Are reading and spelling phonologically mediated? Evidence from a patient with a speech production impairment

In this paper, we report the case of a patient, PS, who shows excellent comprehension of the meaning of words despite making a large number of phonemic paraphasias on tests of oral reading, picture naming, auditory repetition and in spontaneous speech. In a series of experiments, we attempt to demonstrate that his preserved reading comprehension could not reasonably be explained in terms of the theory of phonological mediation during reading (cf. Van Orden, 1987). Although he is good at comprehending the meaning of written homophones, the critical finding is that PS is unable to produce both meanings of a homophone when presented with the written form of one version of the homophone. That is, he is unable to make a response such as ''inherits'' in response to the written word ''air'' (heir). Although phonological mediation may play some role in the way in which meaning is accessed from print by normal readers, the results of this study suggest that it is not obligatory (cf Coltheart, Patterson, & Leahy, 1994). A similar argument is advanced with respect to PS's relatively well-preserved written spelling.