American Scientist

Christine Kenneally reviews the current state of the field in her new book. An experienced science journalist with a Ph.D. in linguistics, she is well qualified for the task. The focal point for her discussion is a high-profile symposium on the evolution of language held in 2005 at Stony Brook, New York, where many of the leading players met and gave talks, but her reading and interviews range more widely. The First Word is almost certainly not the last word, but it does provide a lucid, readable, comprehensive account of the different ideas that are now current.

One figure who continues to exert a major if not always benign influence is Noam Chomsky, the dominant linguist of the past half-century, who made a rare appearance at the symposium. It might be said that Chomsky actually helped prolong the ban, because he has long argued that one simply cannot know how language evolved and has even suggested that language may not be the product of natural selection. This position was first explicitly questioned by Steven Pinker and Paul Bloom (who were not present at the symposium but are appropriately included in Kenneally's story) in a classic article published in 1990, in which they retained much of the Chomskyan stance but made a strong case for the incremental evolution of language through natural selection. Much of the subsequent development has been in more direct opposition to Chomsky and seems set to redefine the nature of language itself. As the book relates, Chomsky appeared at the symposium only to give a public address, which I and many of the others in attendance found largely incomprehensible; he arrived and departed without engaging with any of the other speakers.

Chomsky first achieved prominence with his 1957 book Syntactic Structures, which argued that syntax could not be explained in terms of associations. However, theorists such as Simon Kirby and Morten Christiansen have made considerable progress toward developing connectionist theories; language may after all depend on learning principles, and not on some innate  language-acquisition device. Sue Savage-Rumbaugh is accorded her rightful place as a pioneer in the effort to discover the linguistic abilities of apes, but the notion of a continuity between ape language and human language remains implacably opposed by those who retain at least vestiges of Chomskyan theory, a group that includes Steven Pinker, Ray Jackendoff and Derek Bickerton.

Kenneally also describes the idiosyncratic work of Luc Steels, who has established artificial robot-inhabited worlds in which languagelike structures arise spontaneously. As Jim Hurford, another prominent symposiast, remarked, we may be witnessing the demise of the Chomskyan notion of universal grammar, the supposedly innate structure, unique to humans and peculiar to language itself, that is said to underlie all languages. Instead, language may depend on more general cognitive abilities.

One complaint I have is that most of the experts discussed in the book seem to equate language with speech. (Kenneally herself writes that speech "is crucial to language.") A partial exception is Michael Arbib, who grounds language evolution in manual gestures, building a scenario in which an intermediate form of communication, which he calls proto-sign, forms the scaffold for the incorporation of vocalization in an ascending spiral toward full syntactic speech. Arbib draws on so-called mirror neurons, first discovered in the monkey brain, which respond both when the subject makes grasping movements and when it observes the same movements made by others. These neurons are found in areas homologous to speech areas in the human brain and seem to provide a natural platform for the evolution of language. Mirror neurons, though, may be an overworked commodity in modern evolutionary and cognitive theory, providing convenient explanations for anything from language to imitation to theory of mind.

Yet language can consist of a combination of hand gestures and facial expressions rather than vocalizations. Even Arbib appears not to recognize that signed languages are indeed true languages; he seemingly confuses them with pantomime. My own view, which I presented at the symposium and which is mentioned in the book, is in fact similar to Arbib's but allows that language might well have evolved as a sophisticated form of ritualized and grammaticalized gesture before the eventual takeover by vocalization; even now manual gestures are woven into our speech. Arbib is quoted as saying "It's hard to build up a rich tradition just through gesture," but a visit to Gallaudet University in Washington, D.C., where the language of instruction is American Sign Language, might persuade him otherwise. Missing from The First Word are sign-language experts such as Ursula Bellugi, Karen Emmorey, David F. Armstrong, Sherman Wilcox and the late William C. Stokoe.

Given the book's breadth of coverage, such an omission is all the more surprising. But that is perhaps a minor quibble. Kenneally's dilemma is that, although she found herself excited by the ideas she encountered, many of them are mutually incompatible, so no clear pattern emerges.Nevertheless, she writes in an engaging, chatty style, and readers will gain a broad understanding of what language is about and how it might have evolved. She ends in a rather gimmicky fashion by asking various researchers their opinion as to whether and how language would evolve in a boatload of babies shipwrecked on the Galápagos Islands but provided with the sustenance to survive and thrive. My response? They might well develop language, but they'd surely all have different theories as to how it happened.