The Nominative & Accusative and Their Counterparts (review)

Whereas a considerable number of typological studies have been devoted to the problem of ‘ergativity’, its structural counterpart, namely ‘accusativity’, has not yet experienced an analogous treatment. Hence, any book that aims at presenting studies in ‘accusativity issues’ from a typological perspective should be welcomed. The present volume importantly contributes to this perspective. Nevertheless, the reader should note that it does not by itself offer an in-depth study of accusativity as such. Rather, it starts from a semasiological determination of the two case forms that are thought to play the most crucial role in the constructional patterns of accusativity: nominative and accusative. Hence, the conception of the volume is based on the assumption that nominative and accusative case forms ‘both constitute fundamental arguments of the verb’ (1). All ten articles in the volume aim at elaborating the functional scope of these two case forms, although the reader should not expect that all articles lay the same weight on them. In addition, the book’s ‘form-to-function’ approach guarantees that the reader obtains only a partial, albeit highly illuminating, picture of the universe of accusativity.

A well-written ‘Introduction’ by the editors (1–14) guides the reader through the book, summarizing both the fundamental claims of its authors and the individual articles. The first of the ten papers (by Michael Herslund) deals with ‘romance transitivity’ (15–39). Starting with data that mainly stem from Old French, Old Provençal, Spanish, Italian, and Rumanian, the author elaborates a highly important scale of ‘object features’ in the Romance languages, introducing the notion of ‘object zone’ to describe formal and functional variation within the functional domain of ‘object’ (incorporation < neutral < supertransitive). Ludo Melis continues the discussion of object features in his article ‘Objects and quasi-objects: The constellation of the object in French’ (41–79), suggesting four layers of the ‘object domain’ in French (lexicogrammatical, categorial, syntactic, and semantic). Nicole Delbecque’s contribution, ‘A construction grammar approach to transitivity in Spanish’ (81–130), also concentrates on the object zone, proposing an extremely important interpretation of the a/Ø-opposition of Spanish ‘objects’ from the viewpoint of constructional grammar which ultimately relates to the generalized force-dynamics approach in cognitive grammar.

The following two articles turn to Germanic. Kristin Davidse discusses the ‘Nominative and oblique in English’ (131–73), and Luk Draye elaborates on aspects of functional case merger and word order features in German (‘Aspects of nominative and accusative in German’, 175–200). In her article ‘The source-path-goal scheme and the accusative in interaction with the genitive in Polish’ (201–25), Zofia Kaleta turns away from the otherwise semasiological perspective taken in the volume. Positing the well-known source-path-goal schema, she examines the extent to which the two case forms genitive and accusative are involved in the formulation of the path-goal subschema.

The remaining four articles extend the discussion to non-Indo-European languages. Eugene Casad elaborates on the question of ‘Objects, verbs and categories in the Cora language’ (227–64). Casad’s interpretation of the Cora object zone is an extremely valuable contribution to the problem of how properties of grammatical relations can be made accessible to typological description. In his article ‘Ergativity and accusativity in Basque’ (265–84), Larry Trask addresses the well-known question of how to determine the ‘subject’ in Basque. Making a crucial distinction between morphological and syntactical ‘transitivity’, the author arrives at the conclusion that Basque basically has ‘accusative’ syntactic properties embedded into an ‘ergative’ morphology. Bill McGregor’s contribution, ‘Ergative and accusative patterning in Warrwa’ (285–317), challenges at least in part the standard hypothesis about the primitiveness of the S, A, and O functions. Accordingly, Warrwa, a typical ‘split language’, is organized along two ‘experiential’ parameters that oppose...