Strijkers, K. 1, 2 & Costa, A. 2
1 Universitat de Barcelona
2 Universitat Pompeu Fabra
A mayor endeavour in the study of bilingual speech production is to uncover the processing differences between a bilingual’s first (L1) and second language (L2). Extensive fMRI research seems to point out, at least for highly proficient bilinguals, that the same brain areas are activated during L1 and L2 speech. The small variations between a bilingual’s two languages, if any, seem to reflect a stronger engagement of frontal areas during L2 speech, which has been related to late stages in the processing stream such as phonetic computations (e.g., Indefrey, 2006). However, from behavioural studies we know that bilinguals are substantially slower in naming pictures in their L2 compared to their L1, even for high-proficient bilinguals (e.g., Ivanova & Costa, 2008). In the present study we explored when L1 speech produces processing differences compared to L2 speech by recording the event-related brain potentials (ERPs) while highly proficient Spanish-Catalan bilinguals had to overtly name pictures either in L1 or L2. The rationale was that the ERP technique with its fine temporal resolution might reveal differences which could not be traced with the slow BOLD response in previous fMRI studies. Results demonstrated an early divergence in electrical brain activity between L1 and L2 at the P2 component (onset ± 190 ms). In fact, the results perfectly mimicked the early P2 modulation found for lexical frequency as demonstrated both in the present study as in previous work (e.g., Strijkers et al., 2010). We conclude, in contrast to fMRI findings, that processing differences between L1 and L2 occur early at the lexical level, probably reflecting an increase in attentional resources during L2 speech. Furthermore, the data support the view that the slower naming latencies in L2 compared to L1 are the consequence of a frequency effect between the two languages, rather than an inhibitory effect.