Understanding of the mechanisms of mindfulness in the perspective of the predictive coding model

Abstract

Mindfulness affects a variety of psychological problems, and many psychological concepts have been introduced to explain the mechanism of these effects. However, a unified understanding has not been attained. In this paper, we discuss the mechanism of mindfulness from the perspective of the predictive coding model, which has been attracting attention in recent years as a potential unified theory to explain various brain functions. Based on our review, we found that mindfulness is characterized by (1) predominantly minimizing prediction errors via perceptual inference, which updates predictions based on sensations rather than by active inference (which attempts to modify sensations to match predictions); (2) increasing the precision of sensory input by attention, and (3) decreasing the precision of prior predictions by not weighting past experiences. These features may lead to rapid minimization of prediction error and an accurate understanding of the inner and outer environment. By redefining mindfulness in the predictive coding model, we discuss the possibility of understanding the mechanisms of mindfulness from a unified perspective that explains human perception, behavior, cognition, and emotion.