Dying words: Endangered languages and what they have to tell us (review)

This very well-written book examines the problems of language endangerment and death, and their far-reaching consequences for linguistics, anthropology, and several related fields. Drawing on an encyclopedic range of examples, Nicholas Evans establishes the basic link between linguistic diversity and population (Ch. 1), the diversity of recording technologies (from cuneiform to digital video, Ch. 2), the surprising diversity of grammatical systems (Ch. 3), the embedding of language in social cognition (Ch. 4), the role of linguistic evidence for understanding prehistory (Chs. 5 and 6), the light shed by spoken languages on ancient scripts (Ch. 7), the resurgence of research on linguistic relativity (Ch. 8), the expressive power and fragility of oral cultures (Ch. 9), and the challenges facing documentary linguistics today (Ch. 10). The treatment of these topics is brief but erudite and full of fascinating facts from comparative linguistics, the histories of languages, writing systems, and linguistics (from Pāṇini to the present). The tone is one of loss, given the stories of last speakers and lost languages. But this is always tempered by the excitement of discovery, a palpable respect for languages, their speakers, and the scholars who study them, and a passionate optimism that there is great work to be done. Overall, this is a book that should be widely read by linguists, linguistic anthropologists, and anyone concerned with language in its total context of mind, society, history, and nature. The extensive examples (from many scores of languages), excellent figures, index and bibliography, and consistently lucid arguments make this a superb book for graduate or upper-level undergraduate seminars.

In the opening chapter of Dying words (DW), E points out that roughly ten millennia ago (on the threshold of sedentism), about ten million humans spoke between three and five thousand languages (a number derived by estimating one language to each hunter-gatherer group of circa two thousand speakers, the rough upper limit of sustainability). The figure ten million represents 0.5% of today's global population, which speaks about six thousand languages. Even as rough estimates, these proportions make E's point forcefully: there has been a vast decline in the number of different languages relative to human population. Moreover, the distribution of languages to world regions is extremely uneven, with global languages spoken (in various forms) across areas of huge extent, and other small areas, which E calls 'hotbeds of diversity', showing many hundreds of endogenous languages (see table 1.2, p. 17). What is most important about this deceptively simple observation is the way it ties together four elements that will make up the backbone of DW: (i) population size, a purely numerical value, (ii) geographical extent in terms of area, ecological features, and natural resources, (iii) mode of subsistence, with a focus on hunter-gatherers, for whom geographical extent constrains population size, and (iv) language diversity, especially the number of distinct languages that correspond to the population groups. By way of resource availability, mode of subsistence places pressure on language diversity in a compression effect, such that smaller populations in a greater number of groups lead to greater numbers of languages. This effect is derailed with the explosion of human population through sedentism, the industrial revolution, and modern technology, leading to what might be called widespread linguistic monocropping. The demographically driven decline in diversity entails a loss of knowledge, a key problem for research, which E illustrates with ecologically fine-grained vocabulary from Seri (of Baja California) and languages of Arnhem Land (one of his major areas of research).

Each of these four core elements splits into further elements, some of which E develops throughout the book. Population measures raise the question of social organization, if only because the two are entirely different ways of describing collectivity: in a population, every individual counts for just one, and the aggregate is structured by age and sex only. In a social structure, individuals are neither monovalent nor interchangeable, and we...