In lieu of an abstract, here is a brief excerpt of the content:

854LANGUAGE, VOLUME 73, NUMBER 4 ( 1997) The dative: Vol. 1, Descriptive studies. Ed. by William Van Belle and Willy Van Langendonck. (Case and grammatical relations across languages, 2.) Amsterdam: John Benjamins, 1996. Pp. xvii, 455. Cloth Hfl 225.00. Reviewed by Martin Haspelmath, Free University of Berlin/University of Bamberg This book consists ofeleven papers on the dative (or the indirect object) in individual languages, preceded by a short introduction by the editors. The areal and typological diversity of the languages is not overwhelming: four Romance languages (Latin, by Willy Van Hoecke, 3-37; French, by Ludo Melis, 39-72; Spanish, by NicoleDelbecque and Béatrice Lamiroy, 73-1 17; Portuguese, by Rosane de Andrade Berlinck, 119-51), four Germanic languages (German, by Luk Draye, 155-215; Dutch, by the editors of the book, 217-50; Afrikaans, by Leon G. de Stadler, 25 1 -88; English, by Kristin Davidse, 289-338), two further Indo-European languages (Polish, by Brygida Rudzka-Ostyn, 341-94; Pashto, by Wojciech Skalmowski, 395-406), and Orizaba Náhuatl, by David Tuggy, (407-52). If the purpose of the book were to facilitate cross-linguistic comparison of the dative, one would expect the editors to provide a definition of this category that is potentially applicable to any language. Clearly, a morphological definition can only identify specimens of the dative case within a single language, so it is of little help. A purely semantic definition (e.g. in terms of semantic roles such as 'recipient' or 'goal') is more readily suited for typological comparison, but for this purpose the traditional case-label 'dative' seems slightly odd. Furthermore, the expressions singled out by such a definition would be very heterogeneous formally, and no languageparticular category would correspond exactly to such an abstract definition. The solution seems to be to use the term 'dative' for the class of relational markers that have the expression of the recipient role as one of their core uses, i.e. a definition at the level of the gram-type (see Bybee et al. 1994:3 for this notion). However, the strengths of the volume are not m a sophisticated methodology for typological comparison but in detailed descriptions of individual languages. In the introduction and in the eleven chapters, the problem of cross-linguistic comparability is simply ignored, and toward the end of the introduction, the editors in passing give an honest definition of 'dativity' as understood in the book: 'the meanings associated with the dative case in, for instance, Indo-European'. Thus, the indirect-object constructions in English, Dutch, and Afrikaans and the applicative verbs of Náhuatl also find a place in this collection. Most of the papers were written by linguists working at the University of Leuven, and their theoretical outlook is a mixture of European functionalist structuralism and Langackerian cognitive linguistics (the latter especially in the papers by de Stadler, Rudzka-Ostyn, and Tuggy). From a sociological point of view, this is quite interesting: European structuralism survived intact in Belgium longer than elsewhere, perhaps because of the dominance of the French language . Until not too long ago, a volume like this would have been published in French or Dutch. Now Belgian linguists have entered the international arena, and they seem to find the postChomskian cognitive linguistics more congenial to their tradition, thus skipping the stage of Chomskian generative grammar. While the individual papers discuss morphological, syntactic, and semantic issues, it is at the level of semantics that we expect the papers and the facts presented in them to be the most readily comparable. In the structuralist tradition, some of the papers try to arrive at an overall characterization of dative meaning: for Van Hoecke (on Latin), 'the dative serves as the limit of the predicate in the sense that it indicates the ultimate term to which the action or process referred to tends' (3 1 ), and Draye (on German) suggests that 'the dative function can be defined in terms ofa reference point function' (209). Rudzka-Ostyn, discussing the Polish dative, develops an elaborate hierarchy of increasingly schematic values and ends up with a description in terms of the notions of endpoint, recipient, and landmark. It can easily be seen that...

pdf

Share