Search

. 'GSGM' builds no syntactic structures; rather, semantics and morphosyntax are both generative and run in parallel, and so have no need of mediating syntactic structure. In this environment, positing different modes of embedding for NPs and clauses, as in Williams 2003, affords new minimal accounts of movement, scope, agreement, and the bounds on them. Syntax/morphosyntax is instantiated as a single parameterized function ('Combine'). Combine derives both mirroring and nonmirroring structure, as any theory must, but in distinctive 'size-relative' patterns determined by its parameters. [ABSTRACT FROM AUTHOR]

Historically, control in generative grammar has fallen within the province of syntactic theory. One primary reason for this is that Mainstream Generative Grammar (MGG) has imposed a strong uniformity criterion on analyses as a measure of their explanatory adequacy. One aspect of this uniformity criterion, which we call Interface Uniformity, holds that the syntax-semantics interface is maximally simple, in that meaning maps transparently into syntactic structure, and that it is maximally uniform, so that the same meaning always maps into the same syntactic structure. It follows from Interface Uniformity that a nonfinite VP has a syntactic subject that is assigned its external θ-role. We argue that this view is most sustainable if one does not take into account the full richness and complexity of control phenomena, but treats control strictly in terms of complementation. When a fuller range of phenomena is taken into account, it appears that it is preferable to "turn over control" to the semantics, which is better equipped to capture the facts. We outline how to formulate the syntax-semantics interface so as to get the semantic facts to line up properly with the syntactic facts. This analysis of the interface extends naturally to raising. [ABSTRACT FROM AUTHOR]