Rochelle S. Newman
James R. Sawusch
Dept. of Psychol., SUNY at Buffalo, Buffalo, NY 14260
Previously, results were presented showing that varying the duration of a distal vowel can produce assimilatory, rather than contrastive effects on perception. In a /shwas/--/chwas/ series, a longer vowel yielded more /sh/ responses rather than more /ch/ responses. Varying /w/ duration produced the standard contrast effect. The assimilative effect may reflect a chaining of contrast effects: the long /a/ made the /w/ seem shorter, and the ``shorter'' /w/ made the initial segment seem longer. Alternatively, the lack of a distinct acoustic boundary between /w/ and /a/ may have produced these results, in which case results should be very different for a voiceless stop followed by a vowel. In new experiments, a similar series was made ranging from /chkas/ to /shkas/, and the /k/ and /a/ durations were varied separately. The /k/ produced the usual contrast effect while the nonadjacent /a/ had no effect. These results and their implications for the understanding of rate normalization will be discussed. [Work supported by NIDCD Grant No. DC00219 to SUNY at Buffalo and an NSF Graduate Fellowship to the first author.]