Neuroimaging of Language: Why Hasn't a Clearer Picture Emerged?

Two broad questions have driven dozens of studies on the neural basis of language published in the last several decades: (i) Are distinct cortical regions engaged in different aspects of language? (ii) Are regions engaged in language processing specific to the domain of language? Neuroimaging has not yet provided clear answers to either question. In this paper, we discuss one factor that is a likely contributor to the unclear state of affairs in the neurocognition of language, and that, in our opinion, has not received sufficient attention in the recent literature. In particular, fMRI studies of language have relied, almost exclusively, on group analyses, in which data from multiple individuals are co-registered to and analyzed in a common space. We argue that this approach can obscure functional specificity because of the anatomical variability across individual brains, and we advocate the use of an alternative approach - the functional localization approach - that circumvents this problem.