Neural mechanisms of language control in bilingual speech production
Bilinguals have a remarkable ability to coordinate their languages during speech production; they can switch language with ease, but can also stay in a single language if required. What are the neural mechanisms underlying such language control? It has been proposed that bilingual control is supported by a range of cognitive processes (e.g. inhibition), with recent neuroimaging evidence pointing towards the engagement of executive-control brain regions during language switching. Building on these existing knowledge, the present thesis aimed to further investigate the exact manner in which language control takes place in the bilingual brain, focussing on three particular aspects: scope, timing, and demands for control under different circumstances.
Chapter 2 examined the scope of language control, by looking at whole-language and item-specific control side by side in a modified language-switching paradigm. The role of the pre-supplementary motor area (a key brain region for domain-general inhibitory control) was then probed using transcranial magnetic stimulation (TMS). An overall performance decrement following the disruption of this brain region revealed its essential role in general speech execution, while no reliable evidence was found for its specific involvement in either whole-language or item-specific control. Chapter 3 investigated the timing of language control, by dividing the language-switching process into two separate stages (preparation and production). Brain activities were recorded using magnetoencephalography (MEG), and reconstructed into source activities in pre-selected regions of interest. The findings suggested that different control processes took place during the preparation stage and production stage, supported by the left and right inferior frontal gyrus, respectively. Chapter 4 examined how the demands for language control differed between natural and forced switching. The effects of language switching and mixing were compared across three different contexts, with varying levels of external constraint on language selection. The behavioural and MEG results indicated that freedom of language selection and consistent language choice for each given concept were both factors that contributed to making natural switching easier than forced switching.
Taken together, the findings in this thesis expand our current knowledge on the neural mechanisms of language control in bilingual speech production, particularly regarding whole-language and item-specific control, preparation and execution of a language switch, and reduction of control demands in natural switching. These new findings deepen our understanding of bilinguals’ skilful language coordination in daily life.