Font Size: a A A

Agreement at the Boundaries: Synchronic and Diachronic Approaches to ϕ-Agreement in the Left Peripher

Posted on:2018-11-20Degree:Ph.DType:Dissertation
University:Cornell UniversityCandidate:Courtney, Sarah GrayFull Text:PDF
GTID:1446390002986494Subject:Linguistics
Abstract/Summary:
This dissertation examines complementizer agreement (CA) phenomena in which ϕ-features appear on a complementizer, clause-linking marker, or otherwise, syntactically speaking, at the C0 position.;This dissertation will argue that CA is in fact a straightforward output of the syntax module under standard Minimalist assumptions, and that the analysis of CA requires that we simplify rather than complicate our understanding of the probe-goal relationship. CA may be the result of a uϕ-probe at C 0 acting alone, agreeing with a closest goal in situ. More common are cases where CA in relative clauses results from the combination of Agree and movement into specC.;An independent ϕ-probe at C0 is both synchronically necessary, and also to be diachronically expected given the source constructions. I argue that the goals available for probes at C0 are fed into the closest goal position by the lower structure and that argument structure---e.g., the placement and feature checking of subjects and objects---and information structure---e.g., the raising of Topics---may feed arguments and their ϕ-features into the path of C0's probes and yield CA. Cross-linguistic differences in the ability of non-subjects to agree at C follow straightforwardly from differences in the reusability of ϕ-features in different languages (cf. Carstens 2003).;Diachronically, having u? probe at C0 is the natural output of syntactic directionality (as argued for by, e.g., van Gelderen 2009). ϕ-features of source constructions influence the ϕ-features found in their descendants; upward- (Bantu) and downward- (Germanic) agreeing CA are the outputs of different diachronic developments. One has its source in the inherent ϕ-features of a pronoun (goal reanalyzed as probe), while the other is a reanalysis of a verb as a complementizer (T-to-C reanalysis).;I propose that CA---while typologically exotic---is syntactically normal. Accounting for CA with a normal Agree relation solves several theoretical issues for the C-T relationship and provides valuable insight into the nature of probes and the behavior of Agree.
Keywords/Search Tags:&phiv
Related items