ABSTRACT A central question of linguistic theory is the relationship between syntax, the formal organization of the meaningful elements of language, and semantics, the system of rules and principles for interpreting the meaning of a syntactically organized utterance. When one makes a serious effort at formalizing these components of a language, a number of paradoxes and quirks arise to frustrate any given systematic choice as to their structure and relationship. Developing and improving models of language which surmount these obstacles to give fuller and more coherent accounts of the full range of linguistic phenomena is the trade of syntactic and semantic theoreticians, and the aim of the present project. One of the fundamental choices to be made in constructing a model of language is to determine whether the semantic interpretation takes as its input structure the surface syntactic structure of an utterance, or whether there is some more abstract underlying syntactic structure which serves as the basis of semantic interpretation. Both possibilities have been explored extensively in the literature, and both have advantages and disadvantages in the naturalness with which they account for linguistic data, so the question remains open. The present project takes as its premise a model which directly interprets surface syntactic structures, and will examine a number of the unresolved problems which such a model confronts. Some recent work by the Principal Investigator and others has suggested new approaches to these problems, which the PI will explore in detail. The problems relate generally to the interaction of variable binding with extraction and with certain anaphoric phenomena such as VP deletion in English. The phenomena to be examined will include the behavior of reflexives and bound pronouns in wh-questions, pseudo-clefts, and the 'tough' construction; the account of sloppy identity in VP deletion and 'paycheck' pronouns; the analysis of antecedent contained deletion; and the interaction of weak crossover with all of these constructions.
