Fragments and left-edge ellipsis: The division of labour between syntax, semantics, and prosody

Abstract

This chapter provides a unified treatment of fragment ellipsis and left-edge ellipsis in English. Both processes are argued to be cases of ellipsis at the syntax–prosody interface, very late in the derivation. A formalization is given in terms of an optimality-theoretic approach to the syntax-prosody interface. Left-edge ellipsis is motivated by a constraint favoring Strong Starts to prosodic constituents, while fragment ellipsis occurs because of the conflicting requirements of different constraints on linearization of focused constituents. In both cases, the optionality of ellipsis is captured by variable re-ranking of the (ellipsis-disfavouring) Max constraint with respect to the markedness constraints which favour ellipsis. On this view, although syntactic processes (in particular LF-movement of foci) feed fragment ellipsis, there is no syntactic licensing of fragment ellipsis as such. The chapter also discusses the role of Recoverability in constraining the interpretation of these elided structures and ensuring that ellipses are not over-generated.