Thesis Proposal: A Plan-Based Approach to Conversational Implicature.

Abstract

Implicatures are those propositions that must be inferred from an utterance in order to make sense of it in context, but which are not truth conditions or entailments of that utterance. Conversational implicatures are those that can be explained by general principles rather than lexical choice. Conversational implicature was defined in philosophy and discussed in linguistics, but lacks an adequate account in any discipline. For the computational linguist, this means that much information conveyed indirectly in discourse has no computational model. Knowledge of goals and plans is necessary for the computation of some implicatures, and very useful for a much larger class of implicatures. This model uses a set of inference rules about STRIPS-style plans for implicature computation. It incorporates a computational model of speech acts (based on propositional attitudes) and surface speech acts (based on linguistic features). An implementation is proposed and related to other research in computational linguistics.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jun 30, 1987
Accession Number
ADA189201

Entities

People

  • Elizabeth Hinkelman

Organizations

  • University of Rochester

Tags

Communities of Interest

  • Biomedical

DTIC Thesaurus Topics

  • Artificial Intelligence
  • Classification
  • Cognitive Science
  • Computational Linguistics
  • Computational Science
  • Computations
  • Computer Science
  • Health Services
  • Language
  • Linguistics
  • Lisp Programming Language
  • Literature
  • Natural Languages
  • New York
  • Psychology
  • Reasoning
  • Recognition

Readers

  • Computational Linguistics

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - Machine Translation