English vowel perceptual training in Cantonese-speakers learning English as a second language : a comparison between purely auditory full-set training and auditory plus visual and articulatory input full-set training

Abstract

Purpose: Studies on speech perception training have shown that structured and intensive auditory training using a full-set (i.e. targeting all vowels) of highly variable stimuli with corrective feedback can improve L2 learners’ perception of non-native vowels. Some theories suggest that information about the articulatory gestures from multisensory input is critical for speech acquisition. Present study compared the effectiveness of two full-set training approaches, one included explicit instructions and visual information on the articulatory features of AE vowels (AVA group) and the other without (AO group), on improving AE vowel perception in Cantonese learners of English. Method: Twenty-three native Cantonese-speaking adults were trained for 10 sessions on 10 American English monophthongs using one of the two approaches. Participants’ performance in two pre-tests, an interim test and a post-test was compared. Results: Results indicated that both training approaches were effective. Participants in the AVA group received fewer auditory training than the AO group but achieved the same training effects. Training effects were generalized to untrained words and novel speakers in both groups. Conclusions: Both training approaches were equally effective but training with articulatory features was more efficient in improving Cantonese learners’ perception of American English vowels.