知覚学習システムを利用した促音の聴取訓練による学習効果の検証 -長短母音の位置およびピッチパターン, 促音の後続子音種を中心に-

Abstract

To establish an effective perceptual training method for non‐native listeners, the present study investigated the effectiveness of perceptual training on the distinction between single and geminate consonants for Korean‐speaking learners of Japanese. In particular, the present study examined how the effectiveness of perceptual training is influenced by the following factors, which are known to affect perception of phonemic length contrasts by non‐native listeners: type of consonant involved in consonant length contrast, position of vowel length contrast, and accent type (HL and LH). Two different types of perceptual training were conducted. One was training using stimuli contrasting in single and geminate consonants spoken at a single (normal) speaking rate (fixed‐rate training). The other was training using stimuli spoken at three (fast, normal, slow) speaking rates (mixed‐rate training). The results showed that perceptual training improved the learners’ ability to identify consonant length contrasts for both the fixed‐rate training group and the mixed‐rate training group. In addition, there was no significant difference in accuracy between plosive consonants and fricative consonants before training; also, both types of consonant significantly improved in accuracy after training. However, vowel length contrasts showed differences in accuracy depending on accent type and within‐word position. Specifically, vowel length contrasts which appeared in word‐final position and which appeared in words whose accent type was falling (HL) were the most difficult to identify for Korean‐speaking learners; these words did not significantly improve in accuracy even after intensive perceptual training. These results suggest that perceptual training of consonant contrast helps Korean‐speaking learners’ perception ability of length contrast limited extent.