Transcoding Sound to Spelling: Single or Multiple Sound Unit Correspondence?

We report a detailed analysis of the written spelling of an aphasic patient, KT, in whom no viable comprehension could be demonstrated despite accurate repetition and fluent speech (transcortical sensory aphasia). His spelling con formed in all respects with the definition of an orthographic/ lexical dysgraphic such that his writing of regular, orthographically unambiguous words and nonsense syllables was entirely satisfactory. Lexical variables such as word class or word frequency had no effect on his performance. Irregular or ambiguously spelled words were selectively impaired. An independent measure of orthographical ambiguity was defined in relation to the frequency with which a phoneme was represented by a particular grapheme and the number of alternative graphemes for the same sound. For the majority of sounds studied, KT was able to demonstrate the use of all the most frequent alternative transcoding units for each sound but he did not use them in the same frequency with which they occurred in written English. Thus his performance bore little or no relationship to predictions based on single phoneme/ grapheme counts and there was unequivocal evidence for his ability to use multiple soundunit correspondences.