Salish Applicatives by Kaoru Kiyosawa and Donna B. Gerdts (review)

First, a little history. In 1998, the late M. Dale Kinkade gave his final seminar on comparative and historical Salish Linguistics at the University of British Columbia. It was a bittersweet occasion for him: though he was happy to escape the institutional side of academic life, he already knew by then that he would not live to complete the comparative morphology of Salish that he had been compiling on and off for the previous twenty-five years, and he was anxious that the next generation of Salishanists would take up where he had left off.

Kaoru Kiyosawa attended that seminar as a graduate student, and was inspired by Kinkade’s work to write her Ph.D. dissertation on the comparative morphology of applicatives, under the supervision of Donna Gerdts at Simon Fraser University. Six years after Kinkade’s death, Kiyosawa and Gerdts joined forces to complete one part of his legacy, with this comprehensive overview of the grammar of of applicatives in Salish. The book is fittingly dedicated to Kinkade’s memory.

Applicatives are a prominent and comparatively well-described part of the morphosyntax of every Salish language: they are easy to find in textual materials, and most descriptive grammars devote at least some space to discussing them. Furthermore, they are far more highly developed in Salish than anywhere else in the Pacific Northwest; and in a broader typological perspective, it is unusual to find a large and relatively homogeneous language family with such a variety of applicative morphemes, making for a potentially fascinating study in "micro-parametric" variation. (The obvious comparison is to Bantu, which has generated a substantial proportion of all the syntactic literature on applicative syntax.)

There are two main empirical foundations to the current study: a one-thousand-sentence database compiled by Kiyosawa from existing sources, including data from twenty of the twenty-three Salish languages (the extinct and poorly documented Pentlatch, Quinault, and Lower Chehalis were excluded for lack of examples); and fieldwork carried out by Gerdts on Island Halkomelem. The primary thrust of the work is comparative-typological, though attention is paid to the diachronic dimension, building on Kinkade’s reconstruction of Proto-Salish applicative morphology. Following introductory material on Salish and a useful overview of the morphosyntax of applicatives across the family, chapters are devoted to each of the two main classes of applicative, which the authors term "relational" and "redirective," more or less in line with established usage amongst Salishanists. For each class, the authors explore two issues: the nature of the verbal "base" to which the applicative suffixes attach, and the thematic role of the applied object. Further chapters treat "exceptional" uses of applicatives, the relation between applicatives and transitivity, combinations of applicatives with each other and with other (de-)transitivizing morphemes, the discourse functions of applicatives (particularly in relation to animacy and topicality), and the place of the Salish system in a general typology of applicative morphosyntax.

The main strength of the book lies in its broad empirical coverage, which provides a clear and comprehensive picture of major trends in applicative morphosyntax across the family, as well as highlighting some interesting outliers (with Bella Coola representing, unsurprisingly, the least typical pattern). Overall, the level of scholarship is high, the writing lucid, and the exposition clear. I was also impressed by the organization and layout of the book; the maps in particular are amongst the best I have come across in any volume on Salish. I would not hesitate to recommend the introductory part of the book as an excellent short guide to Salish morphosyntax for those unfamiliar with the family.

Perhaps inevitably, however, breadth of coverage comes somewhat at the expense of depth. Part of the problem here is methodological: the database which provides the principal source of evidence for all the languages, save Halkomelem, is subject to the usual limitations of corpus-based work, particularly when such a small database covers...