INDIVIDUAL DIFFERENCES IN RESTING-STATE THALAMO-CORTICAL CONNECTIVITY WITHIN THE LANGUAGE NETWORK OF EARLY BILINGUALS

What differentiates the capacity to perceive a phoneme from failing to do so? Current neurocognitive models support that speech processing involves an integrated pipeline linking subcortical relay stations to a dedicated cortical language network (Lee, 2013; Fedorenko et al., 2024). In this study, we investigated the intrinsic functional connectivity of specific regions within the language network in early, highly proficient bilinguals who vary in their ability to discriminate a nonnative-language phoneme contrast /e/--/$\varepsilon$/.Using resting-state fMRI ROI-to-ROI connectivity analysis, we found stronger connectivity between the the left medial geniculate body (MGB) and the bilateral angular gyri (AG) in the group of bilinguals that achieve better perception of the nonnative contrast. This aligns with evidence that the left MGB is an adaptive relay sensitive to the rapidly varying spectrotemporal features of speech (von Kriegstein et al., 2008, Diaz et al., 2018). Although the AG supports sound-to-meaning mapping, it patterns distinctly from the core language network by responding to non-verbal information and showing insensitivity to syntactic complexity (Burgaleta et al., 2014; Fedorenko et al., 2024). In this circuit, the present findings suggest that the AG likely acts as a critical phonological interface where functional features reflect more robust categorical representations of phonetic contrasts. The absence of right MGB connectivity reflects established left-hemisphere dominance for linguistic computations (Lee, 2013). Identifying these pathways in task-free data confirms these nodes as stable ``natural kinds" embedded in the brain’s baseline architecture (Fedorenko et al., 2024).