Search

A number of investigators have demonstrated that patients with pure alexia comprehend briefly presented words which they are unable to explicitly identify. We suggested previously that these patients may read by means of two distinct procedures: a laborious letter-by-letter method and a 'whole-word' procedure which, at least initially, does not support explicit word identification. We report a test of this proposal in a patient with pure alexia. We reasoned that if the patient had access to two distinct and incompatible procedures, he might be induced to switch from one to the other by changing task demands. We found that when instructed to name words, the patient employed a letter-by-letter strategy; in contrast, when instructed to make lexical decision or semantic judgements about rapidly presented words, he appeared to use a 'whole-word' strategy. These data support the hypothesis that two distinct procedures are available to this patient. We argue, further, that is necessary to suppress use of the letter-by-letter strategy to demonstrate whole word reading capability in pure alexics, and that failure to do so may account for negative findings in other cases reported in the literature.

We describe 4 patients who developed pure alexia after infarctions of the left cerebral hemisphere. All subjects employed a letter-by-letter strategy (with varying degrees of success) to explicitly identify visually presented words. Although all 4 subjects explicitly denied that they could identify briefly presented words, they all performed significantly better than chance on lexical decision and forced-choice semantic categorization tasks with briefly presented words which they could not explicitly identify. Three subjects regained the ability to explicitly identify briefly presented words; these subjects were more accurate with nouns than functors and words of high as compared with low imageability. Additionally, these subjects were impaired in the processing of suffixes. These data are not accommodated by the 'disconnection' account of pure alexia but are more consistent with the hypothesis that reading in these patients is mediated by the right hemisphere.