Abstract
A plain, blank canvas doesn’t look very beautiful; to make it aesthetically appealing requires adding structure and complexity. But how much structure is best? In other words, what is the relationship between beauty and complexity? It has long been hypothesized that complexity and beauty meet at a “sweet spot”, such that the most beautiful images are neither too simple nor too complex. Here, we explore this connection experimentally, by taking advantage of an information-theoretic approach to shape representation. We algorithmically generated a library of smooth 2D polygons, and determined their complexity by computing the cumulative surprisal of their internal skeletal structure — essentially quantifying the amount of information in the object. We then stylized these shapes as “paintings” by rendering them with artistic strokes, and “mounted” them on framed canvases hung in a virtual room. Subjects were shown pairs of these mounted shapes (which varied in their skeletal complexity), and were asked to choose which shape looked best in a given room (by previewing how each painting appeared on the wall). Remarkably, subjects preferred paintings that were neither too simple or too complex, such that moderately complex shapes were chosen as the most attractive paintings. Follow-up experiments generalized this result to a variety of different rooms (such as a kitchen, exhibit hall, or bedroom) and canvases (including different sizes and aspect-ratios). These results suggest a quadratic relationship between aesthetics and visual complexity, and demonstrate the utility of information theory for exploring for exploring surprisingly high-level aspects of our visual experience.