Ellipsis in Discourse : Implications for Linguistic Analysis By Computer, the Child's Acquisition of Language, and Semantic Theory

Abstract

This paper explores the basis for semantic interpretation of naturally occurring English discourse. Findings concerning linguistic analysis of discourse by computer and the child's language imply the necessity of distinguishing between language and the use of language. This is necessary in order to separate competence, the ability to produce novel sentences which conform to the syntactical rules of the language, from abilities with respect to language in its ordinary uses which cannot be subsumed under the acquisition of syntax. Aquisition of the language in its ordinary uses is found to be a necessary condition for the semantic interpretation of discourse. What characterizes natural discourse is the prevalence of various kinds of ellipsis. A categorization of these is made and bases of their interpretation proposed. Conditions regarding syntactic structure which are necessary and sufficient for interpretation of discourse are examined. It is proposed that the lexicon for the deep structure of a transformational grammar adequate to the interpretation of discourse is to be regarded as an index and "book" rather than a dictionary.