Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Effects of bilingual experience on the slope of aperiodic activity in EEG signals during anticipation of imminent words in sentence contexts
Poster B50 in Poster Session B, Friday, October 25, 10:00 - 11:30 am, Great Hall 4
Katherine Sendek1, Tamara Swaab2; 1University of California - Davis
The presence and utility of lexical semantic prediction in bilingual L2 has been debated, with many studies showing decreased, delayed, or non-existent effects of prediction in L2. Competing theories suggest that the differences are either due to L2-specific differences in experience or to a general effect of bilingualism, wherein differences are due to the presence of two languages. However, these studies have typically compared L2 in bilinguals to monolinguals. Thus, the conflicting findings on prediction during sentence processing may be due to a conflation of bilingual and L2 effects. Studies that investigated prediction in both L1 and L2 have shown slightly delayed prediction effects in both languages, suggesting that there is an effect of bilingualism on prediction, rather than an L2-specific effect. In the present study we examined effects of language experience in Mandarin-English bilinguals using EEG. The goals of the study were to 1) isolate the potential impact of bilingual experience on prediction, and 2) examine if neural and cognitive individual characteristics contribute to modulation of predictive processing in bilinguals with different language experience. We measured the amplitude of the N400 to critical words that were either highly predictable or not in the context, and the slope of the aperiodic activity prior to the onset of the critical words. This is broadband EEG measure has been shown to be sensitive to neural signal efficiency or connectivity, with steeper slopes indicating greater efficiency. The steepness of the aperiodic slope is furthermore predictive of the size of the N400 effect – greater N400 effects of cloze probability are predicted by a steeper slope. In the present study, we recorded EEG during English sentence reading for English monolingual, English-Mandarin bilinguals, and Mandarin-English bilinguals. We hypothesized that if differences in language prediction were due to L2-specific factors, then we would see more similarity between the English monolinguals and English-Mandarin bilinguals compared to the Mandarin-English bilinguals. In contrast, if the differences are due to general bilingual effects, then the two bilingual groups would be more similar compared to the monolingual group. In this ongoing study, participants (N= 10 monolinguals, 14 bilinguals) read 80 sentences (40 high cloze). Our preliminary analysis showed a significant N400 effect of cloze probability for the monolingual group (t(9) = -2.72, p<.05) and the combined bilingual groups (t(13) = -7.36 , p<.05); a significantly steeper aperiodic slope for the bilingual groups relative to the monolingual group, and an interaction between aperiodic slope and group, wherein aperiodic slope was only a significant predictor of the size of the N400 effect for the monolingual group. Overall, our N400 results suggest that Mandarin-English bilinguals predict imminent words in sentences contexts. Furthermore, the increased neural connectivity suggested by the steeper aperiodic slope in bilinguals suggest neural adaptations associated with managing two languages. However, the lack of a predictive relationship between aperiodic slope and the N400 effect in bilinguals suggests that bilingual neural processing mechanisms during language comprehension may differ from those of monolinguals.
Topic Areas: Multilingualism, Methods