My first book of Generative Anthropology, The Origin of Language (TOOL), was published in 1981. At that time, there was not a great deal of literature about language origin. Everyone quoted the ban on the subject imposed in 1866 by the Société Linguistique de Paris. There were a few seekers for the Ursprache, such as Alexander Marshack and Mary Le Cron Foster; Gordon Hewes suggested that human language began with gesture (as I recall, one of his pieces of evidence was that the palm of the hand–whose visibility at a distance is essential for a gesture-language–is never darkly pigmented even among those with the darkest skin). Derek Bickerton’s Roots of Language, which uses pidgins and creoles to make the case for a universal basic syntax, also appeared in 1981.

The originary scene in TOOL was closer to Girard’s model than more recent versions: the moment of the “aborted gesture of appropriation” that becomes the first linguistic sign was situated after rather than before the sparagmos, making deferral a consequence of mimetic violence. I made no attempt in TOOL to associate the hypothetical moment of origin with palaeontological data, and made little reference to primatology. I thought, naively as it turned out, that once the model of origin was supplied–and I was convinced I had supplied it–the empirical details could be worked out by others more familiar with the terrain.


How have my ideas changed since 1981? I still believe not only that my model of language origin is the most parsimonious but that it alone explains the connection between language and religion–a subject virtually ignored since the days of Max Müller. But I have become less utopian concerning GA’s potential for displacing the methods and perspectives of social science. Originary thinking is of value in explaining cultural forms and ideas and as a critique of the conclusions that may be drawn from empirical research; it is neither a program for such research nor a substitute for it.

What of the originary scene itself? In TOOL and for some time afterward, I conceived the scene as an event for which we could of course not expect to find evidence but that we could attempt to reconstruct by fleshing out our hypothesis concerning human culture (“the deferral of violence through representation”) with whatever empirical knowledge we might acquire of the life of our ancestors. As we continue to revise our understanding of humanity’s prehistory, it seems increasingly appropriate to consider the scene as a heuristic construction–and yet not a simple fiction like the social contract or John Rawls’ “original position.”

Given the basic presupposition that the signs of human language are derived not from appetite but from its deferral, the originary event, understood as a “little bang,” the first hesitation that converts a gesture of appropriation into a sign, must have occurred. But the originary hypothesis is stronger than the tautological assertion that, given that C lies on the path between A and B, if you go from A to B there is a moment at which you are at C. This weak notion of a moment of origin does not suffice to explain the scenic nature of human culture. If I emphasize religion in tandem with language, it is because all manifestations of religion take place explicitly on an external or internal scene. Indeed, one could define religion as the most general mode of representation that is explicitly scenic. The simplest understanding of God is as the Being that maintains the permanence of the scene of representation; the consolation of prayer is that, in whatever solitude or misery one finds oneself, its interlocutor is conceived as always on the scene.

The minimal originary hypothesis need not postulate a watershed event that would radically transform relations among the members of the group. The equalitarian configuration of the scene of representation, with the sacred object at the center and the human participants on the periphery, need not replace all at once the pre-human configuration of “pecking-order” power relations. But we can say that, however fleetingly, the event-nature of the event must have imprinted itself on the consciousness of the participants, else there would be no cultural scene–and no history–at all. However minimal, the little bang must nonetheless be understood as an event.

In the Spring 2000 GA seminar, Antoine Philippe of UCI made an insightful point about this event: it is an event in which nothing happens. Projected onto the worldly plane, the conversion of a gesture of appropriation into a “vertical” sign turns an action into a non-action. The first event is a non-event; the first memorable scene is constituted by the uneventfulness, however temporary, that takes place on it. This emptiness, Sartre’s néant, is the space-time ofrepresentation, which is the ground of the specifically human freedom that Sartre finds there.


The most serious misconception about GA is that what I call the “minimal” requirement of an originary hypothesis is rather a violation of the rule of parsimony than an expression of it. Discussions of morality or religion tend to begin either in medias res, with the data–which may be obtained largely from thought-experiments–or with an a priori model, motivated only intuitively, that is subsequently applied to this data to see if it fits. In the place of an originary hypothesis, the author customarily refers to the état présent of the field and/or to classical texts. This procedure may be parsimonious in minimally disturbing the terrain of previous discussions, but surely not in minimizing its ontological presuppositions.

Even when philosophers discover aspects of language that have clear anthropological roots and implications, such as Austin’s notion of the “performative,” these aspects are taken as given and never traced back to an originary model. Yet “doing things with words” is a cultural activity; any model of the phenomenon that ignores its origin and cultural function is incomplete. How did it become possible to use words to marry, to christen, to promise? What is the most parsimonious model of the human that explains these phenomena? It is in answer to questions such as these that the minimality of originary thinking is made evident.

Just about every thinker since Marx has denounced metaphysics and offered a way to replace it, or, more prudently, to “deconstruct” it. In comparison with what has gone before, originary thinking offers the clearest definition of the metaphysical and of our limited but real capacity to “comprehend” it within an anthropology. Metaphysics is best understood as the mode of thought that recognizes no utterance-form more elementary than the declarative sentence orproposition. Metaphysical discourse is sufficient for analyzing the natural world, which is not dependent on our representation of it; this discourse cannot adequately describe the human world because it has effaced the trail that leads back to its own anthropological origin.

I believe that in the uniformitarian spirit of science, the emergence within the natural world of the cultural world of representations–entailing the often implicit but always operative dualism of our reflections on human ontology–should be made the object of an explicit hypothesis. Within the Humanities, this is even more crucial in the domains of ethics, religion, and philosophy in the traditional sense than in those of literature and art, whose works are only effective if they contain so to speak a “portable anthropology” within themselves. The originary hypothesis obliges us to understand the human faculty of representation not as an inexplicable given but as generated within a world where it did not exist. This generative model provides an anthropological basis for ethical concepts such as the categorical imperative, which we are otherwise asked to accept on faith, by grounding these scenic phenomena in a hypothetical originary scene where the relationships among the participants are made explicit. Similarly, this heuristic makes intelligible both the “supernatural” entities of religion and the imaginary constructions of esthetic culture, where the scenic is abstracted from its relationship to the everyday world temporally as well as spatially. The originary hypothesis permits the transcendence or Aufhebung of the monism-dualism, body-mind debate that circles constantly around the same tired dilemmas.

GA has been around for twenty years now. Its survival into the next generation will depend on whether you, dear reader, find that it helps you to better perform your task in our endless human enterprise of self-understanding.