The role of L2 AoA on language switching costs: Only early bilinguals feature symmetrical language switch costs

2010

Abstract  Language-switching tasks have shown that low- and high-proficient bilinguals display different patterns of language switching, the former obtaining asymmetrical switching costs (larger switching costs from L2 to L1 than vice versa: Meuter & Allport, 1999; Costa & Santesteban, 2004) while the latter show symmetrical switching costs (same switching costs from L2 to L1 and L1 to L2: Costa & Santesteban, 2004; Costa, Santesteban & Ivanova, 2006; Ekert-Centowska, 2006; Kroll et al., 2008). In this paper we report on the results obtained in a language switching task and investigate whether these patterns depend not only on the L2 learners’ proficiency (Green, 1998; Costa & Santesteban, 2004), but also on the age of L2 acquisition (L2 AoA). We tested a group of 12 highly proficient Spanish – Basque bilinguals who had started learning their L2 (Basque) after age 12 (late learners). Using the same materials as in Costa et al. (2006), participants were asked to alternate naming 10 pictures of common objects in their L1 and L2. Our results revealed larger switching costs from L2 to L1 than vice versa. Thus, as low-proficient bilinguals, the high-proficient bilinguals in our study featured asymmetrical switching costs when switching between their L1 and their L2. We compared our results with those from the Spanish-Basque bilinguals (early learners) tested in Costa et al. (2006) and we concluded that the language switching cost patterns of early and late Spanish-Basque high-proficient bilinguals is significantly different: symmetrical vs. asymmetrical patterns, respectively. The differences we have found between both groups (early and late Spanish-Basque highly proficient bilinguals) suggest that AoA is a relevant factor in the type of language switching mechanism employed by bilinguals. To be able to shift from an inhibitory mechanism to a language-specific selection mechanism, early AoA and high-proficiency in L2 are required


Mail Portal

LOGIN @laslab.org
powered by Google