Effects of unilateral brain damage on the processing of speech sounds

Abstract

This study sought to determine whether phonetic transformations of acoustic signals would be accomplished largely within the language-dominant left hemisphere. An analysis of error patterns on dichotically presented CVs was made on three groups of right-handed native English speaking subjects. The performance of left brain-damaged subjects, right brain-damaged subjects, and normal controls was compared in the following conditions: (1) On trials in which the members of a pair of stimuli shared a phonetic feature (either place or voice), and (2) on trials in which the members of a pair contrasted on both features. It was found that normal and right-hemispheric subjects profited from shared phonetic features. Left-hemispheric subjects, in contrast, made relatively little use of feature sharing, and to the extent that they did, showed an advantage only for shared voice. The findings are discussed in terms of two levels of processing (auditory and phonetic) and in terms of hemispheric specialization.