DETAILS OF PROJECT

During second language (L2) acquisition neural representations of first language (L1) and second language assimilate. In the reading network, this assimilation leads to a shift from effortful processing to automated reading. Longitudinal studies can track this transition and reveal dynamics that might not become apparent in behavior. Here, we report results from a longitudinal functional magnetic resonance imaging (fMRI) study, which tracked functional changes in the reading network of beginning learners of Greek over one year. We deliberately chose Greek as an L2 that would have similar orthographic transparency but a different alphabet than L1 (Polish). fMRI scans with lexical and semantic decision tasks were performed at five different time points (every ~3 months). The results show comparable activation time courses in both tasks. Classical language areas (left inferior frontal gyrus, left precentral gyrus, and bilateral supplementary motor cortex), and cognitive control areas (inferior parietal lobe) showed stronger activation in response to Greek words compared to Polish after the first months of instruction. Task-related activity in the reading network remained constant throughout the remaining 6 months of learning and was also present in a follow-up scan 3 months after the end of the course. Our findings support the notion that reorganization of language networks is achieved after a relatively short time of L2 instruction. For the reading network, this reorganization might depend on successful grapheme-to-phoneme mapping that occurs in the first months of learning. We also demonstrate that cognitive control areas monitoring conflict between languages are recruited in L2 reading at low proficiency levels. Differences between L1 and L2 networks, when compared directly, become reduced after a sustained learning period, confirming the convergence of the reading networks over time.