Self Relevance and Aesthetic Appeal

While a degree of aesthetic appeal can be predicted from image features, a large proportion of variance in aesthetic ratings differs from person to person, particularly for visual art. We hypothesize that this is due to the ability of visual art to speak to a person depending on their lived experience. In a series of behavioral and brain imaging experiments, we are exploring the role of self-relevance, the degree to which an artwork resonates with one’s self-construct, in predicting aesthetic appeal. We are using artificial intelligence algorithms to generate custom artworks for individual participants, allowing us to manipulate the degree to which an artwork contains self-relevant content, independent of stimulus properties or artistic style.