Search Abstracts | Symposia | Slide Sessions | Poster Sessions
Dissociating Proactive and Reactive Language Control in Bilinguals: MEG Evidence from Delayed Picture Naming
Poster Session D, Thursday, October 1, 4:30 - 6:30 pm, Wangari Maathai
This poster is part of the Sandbox Series.
Jihyun Park1, Mana Tanifuji1, Shinji Nishimoto2, Hiromu Sakai1; 1Waseda University, 2The University of Osaka
Bilingual language switching requires speakers to select the intended language while suppressing interference from the non-target language. A central question is when this control begins. It may operate proactively, when speakers prepare the target language in advance, or reactively, when they select the target language after competing information has become active. Although previous multivariate EEG evidence suggests partly overlapping time courses for these processes, their spatial profiles remain unclear (Zhang et al., 2023). To address this issue, we used magnetoencephalography (MEG) in late Chinese–Japanese bilinguals. We examined the spatiotemporal properties of proactive and reactive language control. The delayed picture-naming paradigm separated object processing, language selection, and response preparation from overt articulation. This design allowed us to track language-control processes before speech production while reducing speech-related artifacts. We tested three cue-timing conditions: simultaneous, cue-first, and picture-first. In the simultaneous condition, the picture and language cue appeared together, resembling standard language-switching tasks. In the cue-first condition, the language cue appeared before the picture, testing whether speakers could prepare language selection before object-related lexical information became active. In the picture-first condition, the picture appeared before the language cue, testing reactive control after object-related lexical information had been activated. Switch and non-switch trials were compared separately for Chinese (L1) and Japanese (L2) naming. Source activity was reconstructed from MEG signals using minimum-norm estimation (https://mne.tools/stable/index.html). Source-space time courses were extracted from predefined regions of interest. These regions were selected based on previous neuroimaging studies and models of bilingual language control (Abutalebi & Green, 2007; Luk et al., 2012). They included the anterior temporal lobe (ATL), anterior cingulate cortex (ACC), inferior and middle frontal gyri (IFG and MFG), superior frontal gyrus (SFG), premotor cortex/supplementary motor area (premotor/SMA), and supramarginal gyrus (SMG). Preliminary results showed clear delay-period neural activity. In the simultaneous condition, switch trials elicited stronger activity than non-switch trials. Clear differences appeared in the left IFG, left SFG, bilateral ATL, and bilateral ACC. In the cue-first condition, delay-period switch–non-switch differences were mainly limited to the right ATL. This suggests weak proactive language-control activity before object-related lexical information became active. In the picture-first condition, switch-related differences appeared in both delay periods, after picture appeared and after the language cue. This pattern may indicate that object-related lexical information was already active before the cue, and that reactive language control emerged once the target language was specified. These findings suggest that cue timing modulates the engagement of proactive and reactive language control. In addition, switching direction shaped the neural response. Switching from L1 to L2 elicited stronger activity than switching from L2 to L1. For late bilinguals, this asymmetry may indicate that activating and maintaining the weaker L2 was more costly than releasing inhibition on the dominant L1. Overall, these results provide MEG evidence that proactive and reactive language control can be partly dissociated in space and time.
Topic Areas: Multilingualism, Language Production