Thomason’s book is an excellent exploration of the linguistics of language contact. She does not focus on the sociolinguistics of language contact but concentrates on the linguistic results of language contact and the processes by which they come about. She explains what she means by language contact and reminds us that language contact is the norm and not an unusual phenomenon, as some North Americans may feel. No language, she says, has developed in total linguistic isolation.

A partial list of chapter titles gives some indication of topics covered: ‘Contact onsets and stability’, ‘Contact-induced change: Results’, ‘Contact-induced language change: Mechanisms’, ‘Contact languages I: Pidgins and Creoles’, ‘Contact languages II: Other mixed languages’, ‘Language death’, and ‘Endangered languages’.

Two key concepts that T develops are ‘imperfect learning’ and ‘shift-induced interference’ (the index helpfully listing the many pages where these are used). T differentiates the results of language contact when there is ‘imperfect learning’ or bilingualism, pointing out several differences. For example, the ‘mixed languages’ Michif and Mednyj Aleut developed in communities where speakers controlled both languages, but pidgins arise where there is imperfect learning.

T presents a number of intriguing claims that scholars will want to evaluate in light of their own research, such as evidence of language interference from a second language back into speakers’ first language, evidence against Derek Bickerton’s hypothesis as an explanation of creole genesis, the claim that codeswitching is not a major mechanism of shift-induced change, and a recognition of multiple causes in cases of contact-induced change.

T’s discussion of Charles A. Ferguson’s Ethiopian Language Area (Sprachbund) should be tempered by Mauro Tosco’s ‘Is there an “Ethiopian Language Area”?’ (Anthropological Linguistics42.329–65, 2000).

With a 25-page glossary, a helpful index, and clearly written text, the book is useful as a textbook for graduate students (who have had an introduction to sociolinguistic concepts) but wrestles with enough cutting edge questions to challenge mature scholars as well.

The editors have tried a format that was apparently meant to be friendlier to the less scholarly reader by avoiding the citation of sources in the text. Instead, each chapter closes with ‘Sources and further reading’, where T suggests additional useful reading on the subject and then conversationally mentions all of the authors she has quoted or alluded to in the course of the chapter, citing page numbers where applicable. This, together with nearly invisible section headings, was presumably done to make the text flow more continuously. But this is not a novel. I trust this experiment will be recognized as a failure, and the publishers will return to the established practice of [End Page 672] linking citations to their actual references in the text. A second edition should fix these problems (and one certainly does hope for an updated edition in the future), as T provokes new study to answer some of the questions this book raises and leads readers to rethink previously overlooked relevant points in the literature.