Between hallucination and reality: Predictive processing and the problem of consciousness

Abstract

The predictive processing theory (PPT) is one of the most influential contemporary approaches in cognitive science, with the ambition not only to explain perception and behavior but also to account for the fundamental properties of consciousness. Grounded in Bayesian inference PPT portrays the brain as a hierarchically organized system that continuously generates and updates predictions about both the external and internal world. This paper examines the theoretical foundations of PPT, and analyzes the explanatory scope of the theory in the context of consciousness. Special attention is given to the problem of distinguishing between conscious and unconscious processing. Two leading hypotheses are explored: consciousness as a perceptual hypothesis updated by new sensory information, and consciousness as the hypothesis with the highest posterior probability. Through the analysis of empirical phenomena such as bistable perception, hallucinations, and blindsight, the paper shows that PPT may offer necessary conditions for the emergence of consciousness, but the question remains whether these conditions are also sufficient. It is concluded that while PPT provides a robust functional framework for understanding conscious perception, it must further clarify the phenomenal aspect of conscious experience in order to overcome the existing limitations in explaining consciousness.