Abstracts Language, Literature, and Linguistic

Conditions on logical form derivations and representations

by Hidekazu. Tanaka

How is the logical form of a sentence expressed in natural language? This thesis examines in detail wh-questions and negative polarity items in Japanese and English in an effort to pin down a number of issues related to this question. Chapter one introduces some of the basic notions of current syntactic theory within which the discussion in this thesis takes place. The chapter also contains basic syntactic properties of wh-questions and negative polarity items in English and Japanese. Chapter two advances a cooccurence restriction condition on wh-questions and negative polarity items. The condition to be introduced is referred to as the Linear Crossing Constraint (LCC). Assuming with Saito (1992) that scrambling can be undone at the level of logical form, it is argued that the LCC applies to the surface form of a sentence. Various consequences that follow from the LCC are also discussed. Chapter three argues that wh-phrases and negative polarity items undergo movement in the logical form component of grammar. The discussion in this chapter is dependent on the scope facts involving these grammatical constructions. Chapter four is concerned with the Subjacency Condition. Nishigauchi (1992) proposes that movement in the logical form component is constrained by the Subjacency Condition in much the same way as movement is in the overt component. It is shown that the relevant sentences pointed out by Nishigauchi should be accounted for by a condition on logical form representations. Chapter five deals with why certain instances of scrambling can be undone in the logical form component but others cannot, as observed by Takahashi (1993). The aim of Chapter six is to develop an account of the distribution of adjunct wh-phrases, such as why and naze. It is pointed out that naze shares a number of characteristics in common with negative polarity items and floating quantifiers. I argue that there is only one specifier position per functional head. Chapter Seven extends the theory developed in Chapter six to another set of data. It is argued that the distribution of floating quantifiers can naturally be captured under the proposed theory. The final chapter concludes this thesis by pointing out some consequences of this theory.