Regardless of the specific imaging method used, the task of elucidating the neural basis of speech perception by relating areas of the auditory cortex to specific speech processing functions poses an enormous challenge. First, the auditory cortex is anatomically very dense, extending over a region of approximately 40 centimeters, while containing 12 or more cortical fields, some of which may have linear dimensions of 10 to 20 millimeters. Second, acoustic processing is highly distributed in nature; unlike vision, it does not present a clearly recognizable modularity of function (there exist no auditory counterparts of the visual areas for color and motion), with many sound features being encoded before the signal reaches the cortex. Moreover, it involves a complex array of stimulus features (for example, rate, frequency, and amplitude), dynamically arranged through varying temporal patterns, and cross-modally related to articulatory gestures (such as rounding or pressing the lips together). Finally, mapping speech-specific functional areas requires contrasting brain activity in speech perception against a baseline condition, defined by an equally “complex” nonspeech acoustic stimulus.

These factors explain why functional neuroimaging has, until recently, failed to show the left hemisphere specialization for speech processing predicted from the clinical literature. In their challenging paper, Scott and Wise guide us through the intricacies of neuroimaging studies of speech processing, to offer new and exciting perspectives. The upshot of this fascinating tour is that, after controlling for acoustic complexity, neuroimaging can indeed provide evidence for left hemisphere lateralization for the processing of intelligible speech. Intriguingly enough, the cortical regions involved are very similar to monkeys’ cortical regions activated by the calls of other monkeys.