The logico-linguistic evidence underlying Montague's language descriptions

Abstract

Richard Montague's analyses of certain fragments of English have deservedly received much attention as the first sophisticated attempts to use model theoretic semantics for natural language description. They are especially interesting in that they exploit for linguistic purposes the apparatus of intensional logic developed in the 1960's by Kripke and others. In view of the influence they have had it is natural to ask: How accurate are these language analyses, regarded as empirical descriptions of actual English? And in particular, how successful is their use of the intensional apparatus? Montague presents detailed syntactic and semantic rules for three fragments of English. The first, which I shall call Fragment I, is described in his (1970a); the second, or Fragment II, in his (1970b); and the third, Fragment III, in his (1973). I wish to examine here the descriptive accuracy of his semantical rules for these fragments, those for Fragment I in some detail and those for Fragments II and III more briefly. Some prior acquaintance with the three relevant papers will be assumed; they are conveniently available in Thomason (1974), to which my pagination will refer. I should warn the reader that I plan to make the examination of Montague's work a vehicle for illustrating my own ideas about how evidence may be brought to bear in testing linguistic hypotheses.