Linguistic versus non‐linguistic processing of speech prosody in dichotic listening

Abstract

Linguistic and non‐linguistic information processing of speech prosody are studied using two dichotic listening tasks, a Word task and a F0 task. During the Word task, subjects were required to identify either right‐ or left‐ear stimulus from two‐syllable homophonic words presented with different pitch accents. During the F0 task, subjects were required to identify either right‐ or left‐ear stimulus from F0 partials extracted from the words used in the Word task. The correct percent of responses was high and RT was short for high familiarity words presented to the right ear rather than the others under the Word task, while no such differences were found under the F0 task. RT of the Word task was shorter than the F0 task. These results suggest that the processing of linguistic speech prosody under dichotic listening conditions is based on the interactive auditory and linguistic neural resources with right‐ ear or left‐hemispheric dominance, and are faster than auditory F0 pattern identification. The tasks developed can be used to detect central auditory processing disorders.