Predictive Processing

Predictive processing enters the depth-psychology corpus not as a peripheral technical curiosity but as an organizing framework capable of reinterpreting core psychodynamic constructs — perception, emotion, the unconscious, and the therapeutic relationship — in terms of hierarchical generative models and Bayesian inference. The most sustained engagement comes from Lisa Feldman Barrett, whose constructionist account of emotion depends entirely on the thesis that the brain is fundamentally a prediction machine, continuously generating and refining simulations against incoming sensory data. Daniel Siegel extends this logic developmentally, arguing that early relational experience sculpts precisely those predictive models that subsequently filter all perception and emotional life. Hugh McGovern presses the framework into explicitly Jungian territory, proposing that archetypal images emerge through hierarchically structured generative models whose eigenmodes are selectively liberated by psychedelic compounds. The insular cortex literature — Hassanpour, Paulus — finds predictive processing indispensable for understanding interoceptive signalling and its disruption in psychiatric conditions, particularly addiction. Wolfram Schultz’s dopaminergic reward-prediction-error research provides the subcortical substrate upon which much of this edifice rests. Key tensions involve the adequacy of Bayesian mechanics to capture the symbolic and imaginal richness of depth-psychological phenomena, and whether minimizing prediction error constitutes a genuine account of meaning or merely a formal redescription of it.

In the library 20 passages

the brain runs internal models that function as Bayesian filters for incoming sensory input, driving action and constructing perception and other psychological phenomena. This hypothesis is often called predictive coding.

Siegel synthesizes the neuroscientific consensus that Bayesian predictive coding — wherein top-down predictions are continuously compared to bottom-up sensory signals — constitutes the brain’s fundamental operating principle, with direct implications for how relationships shape psychological development.

Siegel, Daniel J., The Developing Mind: How Relationships and the Brain Interact to Shape Who We Are, 2020thesis

Dig deeper with Sebastian →

your experience right now was predicted by your brain a moment ago. Prediction is such a fundamental activity of the human brain that some scientists consider it the brain’s primary mode of operation.

Barrett advances the central claim of her constructionist emotion theory: the brain’s primary function is prediction, not reactive stimulus-processing, and all experience — including emotion — is a predicted construction.

Barrett, Lisa Feldman, How Emotions Are Made: The Secret Life of the Brain, 2017thesis

Dig deeper with Sebastian →

A prediction originates as a multisensory summary, representing the goal of the concept, in a portion of the interoceptive network known as the default mode network.

Barrett locates the origin of predictive processing within the default mode network, linking conceptual generation to interoceptive prediction and providing a neural architecture for the concept cascade that underlies emotion construction.

Barrett, Lisa Feldman, How Emotions Are Made: The Secret Life of the Brain, 2017thesis

Dig deeper with Sebastian →

The interoceptive network issues predictions about your body, tests the resulting simulations against sensory input from your body, and updates your brain’s model of your body in the world.

Barrett specifies that interoception itself operates through predictive processing, with dedicated brain regions continuously anticipating and correcting bodily states — a mechanism fundamental to her theory of how feelings and emotions are made.

Barrett, Lisa Feldman, How Emotions Are Made: The Secret Life of the Brain, 2017thesis

Dig deeper with Sebastian →

the manifestation of these simulations occur via the mechanisms of hierarchical generative models. These representational systems are experientially dependent and emerge due to a complex interplay between innate cognitive phenotypes and developmentally contingent environmental feedback.

McGovern argues that Jungian archetypal simulations are instantiated through hierarchical generative models consistent with predictive processing, linking Bayesian model selection and free energy minimization to the emergence of archetypal representations.

McGovern, Hugh, Eigenmodes of the Deep Unconscious: The Neuropsychology of Jungian Archetypes and Psychedelic Experience, 2025thesis

Dig deeper with Sebastian →

archetypes emerge because of recurrent evolutionary pressures, and associated patterned sensory encounters and behavioral practices, through human phylogeny. The Free Energy Principle and predictive processing

McGovern frames the entire neuropsychological account of Jungian archetypes within the Free Energy Principle and predictive processing, positioning these frameworks as the mechanistic bridge between evolutionary pressures and archetypal phenomenology.

McGovern, Hugh, Eigenmodes of the Deep Unconscious: The Neuropsychology of Jungian Archetypes and Psychedelic Experience, 2025thesis

Dig deeper with Sebastian →

activation in more anterior portions of the insula during the anticipatory period, before the onset of isoproterenol action, consistent with a predictive processing role.

Hassanpour identifies a functional dissociation within the insula whereby posterior regions handle afferent sensory processing while anterior regions perform predictive processing, providing direct neuroimaging evidence for the EPIC model’s hierarchical interoceptive architecture.

Hassanpour, Mahlega S, The Insular Cortex Dynamically Maps Changes in Cardiorespiratory Interoceptionthesis

Dig deeper with Sebastian →

the dysgranular mid and granular/hypergranular posterior insula, regions that both have the requisite cytoarchitectonic structure to compare afferent interoceptive signals arriving via the thalamus with interoceptive prediction signals arriving via the agranular anterior insula.

Hassanpour provides cytoarchitectonic support for the predictive coding account of interoception, specifying that the laminar structure of mid and posterior insula enables the comparison of prediction signals against ascending interoceptive afferents.

Hassanpour, Mahlega S, The Insular Cortex Dynamically Maps Changes in Cardiorespiratory Interoceptionsupporting

Dig deeper with Sebastian →

Prediction errors aren’t problems. They’re a normal part of the operating instructions of your brain as it takes in sensory input. Without prediction error, life would be a yawning bore. Nothing would be surprising or novel, and therefore your brain would never learn anything new.

Barrett rehabilitates prediction error as a necessary engine of learning and novelty rather than a system malfunction, elaborating the functional logic of predictive processing for lay and clinical audiences.

Barrett, Lisa Feldman, How Emotions Are Made: The Secret Life of the Brain, 2017supporting

Dig deeper with Sebastian →

The brain in general is a predictive organ, preparing for what comes next. The ‘social brain’ is really the predictive brain, which develops as a function of social experience aimed at allostasis regulation.

Siegel, drawing on Atzil and colleagues, argues that the social brain is fundamentally a predictive brain shaped by attachment relationships for allostatic regulation, integrating predictive processing with relational developmental theory.

Siegel, Daniel J., The Developing Mind: How Relationships and the Brain Interact to Shape Who We Are, 2020supporting

Dig deeper with Sebastian →

‘Reward prediction error’ then means the difference between the reward I get and the reward that was predicted. Once I get the same can again and again for the same button press, I get no more surprises; there is no prediction error, I don’t change my behavior, and thus I learn nothing more.

Schultz provides the canonical dopaminergic account of reward prediction error, establishing the subcortical neurochemical substrate that underlies broader predictive processing frameworks and reinforcement learning models.

Schultz, Wolfram, Dopamine reward prediction error coding, 2016thesis

Dig deeper with Sebastian →

Processing of prediction errors rather than full information about an environmental event saves neuronal information processing and, in the case of rewards, excites neurons with larger-than-predicted rewards.

Schultz demonstrates that the dopaminergic system encodes prediction errors rather than absolute reward values, providing the biological efficiency argument central to understanding how predictive processing minimizes computational load.

Schultz, Wolfram, Dopamine reward prediction error coding, 2016supporting

Dig deeper with Sebastian →

our brain continually predicts upcoming events based on implicit knowledge. Dopamine signals violations of expectations, or prediction errors, driving learning to update expectations.

Schoeller applies predictive processing to aesthetic experience, arguing that the pleasure of chills and musical engagement arises from the rewarding resolution of prediction errors, linking dopaminergic signalling to aesthetic surprise.

Schoeller, Felix, The neurobiology of aesthetic chills: How bodily sensations shape emotional experiences, 2024supporting

Dig deeper with Sebastian →

the relinquishment of typical top–down, hierarchical causation in the brain and the liberation of bottom-up flow may mean that phenomenological (archetypal) visions can shimmer into perceptual awareness.

McGovern proposes that psychedelics disrupt the normal top-down predictive hierarchy, allowing suppressed archetypal generative models to surface into awareness — an application of predictive processing to the phenomenology of psychedelic states.

McGovern, Hugh, Eigenmodes of the Deep Unconscious: The Neuropsychology of Jungian Archetypes and Psychedelic Experience, 2025supporting

Dig deeper with Sebastian →

the prediction might unpack into details of Uncle Kevin’s appearance. These details are themselves predictions based on probabilities, so your brain can compare the simulation to actual sensory input and compute and resolve any prediction error.

Barrett illustrates the cascading, hierarchical structure of predictive processing through the unfolding of conceptual predictions from multisensory summaries down to primary sensory cortex, elaborating the mechanics of the concept cascade.

Barrett, Lisa Feldman, How Emotions Are Made: The Secret Life of the Brain, 2017supporting

Dig deeper with Sebastian →

The enclosed particles behave as if inferring, using Bayesian logic, the nature of the changing environment which lies outside the boundary. Friston suggests these internal states and their Markov Blankets are models of prototypical biological systems.

Mizen situates Friston’s Free Energy Principle — the theoretical foundation of predictive processing — within a psychodynamic account of self and soma, using the Markov Blanket formalism to describe how biological systems maintain internal equilibrium through Bayesian inference.

Mizen, C. Susan, The Self and alien self in psyche and somasupporting

Dig deeper with Sebastian →

That future unit will contain expected feelings, that is, predictions made by internal models of the world and of others’ behavior based on heredity and experience. If the present conditions do not match the expected conditions, the dissonance can immediately produce increased salience.

Craig situates predictive processing within his homeostatic interoceptive model, showing how internal models generate affective predictions and how mismatches produce the salience signals that drive attention and learning.

Craig, A.D. Bud, How Do You Feel? An Interoceptive Moment with Your Neurobiological Self, 2014supporting

Dig deeper with Sebastian →

Mental models are basic components of implicit memory. Our minds use mental models of the world in order to assess a situation more rapidly and to determine what the next moment in time is most likely to offer.

Siegel introduces the concept of mental models as predictive schemas encoded in implicit memory, prefiguring the formal predictive processing account and linking it to attachment-based developmental theory.

Siegel, Daniel J., The Developing Mind: How Relationships and the Brain Interact to Shape Who We Are, 2020supporting

Dig deeper with Sebastian →

The concept is constructed in the moment, ad hoc. And among these myriad instances, one of them will be the most similar (by pattern matching) to Sophia’s current situation. That’s what we’ve been calling the ‘winning instance.’

Barrett’s account of concept construction as on-the-fly pattern matching across a population of instances is an applied elaboration of predictive processing, illustrating how the brain selects the best-fitting prior model for incoming sensory data.

Barrett, Lisa Feldman, How Emotions Are Made: The Secret Life of the Brain, 2017supporting

Dig deeper with Sebastian →

Related terms