American Scientist

Origins of Language: Constraints on Hypotheses. Sverker Johansson. xii + 345 pp. John Benjamins Publishing Co., 2005. $114.

After a long period of disrepute, the evolution of language has recently reemerged as an important focus of scientific discussion—a shift fueled both by exciting discoveries in paleoanthropology and by rapid progress in understanding the neural and genetic mechanisms involved in human language. The complexities of these domains combine to make the core questions of biolinguistics some of the most challenging scientific issues of our time: What is it about our brains that allows humans, but not other animals, to freely convey arbitrary thoughts to one another? What genes underlie these differences? How, and why, did the special neural mechanisms behind language evolve in our species?

Despite rapid progress and a growing consensus that such questions are scientifically answerable, major obstacles stand in the way of resolving these matters. These roadblocks include barriers to interdisciplinary communication (exemplified by the range of meanings workers in different spheres have attached to such words as language or symbol) and a pervasive tendency to propose oversimplified, single-cause hypotheses. These and other difficulties have made biolinguistics a highly contentious field, riddled with mutual misunderstandings.

Sverker Johansson, a physicist by training, comes into this area of study as an outsider, one who is attempting to gather together the many strands of data and theory relevant to the study of language evolution. His fresh, enthusiastic view and clear, pragmatic approach are what one might expect of a physicist. He lists the main problems and proposed solutions, briefly reviews the various sources of data and in conclusion fills in a 4 x 8 summary table with estimates of the likelihood of the various theoretical possibilities. I can't help admiring the approach, because most single-author books in this field defend the writer's own new solution to the problem of language evolution; few dispassionate surveys are available.

In his search for information, Johansson casts his net wide, and the voluminous selection of references presented (80 pages' worth, in small type) is, by itself, an impressive contribution. For anyone who believes there is little data relevant to theories of language evolution, this book will serve as a useful corrective.

The downside is a distinct trade-off between breadth and depth. Many of Johansson's reviews are lists, of the form "X said this, but Y said the opposite," giving little detailed critical analysis of either the disputes themselves or the types of investigation required to resolve them. The reference list is something of a hodgepodge, with popular books, journalists' summaries and unpublished Web documents receiving equal billing with peer-reviewed scholarly articles. Worse, data that have been thoroughly discredited are sometimes repeated uncritically, including some proposed fossil indicators of speech, such as the size of the hypoglossal canal (formally withdrawn by its originators). Given that such data figure quite prominently in Johansson's final judgment that speech must have arisen early in human evolution, these oversights end up being rather significant for the overall conclusions.

Champanzees certainly have important ingredients necessary for human language (for example, the ability to pair arbitrary signals with meanings), but they are still not discussing philosophy or even what they had for dinner yesterday.

• Facebook
• Twitter

An example of the dangers of an uncritical attitude is Johansson's treatment of the "ape language" controversy, which provides a good lesson about how interdisciplinary misunderstandings and overblown rhetoric (on both sides) have almost hopelessly crippled this field of study. Relying heavily on popular books by ape-language researchers, Johansson concludes that apes have latent capacities for language that approach the abilities of modern humans. Quoting Sue Savage-Rumbaugh's question "Why are we afraid of apes with language?," Johansson asks, "Would not our world be richer if they did have minds and language?" Well, yes, it would be wonderful if apes (or dolphins, or dogs, for that matter) had language in a full human sense, so we could discuss with them their ideas, memories, feelings about life, and so on. But unfortunately we can't, because none of these species has, or can be trained to have, a language in anything like this rich sense—regardless of our hopes or fears. We should not discount the notable and important achievements of chimpanzees and other species; nor should we exaggerate their accomplishments and thus underestimate the evolutionary distance modern humans have traveled since our last common ancestor with chimpanzees. Chimpanzees certainly have important ingredients necessary for human language (for example, the ability to pair arbitrary signals with meanings), but they are still not discussing philosophy or even what they had for dinner yesterday. Although Johansson cites both sides of this debate, he comes down sharply on one side of it and unfortunately gives readers little basis for judging the scientific data for themselves. Such gaps make the book hard to recommend as an authoritative introduction to the field. A multi-authored volume such as Language Evolution, the recent survey edited by Morten H. Christiansen and Simon Kirby (Oxford University Press, 2003), would be a better option.

Another troubling aspect of Origins of Language is that Johansson sometimes appears to misunderstand hypotheses that he dismisses. A prominent example is the treatment of linguist Noam Chomsky, whose influential idea of an innate basis for the human capacity to acquire language is treated in a derogatory tone throughout the book. As with "ape language," Chomsky's hypothesis of a Language Acquisition Device has occasioned much controversy, and Johansson should have dispassionately reviewed the pros and cons of this idea rather than recycling the various misinterpretations of Chomsky perennially erected by his opponents: that language is "monolithic," that language couldn't have evolved and so on. Many of these misinterpretations have been recently and forcibly rejected by Chomsky himself, in papers that Johansson cites as "important"—so why continue to criticize him based on these misconceptions?

Various other interesting hypotheses receive similarly short shrift, including a possible link between the evolution of music and language, the "holistic protolanguage" theory that complex phonology might have preceded complex syntax, and the idea that language evolved first in the service of thought and was only later put to use for communication. Although each of these ideas has advantages and disadvantages, none of them is simply silly: Each makes testable predictions, and all deserve a more balanced and insightful treatment than they receive here. Thus at least some of the "constraints on hypotheses" presented in this book seem to result more from a failure of imagination than from rejection forced by current data.

Despite these criticisms, Johansson's book—the first book-length, single-authored attempt I know of to synthesize this fascinating and rapidly growing field—makes several important contributions. First, it offers a good overview of the many relevant strands of data—from fossils, from genes, from brain imaging, from animal communication. Although the quality of Johansson's reviews varies, the very attempt to synthesize these diverse scientific fields is admirable and useful, and the flaws stand as a challenge for future scholars to try to do better. More important, the basic scientific attitude embodied in the book—survey the hypotheses, survey the data and then combine them to exclude some hypotheses—is certainly the correct way forward for the field. It is rather surprising that no one has done it before, and anyone interested in entering the field of language evolution should look to this book for an overview of some of the important debates.