A-Probes, Case, and (In)Visibility
Gabriela Alboiu
August 2010

This article argues for a dependency between structural Case and phasal domains and against Case values as intrinsic properties of (C)-T and (v*)-V. Rather, Nominative or Accusative values are derived compositionally from properties of the entire Probing domain: (i) NOM occurs whenever the Probing domain is specified as [uD, u-phi], while (ii) ACC is assigned if the Probing domain is specified as [uD]. In the absence of an A-Probe, a DP’s [uCase] feature automatically deletes upon Transfer but no Case value is assigned, so DP lexicalization fails. The presence of a [uCase] feature is assumed on all DP arguments, whether null or overt. However, after Case valuation, DPs with inherent intensions and extensions will be lexicalized but variables, such as PRO, will not. The analysis focuses on DP subjects (both lexical and PRO) in non-finite CPs, and relies on availability of null expletive 'pro' as a UG primitive. It assumes Chomsky’s Feature Inheritance Model (Chomsky 2007, 2008, Richards 2007), default Case as in Schütze (1997, 2001), as well as Distributed Morphology (Halle and Marantz 1993, Embick 2007). It aligns with views where the Case Filter, while syntactically relevant (Legate 2008), is a PF constraint (Lasnik 2008, Sigurðsson 2008).
Format: [ pdf ]
Reference: lingbuzz/001163
(please use that when you cite this article)
keywords: case, phases, agreement, expletive pro, pro, non-finite cps, lexical subjects, syntax
Downloaded:443 times


[ edit this article | back to article list ]