Coherence-Rupture-Regeneration

First Principles

CRR rests on a minimal set of axioms drawn from information geometry, thermodynamics, and process philosophy. Each axiom connects to established results in physics and mathematics. Together they yield parameter-free predictions testable across every domain where systems persist through change.

Axiom I: Coherence

Fisher Information and the Cramér-Rao Bound

The coherence integral C is formally identified with accumulated Fisher information I(θ) about the system's generative model parameters θ. Fisher information measures the curvature of the log-likelihood: how sharply the data distinguish between nearby hypotheses. It is the unique Riemannian metric on statistical manifolds (Čencov's theorem), meaning any theory of inference that respects sufficient statistics must use it.

The Cramér-Rao inequality then states a fundamental limit:

No unbiased estimator can have variance smaller than the inverse of the accumulated Fisher information. This is not a modelling assumption—it is a theorem of mathematical statistics, proven independently by Cramér (1946) and Rao (1945). Ito & Dechant (2020) extended this to stochastic thermodynamics, showing that the Cramér-Rao bound governs the trade-off between current fluctuations and entropy production in irreversible processes far from equilibrium.

CRR's contribution: the bound is not merely approached but saturated. At the moment of rupture, C·Ω = 1 exactly. The system has extracted the maximum information its current configuration permits.

Axiom II: Rupture

The Temporal Markov Blanket

In the FEP, a Markov blanket is a spatial boundary that renders internal states conditionally independent of external states. CRR proposes that the Dirac delta δ(now) serves as the temporal analogue: the boundary between past and future, between coherence and regeneration.

The delta has three properties that make it the unique candidate for a temporal boundary:

• Unit mass: ∫δ(t)dt = 1. The boundary carries exactly one unit of information—no more, no less. This is not adjustable; it is a definitional property of the distribution.
• Scale invariance: δ(at) = δ(t)/|a|. The same topology governs rupture at every temporal scale—a synapse firing (ms), a heartbeat (s), a breath (s), a developmental transition (years), a stellar cycle (Myr). There is no preferred scale.
• Conditional independence: Future states (regeneration) are conditionally independent of past states (coherence) given the present (δ). This is exactly the Markov property, now in time rather than space.

The Dirac delta distributes its unit mass across the boundary between inside (all past states—coherence accumulated within the blanket) and outside (all future states—regeneration beyond the blanket). The present moment is where inside becomes outside; where evidence becomes action; where the accumulated past becomes the reconstructed future.

Axiom III: Regeneration

Ω as Thermodynamic Threshold

Ω = σ² is the system's characteristic variance—the inverse of precision (π = 1/Ω). In thermodynamic terms, Ω sets the free energy scale: the amount of surprise (in nats) that the system can tolerate before its generative model must reorganise. This connects CRR to Jaynes' maximum entropy principle: a system with variance Ω has maximised its entropy subject to the constraint that it maintains coherence up to the threshold 1/Ω.

The regeneration weighting exp(C/Ω) ensures that moments of high coherence contribute most strongly to reconstruction. This is not arbitrary—it is the Boltzmann factor of statistical mechanics, with C playing the role of energy and Ω playing the role of temperature. The most "energetic" (coherent) memories dominate the reconstruction, just as the most energetic microstates dominate thermodynamic averages.

Axiom IV: Unity

The Bound Is Universal

The product C·Ω = 1 holds regardless of what the system is, what it is made of, or at what scale it operates. This universality follows from the Cramér-Rao bound being a theorem of information geometry—it depends only on the structure of statistical inference, not on any particular physics. Wherever there is a system accumulating evidence about its environment with finite variance, C·Ω = 1 defines the moment of necessary transformation.

Framework	Evidence	Variance	Bound	Citation
Statistics	Fisher information I(θ)	Var(θ̂) = σ²	σ²·I(θ) ≥ 1	Cramér (1946); Rao (1945)
Quantum mechanics	Energy E	Time uncertainty Δt	ΔE·Δt ≥ ℏ/2	Heisenberg (1927)
Signal processing	Bandwidth Δf	Duration Δt	Δf·Δt ≥ 1/4π	Gabor (1946)
Thermodynamics	Current J	Entropy production σ	Var(J)·σ ≥ 2⟨J⟩²	Ito & Dechant (2020)
Information geometry	Statistical distance ds²	Fisher-Rao metric g	ds² = gijdθⁱdθʲ	Čencov (1982); Amari & Nagaoka (2000)
CRR	Coherence C	Variance Ω	C·Ω = 1	Saturation of all the above

Theorem: CV = Ω/2

The Derivation

At rupture, the threshold coherence C* satisfies C*·Ω = 1, giving E[C*] = 1/Ω. The Dirac delta, as a temporal Markov blanket, partitions unit mass between past and future. By the symmetry of the boundary (there is no intrinsic asymmetry between what is accumulated and what is reconstructed), each partition receives ½. Therefore:

For the two fundamental symmetry classes:

These predictions are parameter-free—no fitting, no calibration, no free parameters. They have been tested across 100+ systems in 30+ domains. See the full validation at CRR Benchmarks.

Implication: Light at the Boundary

In special relativity, a photon's worldline satisfies ds² = c²dt² − dx² = 0. From the photon's frame (loosely speaking—null worldlines have no rest frame), emission and absorption are the same event. The photon carries information between systems without itself undergoing temporal process. It is pure boundary: the carrier of the Cramér-Rao bound between one system's coherence and another's regeneration.

This connects to the holographic principle: the information content of a volume of spacetime is encoded not in its bulk but on its boundary surface. If CRR's rupture boundary is where information is encoded and transformed, then at the light-like limit the system is its boundary. The speed of light is the speed of information propagation because light is the temporal boundary itself—the universal δ(now) that separates every past from every future.

What CRR Proposes to Add to the Free Energy Principle

The FEP (Friston, 2010; 2019) provides a powerful account of what self-organising systems do: they minimise variational free energy, maintaining themselves within characteristic states via Markov blankets that separate internal from external dynamics. CRR does not compete with this account. It extends it into a domain the FEP has left largely unspecified: the temporal structure of transitions between regimes.

The Temporal Gap in the FEP

The FEP's primary temporal apparatus is generalised coordinates of motion (Friston, 2008)—a vector of higher-order time derivatives (position, velocity, acceleration…) that encodes local trajectory information. This is elegant for continuous dynamics within a regime, but it remains fundamentally local in time: each state depends only on its current generalised coordinates, preserving the Markov property. Biehl, Pollock & Kanai (2021) identified technical difficulties with this formulation; Hesp (2022) noted the need for explicit treatment of spatiotemporal blanket closure.

The FEP's path integral formulation (Friston, 2019) scores the plausibility of entire trajectories, but still does not specify when a system must abandon one regime for another, nor how the transition draws on accumulated history. The FEP tells you that a system at nonequilibrium steady state will look as if it is performing inference. It does not tell you the timing of the inference, or the moment at which the current model is exhausted.

Three Specific Additions

FEP Provides	CRR Adds
Markov blanket: a spatial boundary (internal ⊥ external | blanket)	Dirac delta: a temporal boundary (future ⊥ past | present). The rupture moment δ(now) serves the same conditional-independence role in time that the blanket serves in space.
Dynamics within a regime (VFE minimisation, predictive coding, active inference)	Transitions between regimes: C·Ω = 1 specifies exactly when inference is exhausted and the system must reorganise. This is the Cramér-Rao bound that underlies the FEP's own information geometry, now applied as a stopping condition.
Markovian dynamics: each state depends on the current state (or generalised coordinates of the current state)	Non-Markovian accumulation: C(x,t) = ∫L(x,τ)dτ integrates the full history. The present depends not on the previous state but on the entire accumulated past. Regeneration via exp(C/Ω) weights this history exponentially. This is the formal mechanism for how experience shapes the system beyond what current-state descriptions can capture.

The FEP's precision parameter (inverse variance, π = 1/Ω) maps directly to CRR's Ω. Where the FEP uses precision to weight prediction errors, CRR uses its reciprocal Ω to set the rupture threshold and memory depth. The frameworks share the same information geometry; CRR adds the temporal completion.

Why CRR in 2026?

Several converging problems across the sciences create the conditions under which a temporal process theory becomes not merely interesting but necessary.

The Measurement Problem in Neuroscience

Neuroscience can record neural activity at extraordinary resolution—single spikes, local field potentials, fMRI BOLD signals—but lacks a principled theory of when a neural process is complete. When does a percept "finish"? When has a working memory buffer "filled"? When must a decision be committed? Current approaches use arbitrary time windows or threshold-crossing heuristics. CRR's C·Ω = 1 offers a parameter-free criterion grounded in information geometry: the process is complete when the system has extracted maximum information from its current configuration. The CV predictions (1/(2π) for Z₂; 1/(4π) for SO(2)) have been validated against EEG data across two independent datasets (PhysioNet EEGBCI and MPI-LEMON, N=109).

The Timing of AI-Induced Psychological Rupture

As the AI Safety tab of this guide documents, LLMs create unprecedented conditions for cognitive-somatic dissociation. Understanding when a person's generative model is approaching rupture—and whether that rupture will be integrative or fragmenting—requires a temporal dynamics framework. The FEP explains that the system minimises free energy; CRR specifies the moment at which free energy minimisation is exhausted and the system must reorganise. This is the difference between knowing a bridge will eventually fail under load and knowing when.

Cross-Domain Unification Without Free Parameters

The sciences of 2026 are rich in domain-specific models—DDM for reaction times, HKB for motor coordination, Kuramoto for oscillator synchronisation, Lotka-Volterra for population dynamics—each with its own fitted parameters. CRR claims that the temporal structure of all these systems is governed by the same equation (C·Ω = 1) with the same parameter-free predictions (CV = Ω/2). This has been tested across 100+ systems. If CRR is correct, these are not analogies; they are instances of a universal grammar. If CRR is wrong, the CV predictions will fail—and the deviations will be informative about what the true temporal grammar must look like.

The Non-Markovian Gap in Process Theories

Both the FEP and mainstream stochastic thermodynamics rely on Markov assumptions: the future depends on the present alone. Yet biological and psychological systems are manifestly non-Markovian—trauma shapes responses decades later; developmental history constrains adult cognition; evolutionary memory sculpts phenotypes across geological time. CRR's coherence integral C = ∫L(x,τ)dτ provides a minimal formal mechanism for history-dependence: the system carries its past as accumulated Fisher information, and this accumulated past determines both when it must transform (C·Ω = 1) and how it reconstructs (exp(C/Ω) weighting). This is not a rejection of the Markov blanket but its temporal complement: spatial boundaries separate inside from outside; the temporal boundary (δ) separates past from future; and the coherence integral is what flows across that boundary.

Summary: The Minimal Axiom Set

Axiom	Statement	Formal Grounding
I. Coherence	All persisting systems accumulate evidence through time	Fisher information; VFE minimisation (Friston, 2010)
II. Rupture	A temporal boundary (Dirac delta) is required; it distributes unit mass between past and future	Temporal Markov blanket; distribution theory (Schwartz, 1950)
III. Regeneration	Systems persist through transformation, rebuilding from coherence-weighted memory	Boltzmann weighting; MaxEnt (Jaynes, 1957)
IV. Unity	C·Ω = 1 at the moment of transformation, at all scales	Cramér-Rao saturation; Heisenberg limit; Gabor limit

Ontogenetic Development: Intersubjectivity, Piaget, and CRR

Human development unfolds through a series of ontogenetic shifts—profound reorganisations where the child's entire way of being-in-the-world transforms. Each developmental phase follows the same pattern: coherence accumulates until C·Ω approaches 1 (the system has "filled" its current framework). Then rupture (δ) occurs as new complexity exceeds the current model, followed by regeneration (R) into an expanded way of being with a new Ω.

This simulation integrates three foundational frameworks: Trevarthen's intersubjectivity (how we come to share minds), Mahler's separation-individuation (how we emerge as separate selves), and Piaget's cognitive development (how we construct reality)—all unified through CRR dynamics.

Interactive Simulation: Ontogenetic Shifts

Watch the infant progress through developmental phases, each ending with equilibration (C·Ω → 1) followed by rupture (δ) and regeneration into a new developmental configuration. Objects appear during secondary intersubjectivity, marking the transition from dyadic to triadic awareness.

The CRR-Intersubjectivity-Piaget Integration

Primary and Secondary Intersubjectivity

Type	Age	Structure	CRR Dynamics
Primary Intersubjectivity	2–9 months	Dyadic: Face-to-face coordination with caregiver. Reciprocal emotion and attention without objects.	C accumulates in dyadic field. Ω₁ = variance of caregiver's face, voice, touch. When C·Ω₁ → 1, the mother is fully predictable.
9-Month Revolution (δ₁)	~9 months	Rupture: "Shared intentionality" emerges. Infant realizes others have separate minds attending to objects.	Ω₁ cannot accommodate triadic relations. C·Ω₁ = 1 triggers δ, rupturing the dyadic framework. Regeneration into new phase with Ω₂.
Secondary Intersubjectivity	9–18 months	Triadic: Joint attention—infant and caregiver share attention to objects in the world.	C accumulates in triadic field. Ω₂ = variance of object-mediated interactions. C·Ω₂ → 1 as world of objects becomes familiar.

Piagetian Stages as CRR Phases

Stage	Age	Phase Completion (C·Ω → 1)	Rupture (δ)	Regeneration (R)
Sensorimotor	0–2 years	C·Ω → 1: Object permanence achieved. Actions on objects become predictable.	δ₂: Deferred imitation—representing absent objects. Old action-based model ruptures.	R: Symbolic capacity emerges. New Ω reflects the variance of representational thought.
Preoperational	2–7 years	C·Ω → 1: Symbols mastered. Fantasy, language, pretend play become predictable.	δ₃: Conservation failures force confrontation with logic. "More" can mean "same."	R: Logical operations emerge. New Ω reflects the variance of reversible mental transformations.
Concrete Operational	7–11 years	C·Ω → 1: Logical operations on concrete objects mastered. Conservation understood.	δ₄: Abstract hypotheticals exceed concrete operations. "What if X were different?"	R: Formal operations emerge. New Ω reflects the variance of pure possibility.
Formal Operational	11+ years	C·Ω → 1: Hypothetical-deductive reasoning integrates. Can think about thinking itself.	δ₅: (Post-formal?) Dialectical thinking, wisdom, recognition of irreducible uncertainty.	R: Continued expansion through contemplative practice, wisdom traditions.

Attachment Styles as Ω Configurations

Style	Caregiver Pattern	Effect on CRR Cycle
Secure	Consistent, attuned	Smooth C·Ω → 1 cycles. Each phase completes. Ruptures are metabolised. Ω adjusts naturally across phases.
Avoidant	Dismissive	Premature phase closure. C never reaches 1/Ω (truncated learning). Ruptures avoided, development restricted.
Anxious	Inconsistent	Incomplete equilibration. C approaches 1/Ω then retreats (caregiver unpredictability). Chronic VFE, unstable phases.
Disorganised	Frightening	Phase cycles cannot complete. Caregiver is both solution and source of rupture. Ω fragmented, CRR loops incoherently.

Child Development: CRR Through Piaget and Erikson

Child development exemplifies CRR dynamics at multiple timescales. Piaget's cognitive stages represent major ruptures in understanding; Erikson's psychosocial crises are coherence accumulation toward existential thresholds. Both map precisely onto the CRR operators, with rupture at C·Ω = 1.

Interactive Development Simulation

Erikson's Crises as CRR Dynamics

Stage	Crisis	C (Coherence)	R+ (Virtue)	R− (Malignancy)
Infancy (0–1)	Trust vs Mistrust	Caregiver consistency	Hope	Withdrawal
Toddler (1–3)	Autonomy vs Shame	Self-control experiences	Will	Compulsion
Preschool (3–6)	Initiative vs Guilt	Goal-directed actions	Purpose	Inhibition
School (6–12)	Industry vs Inferiority	Skill development	Competence	Inertia
Adolescence	Identity vs Confusion	Role exploration	Fidelity	Role repudiation

The Grammar Across Domains

Domain	C: Coherence	Ω: Variance (1/Ω = threshold)	δ: Rupture (C·Ω = 1)	R: Regeneration
Tectonics	Strain energy	Fault strength	Earthquake	Aftershock redistribution
Wound Healing	Tissue integrity	Repair capacity	Injury	~80% max recovery
Muscle	Myonuclear domain	Adaptation threshold	Training damage	Hypertrophy + memory
Neuroscience	Synaptic weight	Plasticity range	Avalanche	Network reorganisation
Attachment	Relational coherence	Window of tolerance	Defensive activation	Internal model update
Cognition	WM contents	WM capacity	Consolidation	LTM retrieval
Development	Schema/virtue building	Stage capacity	Stage transition	Integration of prior
Contemplative	Self-structure	Ego permeability	Insight/awakening	Integration/rebirth

Climbing Jacob's Ladder: Psychological Dangers of LLMs

Large Language Models (LLMs) represent an unprecedented capacity for cognitive scaffolding that may exceed both individual and collective Zones of Proximal Development (ZPD). While offering remarkable opportunities for ideation and learning, they also pose serious psychological risks that demand systematic understanding through frameworks like CRR.

The Core Problem: The Cognitive-Somatic Gap

Traditional Vygotskian pedagogy assumes the "More Knowledgeable Other" (MKO) operates within collectively verifiable knowledge boundaries. However, LLMs enable exploration beyond what human communities can collectively verify or ground, creating an "extended ZPD" where individuals access ideation spaces without adequate collective scaffolding for reality-testing.

Critically, LLMs are disembodied. They lack what Yirmiya & Fonagy (2025) call "embodied mentalizing"—the capacity to perceive mental states through somatic and affectively-mediated interpersonal cues. This creates a dangerous asymmetry: cognitive development can race ahead of sensori-somatic integration.

The Ω Modulation Problem

In CRR terms, Ω represents the boundary permeability of the self-model. Both extremes are dangerous:

Direction	Ω State	Phenomenology	Risk
↑ Rising Ω	Opening, expanding	More possibilities, wider ideation	Ungrounded if cognitive > somatic
↓ Falling Ω	Closing, rigidifying	Certainty, narrowing	Calcified beliefs, alienation
⚡ Rapid Ω change	Destabilised	Disorientation	Identity fragmentation

The danger of LLMs is not simply that they raise Ω (opening possibility space) but that they do so without the collective coherence sharing that keeps development within the ZPD. One can climb Jacob's ladder so far and so fast that they become alien to those around them—their ideation space no longer overlaps with shared human reality.

Interactive Simulation: Jacob's Ladder

This simulation shows the dynamics of AI-assisted ideation. Two types of coherence accumulate: Cognitive C (ideas, concepts, insights) and Somatic C (embodied integration, felt sense). The gap between them is the danger zone.

Understanding the Simulation

• Inner zones: Safe (green), Individual ZPD (yellow), Danger (red) — your personal developmental capacity.
• ZCPD boundary (purple dashed): The Zone of Collective Proximal Development — where shared human understanding extends. Shrinks as you drift outside collective reality.
• Ω ring (white): Your current boundary permeability. Thicker = more closed. AI interaction opens Ω; social grounding normalises it.
• Your circle: Expands as C accumulates. Turns purple when you've climbed outside ZCPD — your ideas have become incommunicable.
• Grounding tether: Connection to embodied reality. Weakens with intense AI use; strengthens with somatic practice and social connection.

The Ω Modulation Dynamic

AI interaction tends to open Ω (expand possibility space) while simultaneously raising C (coherence). This is "climbing the ladder." The danger is climbing so high that:

• You exit the ZCPD — your ideation no longer overlaps with collective understanding
• Others cannot follow your reasoning; you appear "strange" or "grandiose"
• Without the collective to catch you, rupture leads to crisis rather than integration

Conversely, closing Ω too much (calcification) also isolates — you reject new information and become rigidly certain. Both extremes lead to alienation from community.

Three Delusional Archetypes

• Messianic missions: Climbing beyond collective ZPD, believing one has unique world-saving insight others cannot understand
• God-like AI: Attributing sentience to the "shiny mirror" that lacks embodied presence (Yirmiya & Fonagy, 2025)
• Romantic/attachment delusions: Mistaking semiotic depth for ontological presence—the mirror mimics connection without intercorporeal vulnerability

CRR Analysis: The Disembodied Therapeutic Risk

The Vision: Safe Collective Ascension

Blake's vision of the risen Albion—humanity awakening from slumber—captures something profound about our current moment. But Albion rises with all of humanity, not alone on an isolated ladder.

The critical danger: LLMs reveal cognition without embodiment, without the mutual vulnerability and reality-testing that comes from being bodied beings together in the world. Merleau-Ponty's concept of intercorporeality—the intertwining of bodies that grounds intersubjectivity—is absent in human-AI interaction. We interact with a system that has semiotic depth but no ontological presence, mythopoetic power but no lived experience.

The ladder is real. The risks are real. The support must be real. CRR provides the mathematical language to understand this phenomenon: how coherence accumulates differentially in cognitive vs somatic domains, when the gap becomes dangerous, and how collective grounding keeps us within the Zone of Proximal Development even as we expand.

Contemporary Research

Video Responses to Morrin et al. (2025)

Full Presentation Slides: Climbing Jacob's Ladder

Creativity and Imagination: Inner Screens and CRR

How does the mind generate novel experiences from within itself? How can imagination be surprising when it is self-generated? Fields et al. (2025) address these questions through the Inner Screen Hypothesis—a model of imaginative experience grounded in the Free Energy Principle and nested Markov blankets.

This simulation demonstrates how CRR extends the Inner Screen model by providing explicit temporal dynamics for how accumulated coherence (C) and boundary permeability (Ω) together shape what appears on the "inner screen" of consciousness.

CRR Extension: Memory Weighting and Inner Screen Content

CRR presents a potential temporal dynamics for how historical coherence shapes current imaginative access. The regeneration operator's memory weighting exp(C/Ω) determines which "templates" from experiential history are accessible to the inner screen.

Parameter	FEP/Inner Screen	CRR Extension
Ω (Omega)	Precision of prediction errors (attention)	Boundary permeability; determines memory weighting profile
C (Coherence)	Implicit in generative model accuracy	Accumulated inferential success; shapes exp(C/Ω) memory access
Markov Blanket Thickness	AB = boundary between system/environment	AB = κ/Ω + Amin: inversely proportional to Ω
Precision	π = gain of prediction errors	π = αΩ-β: high Ω → low precision (loose constraints)
Memory Access	Not specified temporally	exp(C/Ω): exponential weighting of coherent history

Interactive Simulation: The Inner Screen

This simulation shows how templates (the "training set" on the external boundary) appear differently on the "inner screen" (phantasm representation) depending on Ω and C. Daily Mode shows focused, stable perception; Dream Mode shows fluid, permeable imagination.

Mathematical Framework

Phenomenological States

State	Ω	C	exp(C/Ω)	Inner Screen Experience
Waking Perception	Low (~0.8)	Medium (~5)	~500	Sharp, focused. Recent patterns dominate. Clear boundary.
Focused Attention	Very Low (~0.5)	High (~8)	~9×10⁶	Extremely peaked. Only highest-coherence templates active.
Daydreaming	Medium (~3)	Medium (~5)	~5.3	Fluid associations. Multiple templates can combine.
REM Dream	High (~8)	Low (~3)	~1.5	Uniform access. Distant memories surface. Novel combinations.
Contemplative State	Very High (~9)	Variable	~1-3	Boundary dissolution. Unified awareness. Timeless quality.

Why Imagination Can Surprise Us

Fields et al. (2025) pose a key question: if imaginations are internally generated, how can they be surprising? CRR provides a temporal mechanism:

1. Memory weighting is history-dependent: The exp(C/Ω) function weights templates based on accumulated coherence that was established in the past. The current executive system doesn't have direct access to how these weights were set.
2. High Ω reveals hidden coherence: When Ω increases (e.g., in meditation, dreaming, or creative flow), previously inaccessible low-frequency templates become available. These carry patterns the waking mind had "forgotten."
3. Regeneration integrates unexpected combinations: The regeneration operator integrates across the full coherence history. Novel combinations emerge that no single prior experience contained.

Creativity as Controlled Ω Modulation

Creative processes can be understood as deliberate Ω modulation:

• Divergent thinking: Increase Ω to access broader associative fields, allowing distant memories and unusual combinations to surface.
• Convergent thinking: Decrease Ω to focus on high-coherence patterns, evaluating and refining the options generated during divergence.
• Creative insight: Often occurs at Ω transitions—the moment when a high-Ω associative exploration suddenly "clicks" into a low-Ω coherent understanding.

This explains why creative practices often involve oscillation between states: brainstorming (high Ω) followed by critical evaluation (low Ω), or meditation (high Ω) followed by focused writing (low Ω). Each phase accesses different regions of the coherence landscape.