A prominent account of face processing suggests that face recognition depends on generic mechanisms involved in processing object classes for which individuals have developed expertise. Many laboratory studies of expertise have used a multi-session training paradigm designed to develop expertise with computer-generated stimuli known as greebles. If the recognition of faces and greebles following training depend on the same mechanisms, impairments with faces should be accompanied by impairments with the acquisition of greeble expertise. Contrary to this prediction, we present two cases of acquired prosopagnosia who exhibit normal greeble learning. Florence (female, 29, with a right anterior temporal resection for epilepsy) and Herschel (male, 55, with right occipitotemporal lesions following several strokes) completed an eight-day greeble training procedure used in previous studies. Their accuracy and response times were similar to those of age-matched control participants. In addition, by the end of the training procedure, both Florence and Herschel fulfilled the criterion expertise researchers claim signals successful acquisition of greeble expertise: comparable response times for greeble recognition at the family and individual level. As expected, Florence and Herschel failed to match controlsa?? learning profile in a follow-up training procedure with faces, demonstrating a dissociation between face and greeble expertise. In addition, Herschela??s lesion disrupted his right fusiform face area (FFA) so his results show that greeble learning can occur without an intact FFA. In sum, our findings are inconsistent with claims from the greeble literature challenging face-specificity, and indicate that distinct mechanisms are used for face recognition and the object recognition processes used in greeble training procedures.
展开▼
