Brain potentials reveal activation and inhibition of the non-target L2 in bilinguals’ L1 context

Aims and objectives/purpose/research questions: Using behavioral measurements and event-related potentials (ERPs), this study investigated the temporal dynamics of Chinese-Japanese bilinguals’ non-target second language (L2) activation and inhibition when they process their first language (L1) embedded with Chinese-Japanese homographs. Design/methodology/approach: Both Chinese-Japanese bilinguals and Chinese monolinguals were recruited to perform an orthographic judgment task on Chinese word pairs. In these pairs, priming words were either interlingual homographs or non-homographs. Target words were semantically related or unrelated to the priming words’ Chinese meanings. Crucially, Chinese-unrelated pairs with homographic primes were semantically related in Japanese. Data and analysis: Behavioral and ERP data were analyzed using mixed-effects models. The key focus of the analysis was on the fixed effects of semantic relatedness (related/unrelated), priming word type (homograph/non-homograph), group (bilingual/monolingual), and their interactions. Findings/conclusions: The results revealed a more positive P200 amplitude for semantically related word pairs relative to unrelated ones in bilinguals, indicating early activation of homographs’ L1 meanings during the early stage of processing. Comparable N400 amplitudes were observed in the bilingual group for pairs containing homographic primes, suggesting co-activation of both L1 and L2 meanings of these homographs. Subsequently, L2 meanings of homographs failed to trigger a significant LPC effect in bilinguals, implying that the L2 meaning may be inhibited at the late processing stage. Originality: This study provides ERP evidence that non-target L2 undergoes a dynamic process of mid-stage activation followed by late-stage inhibition during L1 processing, even across typologically distinct languages. Significance/implications: These findings support two key bilingual lexical access models: the Inhibitory Control (IC) model and the Bilingual Interactive Activation Plus (BIA+) model, both of which posit that late-stage L2 inhibition optimizes target language processing. By documenting temporal dissociation between L2 activation and inhibition, the study highlights adaptive cognitive control mechanisms in bilinguals, revealing how cross-language activation is managed over time to balance semantic integration and task demands.