BOOK NOTICES 621 Swann does not merely adopt a feminist stance in making her suggestions for teaching; she balances feminist advocacy with critical questioning , acknowledging the issue, for example, of whether 'equality' is always desirable. She concludes that equal opportunity initiatives need to be considered in the context of each school and classroom. GiWi, boys, and language is structured more like a textbook than a research study. It makes heavy use of chapter summaries, repeats key ideas, and surveys a wide array of research conducted over the past fifteen years. It is therefore easily readable and would be an appropriate text for undergraduate students in linguistics and education. [Ruth E. Ray, Wayne State University .] Gender and conversational interaction. Ed. by Deborah Tannen. (Oxford studies in sociolinguistics.) New York & Oxford: Oxford University Press, 1993. Pp. xiv, 327. Cloth $45.00, paper $16.95. Over the past twenty years, the relationship between gender and language has been explored by a variety of methods and approached from different standpoints. The editor characterizes the approach used in this collection as 'ethnographically oriented discourse analysis or, alternatively , interactional sociolinguistics' (3). The book's ten papers, five of which have appeared elsewhere, are grouped into four parts. In Part I, 'Talking among friends' (15-80), the first two papers focus on junior high school and high school girls' talk. Among the findings by Donna Eder is that the girls of the group she studied were able to detach themselves from traditional views of feminine behavior through humor and verbal play. Penelope Eckert discusses how the girls in her study group negotiated behavioral norms and balanced through talk the conflicting needs for popularity and independence . Barbara Johnstone, who analyzes spontaneous stories based on personal experience told by white middle-class midwestern men and women, seeks to determine the sources of power the storytellers draw on to overcome dangerous situations. According to the author, because women often consider themselves powerless individuals, the source of their power is the community, its resources being tapped through discourse. By contrast, power for men comes from individual actions in opposition to the actions of others. Part II, 'Conflict talk' (81-162), also consists of three papers. The first two are based on research conducted among American schoolchildren . Amy Sheldon shows that the gendered nature ofchildren's peer talk may be established as early as the age of three, with boys tending to be more heavy-handed and controlling. Next Marjorie Harness Goodwin points up the differences between boys' and girls' dispute stories . Whereas boys tend to deal with the wrongdoings of present participants at the time, girls confront the offending party at some future time. The third paper is based on research among speakers ofTzeltal in the southern Mexican community of Tenejapa. Penelope Brown examines women's discourse in the only setting in which peasant Mayan women may engage in direct confrontation—the court. According to Brown, 'even when [Tenejapan] women are not being polite, characteristic female strategies of indirectness and politeness are manifested in their speech ... [and accordingly] for a deeper understanding of language and gender, we need to take very detailed looks at gender behavior in different situations' (159). The remaining sections, 'The relativity of discourse strategies' (163-227) and 'Critical reviews of the literature' (229-312), comprise two papers each. In Part III, Deborah Tannen demonstrates that 'the intersection of language and gender provides a rich site for analyzing how power and solidarity are created in discourse ' (183). Among the questions Carole Edelsky asks is, 'Under what conditions do men and women interact (e.g., hold the floor) more or less as equals and under what conditions do they not?' (221). That apparently depends on whether whoever has the floor controls it or shares it with others: in the former case men tended to be dominant, whereas in the latter case men talked less than women. The two critical reviews take up interruptions as a means of controlling interactions and gender differences in the amount of talk. According to Deborah James & Sandra Clarke, when it comes to interruptions, the majority of studies do not support any significant difference between men and women—other things being equal. And as to...