NOTES ON SEMANTIC DISCOURSE STRUCTURE

Abstract

The semantic problems of natural language discourse analysis are more serious than the syntactic ones, and much less work has been done on them. It is widely held that a semantic classification of words by their meanings is required for selecting the correct senses of individual words in text, but the implications of this hypothesis have not been fully worked out. It is moreover evident that semantic discourse analysis does not end with disambiguation: understanding a piece of discourse depends on identifying its message, and it is arguable that for this we have to know what the semantic structure of the text as a whole is, where this may not be very directly related to the syntactic structure of its component sentences. It is suggested that a semantic or conceptual classification plays a part here, since understanding the message of a text involves some knowledge of the way concepts may be or are usually combined. In this paper, different aspects of this notion of semantic discourse structure are explored, and their connection with a semantic classification is exhibited.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Mar 03, 1967
Accession Number
AD0648755

Entities

People

  • Karen S. Jones

Organizations

  • System Development Corporation

Tags

Communities of Interest

  • Air Platforms
  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Abstracts
  • Ambiguity
  • Boundaries
  • Construction
  • Corporations
  • Dictionaries
  • Economic Development
  • Education
  • Food
  • Governments
  • Identification
  • Language
  • Natural Languages
  • Robots
  • Security
  • Standards
  • Vocabulary

Fields of Study

  • Linguistics

Readers

  • Computational Linguistics
  • Theoretical Analysis.