DescriptionThis dissertation examines the interaction of agreement and case by investigating case- sensitivity of agreement and differential object marking. I develop a theory of agreement and case where surface agreement and surface case are the outputs of a set of distinct operations distributed across different modules of grammar and ordered in a certain way.
The first part of the dissertation investigates case-sensitivity of agreement. I show that there are three types of languages with respect to case-sensitivity: 1) languages where case blocks agreement totally, 2) languages where case blocks agreement partially, and 3) languages where case does not block agreement at all. I propose that case-sensitivity of agreement is a matter of Relativized Minimality (Rizzi 1990). In languages where agreement is blocked by case (partially or totally), the agreement probe is specified for an underspecified [+nominal] feature and it agrees with the case phrase (KP) (Bittner and Hale 1996; Lamontagne and Travis 1986), which intervenes between the probe and any of the person, number, and gender features on the goal. I show that in languages where case blocks agreement fusion of number and case feeds agreement. I develop a theory of two-step agreement (Arregi and Nevins 2012; Bhatt and Walkow 2013; Marušič, Nevins, and Badecker 2015) where the first agreement operation (Agree-Link) establishes an agreement relation between a probe and a goal in the syntax by adding a pointer to the probe whereas the second agreement operation (Agree-Copy) dereferences the pointer by replacing it with the actual values of person, number and gender features. Crucially, Agree-Copy applies after Fusion (Halle and Marantz 1993), a post syntactic operation that takes two adjacent heads and turns them into a single head without any internal structure. In languages where case does not block agreement at all, agreement probes are specified for person, number, and gender features and they are introduced separately. This enables them to skip over the irrelevant syntactic objects and establish full agreement with case marked nominals via Relativized Probing (Nevins 2011; Preminger 2014).
The second part of the dissertation explores the interaction of case and agreement by investigating differential object marking. I show that differential object making is not a uniform phenomenon and in some languages differential object marking is the result of Dependent Case assignment (Baker 2015; Marantz 1984) while in others, differential object marking is a result of nominal licensing via agreement (Barány 2017; Béjar and Rezac 2009; Kalin 2017; Levin 2018). When Agree-Copy fails to dereference a pointer introduced by Agree-Link, Agree Case can dereference the pointer by case-marking the goal.
Overall, the dissertation discusses six distinct operations and orders them as follows: Lexical Case ≺ Dependent Case ≺ Agree-Link ≺ Fusion ≺ Agree-Copy ≺ Agree Case. I argue that the first three of these operations apply in the syntax while the latter three are post-syntactic operations.