Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Frequency of second language usage modulates selective attention in bilinguals at the level of lexical and syntactic processing
Poster Session E, Friday, October 2, 11:00 am - 1:00 pm, Wangari Maathai
Nada Stojanovic1,2, Stephen Theron-Grimaldi2, Mirjana Bozic2; 1MRC Cognition and Brain Sciences Unit, University of Cambridge, 2Department of Psychology, University of Cambridge
Bilingualism modulates attentional processing, with bilinguals showing more stable cortical tracking of attended speech and greater resilience to interference compared to monolinguals (Olguin et al., 2019; Phelps et al., 2022). This is thought to reflect flexible redistribution of attentional resources driven by the habitual need to select and suppress competing languages (Green & Abutalebi, 2013; Phelps & Bozic, 2025). A growing body of evidence suggests that these adaptations are dynamically shaped by current L2 usage (DeLuca et al., 2020; Tu et al., 2015). However, language control operates across multiple levels of linguistic representation (phonological, syntactic, semantic), and it remains unclear where usage-dependent effects might emerge, with prior work largely limited to tracking of the speech envelope. Forty-eight early English-French bilinguals, matched on L2 proficiency and learning history, performed a dichotic listening task while EEG was recorded. Participants differed in frequency of L2 usage to form three groups: Active (daily), Moderate (occasional), and Inactive (very sporadic) users (N=16 per group). Participants attended to simple narratives in L1 (English) or L2 (French), while simultaneously presented with intelligible or unintelligible interference. Twelve forward multivariate temporal response function (mTRF) models were fit to features spanning acoustic, phonological, morphological, lexicosemantic, and syntactic levels of speech representation. Linear mixed-effects models tested effects of usage group, attended language, and interference type on neural tracking of attended and unattended streams. All features produced reliable neural tracking above chance in both attended and unattended streams, with the exception of inflectional morphology, consistent with its sparse occurrence in our stimuli. Attended streams were tracked more robustly than unattended streams across all features, with no effect of usage group, consistent with prior backward mTRF decoding of the speech envelope of this dataset (Theron-Grimaldi et al., 2025). In the unattended stream, however, neural tracking of lexical and syntactic features was significantly modulated by usage group, with active users showing stronger tracking than both moderate and inactive users, and no difference between the latter two groups. This effect was independent of attended language and interference type, and was not observed for acoustic or phonological features. These results suggest that usage-dependent attentional adaptation in bilinguals is selectively expressed at lexical and syntactic levels of speech processing, while acoustic and phonological encoding remains comparable across usage groups. This indicates that the bilingual attentional system does not uniformly suppress the unattended stream, but rather engages differentially with levels of the speech hierarchy. Notably, active users showed enhanced rather than reduced tracking of unattended lexical and syntactic content. Further analyses of TRF weights topographies and their temporal dynamics across usage groups, which will be presented at the conference, are expected to shed further light on the nature of this effect.
Topic Areas: Speech Perception, Multilingualism