In last week’s column, I had occasion to remark that the best designed typology of religious phenomena cannot make up for a deficient theory of origin. Scientific studies of, or better put, around the origin of language suffer from a similar deficiency. I hasten to say that these empirical studies include many useful results, whose demonstration is in most cases at least partially independent of any overall hypothesis of language origin. It is surely useful to learn about the descent of the larynx and the evolution of the primate brain. As long as the authors stick to empirically verifiable subjects, they are on solid ground, and only their fellow specialists are qualified to disagree with them. But at some point in most books and articles about the origin of human language, these authors feel obliged to bite the bullet and construct on the basis of their research a hypothesis of how human language originated. Lacking a sense of the paradoxical nature of language origin, they fall into the paradoxes of “supplementary” thinking with far less awareness than Plato or Rousseau. For the most part, these scholars purport to explain what is by their own admission the radically new phenomenon of human language either by age-old gradualistic Darwinian arguments (language enhances fitness) or, even more naively, by a mechanistic reliance on Mendelian ones (language cannot emerge unless a genetically determined language capacity is already present).

Of the abundance of examples, I have chosen six, two each from three recent compilations. In the course of these analyses, I hope to convince you of the intellectual and, indeed, scientific superiority of my own anthropological hypothesis–neither falsifiable nor metaphysical–of the origin of language, religion, art, exchange, desire, and everything else that makes us human.


1. My first example is the article of Tiziano Telleschi, “Origins of language and of society-culture relationships,” in Becoming Loquens (Bochum, 2000), pp. 75-100:

The true break with the world of other animals occurred when, rather than specializing the soma, the evolving living being . . . refined certain aspects of the “neo-mammalian” brain . . . Language, although presenting only a rudimentary structure, was an initial “choice” which was destined to exercise a profound influence on the development of the brain itself, disrupting the conditions previously imposed by natural selection.

By giving a “name” to things by means of “single words,” probably a million and a half years ago . . . , the Hominid of the Lower Paleolithic began to achieve increasing independence from objects. In accordance with arguments put forward by Bronowski and Bellugi, who claimed that in epigenesis language carried out the task of restraining and cleansing the content of communication from the emotional pressures of the limbic system, controlling emotive afferences in order to use them to link past and present experiences . . . and to initiate direct connections between different associative cortical areas of non-limbic types, we can presume that “single words” were able to overthrow the dominance of the limbic system . . . (80-81)

The active agent of the first paragraph is the “evolving living being” that creates language as a result of the evolutionary “choice” of the brain over the body. In the second paragraph, the agent is language itself, which “carried out the task of restraining and cleansing the content of communication from the emotional pressures of the limbic system.” There are no desiring human beings, let alone potentially conflicting ones, in these paragraphs, and for good reason: the emotions that might have led to conflict have been “restrained and cleansed.” A Roy Rappaport (see Chronicle 282), let alone a Durkheim, would recognize that “restraining and cleansing” are precisely the cathartic functions of ritual that since Aristotle’s Poetics have been associated with tragedy. But the “emotional pressures” referred to in this passage are not even remotely conceived as leading to human conflict; their “cleansing” takes place in a wholly cognitive context, as though language removed an emotional filter between the prehuman subject and the object of his cognition. Which is to say that language performs a “cleansing” operation in order to perfect a cognitive system that could not have existed without language. Language is not only an agent, it is a conscious agent, able to plan its cleansing operation with a goal in view; in short, language behaves as though it . . . were making use of language. Which is a lot to expect from an evolutionary abstraction.

Open a daily newspaper and nearly all the headlines, now more than ever, are about interhuman violence of one kind or another and efforts to prevent or control it. But this truth, theorized long ago by the anthropological wisdom of Machiavelli or Hobbes, is virtually absent from contemporary studies of human origins. At the very most, evolutionary advances are described as advantageous in combat between protohuman or human groups; the idea of conflict within the group itself is simply unthinkable. Yet mimetic violence is never really absent; it is merely metaphorized, converted into “emotional pressures” that need to be cleansed from “communication” between members of the group. If, conversely, we understand these pressures as those of mimetic desire, then the second paragraph’s unexplained personification of language metamorphoses into a simplified statement of the originary hypothesis: the conflict occasioned by mimetic desire over a contested object is deferred by the emission of a linguistic sign that represents and in so doing interdicts that object. Needless to say, this hypothesis, in which protohuman creatures rather than “language” or “the evolving living being” are the agents of historical innovation, is implicit in Telleschi’s text only for someone who has already formulated it; it can hardly be said to emerge from it. On the contrary, the opacity of this text is one more demonstration that lacking a generative hypothesis one cannot “cross the Rubicon” between the biological and the semiotic; the solution will never emerge from the data alone.


2. My second example, taken from the same collection, is the article by Merlin Donald entitled “Preconditions for the evolution of protolanguages,” Ibid., pp 101-22. Donald has over many years developed a theory of mimesis that unfortunately lacks any connection to René Girard’s seminal theorization of the subject. Once again, the specter of aggression shows its head, only to vanish without a theoretical trace:

The emergence of mimetic skill would also have amplified the existing range of differences between individuals (and groups) in realms such as social manipulation, fighting and physical dominance in general, toolmaking, tool use, group bonding and loyalty, pedagogical skill, mating behavior and emotional control. This would have complicated social life, placing increased memory demands on individuals; but these communication tools would also have created a much-increased capacity for social coordination, which was probably necessary for a culture capable of moving a seasonal base camp or pursuing a long hunt.

It is important to consider the question of the durability of a hominid society equipped with mimetic skill: adaptations would not endure if they did not result in a stable survival strategy for a species over the long run. Mimesis would have provided obvious benefits, allowing hominids to expand their territory, extend their potential sources of food, and respond more effectively as a group to dangers and threats. But it may also have introduced some destabilizing elements, especially by amplifying both the opportunities for competition, and the potential social rewards of competitive success. (112-13)

In the first paragraph, Donald responds to his own objections. If, on the one hand, increased mimetic skill leads to sharper hierarchical differentiation, on the other hand, the “communication tools” associated with this skill permit a higher level of social coordination. What is unclear is the nature of the creatures we are discussing. Whereas ape societies are hierarchical, all tribal-level human societies are egalitarian. The ability of language to enhance the social coordination of steep hierarchies is realized only at the level of state societies; before that, to judge from the evidence, language use is associated rather with the absence of hierarchy. Could it be that Donald is confusing two different levels of human evolution? That increased mimesis, in leading to language, brings about the dissolution of the hierarchical structures we find in primate societies? Although societies with language can and inevitably do see themselves as cultural wholes, the author speaks of “communication,” “fighting,” “dominance,” and so on in the one-on-one context characteristic of animal societies.

Hence when we arrive at the second paragraph, we are faced with a contradiction that, this time, cannot be dealt with even verbally, but can only be left to us as a problem for the future. It is very Girardian indeed to speak of the dangers occasioned by enhanced “mimetic skill”; enhanced mimetic skill is the independent variable in Girard’s theory of homogenesis. But Donald’s cautionary caveats receive no reply in his own text, and appear therefore as no more than propitiatory gestures.

These paragraphs conclude the section entitled “The cultural impact of mimetic skill distributed in social groups”; the following section, on “Mimesis as a preadaptation for protolanguage,” makes no mention of the problem posed by these “destabilizing elements,” a problem which, however “important” it may be to consider it, is never again broached in Donald’s text. Here as elsewhere, Donald’s discussion of mimesis is handicapped by the absence of any theorization of the difference between one-on-one imitation and the collective mimetic forms of human culture that serve to control rather than to exacerbate competition. This failure in turn reflects a more profound one, which is the failure to recognize the inherently conflictive structure of mimesis. Rather than being mediated by “the opportunities for competition, and the potential social rewards of competitive success,” the “destabilizing elements” to which mimesis is said to give rise are already present within mimesis itself. One can find better anthropological models in Girard’s work before he became interested in anthropology per se than in those of the one anthropologist who makes mimesis the core of anthropogenesis. If Donald would only take the trouble to type the word “mimesis” into Google, he would encounter conceptions of mimesis far more sophisticated than his own, conceptions that he seems never even to have given himself a chance to reject.


I now turn to a second collection of articles on language origin: New Essays on the Origin of Language, edited by Jürgen Trabant and Sean Ward in the Trends in Linguistics series for Mouton de Gruyter (Berlin and New York, 2001).

3. Manfred Bierwisch’s article in this collection, “The apparent paradox of language evolution: can Universal Grammar be explained by adaptive selection?” (pp. 55-79), seeks to avoid a paradox familiar in different forms to students of language origin since Rousseau: since the usefulness of language depends on its being shared with others, it would have to emerge in many minds at the same time–an idea sometimes made to imply that human linguistic capacity originated from a simultaneous genetic mutation.

It seems obvious to me that this problem vanishes if we can construct a plausible hypothesis for the collective origin of language. But those who study the question seem convinced that, whether or not communication is the force driving protolanguage evolution–and a surprising number of scholars think it is not–the only way to speak of the emergence of language is as that of a capacity for language in the individual “phenotype.”

Bierwisch’s solution is to separate “linguistic capacity” from “linguistic knowledge,” the latter, but not the former, being “within the range of linguistic theorising” (71). He then proceeds to propose that a “protolexicon” emerges, relying on Herder for the notion of “reflection” as the origin of what he calls the “stimulus-free and situationally independent assignment of structured signals to conceptual representations” (71). This, of course, is precisely what a hypothesis of the origin of language must explain; here it is taken for granted. Bierwisch’s notes that, once a protolexicon comes into being, it has no fixed, genetic limits as to the number of items it can contain. His hypothesis is that it is the increase in this number that drives the development of syntactic and morphological organization:

I have proposed how to avoid the paradox of language evolution by combining two distinct but interrelated problems, both of which have to be solved anyway: the evolution of the language capacity and the origin of linguistic knowledge. These frequently confounded issues must be clearly distinguished because they depend on fundamentally different conditions affecting the genetic heritage as well as possible knowledge based on it. But it seems that a plausible scenario emerges if they are construed to depend on each other in a non-vicious circle. The capacity to accumulate lexical items could gradually lead to a developmental stage where a random variation indeed leads to an improvement of the linguistic capacity, justifying the urgently desired selectional benefit. (79)It is hard to disagree with the notion that the “language capacity” increases in tandem with the increased use of language rather than appearing all at once through the miraculous acquisition of a “language gene,” or evolving gradually as a purely cognitive capacity that suddenly manifests itself as language–a hypothesis that is either a truism (language can only emerge in a brain that is ready to acquire the mechanism of language) or an absurdity (language is an epiphenomenal communication to others of conceptual connections already present in the individual mind). What Bierwisch does, in fact, is to reduce the problem of language origin to that of the origin of the first “protolinguistic” sign, which is to say, to the problem addressed by the originary hypothesis. In this minimalistic gesture, Bierwisch is in full accord with generative anthropology. Once the sign exists, it will multiply and structure itself syntactically and morphologically, generating Baldwinian (behavior-driven) selection pressure toward an ever-greater language capacity. I think this plausible-sounding argument pays insufficient attention to the non-gradual syntactical gradations of ostensive, imperative, and declarative, but it can be accepted in what the French call its grandes lignes. The missing link, that is, a hypothesis that accounts for the origin of the protolinguistic sign itself, Bierwisch has the modesty not to attempt to supply. However praiseworthy this modesty, the fact remains that one cannot “solve” the paradox of language evolution by proposing a theory that accounts for this evolution only once language, even in its simplest form, has already emerged. The equivalent would be explaining the evolution of life after taking as given the existence of the first living organism.

(Part I of a slightly abridged version of a lecture delivered at Westminster College, Salt Lake City, in March 2003)