You are viewing the SNL 2018 Archive Website. For the latest information, see the Current Website.


Poster E4, Saturday, August 18, 3:00 – 4:45 pm, Room 2000AB

Neural tracking of attended continuous speech in early and late bilinguals

Andrea Olguin1, Mario Cekic2, Tristan A. Bekinschtein1, Mirjana Bozic1;1Department of Psychology, University of Cambridge, 2Department of Computer Science and Technology, University of Cambridge

Directing attention to a single speaker in a multi-talker environment is an everyday occurrence that we manage with relative ease. This selective attention results in stronger neural encoding of the attended speech envelope compared to the unattended one (Ding & Simon, 2014). A prominent but controversial hypothesis in the bilingualism literature states that knowing two languages leads to an enhancement in selective attention (e.g., Bak et al., 2014; Krizman et al., 2012, but see Paap et al., 2015) and that learning a language in early childhood (the putative critical period) enhances plasticity in the neural encoding of speech sound (Krizman et al., 2015). We examined how learning two languages in early childhood (before the age of 6) and after the age of 11 influences the underlying neural mechanisms of selective attention, as compared to monolinguals. We predicted that the extent of bilingual experience influences how auditory interference is managed in a cocktail party-paradigm. To test this hypothesis we tracked the neural encoding of attended continuous speech in 22 monolingual native speakers of English, 22 early Spanish-English bilinguals, and 20 late Spanish-English bilinguals; in the context of different types of acoustic and linguistic interference. Participants always attended to a narrative in their native language while four different types of interference were presented to the unattended ear: a different narrative in their native language, a narrative in a language unknown to the listener, a well-matched non-linguistic acoustic interference (Musical Rain), and no interference. The neural activity was recorded by a dense array 128-channel EEG system and cross-correlated with the speech envelopes for both attended and unattended streams. In addition, to directly compare the patterns of encoding for monolinguals, early, and late bilinguals, we used multivariate Representational Similarity Analysis (RSA) to contrast their attentional encoding early and late in the process. Results showed that there was significantly more robust neural encoding for the attended envelopes than the ignored ones across all conditions in all groups. Critically, we also saw differences between the groups, where native language interference significantly enhanced the encoding of attended speech in monolinguals, but had no such effect in the two bilingual groups. Early bilinguals appear to encode the attended and unattended streams equally when both are in their native language, whereas late bilinguals dissociate native language interference from onset. Taken together, these results demonstrate that top-down selective attention differentially modulates speech envelope encoding in early bilinguals, late bilinguals and monolinguals. They will be discussed in the context of the relevant theories of selective attention and bilingualism.

Topic Area: Control, Selection, and Executive Processes

Back