Abstract
Last year, we introduced DC, a prosopagnosic man who has intact object and word recognition. The aim of the present study is to evaluate whether his face imagery and representation are spared despite his impaired face recognition. Although many have claimed that imagery and recognition depend on the same neural substrates (e.g. Kosslyn et al. 1995, 1999, and O'Craven & Kanwisher, 2000), object recognition can be dissociated from object imagery (Behrmann et al., 1992, 1994; and Servos & Goodale, 1995). DC, and five matched controls (including his brother) were assessed. Participants were asked to recognize 70 pictures of famous individuals. DC could recognize only 37% of them, and controls, 91%. A face imagery test was designed in which face features and their relations had to be imagined and compared. Half of the faces to be imagined were those that DC could not recognized (set of 30), and the other half were those that he could (set of 30). DC's face imagery performance was at 77% correct for both the faces that he previously recognized and those he did not. Controls obtained 77 and 74%, respectively. DC's imagery ability at the Eddy & Glass (1981)'s sentence imagery test, and at the letter imagery test (Behrmann et al., 1994) was normal, and not different from those of controls. Further evaluation (using for example Arcimboldo faces, Mooney faces, and cartoons) show that DC's face representation is intact. We conclude that face recognition can be dissociated from face imagery and representation.