An Evaluation of Alternative Functional Models of Narrative Schemata,

Abstract

Recent research on human memory for text has focused on modeling both the underlying semantic structure of text and the memory structures that encode and interpret this information. This paper evaluates several alternative models for how narratives are encoded, represented, and retrieved from memory. In particular, it addresses four questions regarding the use of narrative schemata in memory: (1) Do narrative schemata bias the likelihood of encoding text information, i.e. does the levels effect in recall reflect differences in the encoding of text propositions? (2) Are the representations for stories hierarchical (as suggested by story grammars) or heterarchical? (3) Does retrieval and recall of propositions from memory depend on a top-down search of the hierarchical memory structure, or can propositions be directly accessed? and (4) Does the memory representation of a text retain the surface information of the text, or is this representation conceptual? These questions define four attributes (encoding bias, memory structure, search process, memory contents), each with two or more values. Different values for these attributes may be combined to form a variety of alternative models for text memory and recall. This paper delineate a number of plausible candidate models within the general 'memory schema' framework and then presents an experiment to comparatively evaluate these models.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jul 01, 1980
Accession Number
ADA100023

Entities

People

  • Frank R. Yekovich
  • Perry W. Thorndyke

Organizations

  • RAND Corporation

Tags

DTIC Thesaurus Topics

  • Acquisition
  • Coding
  • Cognition
  • Comprehension
  • Content Addressable Memory
  • Contrast
  • Data Analysis
  • Educational Psychology
  • Governments
  • Hierarchies
  • Judgment
  • Materials
  • New York
  • Probability
  • Psychology
  • Recognition
  • Terrorists

Fields of Study

  • Psychology

Readers

  • Artificial Intelligence