White matter integrity in the vicinity of Broca's area predicts grammar learning success.

Humans differ substantially in their ability to implicitly extract structural regularities from experience, as required for learning the grammar of a language. The mechanisms underlying this fundamental inter-individual difference, which may determine initial success in language learning, are incompletely understood. Here, we use diffusion tensor magnetic resonance imaging (DTI) to determine white matter integrity around Broca's area, which is crucially involved in both natural and artificial language processing. Twelve young, right-handed individuals completed an artificial grammar learning task, and DTI of their brains were acquired. Inter-individual variability in performance correlated with white matter integrity (increasing fractional anisotropy (FA)) in fibres arising from Broca's area (left BA 44/45), but not from its right-hemispheric homologue. Variability in performance based on superficial familiarity did not show this association. Moreover, when Broca's area was used as a seed mask for probabilistic tractography, we found that mean FA values within the generated tracts was higher in subjects with better grammar learning. Our findings provide the first evidence that integrity of white matter fibre tracts arising from Broca's area is intimately linked with the ability to extract grammatical rules. The relevance of these findings for acquisition of a natural language has to be established in future studies.