Barrios, S.
Cognitive Neuroscience of Language Laboratory/Dept of Linguistics, University of Maryland, College Park, USA
Pallier et al(2001) demonstrated that Spanish-dominant bilinguals, unlike Catalan-dominant bilinguals, process minimal pairs containing Catalan-specific contrasts(e.g. /netə/vs./nεtə/) as homophones. They argue that [1]the conflation of phonological categories causes the lack of sensitivity reported in previous phonetic perception experiments(Pallier et al,1997), as well as homophonous L2 lexical representations. However, there is an alternative explanation for their findings. [2]Due to similarities in rhythmic class and the prevalence of cognates, Catalan input may strongly activate Spanish lexical representations. Thus, Catalan candidates will rarely be selected over native Spanish ones. We report the results of two experiments testing hypothesis[1] while controlling for[2].
Experiment 1 was an AX discrimination task examining the phonetic discrimination abilities of 28 English monolinguals and 28 Spanish-dominant bilinguals, to minimal pairs containing common(i.e. /o-u/) and English-specific(i.e. /a-æ/ and /i-I/) vowel contrasts. Experiment 2 employed Pallier et al’s repetition priming paradigm to investigate word recognition processes in the same participants, and using the same stimuli set.
Experiment 1 showed significant effects of Group and Condition(p<.05), as well as a Group x Condition interaction(p<.05), suggesting Spanish-dominant bilinguals lack sensitivity to difficult English-specific vowel contrasts. In Experiment 2, the effects of Group and Condition were also significant(p<.05). Planned comparisons revealed no difference in the facilitation effects observed for English-specific minimal pair repetitions when compared to exact repetitions for Spanish-dominant bilinguals only, suggesting Spanish-dominant bilinguals have shared lexical representations for word pairs distinguished by English-specific contrasts.
These results replicate previous reports that bilinguals lack sensitivity to non-native contrasts and that bilinguals represent minimal pairs containing difficult non-native contrasts as homophones. Furthermore, by replicating these findings with bilinguals whose L1 and L2 are more distinct, we can be confident that the results reported by Pallier et al are not due to difficulty bilinguals’ may experience as a result of spurious activation of L1 lexical items.