During my first weeks of reading through the recent literature on the origin of language, the progress in scientific research in the various relevant domains did not seem to have resulted in any notable convergence with the positions of originary thinking. I was happy to see that the Chomskian dogma concerning our “language acquisition device” had been largely abandoned, along with the absurd idea of multiple genesis of Homo sapiens, and to learn of the renewed interest in gesture, both as gesticulation and as “signing” (as in ASL and other languages for the deaf). But none of this theorizing came anywhere near the central core of the originary hypothesis: the reconstruction of the founding event of the human.

Derek Bickerton, whose work I discussed in Chronicle 167, recognized the anomaly of the origin of human language and its discontinuity with previous forms of communication. But, on the analogy of the differences he discovered between pidgins–improvised, ungrammatical linguae francae for communication between speakers of other languages–and creoles–pidgins that have become the native tongue of the second generation, who spontaneously impose on them syntactic order–Bickerton splits this origin in two, as (1) the origin of a “protolanguage” of some two million years’ duration, and then (2) about 50,000 years ago, as a result of the maturation of the language-selected brain, the origin of syntactically modern language. This allows him to sidestep the real question of what motivated the first origin, which he implicitly admits he cannot answer.

Thus it seemed that, whatever advances science would make, it would never even begin to understand the need for an originary hypothesis as proposed in my 1981 book, The Origin of Language. This incomprehension, while frustrating, had its compensations; it suggested a nicely dichotomous complementarity between “humanistic” GA and the “positive” biological sciences, cultural anthropologists remaining essentially uninterested in the question of human origin.

Terrence Deacon’s The Symbolic Species (Norton, 1997) puts the situation in a new light. This ambitious overview by a neuroscientist of the emergence, evolution, and present reality of the brain’s language function takes long strides toward the position of the originary hypothesis. It is both disquieting and exalting after all these years to see a biological scientist associate language origin with ritual, interdiction, and the difficulty of making peace among groups of male hunters. The world of positive thought has been unknowingly convergent with the generative theory of mimetic representation; empirical science has moved tentatively toward the positions I elaborated twenty years ago on the basis of an a priori hypothesis of language and culture.

This implies the coming together of the “two cultures,” the scientific and the humanistic, that have remained until recently more separate than ever and that still remain so in important respects. For those of us who are convinced that the middle path taken by Generative Anthropology between the cultural and the natural had produced the first secular synthesis of the positive “is” of nature and the ethical “ought” of culture, this development obliges us to reexamine our intellectual strategy. Thus far, this middle ground, formerly the exclusive terrain of religion, has been approached almost exclusively from the humanistic side. Will that continue to be the case now that connections are beginning to be drawn between the functioning of the brain and the operations of thought? And what of the “soft” social sciences? Girard’s Violence and the Sacred made strong, specific ethnological claims that the ethnologists themselves have largely ignored. Could advances in the study of biological evolution herald a renewal of anthropological research into human origins in the face of the mind-numbing emphasis on cultural “diversity” that has politicized the field and marginalized, even stigmatized, its original evolutionary concerns?


Deacon focuses on language from the standpoint of brain structure. Bickerton understands that the oversized, cortex-heavy human brain is the product of selection for language rather than the other way around, and conceives of the prefrontal cortex as the locus of the “off-line” operations of human language in contrast to the “on-line” processing of other species. But Deacon’s specialized knowledge of the interconnections in the brain between language and perceptual and motor functions allows him to avoid Bickerton’s dichotomized vision of language as with or without syntax. Deacon’s adaptation of C. S. Peirce’s semiotic categories of icon, index, and symbol to the stages of evolution of animal communication and its expression in the brain is a model of the scientific assimilation and application of philosophical thought.

But even more than its author’s semiotic sophistication and neurological expertise, what sets The Symbolic Species above all other scientific writing on language origin is the power of his anthropological intuition. This researcher of brain functions has succeeded in articulating the social nature of language better than any social anthropologist. In speculating on the possible origin of this new form of social activity, he has revived in a new context the powerful intuitions of the nineteenth-century evolutionists, most notably, the link between language and ritual and their function in establishing peace through the interdiction of ethically destabilizing behavior.

To the context of meat-eating and all-male hunting parties familiar to students of the originary hypothesis, Deacon adds a consideration that my own reconstruction of the originary event may fairly be said to have neglected: the necessity of reinforcing the monogamous bond within which our care-demanding “neotenous” children are nurtured. Because the protohuman society depended on meat, groups of male hunters might be absent from home for long periods. In order to assure the stability of the couple and protect the male investment in his offspring, the symbolically guaranteed bond of marriage, based not on mere association but on a true ethical interdiction, would become advantageous for the group as a whole in a way that had never previously been the case in primate societies. This reasoning is not uncongenial to the originary hypothesis. The intensity of the mimetic crisis in which the “aborted gesture of appropriation” became the first “symbolic” sign would surely have been augmented by the fear of each member of the group that not only his life but his genetic future would be in jeopardy if another member of the group preceded him in appropriating the central animal/object.

Deacon’s account is even closer to the originary hypothesis than this powerful but still external ethical motivation suggests. He conjectures that the group’s members learned the new form of association embodied by the sign through the collective repetition in “ritual”–a term whose sacrificial connotations he admittedly appears to ignore. He understands both that the original function of language is ethical and that the crucial ethical problem is “to mediate . . . peace” (403) and that the meaning of language is a social reality not reducible to any individual’s brain state. Deacon even recognizes the power and centrality of religious belief as a feature of the human use of symbols, and although he presents these beliefs as overreachings of our symbol-centered brain rather than as extrapolations from the common origin of language and religion, he makes this common origin more plausible than any other scientific work I have seen. If Deacon does not carry his intuition of the commemorative function of originary language to its radical scenic conclusion, his hypothesis is by no means incompatible with an event of origin.

At the end of his book, as is not uncommon in such endeavors, Deacon speculates on the possibility that we might create artificial beings capable of using language. Dismissing the digital computer as a model of language use–the final chapter contains an enlightening discussion of John Searle’s well-known “Chinese room” model of computer intelligence–Deacon suggests that only a creature possessing “sentience,” the ability to perceive and react to its environment, could be taught to use language as we do. This may some day come to pass. But Deacon’s description of the brain’s Darwinian networks of synapses in which the fittest associations survive to become “knowledge” suggested to me something different: a new justification for humanistic thought.


In recent years, some of the great mathematical conjectures of the past have been proved: the four-color theorem, even Fermat’s “last theorem” that xn + yn = zn has no integer solutions for n>2. But the proofs, rather than being short and elegant, are virtually incomprehensible, have been obtained through exhaustive computations made possible by computers. Command-driven digital computers are not appropriate models for the neural networks of the brain. I allude to these results only as an illustration of the difference between demonstrating and understanding even in the rule-governed world of mathematics. In the brain as Deacon presents it, this distinction becomes a dichotomy. Although we can of course use our brains to reason logically, as I hope I am doing now, the operation of representing by means of a sign the originary central object, and to a non-vanishing degree, every subsequent object, is a self-confirming one, the association with a perceived object not of a mere “index” but of a meaning whose apparently atemporal objectivity depends on the human configuration of mimetic desire that surrounds it. From the standpoint of logical thought, we legitimately speak of the paradoxical nature of this operation, yet from that of the brain’s operation, there can be no “paradox” but only the reinforcement of certain synapses at the expense of others.

This suggests that, however far neuroscience may advance in its description of the operations of the brain, it can never explain the emergence of significance and in particular, of language. This impossibility does not reflect the presence in human language of some mysterious indeterminacy of the kind that led the mathematician Roger Penrose to seek the source of the human sense of free will in quantum undecidability. It is merely an artifact of the incommensurability between our logical use of concepts and the way they are generated in the brain. Were the brain truly, as Artificial Intelligence researchers used to claim, a biological Turing machine, this problem would not arise, but then human consciousness and the use of signs would not have arisen either. But because our brain is capable of reacting to a collective situation of mimetic crisis by transforming an appetitive association into an inhibitory sign, its operation cannot be described in algorithmic terms; its representation of its environment is interdependent with that environment. Whence the need for the paradoxical, culture-based models of “humanistic” thinking to represent this interdependency in terms that can make it understandable, that is, able to become in turn an object of potential desire for these very same human brains.

I do not intend to imply by this that members of the scientific professions are incapable of generating such models, as Deacon clearly comes very close to doing. The need for a separate profession of humanists diminishes every day. Yet we may be consoled by the fact that the tradition within which our thought has been disciplined does not seem fated to lose its relevance. On the contrary, the confluence of humanistic and scientific thought can only make the study of cultural phenomena more exacting. This is the best news in a long time for practitioners of Generative Anthropology.