Differential cue weighting in the acoustics of L1 and L2 English sibilants by Japanese-, Chinese-, and Vietnamese-speaking speakers

This study examines the interaction between L1 influence and L2 target acoustics on differential cue weightings in the production of L2 English fricatives by learners whose native language is Japanese, Chinese, or Vietnamese. These languages include post-alveolar sibilants in their phoneme inventories that differ from the target palato-alveolar ʃ of English, yet they differ from one another in their precise surface realization: alveo-palatal ɕ in Japanese, retroflex ʂ in Vietnamese (northern dialect), with both of these appearing in Chinese. Distinguishing the s∼ɕ contrast relies on F2 differences at the transition into a following vowel, whereas the s∼ʂ contrast depends almost exclusively on distinct measures of CoG (i.e., Center of Gravity). Three groups of participants with Japanese (n = 3), Chinese (n = 3), or Vietnamese (n = 3) as their native language produced English (Condition 1) and native-language (Condition 2) carrier phrases containing nonsense disyllables with medial intervocalic alveolar and post-alveolar fricatives. Logistic regressions reveal F2 to be a primary cue for Japanese speakers in L1 and L2, consistent with direct transfer. Vietnamese speakers demonstrate cue reorganization with some reliance on F2 in addition to CoG, and Chinese speakers exhibit variable performance reflecting the greater number of L1 categories.