Abstract
Recent studies using unfamiliar face recognition tasks have found evidence that pigmentation and shape information are about equally important for face recognition. However, it is possible that the information subjects can discriminate or remember is not the same as what they actually do use to recognize people in the real world. Of course, it is this latter kind of recognition that is of greatest interest. We investigated subjects‘ ability to recognize photographs of their friends. The subject pool consisted of 30 undergraduates who all lived in the same dormitory, and saw one another on a daily basis. These individuals were photographed and later comprised the subjects in a recognition experiment. Two sets of stimuli were created: 1) An averaged face was warped in two dimensions into the shapes of individual faces, and 2) Individual faces were warped into the shape of the averaged face. In set 1, the faces contained the shape cues of the original individuals, but had generic pigmentation cues, while in set 2, the faces contained the pigmentation cues of the original individuals, but had generic shape cues. Subjects viewed these images, and were asked to name the individual whose photograph had been manipulated to produce them. Afterward, subjects viewed the original photographs and named them. Subjects performed better using pigmentation information alone than shape information alone, but better still when both kinds were available (in the original photographs). These results provide evidence that pigmentation is at least as important as shape for face recognition.