2021-06-01 - 2023-05-31 | Research area: Cognition and Sociality
Predictive Processing (PP) is a cognitive framework based on the principle that sentient systems (and specifically nervous systems) strive to anticipate homeostatic imperatives by minimizing prediction errors in immediate action-perception loops and detached, abstract cognition. PP provides the tools and language to describe how generative models or beliefs are constructed and dynamically handled through prediction error minimization. Overweighted high-level generative models or beliefs are increasingly recognized in neuropsychiatric disorders, underpinning maladaptive patterns and habits of movement, behavior, and thought. Understanding which factors might catalyze the flexible revision of generative models is relevant not only for neuropsychiatric disorders but also for maladaptive aggregate human behavior resting on overweighted socially and culturally held beliefs. This aggregate-level likely has emergent properties beyond the sum of individual preferences. The project outlined here aims to explore how PP can help understand and overcome rigid, stereotyped behavioral patterns resulting from overweighted generative models implemented in different, evolutionarily nested levels of cognition. Firstly, existing evidence will be recapitulated and synthesized regarding how generative models are structurally implemented at different levels of cognition, including extended/group/distributed cognition, and how they determine deliberate and automated behavioral trajectories. Secondly, potential parallels, analogies, and metaphors will be explored to understand maladaptive models at supra-individual cognitive levels. This approach will be positioned into a long tradition of extrapolating and understanding the interdependencies between individual and public health and how they interact with the environment. Accordingly, PP might become applicable for sustainable aggregate human behavior on a planet that is increasingly transformed by particular dominant, culturally held belief systems. Such an approach might help to revise conventional beliefs about human nature, from maximizing reward to maximizing feedback-driven adaptive fitness within ecological niches. Finally, specific novel experimental methodologies will be explored that are suitable to investigate belief revision.