Studies of ethnolinguistic variation and studies of regional vowel variation tend to live in separate worlds, owing in part to their different research questions. Ethnolinguistic studies examine the role of linguistic variation in constructing and expressing ethnic identity (e.g., Fought Reference Fought2006), while studies of regional vowel variation focus on large-scale changes (e.g., Labov, Ash & Boberg Reference Labov, Ash and Boberg2006). The finding that ethnic boundaries often impede participation in ongoing changes may also bolster this separation (e.g., Labov Reference Labov and Eckert1991). In this book, a published version of her 2003 University of Michigan Ph.D. dissertation, Anderson attempts to bridge the gap between studying ethnolinguistic variation and analyzing regional vowel variation by examining the social and linguistic conditioning of two phonetic variables in the English of White Appalachians and African Americans in Detroit.

After a brief introduction (pp. 1–3), chapter 2, “Empirical and theoretical background” (4–17), describes the major vowel changes in progress in American English. The Northern Cities Shift operates within an area that includes Michigan, and the Southern Shift is relevant to both White Appalachian and African American migrants to Detroit, though African American English (AAE) is not generally viewed as participating in either of these shifts (Labov Reference Labov and Eckert1991). While less salient components of these shifts, such as the fronting of /u/ and /ʊ/, have not been reported in AAE, monophthongal realizations of /ai/ are a salient marker of both regional (“Southern”) and ethnic (“African American”) identity. This chapter also considers the role of internal and external factors in language change, especially that of language ideology, here defined (14) as “conceptual schemes … used to interpret and understand linguistic variation.”

Chapter 3, “The sociolinguistic and demographic context for the study” (18–34), provides the relevant details of Michigan demography and the history of Southern migration to Detroit. Beginning in World War I and continuing into the middle of the 20th century, White Appalachians migrated to Detroit from West Virginia, Kentucky, Tennessee, and North Carolina, and African Americans migrated from the Deep South and plantation states (as well as some of the Appalachian states). This chapter also discusses relevant aspects of both Appalachian English (AppE) and AAE in the South and in Michigan, where both groups are in contact with Midwestern urban Whites. Before proceeding to the main study, chapter 4, “The pilot study” (35–46), reports on a preliminary study that informed the subsequent research design.

Chapter 5, “Field techniques and acoustic methods” (pp. 47–64), describes the methodological and analytical details of the main study. Ethnographic interviews were conducted with 12 Detroit-area women, divided by generation of arrival (1st vs. 2nd) and ethnic group (White Appalachian and African American). For each of the relevant vowels, measurements of F₁, F₂ and F₃ values were taken at two locations, and the effect of the following phonological context (place and voicing) was noted. These measurements were compared with those of Hillenbrand et al. Reference Hillenbrand, Getty, Clark and Wheeler1995 for 48 Midwestern White women.

The next two chapters, chapter 6, “The high and lower-high back vowels” (65–101), and chapter 7, “The patterning of /ai/” (102–28), present the analyses of the two variables: fronting of /u/ and /ʊ/ and glide-weakening (monophthongization) of /ai/. For fronting, the effects of following context are parallel for AppE and AAE, with less fronting before velars than before alveolars, though AAE has higher rates of fronting before velars than does AppE. This result conforms with that of Nguyen & Anderson Reference Nguyen and Anderson2006, who found the following context to have a greater effect on fronting in AAE than in Midwestern urban White English. A comparison with Nguyen’s (Reference Nguyen2006) real-time study shows the effects of both social class and following context on fronting to have weakened over time. For the more socially salient variable of glide weakening, the effect of the following context is also significant, though stronger for AAE than for AppE. A speaker-by-speaker analysis and a comparison with Nguyen’s (Reference Nguyen2006) real-time study suggest that glide weakening before voiceless consonants is a change in progress for at least some speakers of AAE.

Chapter 8, “The local and supralocal contexts for the patterns of usage” (129–78), uses comments from the ethnographic interviews to contextualize the linguistic results in the local contact situation and in American English more generally. Both AAE and AppE speakers mentioned their ties to the South and their maintenance of Southern cultural practices in Detroit, similarities between Southern Whites and African Americans, differences between Southern migrants and Midwestern Whites, and differences among speech varieties in Detroit. On the basis of these comments, Anderson suggests that the relevant opposition in Detroit is not between African Americans and Whites but rather between Southern migrants (whether African American or Appalachian White) and Midwestern Whites. This view is supported by the linguistic results: While fronting is constrained by (universal) phonetic features and is used by the AAE speakers, and thus does not index regional or ethnic identity, glide weakening is constrained by social rather than linguistic features and appears to index Southern identity (whether White Appalachian or African American). Anderson attributes these findings to dialect contact and leveling subsequent to Southern migration to Detroit. The final chapter, “Conclusions and implications” (179–83), provides some brief comments on the results and discusses the limitations of the study and its implication for future work.

Although I don’t normally use book reviews to comment on typographical errors, in this book they are so pervasive that they frequently impede comprehension. Particularly annoying is the global replacement of {S} by {ʊ} in chapters 6 and 7. I also found the author’s use of the adjective “spectral” (for “spectrographic”) a bit spooky.

The book’s major strength is the number of analytical issues it addresses. Anderson is careful to note the limitations of instrumental analysis and the problem that the lack of methodological agreement in sociophonetics poses for cross-study comparison. Particularly noteworthy is the attention paid to sampling temporal locations within the vowel and to the effects of the phonological context. In addition, rather than relying on absolute or normalized formant frequency values, Anderson uses measurements of the distances between the back vowels /u/ and /ʊ/ and their front counterparts /i/ and /ɪ/ to place fronting within the context of the vowel system. Similarly, her comparison of F₁ and F₂ movement between /ai/ and /ɑ/ allows for the specification of both length and direction of glide in diphthongs.

More importantly, in my view, Anderson’s results provide further impetus to problematize simple, essentialized views of ethnicity as “Black” and “White” (cf. Walker Reference Walker2001). She provides evidence that at least some African Americans do participate in widespread (thus, “not just ‘White-spread’”; 183) vowel shifts. I endorse her call for sociophonetics to move beyond what she labels “the conventional triumvirate of race, gender, and class” (165), and to make use of speaker beliefs about and attitudes toward ethnicity and language, which “cannot be determined simply by correlating linguistic features with social features determined by the researcher” (165). As she admits, a shortcoming of the qualitative, “ethnographic” approach she adopts is that she is not able “to operationalize ideology in an concrete way” (182; emphasis added). Future work in this vein might benefit from the kind of approach taken by Hoffman & Walker Reference Hoffman and Walker2007, in which speaker attitudes toward ethnicity are systematically investigated and operationalized as a separate set of factors. Thus, although this book does not entirely succeed in its stated goals, it does open up a number of important methodological and analytical issues that future work in both ethnolinguistic variation and regional vowel variation needs to address.