Compositional Geometry in Pictures

Abstract

The aesthetic effect of pictures has been suggested to depend in part upon the existence of an implicit or explicit geometrical basis to their composition. In Experiment 1, subjects identified the significant points which might form the basis for such a geometry. In Experiment 2, a different group of subjects expressed preferences either for intact or cut versions of the pictures used in Experiment 1, and for sets of dots based upon the significant points within those cut and uncut pictures. Although subjects showed an overall preference for uncut rather than cut stimuli (in which it was presumed that cutting would have destroyed much of the compositional geometry), both for pictures and for dot stimuli, there was no correlation between the judgments of pictures and dot patterns, either between picture or between subjects, suggesting that compositional geometry was not of aesthetic significance in preference judgments. That conclusion was reinforced by Experiment 3, in which subjects showed no evidence of a preference for synthetic stimuli which were produced so that they had significant geometrical structure.