Dimensions of register variation: A cross-linguistic comparison By Douglas Biber (review)

166LANGUAGE, VOLUME 75. NUMBER 1 (1999) Poizner, Howard; Edward S. Klima; and Ursula Bellugi. 1987. What the hands reveal about the brain. Cambridge, MA: MIT Press. Williams, George C. 1996. Plan and purpose in nature. London: Weidenfeld and Nicholson. Department of Linguistics University of Canterbury Private Bag 4800 Chnstchurch. New Zealand [a.c-mcc@hng.canterbury.ac.nz] Dimensions of register variation: A cross-linguistic comparison. By Douglas Biber. Cambridge: Cambridge University Press, 1995. Pp. xvi, 428. Reviewed by Thomas E. Nunnally, Auburn University Among researchers of register variation, few have more successfully approached the area quantitatively than Douglas Biber, or proffered a more satisfying theory. B developed his own brand of computational and corpus linguistics to move investigation of variation in register from the realm of subjectivity to verifiable quantification. Basically, B argues, registers (functional varieties) vary along several dimensions (i.e. continua), these dimensions themselves, in his words, 'realized by different sets ofco-occurring linguistic features, reflecting different functional underpinnings (e.g., interactiveness, planning, informational focus and explicitness)' (36). In this book, B works from his and his associates' previous analyses of the dimensions of register variation in four languages both to summarize and to significantly extend more than a decade of pioneering work in his multidimensional (MD) approach. As he explains, however, this large study also illustrates how much remains to be done. Modesty ofclaims is to be expected, considering B's well-articulated criticism of other studies in the field that overgeneralized a few binary register or generic differences into blanket statements stereotypically contrasting 'oral' and 'written ' English. B's specific goal is to determine if the MD approach can discover cross-linguistic generalizations (or, loosely, universals) of register variation. B therefore presents this study as no more than a preliminary to discover if register variation across languages holds promise of elucidating cross-linguistic generalizations (i.e. universals). Building upon the MD framework familiar to Language readers from articles by B (1986) and by B and Edward Finegan (1989) and from B's 1988 book Variation across speech and writing, this study compares previously completed MD analyses (with occasional slight modification for better comparison and with some extensions of earlier findings) of English, Korean, Nukulaelae Tuvaluan, and Somali. (The earlier single-language MD analyses themselves were performed by B for English, Niko Besnier for Nukulaelae Tuvaluan, Yong-Jin Kim for Korean, and B and Mohamed Hared for Somali.) To the book's nine meaty chapters and a short conclusion B adds appendices outlining the grammatical structures of Korean (364-80) and Somali (381-402). The first chapter, 'Introduction' (1-26), serves several important functions. It reviews previous research on register (a situationally defined language variety) and computational research into the allied field of sublanguages, introduces B's own MD approach to register variation, and overviews the entire study. B links his quantitative endeavor closely to primary questions of register variation voiced decades ago by Dell Hymes, Charles Ferguson, and others, giving his study credibility and historical continuity. He compellingly discusses four reasons for crosslinguistic analyses of register variation and preparatorily describes the four languages (and some of their cultural correlates) that are compared. Ch. 2, 'The comprehensive analysis of register variation' (27-37), goes on to establish the requirements for 'a comprehensive analytical framework for studies of register variation' (27). Especially important is B's distinction between register markers (linguistic features distinctive to a register) and register features, that is, 'differing quantitative distributions of core linguistic REVIEWS167 features' (28). The quantitative analysis of linguistic features is at the heart of B's MD approach, that is, he attempts to understand register not as the presence or absence of markers but as the varying co-occurrence of features. The chapter then turns to the role of computational analysis, examines various available machine-readable corpora, and ends with a theoretical overview of the MD approach. Ch. 3, 'Sociocultural description of the four language situations' (38-58), explains the background ofthe four languages and the cultures underlying them in more detail and lists the registers identified for each. An occasional discrepancy between the tables and the textual explanation occurs but is not insurmountable (e.g., the Somali register 'analytical...