On the inference of agency in operant action : an examination of the cognitive and neural underpinnings in health and schizophrenia

This dissertation elucidates cognitive and neural underpinnings of the sense of agency, which is the feeling that we are in control of our actions and the subsequent consequences. This consciously accessible sensation of control is pervasive, sometimes subtle, and can even be illusory in nature. Furthermore, the experience of self-agency is socially well-shared and considered to be fundamental to human social functioning. It helps people makes sense of the physical and social world around them, to guide and direct them in their goal-directed actions, and to make sense of who does what when they act and react in a social contact. Understanding the roots and occurrence of the experience of agency, then, is an important target in recent psychological and neuroscientific research. The present dissertation aims to contribute to this endeavor by examining how humans produce the conscious experience of agency. The experience of agency over the consequences of actions can be derived from an inference that we make after we see an outcome. Indeed, agency can be inferred when a specific outcome is matched by pre-activated information that is related to the outcome – for instance, if I had the goal to cause a traffic light to turn green by pure force of thought, and it immediately turns green, I will quickly infer and experience myself as the cause of the green light (the subsequent realization that we could not have caused this does not detract from the initial experience). In self-agency inferences, pre-activated information can either be goal-based, when people engage in goal-oriented behavior, and prime-based, where information can be subtly pre-activated in people’s minds by our environment. The present dissertation provides evidence that goal-based and prime-based agency inferences arise from different cognitive mechanisms: whereas goals use attentional control (i.e., working memory) in order to encode, maintain, retrieve and compare information about goals, prime-based agency inferences follow from automatic cognitive accessibility processes that follow the principles of automatic spreading of activation. A second line of research in this dissertation is the first exploration of the neural underpinnings of self-agency inferences. In multiple studies, it has been shown that parietal and frontal brain regions are involved in goal-based self-agency experiences. This observation, together with other research implicates a brain model where a match between a goal and the subsequent outcome is registered in the parietal lobe, and transmitted to the frontal lobe where conscious self-agency experiences are generated. Finally, I investigate the mechanisms underlying self-agency inferences by studying patients with schizophrenia, who as part of their disease often exhibit agency problems, leading to hallucinations and delusions. For these patients, it was found that they report similar goal-based agency experiences as healthy controls, but that they do not use the same brain areas, leading to the notion that agency processing may be less efficient in these patients. Indeed, when patients with schizophrenia only have primes as input for their agency experiences, they appear unable to do so. Future research may elucidate the specific dysfunctioning mechanisms in patients with schizophrenia