A positive recent development in the quest to extend the methods of biological evolution to the phenomena of human society is the renewed respectability of the related categories of functionalism and group selection. David Sloan Wilson’s recent study of religion, Darwin’s Cathedral (Chicago UP, 2002), is exemplary in this regard. Having played a role in evolutionary biology’s recent abandonment of the dogma of individual selection, Wilson is not afraid to challenge the anthropological establishment’s own version of the dogma, the rejection of (community-level) functionalism, whose “demise . . . in the social sciences bears an eerie resemblance to the demise of multilevel selection theory in biology” (48). Wilson revives Durkheim’s key insights that the function of religion is to reinforce the cohesion of the social order rather than to provide supernatural explanations of the functioning of the natural world; “pragmatic” religious thinking serves a different function from factual thinking. A chapter-length study of Calvinism and briefer discussions of tribal and Judeo-Christian religion provide convincing examples of the functionality of religion in the practical world of human relations, both in the promulgation of society’s ethical laws (a phenomenon well observed by Durkheim) and in the discouraging of “free riders” who would profit from the ethical order to enrich themselves at the expense of the community (a phenomenon that Durkheim does not explicitly discuss). If beliefs in supernatural beings and factually absurd myths have the practical value of discouraging free riders by their very absurdity, then to denounce this absurdity from the factual standpoint is to commit a category error. Wilson recycles Tertullian’s paradoxical credo quia absurdum into a criterion of evolutionary adaptivity.

Wilson’s polemics even with so radically different a mode of thought as rational choice theory, which conceives religion as a quasi-economic exchange relationship with supernatural beings, have a gentle touch. Convinced of “the need to step back and rebuild the social sciences from first principles, making the various subdisciplines consistent with each other and with evolutionary biology (84),” Wilson sees functionalism and multilevel selection as holding out the possibility of unifying the presently disparate social sciences under the banner of a Darwinism broad enough to encompass not merely genetic but cultural evolution.

Yet missing from Wilson’s account, as from Durkheim’s, is any account of the genesis of religious ideas. A biological evolutionist more than anyone should be sensitive to the requirement that functionalism not be divorced from genealogy. To say that an organ has evolved because it is useful is a valid explanation only if one can show from which earlier structure it evolved, and along what path. In biological evolution, although there may be gaps in the fossil record, there is no difficulty in principle in tracing this genealogy. The situation is quite different in the uniquely human domain of representation. But on this subject, Wilson’s intellectual sophistication seems to desert him. His reference to Terrence Deacon’s exclusive attribution to humans of symbolic thought (The Symbolic Species, Norton, 1997) suggests that Wilson has never reflected on the problems posed by the genesis of symbolic thought for his theory of religion (Deacon himself is not guilty of this lapse):

Deacon (1998) has recently argued that symbolic thought sets humans apart from all animals and evolved to enable enforced social contracts such as marriage. If he is correct, it will be an impressive confirmation of Durkheim’s claim that human social life is only possible thanks to a vast symbolism. (54-55)

This is strange language to find in a book about religion! Can humanity’s possession of symbolic language be treated as hypothetical (“If he is correct”) in a book that is largely a study of the functions of religious symbolism? Surely not unless one can show how religious practices and discourses would be understood in the absence of “a vast symbolism.” Lacking this, one would expect Wilson to propose a hypothesis concerning the origin of symbolic language and religion, and to take a stand on their coevality. Durkheim refused in principle to entertain speculative theories on the origin of religion because he saw no way of advancing the question empirically. A biologist who treats the function of religion in the context of evolutionary theory cannot be content with such a dismissal. Wilson does not in fact dismiss the question of origin in the case of either religion or language; he simply never raises it. In this, if in nothing else, he seems content to follow uncritically the standard social-science dismissal of originary scenes.

Wilson’s functionalist approach to religion is a significant improvement over the paralyzing descriptivism that has for several generations dominated anthropological practice. To this mode, typified by Clifford Geertz’ method of “thick description,” which views each society as a unique whole that shares a quasi-esthetic or “structural” unity, Wilson opposes the vision of societies as adaptive systems within which “organs” such as religion share across different cultures a fairly specific common function. Yet Wilson has no method for dealing with the crucial fact that religion and language are closely related manifestations of the same fundamental human ability to create arbitrary or symbolic signs.

Although researchers focused on the rational exploitation of the environment by homo economicus consider it inappropriate to speak of the function of religion even as they analyze its operations, only the most radical cognitivists (who see language as an epiphenomenon of the evolution of the brain–a “spandrel” that subsequently proved useful)–have ever thought to deny that language is functional. On the contrary, the evident functionality of language has long been an obstacle to the explanation of its origin, inspiring a thousand variations on the just-so story about the selective advantage of being able to tell the other members of one’s hunting party about the food on the other side of the hill. The emergence of language having been taken for granted as a functional adaptation, a theoretician such as Roy Rappaport (Ritual and Religion in the Making of Humanity, Cambridge, 1999) proceeds to explain religion as operating to limit the disturbance that this adaptation poses to the social order, whether as a means for limiting the social effect of lying and more generally of speculative thinking, or as an application of language to a preexisting animal propensity to ritual behavior. It is surely a positive step to claim that religion finds a pragmatic use for symbolic thought independent of its function to model the phenomena of the natural world. But the burden of explaining the emergence of religion is thereby only rendered more difficult; if religion is not a mere byproduct of the emergence of symbolic thought, then even the most robust theory of language origin would be insufficient to explain how the early users of language, whose original function was helping us to make sense of the object world around us, ever came upon religious discourse as a tool of social pragmatism.

This difficulty disappears if we hypothesize that language and religion had a common origin in an event where symbolic language was from the outset pragmatic, emerging as a new means of deferring intra-group conflict. If we attribute to originary representation the pragmatic social function of religion, we require no additional hypothesis to stipulate that the factual component of language, eventually embodied in the declarative sentence, proved adaptive in dealing with the natural world. As a result of this adaptation, religious representation, which concentrates on the reinforcement of the social order through the deferral of internal conflict, would increasingly be perceived as separate from the phenomenon of representation in general.


In a whimsical coda that “begins and ends [his] career as an inspirational writer” (233) Wilson offers an “esthetic” paradigm of the different modes of interaction of religious discourse with objective reality that seems meant to persuade us that religious thought, which he has just shown us to be fundamentally utilitarian, also has beauty on its side. Just as with language, Wilson has every reason to include the esthetic in the religious configuration, but he never stops to consider that if religious representations are ethically pragmatic and pragmatic representations are beautiful, one should expect there to be an originary connection between religion, ethics, representation, and beauty. This failure to examine the relationship between religion and language prevents Wilson from articulating within a unified concept of representation his valuable distinction between its pragmatic and factual uses, and a fortiori from giving anthropological concreteness to the Kantian idea that the aesthetic domain stands between religious-pragmatic and factual representation.

Similarly, Wilson sidesteps the question of defining the anthropological status of the “supernatural.” Although he is right to reject the narrow definition of religious discourse (adopted by the rational-choice theorists) as language concerned with supernatural beings, whether or not Buddhism is a valid counterexample, (Wilson himself notes that “[rational-choice theorists] Stark and Bainbridge contend that Buddhism as actually practiced is chock full of gods” [221]), the propensity of religions to posit the existence of supernatural beings cannot go unexplained. The simplest explanation is that this propensity reflects the transcendental nature of representation in general. Whereas the “profane,” everyday use of language ignores its transcendental origin, religion exists to remind us of it, since the transcendence inherent in representation is the guarantee of the ethical order of the community. The simplest form of this reminder is the attribution to the object of sacred representation of the transcendent qualities that are observable only in the sign. Like the aesthetic oscillation between the artwork and its imaginary referent, but not typically within the compass of immediate individual experience, the definition of a transcendent being tends to evoke skepticism, there being no unequivocal evidence for its existence. (That religious skepticism is found only in the post-Renaissance West is a self-serving Enlightenment myth.) This resistance in turn provokes its defenders to a more explicitly extra-rational–and implicitly pragmatic–expression of faith. Arguing about the “existence of God” is no threat to the ethical values guaranteed by transcendence; the debate leads, however slowly, to the convergence of the anthropologies of the opposing sides.

If the origin of humanity can be fully understood only as an event, then the familiar Darwinian mechanism of genetic mutation does not provide an adequate basis for a hypothesis of human origin. The futility of speculations about the appearance of the “language gene” and its distribution within the protohuman group–must the mutation be present in twomembers of the group so that they can converse together?–should make us sympathetic to the grain of truth in Creationism.  Although the attribution of natural consequences to supernatural beings is a category error, the emergence of the exclusively human phenomenon of representation at the point of contact of the supernatural and the natural can be conceived only as a singularity, a memorable scene, an event. The Genesis creation story, whatever its other qualities and deficiencies, exemplifies the event-nature of human origin whereas the Darwinian account does not.

Wilson’s up-to-date conception of evolution markedly includes cultural evolution. Genetic change does not precede cultural change; in principle it follows it. The genetic changes in genus Homo have been dominated by “Baldwinian,” behavior-driven evolution, and the most crucially important human behaviors are those concerned with representation. Far from language being a byproduct of the evolution of the human brain, to the extent that language remains adaptive over a long period, individuals and their brains tend to be selected for their facility with it, whether or not this is centralized in a “language module.” With the appearance of Homo sapiens, all of whose members are more or less equally adapted to language and related forms of representation, the genetic element in human evolution becomes still less significant, and the crucial battles for the survival of the fittest are carried out exclusively on the plane of social organization.


Although the notion of a singular event is not incompatible with Wilson’s extension of Darwinian evolution, it is fully understandable that he does not consider this hypothesis, let alone discuss its consequences for the theory of human evolution.  The fact remains that there is no accepted mode of discourse in either the social or natural sciences in which a hypothetical singularity, by its very nature inaccessible to empirical observation, can be the central organizing principle. Only in religious thought and its secular derivatives in philosophy is such thinking considered permissible.

Wilson’s updated Darwinism can accommodate the brain’s Baldwinian adaptation to representation, but it has no way of specifying that the behavior that is to drive future evolution must emerge on a scene. Although the originary hypothesis must remain outside the domain of falsifiable propositions that constitute Darwinian theory in particular and social science in general, this hypothesis is nonetheless, in a manner analogous to the true but indemonstrable propositions of arithmetic whose necessary existence was demonstrated by Gödel, a necessary unprovable supplement to any empirically based model of human evolution.