Rehabilitation for pure alexia: Efficacy of therapy and implications for models of normal word recognition

Abstract

In an attempt to remediate the deficit in letter processing thought to underlie pure alexia or letter-by-letter reading, a therapy programme was implemented with SI, a 46 year-old woman with pure alexia following a posterior cerebral artery infarction. Prior to the intervention programme, SI's reading and letter processing abilities were characterised in detail, and baseline measurements were taken to rule out spontaneous recovery or fluctuations in her performance. A nine-week, intensive therapy programme was designed to alter SI's report of single letters as a function of serial position in a string from the observed linear function to a more normal M-shaped curve. This involved training SI to apprehend and report the identity of the letter in the final position of a string as string length was increased and exposure duration was decreased. Whereas SI's ability to recognise the final letter improved significantly on post-therapy measures, there was no significant transfer to her word reading performance. The word length effect—increase in reaction time with increasing word length—remained unchanged from pre-to post-therapy, although the intercept shifted downwards. These results raised questions about the relationship between processes involved in single letter report and word reading in normal processing. Findings from rehabilitation studies such as this are not only useful in determining the efficacy of particular intervention techniques but, importantly, shed light on the mechanisms underlying normal cognitive behaviour.