A Theoretical Framework for How We Learn Aesthetic Values

Abstract

How do we come to like the things that we do? Each one of us starts from a relatively similar state at birth, yet we end up with vastly different sets of aesthetic preferences. These preferences go on to define us both as individuals and as members of our cultures. Therefore, it is important to understand how aesthetic preferences form over our lifetimes. This poses a challenging problem: to understand this process, one must account for the many factors at play in the formation of aesthetic values and how these factors influence each other over time. A general framework based on basic neuroscientific principles that can also account for this process is needed. Here, we present such a framework and illustrate it through a model that accounts for the trajectories of aesthetic values over time. Our framework is inspired by meta-analytic data of neuroimaging studies of aesthetic appraisal. This framework incorporates effects of sensory inputs, rewards, and motivational states. Crucially, each one of these effects is probabilistic. We model their interactions under a reinforcement-learning circuitry. Simulations of this model and mathematical analysis of the framework lead to three main findings. First, different people may develop distinct weighing of aesthetic variables because of individual variability in motivation. Second, individuals from different cultures and environments may develop different aesthetic values because of unique sensory inputs and social rewards. Third, because learning is stochastic, stemming from probabilistic sensory inputs, motivations, and rewards, aesthetic values vary in time. These three theoretical findings account for different lines of empirical research. Through our study, we hope to provide a general and unifying framework for understanding the various aspects involved in the formation of aesthetic values over time.

The Aesthetic Self. The Importance of Aesthetic Taste in Music and Art for Our Perceived Identity

“…Other studies have explored aesthetic preference stability and malleability across different time spans, by focusing, for example, on the stability of preferences for geometric figures ( McManus et al, 2010 ), the impact of sensorimotor experience on affective relation to dance ( Kirsch and Cross, 2018 ), or the relative instability for aesthetic ratings for visual art ( Pugach et al, 2017 ) as well as the stability and strength of preferences for music ( Schäfer, 2016 ). There are also more general models of how sensory inputs, motivations, and reward might influence aesthetic preferences over a life span ( Aleem et al, 2020 ). We are currently conducting follow-up studies exploring more directly the occurrence of significant taste changes by letting participants self-report experienced taste changes in their lives followed by an assessment of whether they judge this to have been a change in their identity.…”

Stochasticity, Nonlinear Value Functions, and Update Rules in Learning Aesthetic Biases

“…The first worth mentioning is that it attempts to minimize regret. This minimization holds for both standard reinforcement learning (Sutton and Barto, 2018) and the version here with motivation gating (Aleem et al, 2020). In addition, for the linear value function, the delta rule tends to optimize the trajectory of the free parameters (Aleem et al, 2020).…”

“…This minimization holds for both standard reinforcement learning (Sutton and Barto, 2018) and the version here with motivation gating (Aleem et al, 2020). In addition, for the linear value function, the delta rule tends to optimize the trajectory of the free parameters (Aleem et al, 2020). However, as we will illustrate in the ''Hypotheses Tested in This Article'' sections, this advantage does not apply in general to nonlinear value functions.…”

“…Several theoretical frameworks for aesthetic values have elements of rewardcircuitry and reinforcement learning. Some of these theories are computational (Martindale, 1984;Schmidhuber, 2010;Van de Cruys and Wagemans, 2011;Aleem et al, 2019Aleem et al, , 2020 and some are not (Biederman and Vessel, 2006;Skov, 2010;Vessel and Rubin, 2010;Chatterjee and Vartanian, 2014). Of the computational theories, the only one considering motivation is that of Aleem et al This is also the only theory studying the temporal evolution of learning.…”

“…A potential problem for reinforcement-learning models for the brain is the linearity of many of the most important mechanisms. For example, the model used by Aleem et al (2020) assumes a linear value function (Sutton and Barto, 2018), that is, a linear relationship between sensory inputs and values. Furthermore, this model makes a linearity assumption for the update rule of the value function.…”