Abstract
Previous research has shown that infants demonstrate both featural and configural processing of upright faces. However, these studies used face images with easily identifiable features and parts. Bottom-up image segmentation and feature-based strategies could therefore contribute to the perception of these faces, in addition to configural processes. The purpose of the present experiment was to provide a direct test of holistic face processing in infants using Mooney face stimuli (Mooney, 1957). Mooney photographs are two-toned faces (including cast shadow information) that lack individual features and therefore can only be recognized as a face using holistic processing. Bottom-up image segmentation and feature-based parsing operations cannot operate on Mooney faces; to find any facial feature, such as an eye or a nose, one must first identify the image as a face. On each trial, two identical Mooney faces were presented side by side at 2 degrees eccentricity from fixation, one in the upright and one in the inverted orientation, for two seconds. If infants process upright faces holistically, they should preferentially look at the upright rather than the inverted Mooney face. Duration of looking toward the upright Mooney face was divided by the total duration of looking toward both face patterns to obtain a visual preference score. The results revealed that infants preferred the upright Mooney face significantly greater than the chance level, suggesting that infants are able to use purely holistic processing to identify upright faces.