Neural processes underlaying music reward: unravelling effects of implicit probabilistic knowledge
Abstract
Humans can feel and experience pleasure from a wide variety of stimuli or events. These stimuli encompass a spectrum ranging from primary rewards that satisfy fundamental biological drives, such as food or sex, to secondary rewards that facilitate access to other rewards, such as money, to finally aesthetic experiences, like listening to music. While music does not fulfil the Darwinian imperatives of survival, prior neuroimaging studies have shown that it engages a common reward network than the one observed for primary rewards. These shared neural mechanisms underlying pleasure responses have become a subject of increasing interest in the last years. Indeed, some influential accounts have suggested that music enjoyment is driven by the confirmation and violation of predictions based on prior knowledge. These relationship between pleasure and complexity follows an inverted U-shaped relationship, with moderately complex excerpts being the most preferred. These moderately predictable melodies provide a balance between the acquisition of novel information through surprising transitions, and the confirmation of predictions based on prior knowledge.
The main goal of the present thesis was to assess the neural mechanisms underpinning music perceptual and reward processing, with a particular emphasis on the neural and oscillatory mechanisms that underlie the inverted U-shaped relationship between pleasure and complexity. Prior neuroimaging studies have shown that music listening engages a wide range of neural structures, encompassing both perceptual and reward system regions. However, to date, no study has shown a direct link between the widely acknowledged behavioural inverted U-shaped responses between pleasure and predictability, and the perceptual and reward mechanisms underlying it. Moreover, despite recent evidence showing that low-frequency oscillatory mechanisms, such as theta band, play a crucial role in pleasure and prediction error encoding, no prior study has explored the role of this frequency mechanisms in the inverted U-shaped relationship between pleasure and complexity. To fill this knowledge gap, we designed three studies. In Study 1, a functional localizer was designed with the aim of identifying perceptual and reward regions engaged during music listening at the individual level. This methodological approach enables the overcoming of some of the limitations of group- level analysis, by increasing sensitivity and functional resolution of the analysis. Indeed, by performing analysis at individual level, functional localizers, as well as the associated functional probabilistic atlases, provide a useful tool to facilitate the interpretation and integration of results. In addition, the use of 7T-fMRI enables to increase the funcional resolution of the acquired images, thereby facilitating the identification of subcortical auditory and reward structures that are more challenging to localize using lower functional resolution image acquisition techniques (3T-fMRI). Results suggest that our functional localizer is a reliable and robust tool for identifying the entire auditory pathway regions at individual level. In addition, our findings suggest that this experimental paradigm was also useful to localize cortical and subcortical reward network regions sensitive to music processing. Taken together, our functional localizer may emerge as a useful methodological tool for identifying perceptual and reward regions at individual level, enabling to constraint the analysis to these functional regions of interest. In addition, in order to overcome temporal and economic constraints of performing an independent localizer we also generated a functional probabilistic atlas which contain the prevalence of consistent activations between participants. We expect that both the functional localizer and the probabilistic atlas will facilitate the acquisition of cumulative knowledge, thereby providing novel insights into the regions involved in music processing.
In Study 2 and Study 3 we aimed at studying the relationship between pleasure and complexity. Prior studies have consistently showed an inverted U-shaped relationship between music-induced hedonic response and predictability. In this sense, using artificially generated excerpts based on Western stylistic regularities, in both experiments we explored at behavioural level the relationship between pleasure and complexity. Results confirmed that our experimental paradigm yielded consistent findings, with both studies showing an inverted U-shaped relationship between pleasure and predictability. In addition, in study 2 we explored the neural mechanisms underlying this relationship. Results revealed that both perceptual and reward regions displayed an activity patern resembling an inverted U-shaped distribution, with significant increased activity for melodies of moderately complexity. In addition, the right Nacc, which has been previously shown to be a main reward hub, was the only structure showing that predictability fully mediates the relationship between pleasure and activity in this region. Moreover, in study 3 was assessed the oscillatory mechanisms underpinning the inverted U-shaped relationship, focusing on theta frequency band (4-7 Hz). This oscillatory mechanism has been previously related to hedonic responses and prediction error encoding, suggesting that it may be a potential coordination mechanism for the interaction between distant perceptual and reward regions involved in music pleasure. Results showed that right fronto-temporal theta phase synchronization mechanisms exhibited a significant increase in response to moderately complex melodies. Taken together, these results suggest that both neural and low-frequency mechanisms exhibited an inverted U-shaped pattern, with increased activity for those excerpts that provide an equilibrium between surprises and confirmation of predictions based on prior knowledge.
Overall, the present thesis provides novel insights into the relationship between pleasure and predictability. This is achieved through two studies that assess both neural, perceptual and reward mechanisms involved in this relationship, as well as the oscillatory mechanisms that may support this interaction. Furthermore, it presents a methodological tool that may facilitate future studies in exploring this relationship by enhancing the functional resolution and sensitivity of their analysis or by employing the funcional probabilistic atlas as a reference to interpret their results and integrate them with existint evidence.