The discrimination and recognition of individual visual objects, including faces, words, and common objects, are among the most taxing perceptual challenges confronting observers in their day-to-day life. Not only does the observer need to derive precise information about the various objects under dramatically differing lighting conditions, scales, and vantage points, but the object must also be perceptually individuated from all other instances of that object so that identity can be assigned and the appropriate semantics (and phonology, where relevant) activated. Moreover, all of these processes must be executed accurately and rapidly, notwithstanding the ambiguity of the input arising from the commonality of input features (e.g., all faces have two eyes, a nose, and a mouth in the same spatial arrangement, and all words are made from the same relatively small set of letters). Despite the clear computational challenge associated with object recognition, human observers are remarkably efficient at assigning identity effortlessly and accurately, particularly for faces.
Much recent research has suggested that one way in which this efficiency is achieved is through a division of labor, that is, different classes of input are assigned to different underlying neural systems to mediate the representation of that object type (Downing et al. 2006). At present, there is clear consensus that segregated regions of human ventral cortex are activated differentially in response to different stimulus classes, although the extent to which these regional distinctions are truly domain-specific and exclusive is highly debated.