Dissociating the performance of cortical and subcortical patients on phonemic tasks

Abstract

To assess cortical and subcortical contributions to phonemic processing, patients with left frontal, temporal–parietal, or cerebellar lesions as well as those with Parkinson’s disease were tested on phonemic identification and production tasks. In Experiment 1, patients and controls were asked to identify syllables on both a voicing and place of articulation continuum. Subcortical patients were relatively unimpaired at this task whereas cortical patients were less accurate at identifying the endpoints of both continua and exhibited little evidence of categorical perception. For Experiment 2, controls and patients were asked to produce syllables. Subcortical patients were able to produce contrastive voice onset times (VOTs) for voicing cognates although VOT of the voiceless phoneme was more variable for cerebellar patients. Cortical patients showed greater overlap in the production of both VOT and formant transition intervals. These results are discussed in terms of the type of computations hypothesized to originate from each neural area.