音韻知覚と音韻記憶表象

概要

論文の詳細を見る
Previous studies (Takayama, 2002, 2003) indicated that, in the identification learning of sinewave analogs of stop CVs or vowels, performance under speech instructions using phonemic labels matched to analogs was superior to performance under nonspeech instructions using arbitrary labels, which was better than performance under speech instructions using mismatched phonemic labels. Present study aimed to replicate these findings using analogs of other speech sounds, Japanese "わ" and "ら" in the first experiment and American English /la/ and /ra/ in the second experiment. It was predicted that the superiority of matched speech instructions over other instructions (and nonspeech instructions over mismatched speech instructions) should be found in learning on analogs of native syllables, but that, in learning on analogs of nonnative syllables, no difference should be among instructions groups. These predictions were mostly confirmed. The hypothesis that a phonemic perception is a process in which perceptual representation of signals is matched to phonemic representation in long-term memory was discussed.
近畿大学の論文
2005-12-20