Reader Orientation — Predictive Processing Extension
For: Absorption as Predictive Constraint: A Mechanistic Extension of Predictive Processing for Durable Human Change
How to Read This Paper
This paper is written for readers already familiar with predictive processing, active inference, or hierarchical Bayesian models of cognition. It does not reintroduce those frameworks.
Instead, it addresses a specific limitation within them:
why prediction error so often fails to produce durable updating of higher-level models such as habits, beliefs, and identity.
Readers should approach this paper as a theory-level extension, not as a proposal for a new computational architecture, state of consciousness, or applied method.
The Problem This Paper Addresses
Predictive processing successfully explains how cognition operates as inference. However, it remains underspecified with respect to a persistent empirical observation:
Prediction error is frequently generated without producing lasting model revision.
Across learning, psychotherapy, and behavior change:
– disconfirming evidence is often acknowledged without updating
– insight occurs without behavioral change
– effortful practice fails to consolidate
Conversely, relatively modest experiences sometimes produce disproportionate and enduring effects.
This paper addresses that gap.
Core Contribution
The central claim of this paper is that:
Absorption functions as a predictive constraint that modulates when experience is treated as evidence capable of revising higher-level predictive models.
Absorption is not introduced as:
– a novel state of consciousness
– a phenomenological depth variable
– a new computational primitive
It is framed as a constraint on inference that:
– stabilizes a dominant predictive stream
– reduces competition among alternative interpretations
– increases the likelihood that prediction error is integrated rather than dismissed
This alters how existing predictive processes operate, not what those processes are.
What This Paper Claims
Specifically, the paper argues that:
– Absorption attenuates competition among predictive models
– Prediction error under absorbed configurations is more likely to revise higher-level models
– Variability in learning outcomes reflects differences in absorptive constraint, not just error magnitude or precision
– Repetition under absorbed conditions produces more durable updating than isolated high-intensity experiences
– Individual differences affect rate of updating, not the ceiling of possible change
What This Paper Does Not Claim
This paper does not propose:
– a replacement for predictive processing
– a new neural mechanism
– a therapeutic protocol
– a persuasion or influence framework
Absorption is presented as a necessary enabling condition in many cases, not a sufficient cause of change.
Relationship to the UAM Core Paper
The two papers are complementary:
– The UAM Core Paper establishes the general mechanistic framework
– This paper shows how absorption can be formally integrated into predictive processing without violating its core assumptions
Together, they:
– establish the model
– anchor it within contemporary cognitive theory
How to Interpret the Predictions
The predictions in this paper are discriminative, not merely confirmatory.
They are designed to distinguish the absorption-as-constraint account from:
– expectancy-based explanations
– intensity-based explanations
– effort-based explanations
Failure to observe the predicted moderation effects would meaningfully challenge the proposal.
Why This Paper Matters
If predictive processing is to function as a comprehensive account of learning and adaptation, it must explain not only how prediction error is computed, but when it leads to durable revision.
This paper proposes absorption as that missing condition.
Recent Comments