Abstract
Studies of patients with acquired prosopagnosia (AP) have provided invaluable information about human face processing. However, while patients typically complain of familiar face recognition impairments, these studies generally involved processing of unfamiliar faces. Here we conducted a series of 19 behavioral experiments to investigate personally familiar face processing (identification, forced-choice recognition) in PS, a pure case of AP, and her colleagues using faces of the children they supervised in a kindergarten. Stimulus manipulations included variations in the availability of external, color and spatial frequency information, as well as stimulus similarity and spatial location of diagnostic information using anti-caricatures and the response classification technique ‘Bubbles’, respectively. Further experiments assessed the extent to which facial information was perceived inter-dependently (e.g. whole-part advantage, composite face effect, global facial geometry). Across experiments, PS relied more heavily on the availability of local details as compared to normal controls, and exhibited deficient processing of the overall facial configuration of familiar faces and of the eye region. Although the use of familiar faces led to particularly robust effects and dissociations between the patient and normal controls, these findings parallel observations in PS and other AP patients made previously using unfamiliar faces. We suggest that PS’s deficient representation of the eye region arises because it contains multiple features over a small space, and its diagnosticity is heavily dependent on the ability to integrate this information simultaneously (i.e., holistic perception). These observations support the hypothesis that AP is associated with a loss of holistic face perception, which affects both the processing of unfamiliar and familiar faces. More generally, they do not support the view that familiar an unfamiliar face processing differ qualitatively in AP.
Meeting abstract presented at VSS 2015