Everyone recognizes that there is a qualitative difference between human language and animal communications systems, but while the methods of empirical science can conceive of emergent structures, they have no way of conceiving transcendental ones. This is the crux that demonstrates the need for Generative Anthropology as a way of thinking distinct from religion, philosophy, or empirical science.

It is impossible, as they say, to prove a negative, and so, as Karl Popper pointed out, one can never empirically demonstrate a universal hypothesis. A black swan can falsify the thesis that all swans are white, but no number of white swans can prove it. In somewhat the same way, it is impossible to demonstrate the necessity of a transcendental hypothesis. In the real world, the dichotomy between one- and two-dimensionality, which allowed Claude Lévi-Strauss (CLS) to claim (see Chronicle 628) that language and signification must have appeared all at once, can always be described as a qualitative modification of an earlier state. Writing not as an anthropologist but as a philosopher, CLS was able to refer to this difference in transcendental terms by removing it entirely from the empirical sphere, with no attempt to create a plausible framework for its emergence. But once you come to refer to worldly creatures, the emergence of a new idea can inevitably be related to the ideas that were in their brains previously, and the “absolute” leap can be reduced, in Darwinian terms, to the result of an incremental physiological/behavioral change. Hence doing away with the “naïve” idea of a punctual scene of origin has been the implicit burden of the dozens of articles and book chapters that have appeared in recent years describing the emergence of aspects of language out of animal interactions and/or the evolution of the brain.

As I said at the beginning of this series, this is well and good; scientists shouldn’t even try to demonstrate a punctual origin for language. There is no point wasting time in arguments over whether language “really” begins with a “first” sign, or whether it’s only “pre-language” until it has declarative sentences, let alone a full set of syntactic markers. Transcendence is a theoretical model, not a “fact.”

Hence neither is there any point in my going through these articles one at a time to point out how each one fails to account for language’s transcendental nature. To my knowledge, the closest a scientist has come is Terrence Deacon, in his 1997 The Symbolic Species. Deacon distinguishes, in C. S. Peirce’s terms, the symbolic signs of true semiotic systems from the indexical signs used by animals, and furthermore insists that they do not derive from them, being located in different parts of the brain. Deacon even includes a chapter—the longest in the book—describing how language might have originated when males went off to hunt and wanted to insure the fidelity of their female partners with a “marriage” ceremony. Yet when I wrote to Deacon expressing admiration for this scenario, he pooh-poohed it as mere speculation.


Given that it is nonetheless useful for me to provide one more example of the scientific opposition to the language Rubicon, rather than going through a collection of articles that share a theme but not a thesis, I have chosen to discuss a recent book by a single author. Michael Corballis’ The Truth about Language (Chicago, 2017), a volume destined for the general public, is not afraid to reveal to the reader the rather ironic “truth” of the matter.

Mr. Corballis’ (C’s) credentials include a number of scientific publications, mostly dealing with gesture. His principal thesis is roughly as follows: Given that gestures, as opposed to sounds, are in their origin clearly indexical, yet, as the existence of ASL and other sign languages shows, they can be used to create sign-systems more or less equivalent to those of oral language, a gestural as opposed to an oral hypothesis of language origin—a presumption shared by our originary hypothesis—makes it possible to escape the absolute distinction CLS attributes to human language/signification. This conclusion would of course be, in CLS’ terms, a category error; but, precisely, his are not empirical terms.

As an example of the kind of absolute thinking C is trying to discredit, his book opens with a quote from Max Müller, at the head of a chapter entitled “The Rubicon”:

Where, then, is the difference between brute and man? What is it that man can do, and of which we find no signs, no rudiments, in the whole brute world? I answer without hesitation: the one great barrier between the brute and man is Language. Man speaks, and no brute has ever uttered a word. Language is our Rubicon, and no brute will dare cross it.

Müller’s text was delivered (in English) at the University of Oxford in 1861, in a series entitled “Lectures on Mr. Darwin’s Philosophy of Language.” In contrast with CLS, Müller emphasizes not the instantaneity of language’s appearance but the vastness of the distance from the “brutes” that it provides us. But in this particular formulation, the link between language and religion that was so fundamental to Müller’s thinking is not part of the equation. That we alone in the animal kingdom have language is a thesis that can be nibbled away at the edges, something much less easy to do to our unique notion of the sacred, which cannot be understood in the absence of transcendence. For Müller, Durkheim, et al, if not for today’s researchers, these phenomena went together.

I nonetheless find it refreshing to see a book that overtly challenges the Rubicon concept, because the author thereby provides the outline of a falsifiable thesis about language. On the one hand, the transcendental nature of language makes it vain to seek to deny the absolute nature of the distinction that CLS posed so clearly. But on the other, from the social science standpoint, all the articles and books I have been wading through are at least implicitly premised on the usefulness of reducing, not indeed the absolute nature of the difference, but of its worldly manifestations, which are another thing entirely, to their absolute minimum.

We can still figure this in terms of Plutarch’s famous image of Caesar crossing the Rubicon. The Rubicon, which is a real river, has a certain width, which is what makes rivers, as opposed to brooks and streams, difficult to cross. The work of these researchers goes toward reducing the width of the river, by showing how much of the neuro-physiological-interactive capacity that we use for language had already been anticipated by what we can attribute to our pre-human ancestors. But even if the Rubicon could be narrowed to a little trickle, the separation marked by CLS’ distinction, the creation of signs, would remain a red line, the crossing of which cannot be explained in the gradualist Darwinian fashion of these researchers, let alone by the long discredited yet still endlessly repeated nonsense about the irruption of a “language gene.”

The purpose of the originary hypothesis is to provide a plausible scenario for the crossing of this line, one that takes into account, as do none of the social-science discussions, what strikes me as the self-evident fact that only a collective action, a memorable event could inaugurate this new form of communication. But an event is not an adaptation, and is therefore not acceptable to empirical science.

Be that as it may, our hypothetical event explains the crossing itself, whereas the empirical studies explain the narrowing of the river. These two conceptual projects are therefore not in conflict, but complementary, and it is in this spirit that I defend the necessity of the “fourth way” that is Generative Anthropology.


Even in its tendency to go on a bit too long about phenomena of peripheral relevance to language, C’s book deliberately underlines what all those scientific chapters exhibit in the small as a perfectly normal feature of any scientific endeavor, which can never be expected to provide a “final solution.” As C’s mildly debunking title suggests, he wants to show that the river had become over the millennia just a little rivulet, so that crossing it, however significant in its ultimate results, was just another of those incremental Darwinian accidents, such as the one that produced life itself.

Hence C’s book, to put it bluntly, is not primarily about language at all, in the sense that many of the technical studies focus on language-related physiological-neurological-linguistic details. The body of the book deals with functions that we commonly associate with language, but that C wants to assure us are essentially independent of it. In contrast with Deacon’s and CLS’s Peircean distinction between symbolic and indexical signs, he tells us of various brain-based activities: episodic memory, “mind wandering” (daydreaming), mental spatial mapping, toolmaking… that are both shared in part with other species and have something in common with language.

The second of three parts of the book, entitled “The Mental Prerequisites,” includes three chapters, “Thinking without Language,” “Mind Reading,” which discusses the evidence for a prelinguistic “theory of mind,” and finally, “Stories.” This last spends a good deal of time describing various genres of story-telling, from primitive myths to novels and detective stories, even video games. C’s point is that storytelling is an operation that uses language rather than being of language. For example, you can tell a story in pantomime, or in dance, or by drawing a series of pictures, as in a comic book. Language as we know it is in C’s view more a product of storytelling than its prerequisite. (“Narrative does indeed seem to form a large component of our thinking and may well have been the initial spur to language [101].”) If we surmise that some branch of pre-linguistic hominins enjoyed “telling stories” by miming or gesturing, then it is clearly a short step from this to inventing/discovering the formalizations characteristic of language, such as arise spontaneously, for example, in “Nicaraguan Sign Language,” a language newly invented by deaf students not previously exposed to signs. And voilà, language without a Rubicon!

C’s model of language use is never specified, but always seems to be implicitly one-on-one conversation. The notion of a scene of language, even independently of whether it was “the first,” is absent, as is any reference to collective ceremony, let alone worship. When C speaks about telling stories without language, as though any other species told stories in any way whatever, he never makes reference to a center from which the words/music/dance emanate toward a peripheral audience. (Everyone can participate in a dance, but not if it’s supposed to tell a story.)

C defines language as “a form of technology rather than a mode of thought (186).” I’m not sure what it means to call language “a mode of thought” as opposed to a medium of thought, but calling it a “technology” assimilates it to all the adaptations by which living things control their environment: chimpanzees, beavers, even amoebas have “technology” of one kind or another. Rather than a “mode of thought,” language is a mode, a technology if you will, of social communication and control that, in the terms of our hypothesis, serves to avert conflict over a desirable object by deferring its appropriation. Needless to say, none of this is considered here, nor indeed in the entire literature on the question.


No doubt for C, as well as for the authors of the technical articles in these collections, in insisting on an originary event, GA simply doesn’t get it. In their perspective, the idea of a qualitative leap—although C’s model of this is clearly not Müller’s revelation of sun-worship but Chomsky’s “language gene”—is simply naïve, modeled on religious revelation.

The authority of the originary hypothesis derives uniquely from how well, and how minimally, the reader finds that it explains the realities of culture as he knows them: its various “scenes,” from casual conversation and teamwork to religious ceremonies, including the “sacrificial” meals we still share; and how well our account of originary deferral comports with the best attempts to explain the specificity of the human psyche: Girard’s account of mimetic desire, Derrida’s différance and its phenomenological predecessors, notably Sartre’s pour-soi.

In the spirit of Occam’s razor, the hypothesis’ parsimony should be taken not as a mere attribute of elegance but as a presumption of truth. This criterion is classically that of logicians. But today, when even logicians work with “big data,” we should recall that William of Occam was not only a logician but a man of God—which is to say, faith aside, one aware of the indispensability of transcendence.

As was John, who defined “the beginning” by the Logos. So close to Lévi-Strauss, yet so much more complete, for the signs of the Logos also contain the sacred, the transcendental intuition of the unity that binds the sign, its referent, and its source, in the anthropological analogy to the Christian Trinity for which GA’s originary hypothesis provides a minimal formulation.