Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.445907
Title: Modelling graphemic buffer disorder : a connectionist approach
Author: Machtynger, Jonathan
Awarding Body: UCL (University College London)
Current Institution: University College London (University of London)
Date of Award: 2007
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
Abstract:
In studies of dysgraphia, graphemic buffer disorder is considered a distinct syndrome since traditional models of spelling place the graphemic output buffer at the junction of both the lexical and sub-lexical processing paths. Damage to the graphemic buffer often results in well recognised symptoms, but there is evidence to suggest that damage presumed to exist prior to the graphemic buffer such as with deep dysgraphia may also produce errors qualitatively similar to graphemic buffer disorder. We build on an existing connectionist model of the graphemic output buffer and examine how damage inspired by physical impairment found in the nervous system may produce characteristics of both disorders. Since each has been observed to produce some common attributes, we investigate in more detail, and expand on, a claim made by others of a new putative functional syndrome. As part of our investigations, we suggest a methodical and rigorous approach to lesioning connectionist systems. We also critique a number of core design principles associated with the original model and augment its functionality to allow a broader theoretical examination of a number of new areas. These include the production of geminate errors, whether orthography affects a word's propensity for error, and how the model may explain a minimum complexity principle associated with repair strategies in the presence of damage. This thesis models disorders, which are known to produce subtly different, yet qualitatively similar behaviours in different patients, and we assume that a model must be able to produce comparable behaviour. In order to provide a rigorous and structured approach to analysing our results, we create a number of quasi-patients and examine the effect of damage across multiple lesion severities, lesion types, and lesion locations.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.445907  DOI: Not available
Share: