Neural dynamics of voluntary decision-making

Abstract

As we navigate the world, we prepare to make voluntary decisions in relation to the opportunities afforded by the environment. An essential aspect of these decisions is that they are experienced with a sense of free will. In both philosophical and layperson accounts of free will, the "ability to choose otherwise" serves a central role and, further, there is evidence that individuals preferentially choose options that maximise the availability of choice. The underlying reason for this widespread emphasis on choice has been difficult to reconcile with traditional views of free will and decision-making.Embodied theories of decision-making propose that individuals are driven to be in states where a greater number of options are available because this maximises their perceived ability to influence the environment in future states. This view places critical emphasis on the period of time that precedes a decision, during which individuals proactively prepare for the upcoming decision. On this basis, a decision may be experienced as free if the individual believes himself to have multiple potential options available. Studies have shown the ability to predict free choices but little is known about the role potential options plays during the voluntary decision-making process.In my thesis, I investigated the proactive neural processes of voluntary decisionmaking in dynamic environments. In particular, I aimed to test the hypothesis that free decisions are associated with a proactive state in which multiple potential choices are considered. To do so, I carried out a set of three studies where I examined behavioural and functional magnetic resonance imaging (fMRI) data of participants while they performed free decisions in virtual environments.In Chapter 2, I designed an fMRI study to investigate the neural patterns of activity preceding a free decision. I found that frontoparietal and salience networks were more active for free decisions than instructed decisions during an initial period of proactive choice selection. Using multi-voxel pattern analysis (MVPA), I showed that upcoming choices could be predicted from both the motor and visual cortices once participants were able to match their planned colour choice to its spatial location. Notably, the decoding accuracy for free decisions was lower than instructed decisions, indicating less reliable choice representation.This corresponded to eye fixation results showing greater visual attention to the alternative options in the case of free decisions, together indicating that choice consideration may be an ongoing process when individuals freely prepare decisions. iiiIn Chapter 3, I tested whether the state of having alternative options was reflected in neural representations of upcoming choices and, further, if this translated to differences in the subjective experience of free choice. In this fMRI study, I found that the availability of an alternative option could be decoded from patterns of activity in the dorsolateral prefrontal cortex. Moreover, decoding accuracy ...