Teresa Canas Bajo and Steven Shepard
Monday, April 5, 2021
11:10 am - 12:30 pm
Individual differences in holistic processing of Mooney faces
Humans are remarkably sensitive to faces for good reason: they are among the most important sources of visual information we encounter. Faces convey critical social and emotional information; they guide social interactions and our everyday behavior. A large number of studies have suggested that our face expertise is due to the fact that we process faces holistically rather than as a set of separate features. Previous work demonstrates that some individuals are better at this holistic type of processing than others. Here, we show that there are unique individual differences in holistic processing of specific Mooney faces. Mooney faces are two-tone black and white blobs that are readily perceived as faces despite lacking low-level face features. Our results show that Mooney faces vary considerably in the extent to which they tap into holistic processing; some Mooney faces require holistic processing more than others. Importantly, there is little between-subject agreement about which faces are processed holistically; specific faces that are processed holistically by one observer are not by other observers. Essentially, what counts as holistic for one person is unique to that particular observer. The origin of those individual differences remains unclear. One hypothesis is that each observer has a unique family of face templates—a template manifold—which is formed over a lifetime of experience. Faces that are similar to an observer’s particular face templates would have an advantage over faces that differ more from the observer’s templates. In the present study, our goal was to test whether such individual differences in face templates exist. To test this hypothesis, we used a reverse correlation approach to measure individual differences in classification images for Mooney faces. We found that classification images were consistent within each observer but were different between observers. Our findings suggest that humans have consistent and unique face templates that could drive idiosyncratic individual differences in face recognition. Knowing these individual templates may be the key insight needed to develop models of individual observer-specific face recognition, affording a way to predict which faces are recognized by which observers.