Understanding the link between bilingual aphasia and language control

The study of bilingual aphasia is important because we need to be able to recommend treatments consistent with a plausible estimate of the course of recovery. Yet we lack a causal account of recovery patterns. We distinguish between the neural representation of a language network and the regions involved in the control of that network. Contrary to some claims, we argue on the basis of normal data that a single adapted network underlies the representation of more than one language and identify a frontal–(parietal)–subcortical network in its control. In terms of patient data, the broad expectation is that recovery of L1 and L2 will parallel premorbid levels of proficiency where there is no problem of language control. Recent advances mean that such an expectation can be tested on samples of patients rather than by sampling cases reported in the literature. Voxel-based morphometry can be used to relate variations in grey-matter density to variations in task performance. Understanding this relation can then help provide an estimate for future patients of the likelihood of improvement over time or a yardstick against which to measure the effectiveness of any intervention. In addition to this large sample approach, the study of individual cases remains key to achieving an understanding of the connections between representation and control and recovery patterns. We review recent cases of the effects of frontal–subcortical damage in bilinguals and argue that they provide evidence of effects on language selection and control rather than evidence for distinct neural networks underlying the processing of a second language. We conclude that there are good prospects for substantially improving our understanding of recovery patterns and that neuroimaging studies during recovery will provide further constraints on the mechanisms of control.