This dissertation is a generative-grammatical study of the constructions called sluicing and stripping in Japanese. It argues for the thesis, advocated by Hoji & Li (1994) and Fukaya & Hoji (1999), that sluicing and stripping are manifestations of the same syntactic phenomenon in Japanese, i.e., that their derivations involve the same set of formal operations. In order to establish the thesis, this dissertation investigates significantly more involved empirical materials pertaining to island sensitivity and the availability of the sloppy identity reading than previous studies of these constructions.;Chapters 2 and 3 examine the properties of sluicing and stripping in Japanese with respect to island sensitivity. Chapters 4 and 5 investigate their properties with respect to the availability of the sloppy identity reading. It is demonstrated in these four chapters that Japanese sluicing and stripping exhibit the same set of properties in regard to these two aspects. It is then claimed that ellipsis resolution in the case-marked versions of sluicing and stripping involves Constituent Raising in the "antecedent" IP and the copying of the resulting IP onto the ellipsis site. It is also claimed that ellipsis resolution in their non-case-marked counterparts, by contrast, does not necessarily involve such operations and that the copula structure is also a possibility for them. Chapter 6 then argues that the alternative accounts that analyze sluicing as analogous to wh-questions in English cannot capture the clear parallelism between sluicing and stripping demonstrated in the preceding chapters and also that stripping and, as a result, sluicing cannot be reduced to the cleft construction, contrary to what has sometimes been proposed in the literature. This chapter also discusses some implications of the present analysis for English sluicing and fragments. Chapter 7 summarizes the thesis and addresses some outstanding issues. |