Weak Crossover, specificity and LF chains
In this talk I argue that Weak Crossover does not arise with overtly-moved specific wh-elements, while it is systematically present only with quantifiers and operators moved in covert syntax. In order to establish the relevance of specificity in Weak Crossover contexts systematically, a notional and an operational definition of specificity are provided. In particular, a range of constructions affected by the specificity of the extracted phrase are used as tests to support the claim. I develop an analysis that accounts for the refined empirical picture, building on Rizzi's (2001) proposal concerning the form of specific vs. non–specific chains at LF and a condition of non–distinctness of grammatical features for chain links.