Gender-animacy and the morpho -syntax of object clitics in Dutch.

In this dissertation I describe and analyze a variation in the morpho-syntax of phonologically reduced, or DEFICIENT, object pronouns in Dutch. I show that in non-Southern varieties of Dutch (NSV Dutch), the syntactic distribution of deficient object pronouns is dependent on the animacy of the referent of these pronouns. Those with inanimate referents can precede exceptionally Case-marked clause-mate subjects, whereas those with animate referents cannot. Furthermore, deficient object pronouns with animate referents can occur as complements of prepositions, whereas those with inanimate referents cannot. In other varieties of Dutch, deficient object pronouns form a homogeneous class such that their syntactic distribution is not affected by referent animacy (see Zwart (1997)). In this work I propose that the concept 'animacy' is represented in language in terms of the features [animate] and [female] of nominal expressions. I propose that these features are contrastive in grammars where referent animacy affects the syntax of nominal expressions, and non-contrastive in grammars where it does not. In addition, I propose a feature-geometric analysis to account for the relationship between animacy and word gender in languages where animacy features are contrastive. Specifically, I have argued that the animacy features [animate] and [female] are hierarchically organized with respect to the gender features [non-neut] and [fem] such that (i) [animate] entails [non-neut], and (ii) [female] entails both [fem] and [animate] (and thus, transitively, [non-neut]). This feature-geometric analysis constrains possible gender-animacy specifications of nominal expressions in languages, accounting, for example, for the fact that in NSV Dutch pronouns with animate referents cannot be neuter. And finally, I analyze the way in which specification for [animate] affects the syntactic distribution of nominal expressions. Adopting Cardinaletti and Starke's (1999) analysis of pronoun deficiency, I argue that in NSV Dutch, pronouns unspecified for [animate] have a severely deficient formal feature structure, which causes them to move to a higher syntactic position (specifically, to the functional head Sigma, located above the ECM-subject) than do pronouns specified for [animate]. I argue that pronouns specified for [animate] have a mildly deficient formal feature structure, causing them to stay below the ECM-subject in a functional projection AnimateP, where they check animacy features.PhDLanguage, Literature and LinguisticsLinguisticsUniversity of Michigan, Horace H. Rackham School of Graduate Studieshttp://deepblue.lib.umich.edu/bitstream/2027.42/123723/2/3096219.pd