Kyoko Nagao, Kenneth J. de Jong, and Byung-jin Lim
Dept. of Linguist., Indiana Univ., 322 Memorial Hall, Bloomington, IN 47405
Previous results (Nagao et al., 2001) show that non-native listeners exhibit remarkably similar patterns of perceptual resyllabification with English listeners, suggesting that perceptual resyllabification is not a language specific phenomenon. The same listeners tended to identify voiced English tokens as voiceless, in keeping with Japanese voicing categories. In order to determine the degree to which the non-native perceptual resyllabification was due to extensive exposure to English, monolingual Japanese listeners, 8 from an older and 12 from a younger generation, participated in the same experiment. Monolinguals showed perceptual resyllabification of the same tokens as do English listeners', consistent with the previous results. Also consistent with previous results, especially older listeners' responses were more affected by Japanese voicing categories. In addition, older listeners were more likely than English listeners to identify tokens as CV's, while younger listeners were less likely to. The results for older listeners are what is expected of biases toward native categories. The results for the younger listeners, however, seem to indicate a developmental process involving the construction of new prosodic categories. [Work supported by NIDCD and NSF.]