Abstract
Research on experimental aesthetics has been conducted since the nineteenth century. Interestingly, however, few studies have examined the perceived beauty of naturally-shaped objects. In the first experiment, 204 observers were presented with a set of ten snowflake silhouettes that varied in complexity (perimeter relative to area); they were similarly presented with ten randomly-shaped, computer-generated, solid objects that also varied in complexity. For each stimulus set, the observers selected the single snowflake or solid object that was the most beautiful (Fechner's method of choice). The results for the solid objects replicated the findings of earlier research: the most and least complex objects were chosen as the most beautiful. Moderately complex objects were rarely selected. The results for the snowflakes were different. For these visual stimuli, the least complex snowflakes were almost never chosen; only the complex snowflakes were perceived as being most beautiful, with the aesthetic preference increasing with increases in complexity. The distributions of aesthetic preference for snowflakes and solid objects were significantly different (x2(9)= 83.1, p < .000001). The second experiment assessed the perceptual validity of the measures of stimulus complexity used in Experiment 1. Observers viewed the same snowflake or solid object images that were used in Experiment 1 and rated their complexity using a scale from 1 (least complex) to 10 (most complex). The resulting average Pearson r correlation coefficients relating objective and perceived complexity were 0.834 and 0.933 for the snowflakes and solid objects, respectively. Taken together, the results of both experiments demonstrate that there is a systematic and consistent relationship between the human visual perception of beauty and objective measures of stimulus complexity.
Meeting abstract presented at VSS 2016