Evoked and oscillatory EEG activity differentiates language discrimination in young monolingual and bilingual infants

TitleEvoked and oscillatory EEG activity differentiates language discrimination in young monolingual and bilingual infants
Publication TypeConference presentation
Year of Publication2017
AuthorsNacar, L., Guerrero-Mosquera C., Colomer M., & Sebastian-Galles N.
CongressNinth Annual Meeting of the Society for the Neurobiology of Language
PlaceBaltimore, MD, USA

Language discrimination is one of the core differences between bilingual and monolingual language acquisition. Here, we investigate the earliest brain specialization induced by it. Following Bosch and Sebastian-Galles (1997), we hypothesize that bilingual native language discrimination is a complex process involving both language discrimination and identification. We recorded the brain activity of monolingual (N=28) and bilingual (N=28) 4.5-month-old infants using EEG. Infants were passively listening to sentences in their native/dominant language (Spanish or Catalan), and two foreign languages: one of the same rhythmic class (Italian) and one of a different rhythmic class (German). To separate discrimination and identification effects, we defined two different windows of analysis. The early window of analysis (150-250 ms) was selected to explore the response of familiarity to each language by measuring the latencies of the P200 component. Four electrodes (F3, FC5, F4, and FC2) localized on the frontocentral region defined our Region Of Interest (ROI). We run a repeated measures ANOVA with a between factor, Group (Monolingual, Bilingual) and a within factor, Language (Native, Italian, German). The ANOVA yielded a significant Group by Language interaction F(2,167)=3.55 p=0.036. For each of the groups separately, we run a one-way-ANOVA with the factor Language. The ANOVA was significant for the monolingual group F(2,81)= 3.743 p=0.028; but was not for the bilingual group (F<1). Planned paired t-test comparisons revealed a significant difference for the Monolingual group between Catalan and German (t(13)=-2.883 p=0.007) and Italian and German (t(13)=-2.313, p=0.029) but not between Catalan and Italian (t<1). The same analysis for the bilingual group did not reach significance for any of the comparisons. The late window of analysis (400 - 1800 ms) was selected to analyze the identification response reflected in the theta oscillations. Ten electrodes (FC6, CP2, P4, F4, F8, F3, FC5, CP1, P3, F7) localized on the fronto-temporo-parietal regions defined our Region Of Interest (ROI). We ran a repeated measures ANOVA. The main effects of Group and Language did not reach significance (F<1) but it yielded a significant interaction of Group by Language F(2,24) = 15.37; p<0.001. We ran separate one-way ANOVAs for each group on the Language factor. Only the Bilingual group showed significant differences F(2,81) = 3.25 p=0.045. Planned comparisons only showed significant differences for the Bilingual group when comparing the Native language against German (t(13)= 2.30; p=0.029) and the Native language versus Italian (t(13)= 2.17; p=0.038). The results indicate different language discrimination strategies for bilingual and monolingual infants. While monolingual infants show early discrimination of their native language based on familiarity, bilinguals perform additional processing related to language identification. Such additional processing originates in the bilingual group at a higher power in the Theta band for the native language as compared to the foreign languages, reflecting bilinguals’ additional processing of the prosodic properties of the familiar language. This is the earliest evidence found for brain specialization induced by bilingualism.

Type of WorkPoster