Abstract
We feel unusual grotesqueness and distortion when seeing several faces rapidly presented in peripheral vision (FFDE). Here, to reveal the effect of facial parts, we investigate the intensity of FFDE using partially presented faces. In the experiment, a series of face pictures were presented for 10 seconds while switching pictures every 250 ms. We presented 5 conditions: whole-face, upper-half, eye-alone, eye-removed, and contourless-face. Participants were asked to keep pressing a button when they felt distortion of faces and to rate the intensity of distortion and uncanniness after stimulus presentation. We found that the intensity of FFDE depended on the presented facial parts: whole-face > contourless-face > upper-half > eye-removed > eye-alone. The FFDE latency also differed by condition, though it diverged from the order of intensity rating. These results suggest that holistic face processing is involved in FFDE and that there are different determinants of intensity and latency of FFDE.
