Home » Abstract » More distributed neural networks for bilinguals than monolinguals during switching

More distributed neural networks for bilinguals than monolinguals during switching

Kalinka Timmer, John G. Grundy, & Ellen Bialystok (York University, Toronto, Canada)

ktimmer@yorku.ca

bilingualism; task-switching; language switching; event-related potentials (ERPs); mixing cost; switching cost

During task-switching bilinguals outperform monolinguals due to their experience with continuous language-switching. This behavioral advantage is not always present in young adults due to ceiling performance. The present study investigated the underlying processing differences between monolinguals (English) and bilinguals (English-French) during possible similar performance. Further, the underlying mechanisms of task-switching are compared to language-switching with electrophysiological (EEG) measures. Both groups performed a domain general (task-) and language-switching task including blocks with one task (pure) and blocks with two tasks (mix), where the task could be repeated or switched. During task-switching monolinguals and bilinguals demonstrated same mixing (pure vs. mix) and switching (repeat vs. switch) costs. However, within the mixed blocks bilinguals were more accurate than monolinguals. The same behavioral mixing cost was reflected in a more distributed neural network for bilinguals. The same behavioral switching cost revealed earlier processing differences for bilinguals (275 ms) than monolinguals (325 ms). These processing differences could explain the enhanced performance during the mixed blocks. This was supported by the language-switching task that also revealed more distributed networks for bilinguals than monolinguals for the switch cost and additional late (350-400 ms) executive control for monolinguals compared to bilinguals for the mixing cost. Thus, the more distributed networks for bilinguals suggest the integration of verbal and non-verbal control networks during early visual processing (125-175 ms) and later executive processing (225-275 and 325-375 ms).