Concreteness of semantic interpretations of abstract and representational artworks

Abstract

The authors tested two contrasting theoretical predictions to establish whether semantic interpretations of abstract artworks had different lexical concreteness from those of representational artworks. In Experiment 1, 49 non-expert participants provided brief verbal interpretations of 20 abstract and 20 representational artworks. Frequentist and Bayesian Linear Mixed Models showed that the words' concreteness levels were robustly higher for interpretations of abstract artworks than representational artworks. This difference was present regardless of the inclusion or exclusion of function words. Potential diluting or inflating impacts on the effect due to the multi-word responses were examined in Experiment 2, in which 72 new participants provided single-word interpretations for the same artworks. The effect replicated with a larger effect size. The findings suggest that non-expert viewers prioritise establishing what is depicted over seeking deeper meanings if the depicted is not readily established perceptually. The findings are incompatible with the theoretical stance that abstract art has abstract meaning. Instead, the findings are consistent with complex models of aesthetic processing in which meaning may emerge in stages. The effect of art type on the concreteness of meaning is an important, hitherto undiscovered basic finding in empirical aesthetics. Our novel methods enable further research in this field.