Deletion Through Movement

Abstract

What appears to be deletion is, in some cases, the result of syntactic movement out of a phonologically and semantically interpreted domain. Support for this conclusion comes from a phenomenon of question truncation that I call aux-drop. I show that aux-drop questions differ minimally from full questions in that their auxiliary, though present early in the derivation, moves to a position in which it is interpreted neither phonologically nor semantically. I also argue that this deletion through movement is not subject to a recoverability condition. The analysis finds a natural place in a theory where head movement is syntactic (not purely phonological) and spell-out is cyclic. This approach explains the emergence of the factative effect – a type of tense interpretation pattern found in “bare sentences” in, e.g., Haitian Creole and Fòngbè – in an obscure corner of certain languages that normally require full tense specification in finite clauses. My approach finds further support in the patterning of aux-drop questions with VP topicalization and pseudoclefts with respect to morphological mismatch, a pattern of verbal perfect interpretation in the absence of participial morphology.