ABSTRACT
In this article I reanalyze certain superiority phenomena and investigate consequences of the account I propose. In section 1 I briefly review two standard analyses of superiority effects and a problematic type of case (from Lasnik and Saito (1992)) that neither analysis handles correctly. In section 2 I present the Operator Disjointness Condition (ODC), which Lasnik and Saito (L&S) (1992) propose as the central principle accounting for superiority effects (including the problematic cases they present). In section 3 I argue that the ODC can be reformulated as a more natural principle of scope marking: the Scope-Marking Condition (SMC). I then show that the success of this reformulation rests crucially on incorporating the theory of linked chains independently motivated by Chomsky and Lasnik (1993). In section 41 present independent support for the SMC. In section 5 I tentatively propose a new theory of adjunction. In section 61 investigate how subject position in English becomes properly governed by I-raising at LF (a process crucial to L&S’s analysis of superiority). In particular, I seek an explanation for the fact that LF I-to-C movement, but not syntactic I-to-C movement, renders subject position properly governed. I argue that a principled account can be provided within the framework of checking theory (Chomsky ( 1991 ; 1993)). I propose, among other things, that (a) English is a “covert verb-second” grammar, (b) LF representations derived by V-raising are VP-recursion structures resulting from checking and deletion of functional heads and their projections, and (c) index-sensitive head government conditions may be at least partially eliminable and replaced by simpler and arguably more natural requirements on traces (chains), given the derived (VP-recur-sion) constituent structure I propose for checking-induced deletion. In section 71 reformulate certain aspects of the proposed analysis within the bare theory of phrase structure presented in Chomsky (1994).
