Commonsense Aesthetics

Abstract

Aesthetics is an old enigmatic topic in psychology. Here I focus on the beauty of every impression with an emphasis on the visual domain. My approach incorporates new tools (e.g., procedural generation, shape-families, webGL, virtual reality, 3D printing) and multivariate statistics (e.g., mixed-effects regression, PCA) to discover new insights. The purpose of my dissertation is to encourage further research elucidating the influence of shapes in the formation of shared impressions.

The first project focuses on aesthetics in terms of preferences¬–specifically, people's tendency to prefer curvier and smoother shapes. This aesthetic bias is one of the oldest phenomena studied in psychology. My contribution to the existing literature links the aesthetic bias to an objective definition for curvier and smoother shapes. The results suggest that people are sensitive to subtle objective differences in shapes that of which can affect preferences.

The latter two projects focus on aesthetics beyond preferences and more broadly in terms of any impression (e.g., artificial, dangerous, beautiful). I show that impressions of novel objects constitute reliable phenomena that could be further understood by studying shapes in more ways than traditional explored (e.g., curviness). I test stimuli consisting of sets of novel objects maintaining mathematically defined relationships called shape-families because object within them are similar and yet different. Using shape-families, I describe the reliability of object impressions and explore the range of those that may be universally shared.

This dissertation represents a new look at aesthetics and shapes. For example, the prevailing focus of aesthetics today is on preferences or the experience of art. However, this work builds towards a focus on aesthetics in terms of agreement and focuses on a definition of aesthetics that may be generalizable to everyday experiences. Also, when compared to the dominant focus in vision research on studying shapres as it relates to object recognition, this work focuses on impressions. The cumulative results presented in this dissertation demonstrate evidence that there is meaning among the shapes of novel objects.