Summary :
A philosophical theory of language can aim for completeness in either of two senses. It may strive for "horizontal completeness", which would require it to give an account of all parts of language— singular terms, predicates, verbs, adverbs, adjectives, nondeclarative sentences, speech acts, etc. Alternatively, it may strive for "vertical completeness", analysing some linguistic notions in terms of others, but ultimately providing an analysis of language which does not take any semantical or linguistic notions as primitive. Vertical completeness requires that the theory not presuppose notions of meaning, reference, semantical rules, etc., but instead provides analyses for these notions in terms of nonlinguistic and nonsemantical notions. The theory of this book aims at vertical completeness. The objective is to start with nonlinguistic notions and build up a complete theory of language out of them. To assess the extent to which this goal has been achieved, the reader is referred to the final section of the appendix where the path of the analysis is reviewed. The book also attempts to achieve a certain amount of horizontal completeness, but that is more of a task for linguistics than philosphy, and no claim is made to real horizontal completeness in this book. I first began thinking about this material in 1965 in response to the attack on analyticity by Quine and Putnam. The orientation of the work has changed over the years, and it now touches on analyticity only peripherally. Nevertheless, in my mind it lays to rest the ghost of those early concerns. I began serious work on the book in 1975, and a first draft was completed in 1977. The book has subsequently gone through two complete rewritings and several somewhat less extensive revisions, but the philosophical content has remained roughly constant. The changes have been mainly concerned with style of presentation. This book reflects my feeling that is a lamentable lack of precision in most contemporary work in the philosophy of language. This is true even of those works that give the appearance of precision by making heavy use of logic and formal semantics. Such works have tended to ignore the question of how formal semantics is supposed to bear upon more fundamental philosophical questions. I have tried to achieve a much greater degree of precision in the current work. But there is a problem inherent in such a goal. Precision and readability are to some extent incompatible with one another. Readers of early versions of this book were unanimous in their opinion that the book was too hard to read. Thus the bulk of the revisions which the book has undergone have been concerned with striking a reasonable balance between precision and readability. No doubt many will complain that the book is still too hard to read, but I feel that it would be a mistake to sacrifice further precision in the interest of readability.

Copies :