Abstract
Face space is a powerful theoretical framework for understanding the representation of face identity in humans. It has been used to explain several face specific phenomena including the inversion and other-race effects (ORE). The ORE refers to the finding that we are better at discriminating faces of our own compared to other races. Face space predicts that other-race faces will be densely clustered and distinctive, reflecting their unique, poorly discriminated features. For nonhuman primates, the analog of the ORE is the other-species effect (OSE), the finding that conspecifics’ faces are discriminated better than heterospecific faces. The extent to which the OSE is present in nonhuman primates is unclear. Recent studies have suggested that rhesus monkeys may be face-generalists, processing all faces in a similar manner, while the conspecific face has acquired a special status for chimpanzees. The present study utilized a face space framework to directly compare the OSE in nonhuman primates. Rhesus monkeys and chimpanzees were required to discriminate between conspecific and heterospecific faces using a computerized matching-to-sample task. Ninety unique combinations of 10 individuals were presented, including both chimpanzee and rhesus monkey faces representing the conspecific and heterospecific stimuli for each subject species. Subjects’ performance was analyzed using multidimensional scaling and the mean interstimulus distance for each face was calculated. Similar to the ORE in humans, chimpanzees showed significantly better performance discriminating conspecific vs heterospecific faces, and the mean interstimulus distance was smaller for heterospecific vs conspecific faces. Rhesus monkeys showed no significant difference in performance or interstimulus distance between conspecific vs heterospecific faces. Therefore, these data confirm that the conspecific face has acquired a special status for chimpanzees, but not rhesus monkeys. These results suggest that for both chimpanzees and humans, the face processing system becomes specially tuned for the most frequently encountered faces.
Meeting abstract presented at VSS 2012