摘要：This study investigates whether vowel insertion in English consonant clusters produced by Japanese ESL (English-as-a-second-language) learners is due to misarticulation or misinterpretation. Intermediate level Japanese ESL learners read aloud 24 written real and nonsense English words, then, to mimic auditory words produced by a native English speaker. The results showed that the participants inserted a vowel in consonant clusters notably less frequently in the mimicking task than in the reading task, suggesting that the participants can perceive and produce consonant clusters. The participants were also asked to divide each stimulus word into syllables, and they often clearly pronounced extra vowels: e.g. ‘ba-d o -min-ton’ for ‘badminton.’ I conclude that vowel insertion is not because of their inability to articulate consonant clusters, but their misinterpretation that there is a vowel where there is actually not. When they mimicked a native English speaker’s production, they phonetically deleted such vowels, which existed in the first place, while there was still a vowel at their phonological underlying representation or in their mind.
