Rowlett, PA 2002, 'On non-overt specifiers' , ESRI Working Papers in Language and Linguistics, 19 .

Preview

PDF Download (206kB) | Preview

Abstract

I consider non-overt specifiers, in particular two contexts in which they have been posited. First, SpecIP: in finite clauses in nullsubject languages, SpecIP is standardly assumed to be occupied by a null pronominal (little pro) (Rizzi 1982a). Second, SpecNegP: in negative clauses in languages whose sole overt negative marker is associated with NegE, SpecNegP is claimed to be occupied by a null polarity operator (OP) (Haegeman 1995). A specifier, like a complement, is a syntactic dependant of a head. I argue that the null hypothesis is that a head does not have a dependant unless it needs one; a head is capable of ‘doing its job’ on its own, and will therefore be dependant-free, unless it is in some relevant sense lacking, whereby the dependant provides what is missing. In this light, I review the evidence for non-overt specifiers in SpecIP/Spec-NegP and show that the evidence does not stand up to close examination, and that the facts can be accounted for by assuming that the relevant heads can ‘do their job’ without a specifier, and that, consequently, their projections not only have no overt specifier, but actually have no specifier position, either, and therefore no nonovert specifier.

Actions (login required)