ИСТИНА

The paper discusses formal models of structural case assignment and their applicability to Tatar data. We survey two approaches— Chomskyan model of case assignment by functional heads and Marantzian model of configurational case marking — and their application to Sakha, which received much attention in the recent literature. Addressing linguistic evidence from Tatar, we identify a number of contexts where differential case marking is observed. We claim that neither the purely configurational model nor the hybrid model combining Chomskyan and Marantzian modalities of case assignment can account for Tatar data. We argue that the Chomskyan model of case assignment under AGREE can be adjusted in order to successfully account for the full range of Tatar case phenomena. Specifically, we distinguish between different “nominative” forms in Tatar. We suppose that nominative direct objects and nominative possessors are in fact caseless structurally deficient nominals that do not need abstract case in order to be licensed. Nominative subjects, on the other hand, are case-marked and control agreement on the predicative functional head. The nominative form of the postpositional phrase complement is neither a caseless form (as it is available for ezafe-3 nominals) nor a subject case. Since the distribution of genitive and nominative forms governed by prepositions is purely morphological, we propose that they are morphological exponents of an abstract “postpositional” case.