Crowding the face-space: The attractor field hypothesis and within-person variability

Abstract

ABSTRACT Face space theory suggests that faces that are similar to others (i.e., typical faces) are represented in denser regions in face space than distinctive faces. Accordingly, typical face representations can be activated by the same input, leading to mistakenly identifying a person as someone else. A modification of this theory can also accommodate the opposite error in which two images of the same person are mistaken for different people, which results from intolerance to within-person variability. In two experiments, we tested two predictions of the modified theory: (1) greater tolerance of within-person variability should be observed for distinctive faces and (2) the same conditions that increase tolerance to within-person variability should facilitate differentiation of two similar-looking identities. The results support the first of these predictions, but not the second. The findings are interpreted in the context of attention to differences vs. commonalities when learning to distinguish faces from similar-looking others.