An Attempt at a Framework for Semantic Interpretation of Natural Language

Abstract

Due to the universality of natural language, no adequate formal framework for a description of its semantico-pragmatic aspects seems to have been found up to now. Natural language contains not only extensional and intensional contexts, but also sentences on attitudes, quotational (and other metalinguistic) contexts and paradoxes, so that it is a very complicated task to describe its semantics in full. To fulfil this task, it seems necessary to find a connection between the empirical semantico-pragmatic findings of linguistics and the results of intensional logic -i. e. between what Fillmore (1977) calls internal and external semantics. According to investigations combining empirical approaches studying the relationships between meaning and sound with methodological requirements brought by Chomsky into the study of syntax and by Montague into that of semantics, it appears to be necessary to work first of all with a level of linguistic meaning; at such a level the ambiguity of the surface sentences is removed, and every sentence is assigned one or more meanings, which in some cases can be shown to be paradoxical, i. e. to correspond to no proposition or to no truth value in any world. As was pointed out by Sgall, Hajicovä and Panevova (in press), this level can be understood as that of the underlying structure of sentences, i. e. in a certain way it combines the properties of Chomskyan "initial Pmarkers" and "logical form", differing from the latter in that it is understood as language specific and reflecting the structure of natural language sentences: no variables, parentheses, prenex quantifiers or similar devices are present at the level of linguistic meaning; instead of such devices, this level contains the articulation of the (meaning of the) sentence into topic and focus, various kinds of pronouns and similar items, from which the anaBrought to you by | University of California Authenticated Download Date | 6/8/15 3:12 AM

Topic and focus in a formal framework

“…There has also been an attempt to account for TFA in a framework similar to DRT; this attempt is due to Peregrin and Sgall (1986). In this framework, each sentence is associated with a situation-like structure (the "content" of the sentence); the "meaning" of a sentence is then understood as the class of all the embeddings of its "content" into the model.…”

“…These ideas have been in a somewhat more cumbersome way outlined by Peregrin and Sgall (1986) and also by Peregrin (1987). It has been shown that they can be useful for the purpose of accounting for TFA in a dynamic framework based on the DRT-like ideas (although DRT did not serve as its explicit foundation).…”