Understanding face perception by means of prosopagnosia and neuroimaging.

Understanding the human neuro-anatomy of face recognition is a long-standing goal of Cognitive Neuroscience. Studies of patients with face recognition impairment following brain damage (i.e., acquired prosopagnosia) have revealed the specificity of face recognition, the importance and nature of holistic/configural perception of individual faces, and the distribution of this function in the ventral occipito-temporal (VOT) cortex, with a right hemispheric dominance. Yet, neuroimaging studies in this field have essentially focused on a single face-selective area of the VOT and underestimated the right hemisphere superiority. Findings in these studies have also been taken as supporting a hierarchical view of face perception, according to which a face is decomposed into parts in early face-selective areas, these parts being subsequently integrated into a whole representation in higher-order areas. This review takes a historical and current perspective on the study of acquired prosopagnosia and neuroimaging that challenges this latter view. It argues for a combination of these methods, an approach suggesting a coarse-to-fine emergence of the holistic face percept in a non-hierarchical network of cortical face-selective areas.