Interoceptive predictive coding, active inference and the generation of hypnotic suggestion responses

Hypnosis research flourishes when competing theoretical models make testable predictions about significant psychological, physiological and clinical outcomes. Neurocomputational theories are currently unifying theory and research in these diverse domains and are poised to enter the field of hypnosis. Here we explore three current attempts to employ Bayesian predictive coding models to understand the mechanisms used to generate responses to hypnotic suggestions. Predictive coding is shown to be a framework able to generate multiple theorical explanations with testable predictions and distinct consequences for the growth of knowledge and clinical practice. Current theories of dissociated control, absorption and response expectancy can be modelled within this unifying framework. The unique Bayesian mechanism of Active Inference plays a central role in each of the accounts compared and testable neurophysiological predictions are drawn from each model. This provides a foundation for future research programs and successful grant funding to drive the next wave in the understanding of hypnotic suggestion, hypnotic responses and closely related social and clinical phenomena.