Neural and behavioral signatures of language control in highly proficient bilinguals: Evidence from healthy individuals and brain tumor patients

Abstract

The bilingual population can provide a unique insight into bilingual language processing. One area debated in the literature is how a bilingual brain manages the known languages and what happens when damage occurs. The general aim of this thesis was to further our understanding of bilingual language processing and the functional reorganization that may occur due to external circumstances. This thesis extends previously reported findings on time-frequency MEG components that play an essential role in underpinning neural mechanisms during bilingual switching task. It also compares these findings to patients with brain tumors to better understand the impact of the growing tumor and the neural plasticity mechanisms following tumor resection. In this thesis, I also explore the possibility of a non-invasive ¿virtual lesion¿ approach to replicate some of the cognitive deficits when a tumor is located in one of the known semantic hubs trying to differentiate two networks: one including AG, IFG, and preSMA, and another one - MTG and ATL.Overall, the results suggest that highly proficient bilinguals implement a language-independent mechanism involved in cue-based language selection that enables controlled access to the intended lexico-semantic representation in the ATL, possibly through inhibition of the non-target lexical item and/or disinhibition of the target one.In the brain tumor population, the results suggest that when a tumor compromises areas critical for language processing, 1) it affects different languages differently and 2) to compensate for the processing deficits, bilateral AG and DLPFC take over as the main areas for processing.In the TMS experiment the findings suggest that the left AG is involved in processing bilingual language control, whereas the MTG might not be involved to such degree despite previous findings where both AG and MTG are involved. However, in line with previous evidence, this experiment supports the idea that AG and MTG are not constituting the same network and are functionally dissociable.