Shape Conservation and Remnant Movement

Standard (``primary'') remnant movement constructions as introduced by Thiersch (1985) and den Besten & Webelhuth (1987; 1990) differ significantly from other (``secondary'') remnant movement constructions as they have recently been proposed by Kayne (1998), Johnson (1998), and others. This paper shows that a unified optimality-theoretic approach is possible if we assume that whereas primary remnant movement is feature-driven, secondary remnant movement is a repair phenomenon -- movement proceeds in violation of Last Resort in order to respect a higher-ranked Shape Conservation constraint (Williams (1999)). On the conceptual side, a new model of optimization in syntax is proposed: Optimization proceeds locally, not globally, and it affects extremely small parts of syntactic derivations. This move is shown to be both conceptually attractive and empirically motivated.This is authors' final version of the paper. The definitive version was published in NELS 30: Proceedings of the North East Linguistic Society 30 (2000) and is available at http://glsa.hypermart.net/cgi-bin/list.cgi?NELS%NELS30%5%NMüller, M. (2000). Shape Conservation and Remnant Movement. In M. Hirotani, North East Linguistic Society & Rutgers University (Eds.), Proceedings of the North East Linguistic Society 30. Amherst: GLSA.The research reported here was supported by DFG grants MU 1444/1-1,2-1