American Scientist

Lingua ex Machina: Reconciling Darwin and Chomsky with the Human Brain. William H. Calvin and Derek Bickerton. 298 pp. The MIT Press, 2000. $26.95.

In 1866, seven years after the publication of Darwin's The Origin of Species, the Linguistic Society of Paris banned all discussion of the evolution of language. Somehow this seems to have held, with the odd leak, for more than a century, but in the last decade or so the floodgates have opened, as though forced by the pressure that had built up. Even so, the dichotomy between the Cartesian view that language is unique to humans and can only be explained by some miraculous event, and Darwinian notions of continuity between humans and other animals, with incremental change through natural selection, remains as stark as ever.

The authors of Lingua ex Machina are an unlikely pair, both having written previously and extensively on language evolution, but from opposite sides of the divide: Derek Bickerton is a Cartesian linguist and William Calvin a theoretical neurophysiologist. Yet the smiling faces on the dust jacket suggest a pair of dashing fellows who are far from grumpy old men, cheered perhaps by the pleasures of the Villa Serbelloni on Lake Como in northern Italy where they met, in the company of other scholars, to write this chatty book. Each chapter is written by just one of them, with occasional interjections from the other, so that the impression is somewhat of an extended conversation. The result is interesting and often stimulating, although the two sometimes seem to talk past each another, and I don't think they have quite realized their claim of having brought Noam Chomsky (the modern-day Descartes) and Darwin "together at last."

Bickerton is well known for what has been dubbed the "big bang" theory of the emergence of syntax, the critical element that distinguishes human language from all other forms of communication. Several other species, including the great apes, are capable of what he has called protolanguage, which consists essentially of symbols strung together in a way that can make simple sense, but without the refinements that syntax offers. It is the language of pidgin, children under two, second-language learners in the early stages, and, it was once remarked, drunken teenagers. According to the big bang theory, syntax emerged in all-or-none fashion late in hominid evolution, perhaps with the emergence of our own species about 150,000 years ago.

Bickerton argues that syntax has to do essentially with the construction of phrases and clauses, but each of these components by its very definition is dependent on the other, creating what has been called "irreducible complexity," a concept often used in the past (though not by Bickerton) as an argument for intelligent design over natural selection. But in this book Bickerton moves to a much more gradualist account of how syntax might have evolved, even suggesting that the essence of syntax may be discerned in reciprocal altruism among chimpanzees. And nowadays syntax is much simpler than in the early Chomskyan days of deep and surface structure and complex transformational rules. In Chomsky's recent Minimalist Program, the ordering of words in sentences is governed entirely by information stored with the individual words themselves, so that a sentence is essentially self-organizing. The evolution of syntax, then, might well have come less with a bang than a whimper.

Yet Bickerton still holds that protolanguage was around for at least two million years, and reciprocal altruism for even longer, and asks why syntax itself took so long to emerge. At this point Calvin chimes in to suggest that the link might lie in throwing, a suggestion that will not come as a surprise to those familiar with his earlier book, The Throwing Madonna (McGraw-Hill, 1983). Chimpanzees can fling objects, even branches, as a means of defense, but accurate throwing is something that evolved in the hominid line, in part as a result of bipedalism and alterations to the structure of the hand. According to Calvin, throwing has a hierarchical tree structure analogous to that of a sentence.

Bickerton briefly protests that throwing does not possess the open-ended generativity of language, but Calvin has the answer in a complex and somewhat idiosyncratic neurophysiological theory involving hexagonal stacking of neurons and the coding of elements in spatiotemporal patterns that can flit around the brain much as words flit around human discourse. The cortex is assumed to operate as a "Darwin Machine" in which patterns of excitation in the cortex compete for survival. Calvin's neurophysiological speculations are dazzling but too often seem to reduce to metaphors involving choirs, symphony orchestras, grocery stores, fallow fields and even his own back yard, and I'll bet he had most of the other scholars fooled at the Villa Serbelloni.

Calvin recognizes that throwing was probably not the only activity that drove the changes necessary to support syntax, arguing that other ballistic skills and "higher intellectual functions" were probably involved. To my mind, a more general and flexible link between primate behavior and language might be provided by manual gesture, which has a more open-ended quality than throwing. By freeing the hands and arms, bipedalism would have greatly extended the opportunities for expressive gesture in the hominid line. It is not unreasonable to suppose that syntactic language was gestural for much of hominid evolution, bridging that gap of two million years, with vocal control emerging relatively late and eventually assuming dominance, perhaps with our own species, Homo sapiens. I recently had the opportunity to suggest this to Bickerton in person, but he dismissed it with, I am delighted to report, a wave of his hand.—Michael C. Corballis, Psychology, University of Auckland, Australia