REVIEWS189 this sensitive methodological area; and even then, such areas as instrument construction and research design—including the bringing together of ethnographic and quantitative procedures—will remain without proper coverage. I do not think we can afford to educate yet another generation with no more than a teaspoon of familiarity in the areas of research design and data analysis. By doing so, we foster the continuation of the methodological ignorance that has now plagued a generation of sociolinguistic interest, particularly in societal phenomena. Even more serious is F's division of the sociolinguistic enterprise into two separate volumes: this one with virtually no linguistics, and the one to come probably with no sociology. This would be bad enough if all students of the sociolinguistic enterprise were to take a full-year sequence in this area, in which the present text and its presumed sequel could be studied in turn. That would not really be the intellectually most satisfying solution to the problem, which is precisely that of integrating—not separating—the two foci that make up the total sociolinguistic enterprise; but it would be a step forward. As it is, however, there will definitely be many students (indeed, probably a majority) who will study only one of the two texts, and who will therefore inevitably miss half the field. Those whose half-exposure will be restricted to this volume will receive no introduction at all to Hymes, or Labov, or Shuy, or to linguistic and sociolinguistic rules and variables. It remains to be seen what those exposed only to the follow-up volume will need to forgo. I wonder whether it is really necessary to pay that price. Even if both volumes are studied, and even if they become the modal approach to sociolinguistics (a noteworthy step forward from where most courses are today), the true challenge is to present linguistics and sociology together— in interaction, and in progression from micro to macro. I do not say this out of personal pique. Much of my work is mentioned in the volume under review, and I have almost no complaints against how I am interpreted. However, the major challenge has not yet been joined, either in terms of building systematically from lower-order, face-to-face phenomena to higher-order, long-range, national policy (with a series of units and a set of theories that can tie these levels together), or in terms of the inclusion of either sociological or linguistic theory as such. If linguistic theory is to be presented in Volume II, then where oh where will sociological theory be presented? Can a textbook on society be written with no mention of Comte, Durkheim, Weber, or Parsons? This volume (and its companion) may well be the best we have for the time being. It represents an improvement on what existed before, but is still far from what sociolinguistics should be. [Received 18 July 1985.] The social life of language. By Gillian Sankoff. Philadelphia: University of Pennsylvania Press, 1980. Pp. xxii, 373. Reviewed by Shana Poplack, University of Ottawa This collection of 15 articles was written or co-authored over a ten-year period (1968-78), which Sankoff characterizes as a transition from the study of the social context of language use toward the study of how the social nature of language can influence its structure. Uniting widely-cited and classic articles with equally valuable though lesser-known ones, the volume provides an excellent overview not only of S's œuvre, but of the field as a whole. This book represents sociolinguistic scholarship in its broadest sense—incorporating 190LANGUAGE, VOLUME 62, NUMBER 1 (1986) ideas, methods, and data from areas as diverse as the sociology and politics of language, cognitive anthropology, dialectology, lexicostatistics, pidgins and creóles, bilingualism, functions of language use, discourse analysis, and linguistic variation. As a student and practitioner of anthropological, interactionist , functionalist, and variationist approaches to sociolinguistics, S has culled what is pertinent from each and integrated them into a holistic vision—all the more striking when compared to present trends of increasing fragmentation and dissociation among the subfields generally referred to as 'sociolinguistics'. Remarkably enough, the papers gathered here are by and large as timely...