In answer to those who criticize Generative Anthropology for not being empirically based (and without going into the epistemological problems that this concept raises), I prefer to say that GA is a humanistic rather than a social-scientific anthropology. But terminological distinctions can take one only so far. Given the subject of next year’s GASC in Stockholm and the presence of cognitive scientists among the keynote speakers, this seems to be a propitious time to attempt to justify the epistemology and what I hesitantly call the methodology of GA where it differs from that of anthropology as an empirical social science.

The crux of justifying to the scientific world the construction I have called the originary hypothesis is that those who call themselves social scientists, however much they are capable of speculating on the originary moments of their fields of study, consider as serious hypotheses only those that can be empirically tested in the laboratory or in the field. This distinction was brought home to me notably in the brief correspondence in which I engaged early in this century with Terrence Deacon, the author of The Symbolic Species (Norton, 1998), still the most important work on the neuroscience of human language. Deacon’s book includes a long chapter in which he speculates that the original impetus for language came from the conjunction of male hunting parties, a new proto-human activity, with the neotenous nature of our young, which obliges women to devote to their infants a qualitatively longer post-gestational period of intense care than female apes. In Deacon’s scenario, because the investment in children was so great, these hunters required a guarantee that their women remain faithful to them in their absence, assuring them that the children to whom they devoted their efforts were their own, and therefore conceived a ritual of marriage as the occasion for the first use of “symbolic” signs, in Deacon’s Peircean terminology.

I wrote to Deacon praising his attempt to ground language in a social ritual, and saying a few words about my own work. He answered dismissively that these speculations were of no real importance—this despite the fact that the chapter in question was the longest in the book. Because Deacon is a laboratory researcher, as Michael Tomasello is in his own way, he could afford to distinguish between “serious” empirical statements and speculative ones. Thus we may confidently learn objective information about the operations of human language in the brain from the “serious” part of Deacon’s book while regarding his speculative “originary hypothesis” as mere food for thought.

Here then is the key point of divergence between a humanistic and a scientific anthropology. The point of GA is not to generalize from the analysis of hard data, because the origin of human culture, and indeed, that of any human cultural phenomenon, can never be qualified as “hard data.” Even if, as I suggested in a recent French chronique (, we have no real basis for assuming that some divine metaphysic protects human thought from being simulated and perfected by mechanical means, the original human version of it is not simply algorithmic and can only be understood from an interactive, scenic perspective. For us, in a word, “speculation” is the nature of our science. The originary hypothesis is a heuristic, not a more or less accurate account of a discoverable event. But in our context, as opposed to Deacon’s, this makes it all the more indispensable. The point of a humanistic anthropology is to explain the emergence of the qualitative difference between animal and human communication and social organization, and this can be done only by bootstrapping the new human component into the hypothesis itself. The origin of the sign must be illustrated to be understood, and to do this, a fiction must be designed that offers a plausible scenario for this origin, without any hope in our present context (for who knows what historical equivalents of stellar spectroscopy may be discovered if our race lasts long enough) for empirical confirmation.

As last week’s review of Tomasello’s A Natural History of Morality suggests, the separation between the human and the prehuman is not accessible to social science. The origin of the symbolic cannot be understood as a “natural” phenomenon; there is no “natural history of morality” any more than E. O. Wilson’s sociobiology suffices as the basis for a history of human social institutions. The claim that we are indeed biological or simply “natural” creatures is true enough, but about as helpful as the idea that Laplace’s demon’s presumed ability to predict all the universe’s future states demonstrates that we lack free will. At a certain point in his career, Tomasello discovered the distinctive human quality of joint shared attention, but he never fully grasped its unnaturalness. Once individuals can communicate their attentions, they can communicate their intentions, at which point emerges the phenomena of the sacred as the deferral of these intentions, whence significance as other than programmed appetitive attraction/repulsion, and the whole set of human symbolic institutions: rules, laws, customs, rites. None of these are “natural” nor can be derived from one-on-one “partner” relations, which would have been formalized in new ways by the existence of even the most elementary human systems of representation—as Deacon suggests in his marriage example.

Is there a way around this impasse that might permit a mutually respectful dialogue between empirical human/social science and a humanistic anthropology? In the first place, GA is not a philosophical anthropology, however much its overall thrust is similar; it insists that its anthropological etiology of language itself be prior to any subsequent elaboration of concepts, moral or otherwise. Nor is GA tempted to dismiss religion, as in the passage I quoted in Chronicle 519 from Tomasello’s work. Rightly considered, the significant and the sacred are different ways of referring to the same phenomenon, and their divergent histories reflect a specialization among different aspects of this one phenomenon.

Just as in an absolute sense, religious and nonreligious perspectives cannot be reconciled—one is either a man of faith or one is not, one “believes in God” or one does not—so one might say that one is either a believer in “hard science” or in the kind of humanistic science embodied in GA. Yet the point as I see it is not to abolish disagreement but to minimize it around a central core. I believe I have been able to do this with respect to the first dichotomy: most participants in GA activities are in fact religious believers, and to my knowledge, those who are nonbelievers are not hard-core atheists on the model of Richard Dawkins or the authors of books with titles like God Is Not Great. Girard was a believer and I am not, but our modes of thought appeal, on the highest level at least, to the same audience.

The point of the originary hypothesis is to provide a plausible model of an event whose participants would begin as non-humans and would end as humans. Social scientists will no doubt snicker at so naïve an idea; nothing as significant as this transformation can take place all at once; many small steps are required, and so on and so forth. Well, no. There is a binary, absolute distinction between a creature that uses “symbolic” signs and one that does not, because the very use of such a sign is “binary”; either it is a sign or it is not. No doubt in the “real world,” the detection of this first moment would be impossible, and our choice, necessarily arbitrary: how exactly can we determine the first time that an “aborted gesture of appropriation” is understood as a sign? As Adam Katz has pointed out, we certainly can’t assume that even the “first” group of hunter-scavengers all began spontaneously to understand it in this way.

But that is not the point. Just because the “binary” or “digital” phenomenon began in an unclear manner, what is important is that it began, and that its beginning was the inauguration of a new mode of communal communication. The originary hypothesis presents this change in a heuristic fashion, by eliminating empirical distractions and focusing on the significant element of the change. In contrast, a work such as the one I examined last week makes use of any concept it can find to break down the binary passage from prehuman to human, notably in the idea that “morality” first emerged as a relationship between pairs, presumably unmediated by signs. But in the absence of signs, there is no truly human communication, no moral idea, merely moral “attitudes.” And the same is true of any cultural phenomenon. Humans relate in twos only because they first relate to the whole—call it God or merely the human community—and language, which is the chief means of this communication, can only be understood as mediated by this whole. Even when I say “I love you” to my beloved, the signs I am using belong not to “us” but to the virtual community of language users.

The originary hypothesis is not set in stone, and empirical evidence can certainly modify its terms, but not its fundamental binary structure. Philosophers know this—but philosophers are not anthropologists precisely to the extent that they are wedded to the declarative sentence, the proposition. Theologians too know this—but theologians are not anthropologists to the extent that the hypotheses of origin they offer are not minimal. Buddhism, as discussed in the recent Chronicles 515 and 516, comes no doubt as close as any religion gets to understanding the scene of (human) representation as the minimal requirement for being human. But GA is demonstrably the first attempt to wed the understandings of empirical anthropology, philosophy, and religion in accordance with Ockham’s razor to provide a minimalistic understanding of the passage from pre- to full humanity.

In no sense does GA contradict Darwinian biology or any other element of scientific theory. Nor does it dictate in any way the prerequisites for the originary scene to take place. Girard’s idea of “more mimesis” is all that is minimally required; but the conditions under which this “more” reaches the point at which an entirely new mode of social interaction becomes necessary to the species’ survival must be discovered through empirical investigation, not deduced from the minimal scene itself.

But these conditions are ultimately a secondary consideration. In the absence of an originary hypothesis, social science can never truly constitute its object, the human species, as a unique entity. As we have learned after decades of effort, apes cannot really be taught language, let alone religion and art. Or morality either, as Tomasello makes no bones about admitting. Thus humanistic anthropology is fundamentally necessary to understanding the emergence of the human in a way that empirical anthropology is not. The important thing is not to determine the exact moment at which this binary transition occurred, but to understand it ontologically as a binary transition, the introduction into the universe of a new way of relating to one’s fellow humans and to the world.

Although the present appeal is unlikely to succeed where others have failed, my hope is that it will inspire others, the cumulative effect of which can erode the wall that separates our humanistic anthropology from that of social science. Historians of science have become familiar with the idea of “paradigm shifts” to explain “scientific revolutions.” It is time they took a step back and realized that the very idea of paradigms, that is, of thinking in symbolic terms about the elements of our experience, can be understood only as arising in an event that introduced into the world a mode of communication and interaction that had not previously existed, and whose moment of revolutionary change would never be forgotten.