenomena can biological. In fact, Gould’s (1991) thesis is meet the standards of evidence proposed by Buss, D. M., Haselton, M. G., Shackelford. geared toward exapted functions performed Buss et al. (1998). Specifically, they suggest- T. K., Bleske, A. L., & Wakefield, J. C. by the brain to serve psychological rather ed that evidence of special design for a hy- (1998). Adaptations, exaptations, and than biological functions. pothesized function be demonstrated before spandrels. American Psychologist, 53, 533- Additionally, Buss et al. (1998) concluding that any structure or behavior is 548. suggested that Gould (1991) used the term adaptive. As an example, they presented the Gould, S. J. (1987). Freudian slip. Natural exaptation “to cover novel but functionless hypothesis that female orgasm serves the History, 96, 14-21. uses or consequences of existing adaptive function of facilitating sperm trans- Gould, S. J. (1991). Exaptation: A crucial characteristics" (p. 539) My reading of Gould port, for which evidence is reportedly lack- tool for an evolutionary psychology. (1987, 1991, 1997) suggests otherwise: ing. It is thus concluded that female orgasm Journal of Social Issues, 47, 43-65. Structures or characteristics for which there does not serve the hypothesized adaptive Gould, S. J. (1997). The exaptive excellence are no identifiable functions, either current or function. Although instructive, this example of spandrels as a term and prototype. historical, are always spandrels. Thus, the trivializes the difficulty of falsifying, at the Proceedings of the National Academy of term spandrel subsumes what Buss et al. phenotypic level, hypotheses about exapted Sciences, 94, 10750-10755. referred to as functionless by-products. When psychological functions, such as the practice a spandrel is coopted (exapted) for a function, of religion. There are at least two reasons for it becomes an exaptation; there is no need to this. First, as outlined by Buss et al., exapta- Correspondence concerning this comment differentiate between spandrels and tions "carry the additional evidentiary should be addressed to Theodore Beauchaine, functionless byproducts. Ironically, Gould burdens of documenting both later co-opted Department of Psychology, State University (1991) coined the terms exaptation and functionality, and a distinctive original of New York, Stony Brook, NY 11794-2500. spandrel to avoid this sort of confusion. adaptational functionality" (p. 546). Thus, to Electronic mail may be sent to Additionally, although Buss et al. confirm the exapted function of religious [email protected]. (1998) sought to outline the usefulness of the practice, one would be required to terms adaptation, exaptation, and spandrel for demonstrate (a) that enlarged brains were the science of psychology, most of their naturally selected for reasons independent of discussion was restricted to structural rather religious practice, (b) that such brain than psychological levels of analysis.

June 1999 • American Psychologist 439-440