According to the blurb, this book “develops a coherent theoretical approach to style in sociolinguistics” (p. i), but Coupland adds that “[t]his book can be read as a critique of variationist sociolinguistics” (xi). He wants to challenge “its dogged reliance on static social categories, its imputation of identity-values to numerical patterns (quantitative representations of linguistic variation), and its thin account of social contextualization” (xi). As a lapsed variationist himself, Coupland knows very clearly where the bodies are buried and his criticisms of the investigation of style in variationist sociolinguistics are cogent and convincing. He has great fun with Chambers’s improbable invented example of Adonis saw himself in the mirror/Adonis seen hisself in the mirror (Chambers Reference Chambers, Chambers, Trudgill and Schilling-Estes2004:3) but his main target is Labov’s (Reference Labov1966) use of style stratification.

It is, of course, easy to criticize Labov’s approach to stylistic variation with its focus on attention to speech and also his decision to treat speaking and reading aloud as a single dimension, and many others have already pointed out the problems. Coupland, however, goes further and attacks the practice of aggregating the information from a number of speakers “because aggregation rounds down the understanding of stylistic processes” (27, original emphasis). He argues that “[s]ingle-case analyses are more likely to allow an adequate sensitivity to context and contextualization, where we can come to understand what the styling of variation can achieve” (28). The remainder of the book is essentially an argument in favor of single-case studies rather than survey methods. This amounts to a rejection of the variationist approach in favor of a social psychological one.

Coupland points out that “[r]eliance on frequency as the main criterion of sociolinguistic difference introduces a considerable amount of abstraction into the account of variation” (41) and he comments that quantitative methods “keep us at a considerable distance from the primary contextual operations of speaking and from a ‘dynamic’ account of sociolinguistic styling” (41).

Coupland is clearly right to point out the inadequacies in the approach to style in variationist sociolinguistics where the interactional factors outlined by Hymes Reference Hymes1974 are generally ignored. To take an obvious example, few variationist studies based on recorded interviews report on how and where the interviews were carried out or by whom, although it is obvious that age and gender differences between the interviewer and the speaker may affect the kind of speech recorded. (In a rare exception, Pope, Meyerhoff & Ladd Reference Pope, Meyerhoff and Ladd2007 discuss the possible effects of the interviewer on two replications of Labov’s Martha’s Vineyard study.) Moreover, as Coupland points out, interviews provide very limited opportunities for different kinds of speech acts.

As an alternative, Coupland suggests a social constructionist approach:

Coupland also believes that in “late-modernity” the structural factors of region, race, social class, and gender have become less rigid than in the past and that new and more local social categories are coming to the fore: “the hierarchical social world of social class dialects is anachronistic” (104). He sees the need “to broaden the remit for style” (53) by examining “the domain of social relationships” (53). He proceeds to do so by reviewing such approaches as audience design and speech accommodation theory (54–81), with examples from his own work and that of others, though he also points out the limitations of such approaches. Coupland then makes a digression into language ideology (82–105) that is only of peripheral importance for his theory of style.

Chapter 5, “Styling social identities” (106–45) and chapter 6, “High performance and identity stylization” (146–76) are devoted to the notion of “performance” and include micro-analyses of a range of examples. The best of these are Coupland’s detailed examination of Sue’s speech from the Cardiff travel agency study (115–21) and his analysis of a speech by the Welsh politician Aneurin Bevan (156–63). These present instances of micro-analysis that are a model for anyone wishing to carry similar analyses of private or public performances. The other examples pale in comparison.

The final chapter, “Coda: Style and social reality” (177–88), sums up Coupland’s view that in sociolinguistics “it has become less and less satisfactory to work with any simple definition of style” (177, original emphasis). In Coupland’s view, sociolinguistic style “has to encompass the whole field of making social meaning through deploying and recontextualizing linguistic resources” (177). As he admits, “[t]his makes style seem like the whole of discourse” (178) and that, to my mind, is the problem with his attempt to provide a coherent approach to style in sociolinguistics.

Coupland wants to make style encompass all the personal, interpersonal, and situational factors that are present in any speech event. This is admirable in its ambition but unrealistic in practice. There are too many aspects of the situation to consider, including the speaker’s intentions, that cannot be objectively defined. Coupland’s advocacy of micro-analyses foregrounds the subjective interpretation of the investigator, but the criteria for the investigator’s evaluative decisions must come from some objective source or we are left with idiosyncratic opinions. Coupland’s effective analyses of Sue’s and Bevan’s speech come from his knowledge of the relevant speech varieties. In Sue’s case he refers to “the sociolinguistic variables that are most sensitive to social class in Cardiff” (118), and he explains that Bevan’s “dominant voice was South Wales Valleys” (157). You won’t get that kind of knowledge from individual micro-analyses unless you have a range of them and are willing to do the kind of “aggregating” that Coupland otherwise objects to. The hard work of describing social class dialects (anachronistic or not) has been done elsewhere, probably through the use of quantitative methods. In much of his discussion, Coupland has a distressing tendency to refer to such entities as “standard” and “non-standard” (or even worse, “posh” and “Cockney”) as if these were well-defined terms. The use of such labels devalues the detailed analysis of individual speech events.

From Coupland’s perspective I am an unreconstructed anachronistic sociolinguist because I still believe that survey methods using quantitative methods are useful for charting language variation. My complaint is that there appear to be too few studies of this kind being carried out. I also believe that the old-fashioned categories of age, gender, social class, and ethnicity have potentially significant effects on language use, though this will vary from one community to another, as will how each category is defined. I agree with Coupland, however, that quantitative methods are of limited value in tracking stylistic variation, though differing frequencies in the use of certain items may reveal patterns of variation that would not be obvious from micro-analyses (Macaulay Reference Macaulay1999, Reference Macaulay2005). The approach to sociolinguistics that Coupland proposes is a valuable complement to survey methods rather than an alternative to them.

Coupland’s volume would make an excellent textbook for an advanced sociolinguistics class. He writes very clearly and covers a wide range of topics and approaches. There is plenty to discuss and something provocative on almost every page. It deserves the same kind of scrutiny that he advocates for sociolinguistic investigation. I would not want my students to accept his arguments uncritically, but I would expect them to learn a great deal from a close examination of his views.