Twenty-odd years of originary thinking have given me ample occasion to reflect on the resistance it encounters among academics. Personal equations aside, the basis for this resistance would seem to be that, in our time at least, the production of (publishable) research is not visibly furthered by a mode of thinking whose function is to provide a minimal basis for all more specific thought about the human. If everyone could get together and agree on a common ground for such thought, generative anthropology would be the only possible choice, but, even if we grant this less than modest proposition, the fact is that “everyone” isn’t going to get together: the intellectual marketplace, like the economic one, functions through individual exchanges that only incrementally modify the network as a whole. This is not to deny that such a common ground must be found, but to suggest that the sense of its necessity too must propagate incrementally through the network.

The aim of last year’s studies on the modern history of the question of language origin (roughly, Chronicles 175-198) was to demonstrate that, in approaching this issue, thinkers both before and after the great divide that separates twentieth-century science from eighteenth-century speculation end up proposing, albeit in an unconscious, irresolute fashion, versions of what I call the originary hypothesis.

(The transitional nineteenth-century is generally barren in this area, as witness the oft-quoted 1866 declaration of the French Académie des Sciences that hypotheses concerning the origin of language would no longer be entertained. The turn from language to languages is best exemplified by Wilhelm von Humboldt–the subject of an forthcoming Chronicle. For Humboldt, who tried to learn every known language, the focal point of the anthropological study of language is no longer, as for his precursorHerder, the hypothetical diachronic interface between the prelinguistic and the linguistic state, but the empirically accessible synchronic differences of form between language families. In this turn, if not in his models for linguistic analysis, Humboldt anticipates the structural linguistics of Saussure. No doubt Humboldt, whose four categories of agglutinating, incorporating, isolating, and inflected languages are still used today, considered the last type–best exemplified by Sanskrit–superior to the others, but even this“Aryan” quasi-racism should be associated with the synchronic turn of his thought: although nineteenth-century historicism is often linked to colonialism and “orientalism,” the fundamental thrust of racism is not to distinguish among different stages of development but among different kinds of beings.)


A demonstration of the necessity of the originary hypothesis may appear chimerical to those who consider that “real research,” pursued in the absence of any such hypothesis, is closing in on what is for them the essence of the matter–the description of the present state and evolutionary development of the linguistic processes in the brain. But despite the vast scientific progress since the Enlightenment, this scientism suffers from a materialist fallacy that, on the specific point of language origin–and therefore on that of the essence of the human as such–recalls Descartes’ location of the soul’s action on the body in the pineal gland or Gall’s proposal, mocked by Hegel, to divine our mental abilities from the shape of our skulls.

Whether in the era of pre-scientific speculation or in that of empirical research, outside of Generative Anthropology, models of the origin of language without exception rely on an originary hypothesis that remains unaware of itself. The individual studies presented in the earlier Chronicles have noted this fact. The thematization of the originary hypothesis is not a product of technological advances, nor is its relationship to earlier hypotheses of language origin comparable to that between Newton and Aristotle orEinstein and Newton. This thematization depends rather on the historical, which is to say, the ethical experience of postmodernity, the ultimate revelation of which is the impossibility of avoiding the model of exchange as the basis for human relations. It openly avows the leap that attends the creation of the sign-world and seeks to minimize this leap without denying it. To refuse to formulate such a hypothesis is either to deny the discontinuity between animal and human or to subordinate it to the greater discontinuity between human and God.

The originary hypothesis is equally impatient with either secular or religious refusal; conversely, it tolerates equally well secular or religious expression. Its difficulty with religious thought is not the appeal to transcendence as such but the non-minimal nature of this appeal. Since humans may be shown to exist and God or gods cannot, a secular hypothesis of origin might seem “more minimal” than a religious one. But the hypothesis of human origin is not a question that may be posed from outside the human experience; and within that experience, the transcendent Being personified in the Judeo-Christian tradition as God is not detachable from the communal guarantee that makes language and other representational forms possible. We need not “believe in God,” not even “in the foxholes,” to be aware of this dependency.

There is clearly much to be gained, on both sides, by denying the symmetry of the human and the sacred, by either setting God prior to humanity or humanity prior to God–or, like the churchgoing scientist, doing both at once. Setting aside the first choice for the moment, the second has obvious immediate advantages for empirical research. To understand the sacred as a mere artifact allows one to explore its social usefulness while ignoring it as a causal factor. It then becomes possible to draw analogies between animal and human communication while “bracketing” the gap between human representation and its animal precursors. This research could no doubt be carried out, and one day no doubt will be, within the framework of the originary hypothesis; but at its present stage, that empirical scientists’ own hypotheses can be demolished on a priori “philosophical”–read minimal anthropological–grounds, seems irrelevant to their, if not to our, concerns.

Yet in the long term, the truth can never be irrelevant. If indeed language is a social process generated not by some new configuration of the individual brain but by the interaction of a group, then not only can the origin of language not be presented directly in terms of biological evolution, but language as such must be understood as a process of interaction between human beings rather than as a set of connections within an individual brain. The attempts of Derek Bickerton or the far more anthropologically aware Terrence Deacon to model a possible origin for language are not innocent footnote speculations “supplementary” to the serious material. The very awkwardness of these speculations, their throwaway nature, betrays both the necessity of acknowledging the question and the inadequacy of the less than rigorous answers suggested.

To those who seek the secret of language in the brain, I propose as a thought-experiment that, since religion and language emerge simultaneously, one seek the secret of religion in the brain as well. Perhaps there is a region in the brain (in the “right brain,” no doubt) devoted to religious concepts. Whatever the results of neurological research concerning language, clearly the neurological evolution leading to the doubly articulated languages we know is a consequence of the birth of language, not its cause. But this simple fact, even if admitted, is masked by the elaborateness and complexity of this evolution, whereas our common God-concepts (what corresponds in the linguistic domain to the intellectual structures of theology is not language but linguistic philosophy) show no apparent dependence on evolving brain structures. Language, which I called in The Origin of Language “formal” representation, requires a minimal expenditure of energy. Its representation of the (originally sacred) world has adaptive consequences that drive the joint evolution of language and the brain; Deacon points out that in the course of this evolution, the brain and language adapt to each other. Religion, in contrast, is “institutional”; its representations do not interact freely with the world and we have therefore no adaptive interest in increasing their throughput. On the contrary, the cost (Bataille’s dépense) of ritual is a benefit, since the energy we thus expend is diverted from mimetic violence. But the fact that several hundred thousand years of human evolution have not created a religion area in the brain comparable to those of Broca and Wernicke for language–and that we feel no need to seek such a region–sheds light on the fundamental question of where we should situate our “competence,” to use Chomsky’s term, in either language or religion.

Both religion and language are, minimally, forms of representation. Precisely because the object of religious worship, however “primitive,” is Being rather than beings, it must be designated by a sign that persists as the correlative of Being in the absence of the being that first gave rise to both sign and correlative. This generative reasoning relegates to subordinate status the question of where the sacred sign resides in the brain. In the originary scene, an associative or “indexical” sign–the “abortive gesture of appropriation”–acquires a “symbolic” value. Language originates as the generation of a new communally recognized connection, independent of the built-in call system (comparable to laughter rather than speech), yet no mere private Pavlovian association.

This argument, so difficult to accept with regard to language, where syntactical issues loom so large, is rejected only out of inertial prejudice in the domain of the sacred. The often-proposed idea that a genetic mutation led to the emergence of speech is a category error, but one less flagrantly absurd than the idea that such a mutation gave rise to the concept of God. (Imagine the thrill of one day discovering the God gene.)


Today’s empirically driven reflection on these subjects is unconcerned with the internal logic of its hypotheses; the all-important goal is to formulate research programs–and obtain funds–for the systematic collection of information. This is a reality of the academic practice of science, whose general effectiveness in providing benefits to humanity it would be churlish to ignore. In the “early modern” past, however, there was no research establishment focused on work-creation, only a set of individual thinkers. It is among these thinkers that the encounter with the logic of the originary hypothesis can best be tested. The continued reliance of those who today style themselves philosophers on the Enlightenment moral reasoning that culminates in Kant, not to speak of that of Plato and Aristotle, demonstrates that since early modern times at most incremental progress has been made in grounding the fundamental principles of human relations. These Enlightenment thinkers, whose ideas are still felt to be alive today, remain the sole interlocutors willing, tacitly to be sure, to discuss with us the question of human origin.

But beyond the dialogue with texts, my real interlocutors will be those in whom this writing will arouse the long-repressed intuition that the origin of language and religion mustbe thought because it is at the root of anthropological thinking. No doubt humanity has all of human time to assimilate the logic of its origin, but the time is coming when we will need to guarantee the continuation of this time itself through the admission that we all hold this logic in common.