Discourse on the move: Using corpus analysis to describe discourse structure (review)

In the relatively few corpus-linguistics conferences that I have attended, more often than not there has been a panel discussion on whether using electronic corpora is a 'methodology' or a 'field'. The emerging consensus would seem to be that corpus linguistics is a methodology, and one particularly suited for delineating lexicogrammatical patterns of language use, in the pursuit of which ingenuity and intuition may after all play a useful role. On occasion, this has led to a certain kind of incidentalism—'Oh, here is another interesting fact about usage'—thus placing corpus linguists in the category that John Goldsmith (2007) perhaps caustically identifies as 'data fetishists'. Although there is a lot of data in this volume, it is by no means treated incidentally, because the volume breaks new ground by attempting to show how corpus analysis might be used to elucidate and describe the organization of professional discourses, more particularly the scientific research article, the fundraising letter, and the university class. The first and last of these have been the province of the first author, his colleagues, and students at Northern Arizona University, while philanthropic genres have been studied by the other two authors and their associates at Indiana University-Purdue University Indianapolis.

This volume is the twenty-eighth to appear in John Benjamins's prestigious 'Studies in corpus linguistics' series. The series started in 1998, and so has maintained a laudable publishing pace of an average of three volumes a year, and it got off to an excellent early start with Alan Partington's lively and engaging Patterns and meanings. Some of the later volumes have been relatively limited in scope, such as those that have derived from Ph.D. dissertations; others have been more wide-ranging, such as those written or edited by the main authors of this volume. I say 'main authors' advisedly since the authorship of the volume under review is a complicated matter. The authors explain in the preface:

we have been eager to structure the book as a coherent treatment of the subject: an authored book rather than an edited collection of articles. Thus, the three book authors share equal responsibility for revising and editing all chapters, and ultimately the content of all chapters. But, on the other hand, each chapter has different primary authors, including several co-authors in addition to the three book authors for Chapters 1–3, 5–7, and 9. Two chapters were invited, single-authored contributions—Chapter 4 by Kanoksilapatham and Chapter 8 by Csomay.

(xi–xii)

Thus the reader wishing to find the actual authors of many chapters must consult the preface, rather than simply look at the chapter heading.

The cost for this minor inconvenience is more than compensated for by the resulting consistency of purpose, approach, and style, and by the clarity with which methodological issues and procedures are discussed. (Douglas Biber has always been admirably explicit about the methodologies he espouses.)

The opening chapter sets the scene for the rest of the volume. The authors note in particular that corpus-based studies have concentrated on the quantitative distribution of lexical and grammatical features, while discourse analysts have concentrated on the discourse patterns discoverable within a relatively small number of texts drawn from a single genre or from a small set of the same. They conclude, '[a]s a result, we know little at present about the general patterns of discourse organization across a large representative sample of texts from a genre' (11). The chapter then usefully describes in some detail typical procedures that can be adopted for top-down and bottom-up approaches to discourse analysis. Two of the former are then discussed further because these approaches are exemplified in later chapters: the first is 'move analysis' (which a few minds might associate with the present reviewer), and the second is 'rhetorical appeals analysis', which is based on the well-known Aristotelian three-way classification of persuasive...