Do We Need Abstract Case

Abstract

Minimalist approaches often implicitly assume a version of the Case Filter (Chomsky 1981, Vergnaud 1977), whereby, even in languages without morphological case, overt DPs must be Caselicensed. ‘Case’ thus potentially accounts for the distribution of (referential) DPs and motivates phenomena such as A-movement (passivization, raising and, for some, Control). Recent proposals, however, have argued for the parameterisation (Harford Perez 1985, Markman 2009, Diercks 2012) or rejection of abstract Case (Marantz 1991, McFadden 2004, Landau 2006). The time seems ripe, then, for a reconsideration of: (i) the diagnostics for abstract Case, (ii) their cross-linguistic validity (especially in languages without morphological case) and (iii) how they can be accounted for in an explanatory way. In the following, we attempt to address (i)-(iii). Section 2 introduces six Case diagnostics and applies them to six languages with little/no morphological case: Mandarin, Thai, Yoruba, Jamaican Creole (JC), Luganda and Makhuwa. Section 3 provides an account of why Luganda (like some other Bantu languages) appears to lack ‘the abstract Case property’ (in a descriptive sense). Building on this, section 4 then provides a novel account of nominal licensing based on the idea that nominals must be part of a complete phase. Finally, section 5 concludes.