Dissociated Overt and Covert Recognition as an Emergent Property of Lesioned Attractor Networks

Abstract

Covert recognition of faces in prosopagnosia, in which patients who cannot consciously or overtly faces nevertheless manifest recognition when tested in certain indirect ways, has been interpreted as the functioning of an intact visual recognition system deprived of access to other brain systems necessary for consciousness. We propose an alternative hypothesis: That the visual recognition system is damaged but not obliterated in these patients, and that it is an intrinsic property-of damaged-neural networks that they will manifest their residual knowledge in just the kinds of tasks used to measure covert recognition. In support of this, we build a simple recurrent parallel distributed processing model of face recognition and lesion the parts of the model corresponding to visual processing. At levels of damage yielding overt recognition performance comparable to patients described in the literature, the model demonstrates covert recognition in three different tasks: Savings in re- learning correct face-name associations relative to incorrect pairings, semantic priming of occupation decisions on printed names by faces having the same or different occupations, and faster perceptual analysis of previously familiar than unfamiliar faces. Implications for the nature of prosopagnosia, and for other types of dissociations between conscious and unconscious perception, are discussed.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jan 01, 1992
Accession Number
ADA246933

Entities

People

  • Martha J. Fraah
  • Randall C. O'Reilly
  • Shaun P. Vecera

Organizations

  • Carnegie Mellon University

Tags

DTIC Thesaurus Topics

  • Artificial Intelligence
  • Brain
  • Cognitive Neuroscience
  • Cognitive Science
  • Computational Science
  • Computer Vision
  • Identification
  • Information Processing
  • Information Systems
  • Language
  • Neural Networks
  • Neurobehavioral Manifestations
  • Neurosciences
  • New York
  • Psychology
  • Recognition
  • Simulations

Fields of Study

  • Psychology

Readers

  • Brain and Cognitive Science; Experimental Psychology; Cognitive Neuroscience
  • Neural Network Machine Learning.
  • Neuroscience

Technology Areas

  • AI & ML
  • AI & ML - Neural Networks