<b>An acoustic analysis of vowel variation in New World English.</b> By Erik R. Thomas. (Publication of the American dialect society 85.) Durham, NC: Duke University Press, 2001. Pp. ix, 230. ISBN 0822364948. $20.

Reviewed by: An acoustic analysis of vowel variation in New World English by Erik R. Thomas Robert Hagiwara An acoustic analysis of vowel variation in New World English. By Erik R. Thomas. (Publication of the American dialect society 85.) Durham, NC: Duke University Press, 2001. Pp. ix, 230. ISBN 0822364948. $20. This monograph expands the dialectology of English in North America by presenting 192 'case studies' of individual speakers' vowel systems, each summarized as a two-dimensional vowel space. These represent not impressionistic auditory assessments of height and backness, but rather the average first and second formant frequencies (i.e. the frequencies of the two lowest vocal tract resonances, which are generally well correlated with the traditional descriptive dimensions) as measured by the author. Acoustic analytic techniques were applied to audio recordings of each speaker and the measurements used to plot each speaker's vowel production. Following a brief preface (vii–ix),Ch. 1, 'Introduction and methods' (1–14),places the work in the context of a growing body of sociolinguistic investigations making use of acoustic measurements, especially to investigate subtle vowel variations that are difficult to assess via traditional phonetic transcription. Conventions are introduced for labeling vowel categories and contexts (e.g. phonemic symbols and diacritics such as a raised O to indicate a following voiceless obstruent), and the information contained in (and excluded from) the vowel plots is discussed. Ch. 2, 'Vowel variants in New World English' (15–58),provides an overview of the major 'diaphones' of each of the basic North American English vowels and diphthongs. This includes both intrinsic regional variants (e.g. the fronting of /α/ in Northern-Cities Shift speakers) and contextual allophones (such as the typical neutralizations of vowel contrasts before /r/). The case studies begin in Ch. 3, 'Whites from the north' (59–103), which includes speakers from most of the US and parts of Canada. Each speaker's vowel production is represented as a vowel space plot, with each vowel represented as a vector determined by the first and second formant frequencies at points early and late in each vowel. The remaining chapters present vowel summaries from 'Whites from the southeast' (Ch. 4,105–32), 'White Anglos from the south-central states' (Ch. 5,133–60), 'African Americans' (Ch. 6,161–84), 'Mexican Americans' (Ch. 7,185–95), and 'Native Americans' (Ch. 8, 197–200). The absence of a concluding chapter suggests that this work is meant to pioneer and inspire further work that combines traditional sociolinguistic questions with instrumental methodologies. This goal mitigates some of the book's drawbacks, such as the relative difficulty of isolating a particular vowel or context of interest from a complicated vowel diagram, or the uneven distribution of speakers sampled. While three areas (North Carolina, Ohio, and Texas) are extensively covered, speakers from other regions (Canada, the Caribbean, and the western US) represent only a handful of the total. This work makes an important advance in what might be called 'instrumental dialectology'. Its main strengths are that it allows the reader to compare overall patterns in vowel production across an enormous array of different speakers, and it illustrates the acoustic patterning of many well-described vowel alternations and shifts. The book is not bogged down by endless tables of numerical acoustic measurements, statistical approximations of intrinsic variation, or other information which, while interesting, would have hindered its readability and accessibility. Robert Hagiwara University of Manitoba Copyright © 2004 Linguistic Society of America

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.