Discrimination of foreign language speech contrasts: The roles of native language, L2 acquisition and speech production patterns

Abstract

Discrimination of unfamiliar speech contrasts and similarly configured nonspeech contrasts was investigated for French/English bilinguals (native French- and native English-speakers) and monolingual English-speakers. Experiment 1 examined the discrimination of three Thai voice onset time (VOT) contrasts, including a contrast that is phonemic in both languages, a contrast that is phonemic in French but nonphonemic in English, and a contrast that is phonemic in English but nonphonemic in French. This experiment partially replicates and extends Beach, Burnham and Kitamura's (2001) study. It was hypothesised that L2 exposure to the phonetic contrasts would facilitate bilinguals' discrimination of the contrasts. However, the results indicate that L2 experience did not enhance perception of nonnative contrasts, there were no differences between discrimination performances of the bilinguals and the monolinguals. Furthermore, the results suggest that discriminability of nonnative contrasts depended on the acoustic salience of the contrast examined. VOT productions in the nattve-language and in the L2 were also examined to identify relationships between production of VOT and perception of VOT. Bilinguals who produced VOT in their L2 with similar values to native speakers of the L2 also displayed better discrimination of the nonnative contrasts. It was hypothesised that the bilinguals may have learnt to implement these specific voicing features in their L2 through perceptual sensitivity to the features in their L2 speech environment. Experiment 2 tested this hypothesis by examining discrimination of acoustic tone onset time contrasts (TOT). Overall, there was only weak support for this hypothesis. The tendency towards better discrimination of the TOT stimuli from the high proficiency bilinguals, but not of the synthetic VOT stimuli, may be explained, in part. by the fact that the synthetic stimuli discrimination task involved a second presentation of the same phonetic contrasts, or due to the small number of participants in comparison groups for this experiment. The complex relationship between speech production and speech perception is discussed. The stability of discriminatory abilities and the perception of synthetically created speech stimuli were also examined, with results indicating that discrimination of synthetic speech contrasts is similar to discrimination of naturally produced tokens, but that discriminatory of nonnative contrasts is not stable over time. It is concluded that experimental factors, such as methodology, instructions given and participants' criteria for discrimination play important roles in determining discrimination performance as well as native language background. In terms of production and perception, the two skills appear linked, but no specific aspect of production can be said to be sufficient, or indispensable, for good discrimination of nonnative contrasts.