Bilingual switching between languages and listeners: Insights from immersive virtual reality

Abstract

Perhaps the main advantage of being bilingual is the capacity to communicate with interlocutors that have different language backgrounds. In the life of a bilingual, switching interlocutors hence sometimes involves switching languages. We know that the capacity to switch from one language to another is supported by control mechanisms, such as task-set reconfiguration. This study investigates whether similar neurophysiological mechanisms support bilingual switching between different listeners, within and across languages. A group of 48 unbalanced Dutch-English bilinguals named pictures for two monolingual Dutch and two monolingual English life-size virtual listeners in an immersive virtual reality environment. In terms of reaction times, switching languages came at a cost over and above the significant cost of switching from one listener to another. Analysis of event-related potentials showed similar electrophysiological correlates for switching listeners and switching languages. However, it was found that having to switch listeners and languages at the same time delays the onset of lexical processes more than a switch between listeners within the same language. Findings are interpreted in light of the interplay between proactive (sustained inhibition) and reactive (task-set reconfiguration) control in bilingual speech production. It is argued that a possible bilingual advantage in executive control may not be due to the process of switching per se. This study paves the way for the study of bilingual language switching in ecologically valid, naturalistic, experimental settings. 1.1. Language switching: switch costs and reversed language dominance Bilinguals' ability to switch languages on the basis of an external cue has been investigated extensively in studies using the cued languageswitching paradigm (

The Interplay Between Language Form and Concept During Language Switching: A Behavioral Investigation

“…The present results also showed reversed language dominance: faster responses when naming in L2 than in L1 in mixedlanguage blocks. The reversed language dominance has been reported in studies examining unbalanced bilinguals (e.g., Costa and Santesteban, 2004;Christoffels et al, 2007;Gollan and Ferreira, 2009;Verhoef et al, 2009;Peeters et al, 2014;Kleinman and Gollan, 2016;Peeters and Dijkstra, 2018;Wu et al, 2018;Liu et al, 2019;Peeters, 2020). The present study showed a global slowdown of the L1 across semantic conditions in switch trials.…”

On the Connection Between Language Control and Executive Control—An ERP Study

“…In the late time window of the picture-locked ERPs, we also observed a switch-related negativity, which is in line with the findings of Kang et al (2020), Peeters (2020, and Peeters and Dijkstra (2018). According to Kang and colleagues, this pattern represents an N400-like component such that word meaning retrieval is more difficult during switch trials.…”

“…For example, there is no short peak in the picture-locked waveforms that resembles an N2 component, and the difference between switch and repetition trials seems to occur broadly across the scalp rather than being constrained to anterior sites. Not observing a typical switch-related N2 over anterior sites is not entirely surprising, since this component does not seem robust across studies (e.g., Christoffels et al, 2007;Martin et al, 2013;Peeters, 2020;Peeters & Dijkstra, 2018;Timmer et al, 2019;Verhoef et al, 2009). Instead, it looks like the pattern in the early time window of the picturelocked data may reflect the beginning of the pattern that we see in the late time window.…”

“…Stimulus-locked ERPs in the language switching paradigm are less straightforward beyond the N2 component. Some studies have observed a late positive complex (LPC), characterized by a larger positivity over posterior sites for switch trials relative to repetition trials around 400-650 ms after stimulus presentation (e.g., Jackson et al, 2001;Liu et al, 2016;Martin et al, 2013), whereas others have found a widespread late switch-related negativity (Kang et al, 2020;Peeters, 2020;Peeters & Dijkstra, 2018). Task switching studies, on the other hand, tend to find evidence for a P3b component around 400-600 ms after stimulus onset (e.g., Barceló, 2003;Provost et al, 2018;Tieges et al, 2007; for a review, see Gajewski et al, 2018).…”

“…Finally, if different control mechanisms are involved for language and task switching we might also expect a difference in a late time window of the picture-locked ERPs. Task switching studies tend to observe a modulation of the P3b component (for a review, see Gajewski et al, 2018), whereas the language switching literature has shown a switch-related modulation of the LPC (e.g., Jackson et al, 2001;Liu et al, 2016;Martin et al, 2013) or a late switch-related negativity (Kang et al, 2020;Peeters, 2020;Peeters & Dijkstra, 2018).…”