Although it underlies all our scientific and philosophical endeavors, human language ability remains one of science’s greatest puzzles. Generalizations regarding the functioning of language, the mechanisms that enable infants to learn it in the face of limited exposure to data, and its biological evolution are currently topics of vigorous debate. Of these topics, language evolution is perhaps the most speculative, having been delineated as a clear subject of study only in the last century. Why only us, by Berwick and Chomsky, approaches this topic by regarding genetic, neurological, and paleontological evidence on the origins and evolution of language from a computational perspective based on Chomsky’s minimalist program [1].
In this context, by “language evolution” the authors do not mean (historical) changes in the way particular languages are used by their speakers, but rather “evolution of the organisms that use language” (p. 83). This distinction points to a core premise of the book, namely, that language should be regarded as a biological organ on a par with other biological systems. Therefore, discussions on the nature of this cognitive organ and arguments in support of this view drawing on biological and linguistic evidence are allotted a significant portion of the book.
In order for a theory of language evolution along these lines to succeed it needs to be able to relate evolution at the genome level to basic phenotypic characters. But where among the diversity of languages can such characters be found? Berwick and Chomsky propose that this apparent diversity can actually be accounted for by a few underlying principles. They propose that the crucial evolutionary event that gave rise to the language capacity was the evolution of the neurological basis for an operation called “merge.” This operation consists simply of forming a hierarchically structured syntactic object (a set) out of two other syntactic objects. The merge operation, they argue, is uniquely human, and underlies what they call the basic property of human languages, that is, their capacity to “[yield] a digitally infinite array of hierarchically structured expressions.”
This position has far-reaching implications. As a result of the appearance of an internal cognitive operation, language thus understood would not have evolved as a mechanism for communication, but rather as an “instrument of thought.” In support of this hypothesis, the authors contrast the hierarchical nature of linguistic expressions to the constraints imposed by the sequential externalization of such expressions. Though rather technical, these points are discussed very clearly and illustrated with simple motivating examples. In fact, one of the great merits of this book is its accessibility to the nonspecialist. It remains accessible even when it covers the connections between the minimalist account of language evolution and other linguistic theories, and its philosophical background.
The second major theme of the book is the question of how this account fits current evolutionary thought. The initial chapter develops the authors’ views on evolutionary biology and outlines, in this context, the hypothesis concerning the evolution of merge as the cognitive trait responsible for a range of abstract mental abilities, and ultimately for the language capacity. Somewhat uncommonly in cognitive science, the authors reject simplistic selections explanations in favor of a considerably more nuanced view of evolution--one that dismisses gradualism and incorporates stochastic factors, non-selectional mechanisms, genotypic shifts, symbiosis, and morphogenesis. Their dismissal of gradualism is consistent with the hypothesis that the language capacity is uniquely human and arose in a relatively brief period of time between 200,000 and 60,000 years ago. A variety of evidence is reviewed--from birdsong patterns and recent neuroscience findings, to ancient deoxyribonucleic acid (DNA) data--in support of these positions.
This is all very clearly argued, but there are a few weaknesses. The case against gradualism strikes me as over-elaborated. Serious doubts have been cast on the gradualist position since at least the 40s [2]. It seems hardly necessary for a modern book to adduce new arguments. The authors are also careful to emphasize that facts such as the disproportionate role of stochastic factors in small populations (as the early human populations are assumed to have been around the time language arose) are compatible with natural selection. However, natural selection is not central to their argument. While there is evidence to support the notion that language is a “tool for thought,” the idea that such a tool arose from a “simple genetic change” and was selected for the advantages it conferred to the mutant seems a lot less convincing, not least due to inherent limitations of selectional explanations [3].
Yet, although current knowledge of the neurological, mental, and genetic foundations of language remains patchy, Berwick and Chomsky manage to suggest compelling avenues for future research. Their book will certainly provide plenty of food for thought in an area that is likely to attract growing interest in the coming years.
More reviews about this item: Amazon, Goodreads