Over the two years of the project, the research team will carry out several kinds of empirical and theoretical studies of possessive and partitive constructions in a sample of twelve genetically diverse languages. Possessive and partitive constructions are important but relatively unexamined sites for the consideration of crosslinguistic variation in morphosyntax, semantics, and pragmatics, and interactions among these. In the process of carrying out the descriptive groundwork, the team will construct comparable on-line corpora for each of the twelve languages, supplemented by the results of elicitation work with native speakers. The results of these empirical studies will be subjected to formal analyses conducted within the framework of Optimality Theory (OT). These OT analyses of individual languages will then be subjected to further theoretical consideration, providing a sound basis for a set of crosslinguistic generalizations. The typology of possessive and partitive constructions derived from this project can then provide linguistic typologists with a set of hypotheses, based on sound empirical work and a rigorous theoretical framework, that can be tested against analyses of other languages. Finally, the results of this project will provide substantial material for consideration by linguists concerned with finding parallels between clause-level and nominal-level morphosyntactic, semantic, and pragmatic phenomena.
The project outcomes will thus be of several types. The research team's empirical work will result in (a) a cross-linguistically diverse set of electronic corpora which will be made available to other researchers (pending appropriate permissions) and (b) a set of descriptive statements about the structural, semantic and pragmatic properties of possessive and partitive phrases in a diverse set of languages. The research team's theoretical work, made available in scholarly papers, a book and conference presentations, will result in detailed accounts of OT-type universal constraints and their relative rankings in specific languages. These analyses will advance knowledge both about the linguistic dimensions that are important in crosslinguistic generalizations concerning these structures, and about the typological power and predictions of the OT framework. The analyses will also allow comparisons with crosslinguistic work on the clause along a number of dimensions. The project will build capacity for the training of graduate students in linguistics at Boston University and Brandeis University, and the empirical products-the corpora-will be used as teaching materials by the senior personnel during and after the project.
The sample of languages to be studied includes Cantonese, Czech, English, Finnish, German, Icelandic, Korean, Malayalam, Northern Pomo, Spanish, Swedish, and Wolof. Basic descriptive groundwork will contain results of elicitations with native speakers and examples from corpora. OT theoretical analyses will center on (1) effects of person/animacy markedness hierarchies (Silverstein 1976, Aissen 1999) in possessive constructions of various types within and across languages in our sample; (2) the semantics of determiners and their contribution to the "Partitive Constraint," and (3) the various factors contributing to the distribution of case and related phenomena within these constructions, and their relationship to clause-level case patterns.