Talk that counts: Age, gender, and social class differences in discourse. By Ronald K. S. Macaulay. Oxford: Oxford University Press, 2005. Pp. 236. ISBN 0195173821. $39.95.

Reviewed by Charlotte Brammer, Samford University

In Talk that counts: Age, gender, and social class differences in discourse, Ronald K. S. Macaulay applies quantitative analysis ‘to determine to what extent variation in the use of certain linguistic features correlates with extralinguistic categories, in this case, age, gender, and social class’ (8). He situates this text with other variation studies (notably Labov 1966, Wolfram 1969, Fasold 1972, Trudgill 1974, Milroy 1980, Coates 1996, Eckert 2000) and claims a continued interest in ‘the distribution and effects of stable differences … similar to those expressed in Bernstein’s early work (e.g., Bernstein 1962)’ (5).

To explore linguistic variation in terms of age, gender, and social class, M works with two groups of data: (i) Ayr interviews, a recorded collection of twelve interviews with speakers from Ayr, Scotland (reported in Macaulay 1991), and (ii) Glasgow conversations, recorded talk between friends (reported in Stuart-Smith 1999). The Ayr interviews were analyzed for social-class comparisons; the Glasgow conversations were analyzed for age, gender, and social class. Using frequency counts, comparisons with other variationist studies, and some statistical analysis (primarily the Mann-Whitney nonparametric test), M provides good support for his findings. After thoroughly discussing the corpora, his methods, and specific definitions of social class, age, and gender, M unveils his findings in Chs. 6 through 14.

In terms of social class, M identifies differences in frequency of use and variation of application, and reveals as well similarities in the use of specific discourse features. For example, he found that middle-class speakers use passive voice more frequently than working-class speakers do (p < 0.05) and that working-class speakers use dislocated syntax (e.g. she was a very quiet woman my mother) more frequently than middle-class speakers do (p < 0.001) (Ch. 8). In terms of modifiers, M found that middle-class speakers use evaluative adjectives more frequently than do working-class speakers (p < 0.004) (Ch. 10). He also found minimal social-class differences in the frequency of use of you know (Ch. 7) and the use of modal auxiliaries (Ch. 9), and suggests that syntactic variation may not be as salient for exploring social-class linguistic differences as originally thought.

Gender, particularly when linked to social class, may yield the more interesting findings. Age is also an important aspect of his findings. In the Glasgow data, for example, M found that females, especially middle-class women and girls, use you know substantially more often than do males, and that middle-class females are more likely to use you know for emphasis or elaboration, reserving I mean for explanations. In Ch. 7, M proposes that ‘the use of these discourse lubricants is a distinctive part of the discourse style used by middle-class women and that their daughters are learning to follow their example’ (86). Two important discoveries include the females’ more frequent use of coordinate clauses and because clauses (p < 0.05). M found neither modals nor modifiers particularly promising as identifying either gender- or age-specific discourse features. Pronouns are more frequently used by females than males, and M points out that this is consistent with other research, noting that ‘in their use of articles and pronouns the males show themselves to be less interested in people than are the females’ (Ch. 11, 138).

In addition to the highly accessible and engaging discussion, the text’s rich context and extensive bibliography (some thirteen pages) make it an important resource for future studies in discourse and variation analysis.