The center cannot hold: Variations of frame width help to explain the "inward bias" in aesthetic preferences.

Although most visual aesthetic preferences are likely driven by a mix of personal, historical, and cultural factors, there are exceptions: some may be driven by adaptive mechanisms of visual processing, and so may be relatively consistent across people, contexts, and time. An especially powerful example is the "inward bias": when a framed image contains a figure (e.g., a face in profile), people prefer arrangements in which the figure faces inward. Although the inward bias has been studied in many contexts, its underlying nature remains unclear. It may be a function of an image's center (as in the "affordance-space" account, in which people prefer to center the implied functional extensions of objects), or it may be a function of the frame's borders (as in the "looking-into-the-future" account, in which people dislike perspectives on scenes that won't allow them to witness predicted future actions). Here we directly contrast these possibilities using a simple novel manipulation. Observers placed a face (in profile) in a frame to maximize the image's aesthetic appeal, and across observers we varied the frame's aspect ratio. We observed a powerful inward bias, and across frame widths observers preferred an approximately constant positive ratio of space in front versus behind the face. This suggests that the inward bias is driven primarily not by the image's center, but by the frame's borders - and it is consistent with the possibility that certain regions of empty space are prioritized because they are where future actions are predicted to occur.