Effects of Audiovisual Training in the Perception of English [l] and [n] by Cantonese Learners

This research focuses on the perception of two nonnative consonants by adult native Chinese speakers. Native speakers of certain southern dialects of Modern Chinese, e.g., Cantonese, Xinan, Jianghuai, have been reported to have difficulty in learning the contrast of /l/ and /n/ in English or Standard Mandarin, due to the different status and functions of the contrast in their native systems (Schmidt and Kaminski 1997, Li 2006). Li also reported that intensive short-term audio training could only improve Chinese learners’ identification of the sounds moderately. Previous studies have shown that in addition to audio features, visual articulatory information is also integrated into L1 and L2 speech perception automatically and unconsciously (McGurk & MacDonald 1976, Rosenblum, Schmuckler, & Johnson, 1997, Hardison 2003, Hazan, Sennema, Iba, and Faulkner 2005). Our study aims at examining effects of visual information in perceiving nonnative consonants contrasting manners of articulation. Previous literature on consonant perception has reported that visual aid is most significant with places of articulation, and that with manners, except with /r/ and /l/, has received little attention due to lack of research significance. Contrary to the commonly held idea, however, that the articulation of [l] and [n] is visually similar, a pilot study on videos containing production by two female native English speakers of 10 pairs of English words contrasting [l] and [n] syllable initially suggests visual differences in tongue positions in the articulation of the contrast. This study investigates effects of such visual information in perception of the non-native contrast by Cantonese learners of English. Results show that the relatively more distinct visual information can only help Cantonese speakers better perceive the two sounds in certain contexts, and in others cause more confusion.