Intra‐phonemic perception of a synthetic /r/‐/l/ series by musicians and nonmusicians

Abstract

In previous categorical perception studies of synthetic series underlying consonant contrasts, wide individual differences have been noted in subjects' ability to discriminate within‐phoneme‐category acoustic differences. This study explored the relationship between speech discrimination ability and musical training/experience. Twenty very experienced musicians (ten left handed, ten right handed) and 20 musically naive listeners performed three tasks on a synthetic /rak/‐/lak/ speech series: two‐choice identification, open‐set identification, and two‐step and three‐step oddity discrimination. Fifteen of the 40 subjects (nine musicians and six nonmusicians) could reliably categorize the ten stimuli into three or four “phoneme” categories. Overall, discrimination performance was better for musicians than nonmusicians, and indicated that discrimination was better than predicted even from open‐set identification. Results are discussed in terms of correlations among linguistic and nonlinguistic auditory perceptual capacities. [Work supported by NIMH.]