Case assignment on adverbial NPs in Korean

This dissertation investigates the system of Case assignment on adverbial NPs in Korean. First, I examine the issue of what types of adverbials can receive morphological Case in Korean. I question the existing belief that only D/F adverbials can surface with Case-marking and that the adverbials may receive direct Case as long as they are SITUATION DELIMITERS (W&L 1996) indicating the non-delimiting function of Case on non-arguments. I suggest that the addition of the Case particle on adverbial NPs results in different information structures and that it is related to focus interpretation. I propose a Misplaced Marker Hypothesis (MMH) that the Case particle on adverbial NPs could function as an 'operator' that is associated with focus and has the grammar of scope. Second, I investigate the conditions that determine NOM vs. ACC on adverbial NPs. There are two puzzles to solve to maintain the claim that Case on adverbial NPs is structural. First, both NOM and ACC could alternatively be assigned to adverbial NPs irrespective of the predicates. Furthermore, in some limited cases, there is a mismatch between the verbal argument and the adverbial NP. ACC on the adverbial is unexpected when the predicate is considered unaccusative, lacking the ACC-assigning property. I suggest that the answer to the two puzzles is from the fact that Case is not the monopoly of the narrow syntax, but a joint asset of other grammatical representations of the entire sentence. I claim that Case is not just the realization of a narrow syntax, but the realization of other parts of grammar such as semantics and pragmatics. Neo-Davidsonian Event semantics let the adverbials assigned Case not to violate the Case Filter, and Nordlinger's (1998) constructive function of Case makes case morphology construct information about higher phrases in which it is embedded. Because of the traits of non-configurationality, Case morphology in Korean could carry much of the functional load of phrasal syntax determining constituency relations and grammatical functions. I propose that Case Realization in Korean is a complex process that is composed of both syntactic distribution (syntactic case-checking in overt syntax) and semantic licensing (semantic case-checking at LF). As for the syntactic case-checking, structural ACC is assigned to adverbial NPs in a VP-adjoined base-generated position, and structural NOM is assigned to adverbial NPs in a syntactically moved position. As for the semantic licensing of Case on adverbial NPs in Korean due to the traits in terms of non-configurationality, I suggest that Tense, Aspect and semantic property such as theticity could operate as licensing examiners for the full realization of Case. The past tense in T, the telicity in Asp and the categorical judgment represent the 'eventiveness' that is responsible for the unexpected accusative Case feature on the adverbial. I claim that the [+eventiveness] in null head E becomes active when the sentence has the features such as [+past], [+ telicity] and [+ categoricality]. I suggest the 'Feature Percolation' type analysis under the assumptions of the minimalist framework account for 'agreement' between the uninterpretable feature [acc] and the interpretable feature [+F]. The uninterpretable accusative Case on adverbials agrees with the interpretable accusative Case on EP by percolation of the abstract Case assigned to the sentential maximal projection EP. I also suggest that pragmatic encoding is important in that it could decide the semantic judgment type so as to choose the best option for the optimal interpretation.