The SPLIT model of visual word recognition: Complementary connectionist and statistical cognitive modelling

Richard Shillcock, Padraic Monaghan, T Mark Ellison

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

We review three problems in the connectionist modelling of visual word recognition: the restriction of models to monosyllabic words, the difficulty in assimilating fixation data from the reading of continuous text, and the abstractness of the accounts of dyslexic reading. We show how a model of visual word recognition, the SPLIT model, can be anatomically based on the precise splitting of the foveal projection about a vertical meridian. The SPLIT model has a limited instantiation as a connectionist model and a wider instantiation as a conventional statistical analysis. This combination of two modelling paradigms, both based on foveal splitting, gives the best coverage of word recognition phenomena.
Original languageEnglish
Title of host publicationConnectionist Models in Cognitive Neuroscience
Subtitle of host publicationThe 5th Neural Computation and Psychology Workshop, Birmingham, 8–10 September 1998
PublisherSpringer
Pages3-12
Number of pages10
ISBN (Electronic)978-1-4471-0813-9
ISBN (Print)978-1-85233-052-1
DOIs
Publication statusPublished - Jan 1999

Publication series

NamePerspectives in Neural Computing
PublisherSpringer London
ISSN (Print)1431-6854

Fingerprint

Dive into the research topics of 'The SPLIT model of visual word recognition: Complementary connectionist and statistical cognitive modelling'. Together they form a unique fingerprint.

Cite this