Abstract

AToM (A Theory of Meaning) argues that meaning is a dynamical property: coherence under constraint. This paper shows that the Free-Energy Principle (FEP), active inference, Markov blankets, and hierarchical generative modeling already imply a nested coherence architecture. AToM names this architecture, provides an information-geometric vocabulary for analyzing it, and extends it from the individual nervous system to coupled dyads, groups, organizations, and cultures. The central claim is that coherence is equivalent to long-run expected surprise minimisation across scales. Trauma is modeled as a rupture, collapse, or distortion in Markov blanket boundaries; entrainment is modeled as mutual Markov blanket synchronization. A tentative coherence functional is proposed to invite computational implementation. The paper incorporates established references from Friston, Parr, Ramstead, Porges, Siegel, and Henrich, and addresses criticisms of over-universalization by showing that AToM applies strictly within soft-constraint, high-variance domains. The theory concludes with empirical tests suitable for neural, interpersonal, organizational, and cultural data.

1. Introduction: Meaning as Coherence Under Constraint

Meaning has traditionally been located in three domains: the internal mental representations emphasized by cognitive science, the narrative content emphasized by psychology and the humanities, and the symbolic or linguistic structures emphasized by semiotics and philosophy. AToM (A Theory of Meaning) departs from all three by shifting the unit of analysis from content to structure. In AToM, meaning is not a property of symbols, not a property of internal mental pictures, and not a property of story alone. Meaning is the system’s ability to maintain coherence under constraint. Meaning is the ongoing, dynamically updated integrity of a system as it confronts novelty, ambiguity, uncertainty, and perturbation.

Coherence, in this framework, is not a poetic metaphor but a structural invariant: the capacity of a system to remain itself across time despite being perturbed. When we say a person “makes sense,” what we mean is that their predictions, actions, rhythms, and interpretations are internally compatible and stable enough to hold together under load. When coherence is high, experience feels meaningful, intelligible, and navigable. When coherence is lost, meaning collapses into confusion, fragmentation, overwhelm, or dissociation. Meaning is therefore the phenomenological signature of coherence; coherence is the structural requirement for meaning.

This reframing aligns directly with the Free-Energy Principle (FEP), which provides a mathematical account of how biological systems avoid dissolution. Under the FEP, any self-organizing system must minimize long-run expected surprise. Surprise, in the information-theoretic sense, is the degree to which sensory input violates the system’s generative model of the world. A system that does not minimize expected surprise drifts toward entropy and ceases to maintain its form. In AToM language, a system that does not minimize expected surprise becomes incoherent. Thus, free-energy minimisation and coherence maintenance are not two different processes—they are two descriptions of the same underlying dynamic at different levels of abstraction.

In AToM, coherence = expected surprise minimisation expressed across multiple nested scales. A system maintains meaning by keeping its predictions, interpretations, and actions consistent with both its internal state and the external environment. A system loses meaning when the gap between prediction and sensory evidence grows too large, when the internal model fractures, or when cross-scale alignment breaks down. In neural terms, this appears as destabilized oscillations, precision imbalance, or prediction-error cascades. In interpersonal terms, it appears as misattunement, rupture, or relational volatility. In organizational terms, it appears as misalignment, fragmentation of incentives, or breakdowns in communication. Across all these cases, AToM shows that meaning collapses for the same structural reason: coherence fails.

AToM contributes five core advances that build on and extend the FEP while respecting its formal boundaries.

First, AToM provides a unified language for coherence across neural, physiological, interpersonal, organizational, and cultural systems. Humans operate in nested layers of coupling: neurons couple into circuits, circuits into brains, brains into bodies, bodies into dyads, dyads into groups, groups into institutions, and institutions into cultures. The FEP describes how a single system maintains expected-surprise minimization. AToM shows how coherence becomes a cross-scale property when systems entrain to one another. Meaning is thus not merely intra-cranial; it is an emergent phenomenon of nested coherence across layers.

Second, AToM introduces information-geometric tools for quantifying coherence. Coherence curvature (derived from Fisher Information), dimensional stability (how many degrees of freedom remain open), and topological persistence (stability of patterns across scales) allow coherence to be measured directly in linguistic data, physiological signals, group interactions, or institutional dynamics. These tools transform coherence from an intuitive concept into an analytically tractable construct that can be computed, visualized, and tested.

Third, AToM provides an explicit treatment of trauma as a geometric collapse in the coherence manifold. Trauma is not defined primarily by the narrative memory of an event; it is defined by a deformation of the system’s ability to maintain stable, low-curvature trajectories. Trauma shrinks the system’s available degrees of freedom (dimensionality loss), increases curvature (hypersensitivity), creates bottlenecks (attractor traps), and introduces hysteresis (difficulty returning to prior states even when safe). In the FEP, these appear as prediction-error explosions and precision-weighting failures. In AToM, they are expressed as structural distortions in coherence geometry. Both descriptions point to the same underlying phenomenon: the rupture of Markov blanket integrity.

Fourth, AToM identifies entrainment as the primary mechanism by which coherence propagates through systems. Entrainment is the synchronization of rhythms, predictions, and actions across agents. It appears in neural oscillations, physiological coupling, conversational pacing, collaborative behavior, shared rituals, and cultural narratives. Entrainment enables coherence to be restored across boundaries: between two people, between a person and a group, between a group and an institution, and between an institution and a culture. Where FEP describes how a single system maintains structure, AToM describes how multiple systems maintain structure together through mutual prediction-error minimization.

Fifth, AToM establishes a domain boundary that avoids turning the FEP into an over-universal theory. The FEP, taken naively, is sometimes interpreted as a claim that everything minimizes free energy. AToM clarifies that meaning systems operate in soft-constraint, high-variance regimes where multiple coherence geometries are possible. Physics operates under hard invariances; human meaning operates under flexible, adaptive, culturally mediated constraints. Thus, AToM uses the FEP’s mathematical backbone while restricting its application to the domain where meaning, trauma, culture, and interpretation actually live.

This paper demonstrates that the FEP already implies the AToM structure at the neural and organismic levels, and that AToM extends this structure to interpersonal, organizational, and cultural domains without violating FEP mathematics or scope. Coherence, expressed as expected-surprise minimization across scales, becomes the unifying explanatory principle. Meaning becomes measurable. And human systems—from neurons to nations—become legible through the same coherence geometry.

2. The Free-Energy Principle: A Brief Overview

The Free-Energy Principle (FEP), developed by Friston (2010) and expanded by Parr & Friston (2018), Ramstead et al. (2018), and others, provides a unified account of how living systems maintain their organization in the face of constant perturbation. At its core, the FEP states that any system that persists through time—biological, cognitive, informational, or social—must minimize long-run expected surprise. Surprise here does not refer to emotional shock but to statistical deviation: the degree to which incoming sensory signals differ from what the system’s internal model predicts. A system that continuously encounters sensory inputs inconsistent with its internal expectations will drift toward entropy, disintegration, or collapse. A system that successfully maintains its structure must act, update, and interpret in ways that reduce the mismatch between expectation and sensation.

The FEP thus formalizes the intuitive idea that survival requires maintaining coherence: the integrity of a system’s predictions about the world and the world’s feedback into the system. AToM builds upon this foundation by reframing expected-surprise minimization as coherence maintenance and by extending this logic beyond the individual biological organism into interpersonal, organizational, and cultural domains. Where the FEP describes how a single Markov-blanketed agent maintains stability, AToM shows how multiple coupled agents maintain coherence together.

The FEP consists of three essential components:

1. Markov blankets
2. Hierarchical generative models
3. Active inference

Each contributes a layer of structure that AToM generalizes into its cross-scale coherence architecture.

2.1 Markov Blankets

A Markov blanket is the statistical boundary that separates a system’s internal states from the external environment. It partitions the world into four classes:

1. internal states (mu)
2. sensory states (s)
3. active states (a)
4. external states (eta)

Internal states influence sensory states but not external states directly; external states influence sensory states but not internal states directly. Active states influence external states; sensory states influence internal states. This dependency structure defines the system as an entity: anything behind the blanket is “self,” anything outside it is “world.”

This formulation is not merely mathematical; it is ontological. A Markov blanket defines what the system can predict, what it can act upon, and what counts as a meaningful perturbation. The blanket is the interface through which coherence is maintained or lost. AToM interprets this interface as a coherence boundary: a dynamically maintained edge that keeps internal predictions aligned with external pressures.

Trauma disrupts this boundary. Under traumatic load, sensory precision may spike, prediction-error signals may flood internal states, and the statistical independence of internal and external states may break down. From the AToM perspective, trauma is not just an emotional wound—it is a geometric deformation of the Markov blanket itself. When the boundary becomes too rigid, too porous, or too erratic, the system can no longer maintain low-entropy organization; coherence collapses.

Entrainment, in contrast, is the process by which Markov blankets synchronize. When two individuals interact, they form a coupled system: heart-rate variability aligns, gaze timing aligns, vocal prosody aligns, expectations align. These adjustments reduce cross-blanket prediction error. AToM extends this logic upward to groups, organizations, and cultures, all of which can be described as networks of partially overlapping Markov blankets that maintain coherence together through mutual prediction and mutual regulation.

Thus, Markov blankets become the basic unit of coherence geometry. AToM treats them not simply as statistical partitions but as dynamically maintained coherence membranes.

2.2 Hierarchical Generative Models

Markov blankets describe boundaries; generative models describe the internal organization behind those boundaries. Biological organisms rely on hierarchical generative models to predict sensory input at multiple temporal and spatial scales. Fast layers encode immediate sensory patterns (edges, tones, textures), while slower layers encode abstract regularities (objects, social cues, identities, narratives). Higher layers stabilize lower layers through top-down prediction; lower layers inform higher layers through bottom-up error signals.

This nested structure creates a cascade of coherence conditions. For the system to remain stable, each level must maintain internal coherence while also aligning with adjacent levels. This is why disturbances at one layer ripple through the whole system. A mismatch between high-level interpretation and low-level sensation produces prediction-error spikes, emotional volatility, and behavioral destabilization. Similarly, clarity and stability at high levels can calm fluctuations at lower ones.

AToM extends this hierarchical logic across individuals and systems. Just as neurons compose circuits, circuits compose brain regions, and regions compose whole-organism generative models, individuals compose dyads, dyads compose groups, groups compose institutions, and institutions compose cultures. Each level becomes a generative model for the levels below and a constraint for the levels above. Coherence is therefore not merely an intra-brain property; it is an emergent, multi-level property of nested generative architectures.

In AToM’s framing, hierarchical generative models are coherence ladders. They organize the flow of prediction and error across scales. Trauma breaks these ladders by collapsing dimensionality, increasing curvature, or creating bottlenecks. Entrainment repairs them by synchronizing patterns across layers.

The FEP provides the computational logic; AToM provides the structural map.

2.3 Free Energy and Expected Surprise

Expected surprise is a measure of how incompatible future sensory inputs are with the system’s current internal model. Variational free energy is the tractable quantity the system minimizes to approximate this otherwise intractable expectation. Minimizing free energy keeps prediction error low and stabilizes the system’s internal states.

In information-geometric terms, surprise corresponds to curvature. High curvature means the manifold is sharply bent, making the system hypersensitive to perturbation. Low curvature means the manifold is smooth, allowing the system to move through its state-space without catastrophic error. Trauma, anxiety, hypervigilance, institutional breakdown, and cultural fragmentation all correspond to curvature spikes. Healing, stability, trust, alignment, and meaning correspond to curvature smoothing.

Thus, minimizing expected surprise is identical to minimizing coherence curvature. The two descriptions differ only in focus: the FEP describes the computation; AToM describes the geometry.

The FEP tells us why systems maintain coherence.

AToM tells us how coherence is expressed, transmitted, broken, and restored across scales.

Together, they form a unified account of meaning as the maintenance of coherent prediction in a world full of uncertainty.

3. Coherence = Expected Surprise Minimisation

AToM’s central claim is that meaning emerges when a system maintains coherence under constraint, and that this coherence is formally equivalent to minimizing long-run expected surprise. This claim is not metaphorical or rhetorical—it is a structural identity between two descriptions of the same underlying dynamics. The Free-Energy Principle (FEP) provides the computational machinery; AToM provides the cross-scale geometric and phenomenological interpretation.

Under the FEP, a biological system persists by minimizing the discrepancy between its generative model and incoming sensory signals. Under AToM, a meaning-making system persists by maintaining stable, low-curvature trajectories across its internal, relational, and ecological environment. In both frameworks, the system must prevent deviations large enough to threaten integrity. Survival is coherence; coherence is reduced surprise; reduced surprise is free-energy minimisation. The conceptual loop closes.

The mapping between the FEP’s computational quantities and AToM’s coherence geometry is direct and systematic:

1. Variational free energy corresponds to prediction-error accumulation. A generative model that continuously mispredicts its sensory input accumulates free energy and destabilizes. This is mathematically identical to a system whose coherence manifold becomes distorted, jagged, or irregular. In AToM terms, high free energy is high curvature, and high curvature is incoherence.
2. Expected surprise corresponds to future incoherence. Expected surprise is the system’s estimate of how badly its model will mismatch future sensory input. In AToM language, this is the system’s forecast of future coherence failure. If expected surprise is high, the system anticipates that it cannot maintain coherent identity into the future; it loses stability, predictability, and meaning.
3. Prediction smoothing corresponds to coherence restoration. When a system updates its model or adjusts its actions to reduce prediction error, it is smoothing the geometric irregularities on its coherence manifold. The system makes itself more consistent—internally and externally. This is why learning, adaptation, and regulation feel like “things snapping into place.” The geometry becomes smoother; coherence is restored.
4. Trajectory stability corresponds to geodesic coherence. In information geometry, a geodesic is the smoothest possible trajectory through a manifold. A system with coherent internal dynamics follows near-geodesic paths—stable, low-energy trajectories that do not oscillate unpredictably. When a system experiences anxiety, trauma, overload, or relational rupture, trajectory stability collapses; the system is thrown off its geodesic and into a turbulent region of high curvature, where small perturbations produce large deviations.
5. Markov blanket integrity corresponds to coherence boundary maintenance. A Markov blanket is a statistical boundary. In AToM, it is a coherence boundary: the membrane that keeps a system’s predictions aligned with the external world. Trauma, dissociation, sensitization, hypervigilance, and social fragmentation can all be understood as disruptions of this boundary. If the blanket becomes too rigid, the system cannot integrate new information; if it becomes too porous, the system is overwhelmed by noise. Coherence is the active maintenance of this boundary under constraints.

These correspondences demonstrate that coherence is not a loose synonym for free-energy minimisation; it is the form that free-energy minimisation takes in human systems. Coherence is the experiential, geometric, and relational manifestation of the mathematical imperative to minimize expected surprise.

This equivalence becomes most apparent when we look at how systems respond to perturbation. In a stable, coherent state, a perturbation produces a small, proportionate deviation. The system remains close to its expected trajectory and returns smoothly to equilibrium. Under high curvature—high free energy—a perturbation produces large, chaotic deviations, loss of predictability, and breakdown of integrative function. From the outside, this may look like panic, shutdown, disorganization, or instability. From inside the system, it feels like meaning collapse: “nothing makes sense,” “everything is too much,” “I can’t keep up,” “I don’t know who I am.”

People often describe incoherent states using phenomenological terms, but the underlying structure is information-theoretic. When coherence collapses, prediction error spikes, and the system cannot maintain self-consistency. This applies not only to individuals but also to relationships, teams, institutions, and cultures. A couple experiencing chronic misattunement is a two-agent system with rising cross-blanket prediction error. A dysfunctional team suffers from inconsistent generative models among its members. A failing institution loses the ability to coordinate predictions across its departments and stakeholders. A polarized culture loses its shared priors and becomes a high-surprise environment where no group can predict the behavior of the others.

In all these cases, meaning dissolves because coherence dissolves. The system is no longer able to predict itself or be predicted by others. This yields a global sense of uncertainty, instability, and disorder.

Coherence is therefore the central variable that determines whether meaning is gained, maintained, threatened, or lost.

If the system maintains coherent prediction across levels—neural, physiological, affective, interpersonal, organizational, and cultural—meaning expands. If coherence collapses at any level, meaning contracts or fragments.

The FEP accounts for this computationally: a system must minimize surprise. AToM describes how this appears structurally and phenomenologically: the system must remain coherent.

The two frameworks differ in emphasis:

• The FEP is bottom-up: given the mathematics of self-organization, what must be true of a living system?
• AToM is cross-scale: given the phenomenology of meaning, trauma, identity, and culture, what structural invariants unify them?

AToM interprets the FEP as the mathematical engine that drives coherence in all systems capable of maintaining identity under constraint. It asserts that expected surprise minimization and coherence maintenance are not separate processes; they are the same process viewed from different perspectives.

This identity allows AToM to extend coherent free-energy dynamics into domains where traditional FEP applications rarely go: interpersonal synchrony, therapeutic repair, organizational alignment, cultural evolution, mythic compression, and institutional integrity. Each domain involves systems maintaining their internal organization by reducing prediction error—whether that prediction error is neural, affective, social, or symbolic.

When coherence is present, systems move smoothly through their space of possibilities. When coherence is absent, systems oscillate, collapse, or fracture. Meaning, in this sense, is the felt texture of coherence. When coherence holds, meaning feels clear, stable, resonant, and grounded. When coherence fails, meaning feels scattered, chaotic, unstable, or existentially empty.

Thus, coherence is the phenomenological signature and the structural expression of free-energy minimisation. In AToM, coherence becomes the throughline connecting neural dynamics, psychological experience, relational regulation, organizational function, and cultural sense-making. Meaning, viewed through this lens, is simply coherence sustained across constraints and across scales.

4. AToM’s Information-Geometric Foundations

AToM treats coherence not as a metaphor or subjective impression but as a mathematically characterizable structure. To do this, AToM draws from information geometry—the study of how probability distributions form geometric manifolds whose curvature, topology, and dimensionality encode meaningful system properties. Information geometry provides the precision vocabulary needed to describe what coherence is, how it is maintained, how it fails, and how it can be restored. When integrated with the Free-Energy Principle, these tools allow AToM to translate intuitive concepts like stability, alignment, or fragmentation into measurable quantities.

Several constructs are central to AToM’s information-geometric formulation:

1. Fisher Information curvature: the system’s sensitivity to perturbation
2. KL divergence gradient: deformation of predictions under new data
3. Topological persistence (H_k): stability of attractors, loops, bottlenecks
4. Dimensionality: the degrees of freedom accessible to the system
5. Reversibility: how easily a system returns to prior states after deformation

These tools allow coherence to be analyzed with the same mathematical rigor used in statistical physics and variational inference, while remaining domain-appropriate for biological, psychological, relational, and cultural systems. Importantly, these constructs map directly onto FEP variables. FEP provides the bottom-level imperative (minimize expected surprise); AToM provides the mid-level structural geometry through which that imperative expresses itself in real, multi-scale systems.

Fisher Information Curvature

Fisher curvature measures how sharply the system’s probability landscape bends in response to perturbation. High curvature indicates the system is in a sensitive or unstable region—small deviations produce large prediction errors. Low curvature indicates robustness: the system can tolerate variation without destabilizing.

When a system is coherent, its manifold has low curvature; it moves smoothly through its state-space. When coherence is compromised—through trauma, overload, stress, or fragmentation—curvature spikes, and the system becomes hypersensitive. This can be seen in anxiety (tiny cues evoke large responses), PTSD (small triggers evoke disproportionate autonomic reactions), and organizational instability (minor events lead to cascading failures).

Thus, Fisher curvature is the geometric signature of prediction-error sensitivity.

KL Divergence Gradient

Kullback–Leibler divergence measures the difference between the system’s generative model and actual sensory data. The gradient of this divergence describes how the system’s model must deform in order to align with incoming information.

A steep KL gradient means the system must update sharply—it is far from self-consistency. A shallow gradient means the system’s predictions fit the world well.

In coherence terms:

• Steep gradients correspond to incoherence, instability, or identity disturbance.
• Shallow gradients correspond to stability, predictability, and coherence.

Psychotherapeutic breakthroughs often show a pattern where KL gradients soften before the person consciously realizes their own shift. In interpersonal settings, rising KL gradients between two people manifest as misattunement, conflict, or relational incoherence. In cultural settings, rapidly diverging KL gradients across subpopulations manifest as polarization or fragmentation.

KL divergence is thus the coherence deformation metric.

Topological Persistence (H_k)

Topological persistence, derived from persistent homology, measures the stability of topological features (connected components, loops, voids) across multiple scales of resolution.

For AToM:

• H_0 measures whether subsystems remain connected or fragment.
• H_1 measures loops and recurrent patterns—stable or pathological attractors.
• H_2 measures voids and gaps—regions the system avoids or cannot access.

High persistence indicates stable patterns. Low persistence indicates fragile or collapsing structure. A trauma-related attractor—such as a rigid behavioral pattern or intrusive memory—will appear as a highly persistent H_1 cycle; a collapsed relational network will show fragmented H_0 structure; a rigid identity profile may show collapsed or simplified topology.

Topology allows coherence to be understood not only locally (curvature, KL) but globally—across the entire landscape of how states connect, persist, or collapse.

Dimensionality

Dimensionality measures the degrees of freedom available to a system. A coherent system has enough dimensionality to flexibly adapt, reconfigure, and transition between states. Trauma reduces dimensionality: options narrow, behaviors rigidify, and emotional or cognitive repertoire shrinks. Depression also reduces dimensionality, producing low-energy attractors that constrain movement.

Dimensionality reduction is one of the hallmark signatures of coherence collapse.

Reversibility (Hysteresis)

Reversibility measures how easily a system can return to a previous configuration after perturbation. When coherence is high, the system displays elastic stability—perturbations are absorbed, and the system returns to baseline. When coherence is low, the system displays hysteresis—returning to baseline requires more energy than the perturbation that displaced it.

Trauma systems exhibit hysteresis: safety is not enough to restore coherence; the manifold must be gradually reshaped.

Reversibility therefore provides a measure of how deeply coherence geometry has been altered.

Integration with FEP Mathematics

Each of these constructs aligns cleanly with the FEP:

• Fisher curvature = precision weighting and prediction sensitivity
• KL gradient = magnitude and direction of model updating
• Topological persistence = stability of attractors over time
• Dimensionality = model complexity and flexibility
• Reversibility = energetic cost of minimizing free energy after perturbation

The FEP provides the computational rule (minimize expected free energy).

AToM provides the geometric structure through which that rule is enacted.

In this sense, AToM is the geometry of the FEP.

4.1 Plain Text Coherence Operator 

Below is the proposed coherence functional in plain ASCII text, with expanded explanation:

C_hat[q] = integral_over_manifold (

Fisher_curvature(mu)

+ KL_divergence_gradient(q||p)

+ Topological_persistence_Hk(mu)

) d_mu

This operator attempts to capture the total coherence cost across the system’s information manifold. Each term corresponds to a distinct form of coherence stress:

1. Fisher curvature: local sensitivity or instability
2. KL gradient: global prediction deformation
3. Persistence: structural bottlenecks or attractors

The integral sums these across all internal states (mu).

Interpretation:

• Low C_hat means the system’s predictions, topology, and local curvature all remain in a smooth, stable configuration. This is high coherence.
• High C_hat means the system’s manifold contains high-curvature regions, steep KL gradients, or persistent bottlenecks—geometric signatures of trauma, fragmentation, overload, or incoherence.

This operator is not offered as a final or literal physical observable. It is a computational scaffold designed for future implementation using variational inference, multimodal sensing, and topological data analysis. It is intentionally domain-appropriate: rooted in established geometry, compatible with active inference, and extendable to neural, interpersonal, organizational, and cultural data.

AToM therefore uses information geometry not as analogy but as architecture. The coherence operator formalizes meaning as a measurable property of systems navigating constraint—a structural invariant that reveals how coherence is maintained, how it fails, and how it can be restored.

5. Trauma as Coherence Collapse 

In AToM, trauma is not defined by the content of a memory, the presence of fear, or the occurrence of a particular type of event. Trauma is a structural transformation of the system’s coherence geometry. It is what happens when a system’s ability to maintain low-entropy, integrable trajectories collapses under overwhelming constraint. Trauma is not primarily about “what happened,” but about how the system’s manifold deforms in response to what happened. It is a geometric failure mode—a shift into a different region of the information landscape where coherence cannot be sustained without enormous metabolic, cognitive, emotional, or relational cost.

This reframing unifies findings across neuroscience, psychology, physiology, and systems theory. Trauma alters prediction stability (Hohwy), autonomic regulation (Porges), integrative neural pathways (Schore), memory consolidation (Brewin), and relational attunement (Tronick). AToM integrates these under a single invariant: trauma is a collapse in coherence geometry, expressed across the system’s Markov blanket and its internal generative models.

In the language of information geometry, trauma deforms the manifold along five primary axes: dimensionality, curvature, topological connectivity, hysteresis, and boundary formation. These are not metaphors but measurable structural signatures.

5.1 Dimensionality Loss

Dimensionality refers to how many degrees of freedom a system has available. A coherent system can flexibly move across emotional states, cognitive modes, relational contexts, and behavioral responses. Trauma reduces these degrees of freedom. The system constrains itself to a much narrower region of state-space in order to survive overwhelming load.

Clinically, this appears as emotional blunting, avoidance, shutdown, rigidity, compulsive repetition, or narrow, overlearned survival strategies. In dynamical terms, the manifold collapses into a lower-dimensional subspace. Instead of a wide, navigable landscape of possible states, the system becomes trapped in a narrow corridor of “allowed” configurations. Anything outside that corridor produces intolerable prediction error.

This explains why traumatized individuals often appear “stuck” or “frozen.” They are not unwilling to try new patterns; their coherence geometry simply cannot support the dimensionality needed for flexible adaptation.

5.2 Curvature Spike

Trauma increases local curvature on the coherence manifold. Curvature here refers to sensitivity: how sharply prediction error increases as the system moves through its state space. High curvature means small perturbations cause disproportionately large deviations. This is the mathematical signature of hypervigilance.

A traumatized system becomes hypersensitive because its generative model has learned—correctly, in context—that the world cannot be trusted. Precision weighting is altered: threat-related cues receive exaggerated weight, and benign cues are interpreted through a distorted precision landscape. A slammed door, a facial expression, a tone of voice, or an unexpected touch can produce massive internal perturbation.

High curvature is therefore the geometric analog of sympathetic dominance, startle reflex amplification, and generalized anxiety. The system is trying to maintain coherence in a region where the manifold has become jagged and steep—an energetic impossibility.

5.3 Bottlenecking (Topological Collapse)

Trauma introduces bottlenecks in the topology of the system’s state-space. Using the language of persistent homology, trauma increases the persistence of certain attractors (deep ruts of rumination, avoidance, or flashbacks) and collapses other pathways entirely.

These bottlenecks prevent the system from transitioning between cognitive, emotional, or relational states. For example:

• A person may repeatedly loop through intrusive memories (stable H_1 cycle).
• A relational system may collapse into a single repetitive interaction pattern.
• A cultural system may form rigid ideological attractors that block integration.

What looks like “repetition compulsion” is simply the system’s topology becoming constricted. The manifold no longer provides viable pathways to healthier states. The system is not choosing repetition—it is trapped by its own geometry.

5.4 Hysteresis (Irreversibility)

Trauma introduces hysteresis: once the system’s coherence geometry collapses, it cannot simply return to its prior configuration. The energy required to restore the system’s former manifold is far greater than the energy that triggered the collapse. This is why “feeling safe” is insufficient for healing. Safety may reduce incoming perturbations, but it does not automatically reconstruct lost manifold structure.

Hysteresis explains:

• why trauma effects persist decades after the event
• why symptoms remain despite environmental calm
• why repeated reassurance has limited impact
• why recovery requires intensive relational, somatic, and cognitive scaffolding

The hallmark of trauma is not the breakdown itself but the difficulty of returning from it. The system remains in a collapsed configuration because its coherence geometry has been permanently altered.

5.5 Boundary Formation (Dissociation)

Dissociation, in AToM, is a boundary-formation strategy—a way for the system to create new Markov blanket partitions that isolate high-curvature regions from the rest of the manifold. It is not emptiness or absence; it is a protective geometric partition that prevents destabilizing information from spilling into global states.

Dissociation is the system’s last-ditch coherence-preservation strategy. When global coherence becomes impossible, the system maintains local coherence by isolating incompatible subsystems. This yields:

• compartmentalized memory
• emotional numbing
• depersonalization
• derealization
• altered identity states

Each is a different way of constructing new coherence boundaries to avoid catastrophic prediction-error accumulation. Dissociation is therefore not regression—it is an emergency form of coherence engineering.

Trauma Across Scales

A central insight of AToM is that trauma is not exclusive to individuals. The same coherence-collapse patterns appear across systems:

• Dyads experience relational trauma when attunement breaks and cross-blanket prediction error becomes intolerable.
• Families experience trauma when members cannot synchronize emotional or behavioral rhythms.
• Organizations experience trauma when policies, incentives, and communication networks collapse into bottlenecks or high-curvature conflict zones.
• Cultures experience trauma during war, social fragmentation, epistemic breakdown, or rapid technological acceleration.

In every case, the same geometry emerges: dimensionality collapse, curvature spikes, bottlenecks, hysteresis, boundary reformations.

Thus, trauma is a universal coherence failure mode across scales of human organization.

Trauma as Free-Energy Explosion

In FEP terms, trauma is an expected-surprise explosion: prediction error floods the system beyond its capacity for correction. The generative model cannot update fast enough or deeply enough to maintain stability. Sensory precision becomes miscalibrated. Action cannot mitigate threat. Internal states destabilize. Markov blankets deform.

AToM translates this into coherence geometry: the manifold bends, collapses, narrows, and partitions.

The two descriptions are identical in structure but differ in emphasis:

• The FEP describes why the system collapses (unbounded prediction error).
• AToM describes how the collapse appears (curvature, topology, dimensionality).
• The lived experience is what meaning feels like when coherence breaks.

Trauma as a Reconfigurable Geometry (Path to Recovery)

Although trauma reshapes the manifold, it does not make it irreparable. Coherence can be gradually rebuilt. This occurs through:

• relational entrainment (attunement, co-regulation)
• somatic entrainment (breath, movement, vagal activation)
• narrative entrainment (integration, re-interpretation)
• environmental entrainment (consistent, predictable contexts)

These reintroduce smoothness, open new dimensions, and soften curvature.

Healing is not merely psychological—it is geometric reconstruction.

If you’d like, I can continue with Section 6: Entrainment as Mutual Markov Blanket Synchronization (1000 words) or move to another section.

5. Trauma as Markov Blanket Rupture

AToM defines trauma as a collapse in coherence geometry. This has perfect correspondence with FEP failure modes.

5.1 Dimensionality Loss

In trauma, behavioral and cognitive degrees of freedom collapse.

FEP interpretation: major loss of model complexity.

5.2 Curvature Spike

Trauma increases Fisher curvature. Small cues produce large internal responses.

FEP interpretation: precision imbalance and prediction hypersensitivity.

5.3 Bottlenecking

Topological bottlenecks form between cognitive and affective regions.

FEP interpretation: pathological attractors trap prediction error.

5.4 Hysteresis

Systems cannot return to prior states without disproportionate energy.

FEP interpretation: irreversibility in generative model updates.

5.5 Dissociation

Dissociation creates protective boundaries isolating high-curvature regions.

FEP interpretation: reorganization of blanket boundaries to prevent catastrophic surprise.

AToM therefore offers a geometric description of what FEP predicts dynamically.

6. Entrainment as Mutual Markov Blanket Synchronization

If trauma is the collapse of coherence geometry, then entrainment is its restoration. Entrainment is the process by which systems—neural, physiological, interpersonal, organizational, or cultural—synchronize their internal dynamics through mutual prediction-error reduction. It is the mechanism that allows coherence to propagate across boundaries, allowing multiple Markov blankets to function as a coupled, integrated whole rather than as isolated, incoherent nodes.

In the Free-Energy Principle, active inference describes how a single agent minimizes expected surprise by updating beliefs, regulating precision, and acting to reduce prediction error. AToM generalizes this logic: when two or more agents interact, their actions and predictions become interdependent. Minimizing surprise becomes a shared process. Two systems entrain when each becomes part of the other’s prediction-stabilizing environment.

Entrainment is thus mutual free-energy minimisation—expressed as coherence across systems. It is the fundamental mechanism by which meaning becomes social, relational, and cultural.

6.1 Entrainment as Cross-Blanket Coherence

When two Markov blankets interact, they exchange sensory and active states. Each system receives signals generated by the other and updates its internal model accordingly. Over time, these updates reduce cross-blanket prediction error. As their predictions align, the boundaries between the two systems begin operating not as barriers but as coherent interfaces.

This synchronization emerges across multiple channels:

• timing (speech rhythms, motor rhythms, gaze timing)
• physiology (heart rate, respiration, vagal tone)
• affect (emotional resonance, shared valence)
• cognition (shared expectations, joint attention)
• behavior (coordinated action, turn-taking, imitation)

The result is a shared coherence manifold—a region of state-space where neither system experiences disruptive prediction error in relation to the other. This is what humans call connection.

Entrainment is therefore not poetic: it is the structural means by which relational meaning emerges.

6.2 Neural Entrainment: Synchronizing Internal Dynamics

Neural entrainment occurs when oscillatory activity aligns across individuals. This is observed in joint music-making, conversation, shared attention tasks, or cooperative problem-solving. The brain reduces free energy by aligning its temporal structure with predictable external signals—especially signals generated by another brain.

Neural entrainment is the most basic form of inter-agent coupling. It is why synchronized movement increases trust, why shared gaze regulates affect, and why infants tune their neural rhythms to caregivers. It is also why trauma survivors may struggle to entrain—their internal curvature makes synchrony costly or overwhelming.

AToM interprets neural entrainment as the micro-foundation of interpersonal coherence.

6.3 Physiological Entrainment: Autonomic Co-Regulation

Physiological entrainment occurs when heart-rate variability, respiration, electrodermal activity, or vagal tone align across individuals. This is the basis of co-regulation. Two nervous systems stabilize each other by mutually reducing autonomic prediction error. When one person calms or grounds another, they are literally helping reorganize the other’s coherence geometry.

Safety, from an AToM perspective, is a low-curvature relational manifold that agents can synchronously inhabit.

Conversely, in relational conflict, physiological entrainment breaks down. HRV decouples, breathing patterns diverge, precision weighting becomes distorted, and prediction error escalates. The dyad becomes incoherent. This often precedes conscious awareness of rupture, which is why physiological measures are powerful early detectors of relational instability.

Entrainment thus provides the physiological basis of trust, communication, and emotional meaning.

6.4 Interpersonal Entrainment: The Architecture of Attunement

Attunement is entrainment at the interpersonal scale. When two people converse fluidly, understand each other intuitively, or move together effortlessly, their predictions and actions align with low error. Attunement is coherence made visible in interaction.

It relies on:

• microtiming (millisecond-scale turn-taking)
• affective mirroring
• repair cycles (rupture-correction loops)
• predictability and responsiveness
• mutual reduction of uncertainty

The more attuned a relationship, the lower the free-energy cost of maintaining it. Attunement is energetically efficient because the coherence manifold shared by the two people is smooth, predictable, and mutually reinforcing.

This also explains relational trauma: when attunement collapses chronically, the shared manifold fragments, prediction error spikes, and the systems become adversarial or disconnected.

Entrainment is the engine of relational meaning; its breakdown is the engine of relational suffering.

6.5 Group Entrainment: Collective Coherence

Groups form larger-scale Markov blankets with collective generative models. For example:

• A team coordinating a task
• A choir singing
• A military unit marching
• A classroom learning
• A community practicing ritual

In each case, group-level coherence emerges when individuals entrain to a shared rhythm, narrative, or goal. The group reduces its collective prediction error through synchronized behavior, shared expectations, and coordinated action. The group becomes a super-agent with its own coherence geometry.

When group entrainment fails—miscommunication, conflicting incentives, unclear roles—coherence collapses. Prediction error splinters across members, and the group destabilizes.

AToM treats group entrainment as a middle layer between interpersonal and cultural systems.

6.6 Organizational Entrainment: Alignment Under Constraint

Organizations maintain coherence through alignment of processes, incentives, communication flows, and cultural norms. When these are synchronized, the organization behaves as a coherent entity. When they diverge, prediction error accumulates.

Examples of organizational entrainment:

• shared mission and values (high-level priors)
• standardized procedures (middle-level priors)
• coordinated communication rhythms (temporal alignment)
• cross-team synchronization (structural entrainment)

Organizations collapse when their coherence geometry fractures into silos, contradictions, or bottlenecks. This is the organizational equivalent of trauma.

Entrainment provides the stabilizing geometry that allows institutions to maintain identity across complexity.

6.7 Cultural Entrainment: Large-Scale Predictive Alignment

Cultures entrain individuals through:

• shared narratives
• myths
• rituals
• symbols
• law
• norms
• media
• institutions

These elements provide large-scale predictive scaffolding. They reduce social uncertainty by aligning generative models across populations. Cultural coherence enables strangers to coordinate, trust, and coexist without excessive prediction error.

When cultural entrainment breaks—polarization, misinformation cascades, institutional breakdown—coherence collapses at a civilizational scale. Meaning fragments because cross-blanket prediction error becomes unmanageable.

AToM interprets cultural meaning as large-scale coherence.

6.8 Entrainment as the Inverse of Trauma

Trauma is coherence collapse.

Entrainment is coherence reconstruction.

Where trauma reduces dimensionality, entrainment expands it.

Where trauma spikes curvature, entrainment smooths it.

Where trauma forms bottlenecks, entrainment opens pathways.

Where trauma creates hysteresis, entrainment restores reversibility.

Where trauma partitions the manifold, entrainment re-integrates it.

Thus, entrainment is not auxiliary to AToM; it is foundational. It explains how coherence is formed, maintained, and restored across all scales of human organization.

Meaning is not formed in isolation.

Meaning is the emergent product of entrainment.

7. Physics as a Boundary Case: High Symmetry, Low Variability

AToM positions physics not as the template for human meaning but as the boundary case of coherence: the extreme end of the constraint spectrum where variability is minimal, symmetry dominates, and only a narrow range of coherent configurations is mathematically possible. AToM explicitly avoids the common error of stretching physical laws into metaphorical models for psychology or culture. Instead, physics is used as a contrast class—a regime where coherence emerges from rigid constraints rather than adaptive negotiation. This distinction is critical for preventing overreach and for defining the appropriate domain in which AToM operates.

Physical systems exhibit the strongest form of coherence because they operate under hard invariances. The structure of the physical world is determined by symmetries: Lorentz invariance, Poincaré invariance, gauge invariance, conservation laws, variational principles. These invariances tightly define what “coherence” can mean. A coherent physical theory is one that respects these symmetries; an incoherent one violates them and is immediately ruled out. In this regime, coherence is nearly binary. A physical theory either satisfies the symmetry constraints or collapses. There is no middle ground, no gradual degradation, no soft failure modes.

This is fundamentally different from human meaning systems, which operate under soft constraints. Human systems tolerate variability, ambiguity, partial truths, and multiple competing coherence geometries. They adapt to context, history, culture, trauma, and relational dynamics. Physical systems do not. A physical law does not “update” in response to trauma, culture, or negotiation. It does not entrain. It does not collapse into pathological attractors. It does not reorganize its manifold on the fly. It simply holds or fails.

This difference is foundational to AToM’s scope.

7.1 Hard Constraints vs Soft Constraints

In physics, constraints are rigid and non-negotiable. Symmetries determine what forms are possible. For example:

• Noether’s theorem forces conserved quantities from symmetries.
• Gauge invariance restricts allowable field configurations.
• Lorentz invariance restricts how events relate across frames.
• The structure of quantum field theory is fixed by renormalizability and symmetry.

These are absolute. Violate them even slightly and the theory becomes incoherent, inconsistent, or non-predictive.

Human systems do not operate like this. Their constraints are soft, context-dependent, historically contingent, and multivalent. Coherence can be partially preserved, partially lost, or restored through entrainment and reconfiguration. Trauma collapses coherence gradually and can be rebuilt; physical symmetries cannot be similarly bent or reconstructed. A human system can remain meaningful with partial inconsistency; a physical theory cannot tolerate even minor violations of symmetry.

Thus, physics is the simplest regime of coherence: low variance, high invariance, minimal degrees of freedom. Human meaning is the complex regime: high variance, soft invariance, many degrees of freedom.

7.2 Recurrence Options in Meaning vs Physics

One way to understand the gap is to examine recurrence options. In physics, the space of viable solutions is narrow. Laws are universal, timeless, and identical across contexts. There are not many different “stable ways” for a physical system to remain coherent. The manifold of allowable states is heavily constrained.

By contrast, human meaning systems have vast recurrence spaces. A culture may stabilize coherence through ritual, law, myth, narrative, norms, institutions, symbols, or practices. Different cultures use different coherence strategies. Individuals develop different generative models depending on attachment style, developmental history, neurodivergence, and trauma. Organizations maintain coherence through alignment strategies, incentive structures, communication rhythms, and structural coupling. There is no single “right” geometry—many geometries can sustain coherence, and the system may move between them over time.

This multiplicity of coherence geometries is what makes AToM necessary. Meaning systems cannot be adequately modeled by a single set of rigid rules or invariances. They require a framework capable of describing flexible coherence maintenance under uncertainty—something the physical sciences, by design, do not attempt.

Thus:

• Physics = minimal recurrence; coherence is near-unique.
• Human meaning = maximal recurrence; coherence is plural.

7.3 Variational Principles in Physics vs Meaning

Both physics and AToM involve variational principles. In physics, the system follows a path that extremizes action (typically minimizes it). This principle is not chosen; it follows from symmetries. The geodesic in spacetime is uniquely determined by the metric. The coherence trajectory is dictated from outside by invariance.

AToM’s coherence trajectories, however, are emergent. They do not arise from fixed symmetries but from dynamic interactions between:

• internal predictions
• external signals
• relational feedback
• cultural scaffolding
• historical path-dependence
• embodied states
• interpersonal entrainment

Thus, whereas physical trajectories are determined by invariance, coherence trajectories in human systems are negotiated. They are continuously reconstructed through entrainment, learning, and adaptation. Meaning is not a fixed geodesic; it is an evolving solution to the coherence problem under soft constraints.

This distinction protects AToM from sliding into pseudoscientific physics analogies. AToM employs variational logic, not physical variational laws.

7.4 Boundary Conditions and Why Physics Matters

Even though AToM rejects reduction of meaning to physics, physics remains important as a boundary case. It shows the limiting form of coherence—coherence under maximal constraint. By understanding the extreme, we clarify what makes human systems distinctive.

Physics reveals:

• What coherence looks like when constraints are absolute
• How symmetry produces stability
• How invariance structures shape allowable forms
• How manifolds behave under minimal variability

Meaning systems reveal:

• What coherence looks like when constraints are flexible
• How entrainment produces stability
• How multi-agent coupling shapes allowable forms
• How manifolds behave under maximal variability

The contrast is not antagonistic but complementary. Physics displays the lowest-variance coherence geometry; AToM describes the highest-variance ones. The two ends of the spectrum bracket the space of possible coherence architectures.

7.5 Avoiding Overreach: AToM’s Domain Boundary

The Free-Energy Principle is often criticized for being overly universal—“everything minimizes free energy.” AToM avoids this pitfall by explicitly limiting its domain. AToM does not attempt to explain particle physics, cosmology, materials science, or chemical structure. These domains operate under hard symmetries and low-variance manifolds where traditional physical mathematics is more appropriate.

AToM applies to systems characterized by:

• soft constraints
• developmental plasticity
• high variance
• multi-scale coupling
• path dependence
• cultural shaping
• relational dynamics
• multiple viable coherence regimes

In short: AToM applies where meaning exists. Meaning is not present in the electron or the quark; meaning arises only in systems whose coherence must be actively negotiated and maintained across context, history, and relationship.

7.6 Coherence Under High Symmetry vs Coherence Under High Variance

Summarizing the contrast:

Physics (high symmetry, low variance):

• Coherence = invariance under transformation
• Boundaries = fixed and impersonal
• Trajectories = uniquely determined
• Collapse = violation of symmetry
• Restoration = impossible; rules are absolute
• Systems = non-negotiating

Meaning systems (low symmetry, high variance):

• Coherence = integrability across changing constraints
• Boundaries = adaptive and relational
• Trajectories = negotiated, entrained
• Collapse = deviation beyond tolerance
• Restoration = possible through restructuring
• Systems = co-regulating and adaptive

Thus, physics is the boundary case that clarifies why AToM is necessary and where AToM properly applies.

8. Neurodivergence as Precision Coherence Sensing (Expanded ~1000 words)

AToM interprets neurodivergence—especially autistic cognition—not as deficit but as a structural variation in coherence geometry. Neurodivergent systems sample the world differently. They weight predictions differently. They respond to inconsistency differently. These differences are not malfunctions; they are differences in how the coherence manifold is constructed, navigated, and stabilized. In AToM’s framing, neurodivergence is best understood as a precision coherence-sensing phenotype: a mode of information processing that detects breaks, contradictions, or irregularities in coherence geometry long before neurotypical systems do.

This view aligns research from Mottron, Pellicano, Happé, Baron-Cohen, Lawson, Rogers, and many others who show that autistic cognition involves heightened sensitivity to pattern, reduced reliance on smoothing priors, altered precision-weighting, and increased resolution of local environmental structure. From the perspective of the Free-Energy Principle, these differences reflect unique settings of precision parameters—how much weight the system gives to sensory likelihoods versus top-down predictions. From the AToM perspective, these become differences in how curvature, dimensionality, and boundary conditions are experienced and interpreted.

Neurodivergence, in short, is not merely a variation of behavior; it is a variation of coherence geometry.

8.1 Reduced Smoothing, Increased Resolution

Most neurotypical systems operate with strong smoothing priors. They intentionally down-weight small inconsistencies in sensory or social data. This allows them to form stable, coherent predictions even when the world is noisy or partially contradictory. Neurotypical cognition tolerates ambiguity and minor incoherence because it privileges global structure over local detail.

Neurodivergent systems—particularly autistic ones—apply less smoothing. They treat local deviations not as noise but as data. They track micro-structure with higher fidelity and lower tolerance for inconsistencies. This creates several predictable computational effects:

• reduced tolerance for prediction-error noise
• increased detection of micro-perturbations
• sensitivity to broken patterns
• difficulty ignoring inconsistencies in social signals
• enhanced precision in perceptual or conceptual domains

This is not “rigidity.” It is a different coherence strategy—one that preserves fine-grained predictive accuracy at the cost of tolerating less contradiction.

8.2 High Sensitivity to Curvature

In AToM’s geometry, curvature represents how sharply prediction error increases as the system moves through its state-space. Autistic systems experience curvature differently. Because they apply reduced smoothing, curvature spikes appear earlier, sharper, and more salient. A situation that feels “fine” to a neurotypical person—slightly incongruent, slightly chaotic, slightly mismatched—may already be a region of intolerably high curvature for an autistic system.

This explains:

• sensory overwhelm (sensory channels amplify curvature)
• social overwhelm (ambiguous cues produce steep error gradients)
• meltdown or shutdown (curvature spike exceeds system tolerance)
• monotropism (system stays in low-curvature zones to remain coherent)
• insistence on predictability (predictable environments = flat manifold)

Autistic cognition is optimized for coherence maintenance in low-noise, high-regularity environments.

8.3 Dimensionality Differences

Neurodivergent systems often exhibit monotropic coherence: intense focus on a narrow but deep bandwidth of interests or cognitive states. This is typically framed as inflexibility, but AToM interprets it differently. Monotropism reflects a geometry where certain high-dimensional regions feel too unstable or too costly to traverse. The system preserves coherence by operating in regions where curvature is manageable and where predictive precision can remain high.

This narrow-but-deep structure has several advantages:

• extreme expertise acquisition
• stable attention
• deep pattern detection
• low noise in conceptual domains
• increased sensitivity to structural inconsistencies

However, it also means that transitions between states require more energy and that unpredictable environments are more difficult to navigate without coherence collapse.

8.4 Boundary Conditions and Attentional Gateways

Neurodivergent individuals often experience different Markov blanket permeability. Some boundaries are too porous (leading to sensory flooding, emotional contagion, or constant vigilance). Others become too rigid (leading to shutdown, withdrawal, or avoidance). Both are coherence-preservation strategies under environments where prediction error is difficult to regulate.

This variation in boundary dynamics helps explain:

• why autistic individuals may hyper-focus or dissociate under load
• why ADHD systems oscillate rapidly across coherence boundaries
• why sensory gating may be inconsistent
• why interpersonal signals can either overwhelm or fail to register

These are not deficits but variations in how coherence boundaries are managed under uncertainty and perturbation.

8.5 The Role of Trauma in Amplifying Precision

Trauma modifies coherence geometry in all systems, but its impact on neurodivergent systems is often magnified. Because autistic and ADHD systems already operate with high local precision and reduced smoothing, trauma-induced curvature spikes become especially destabilizing.

This produces predictable patterns:

• hypervigilance becomes massively amplified
• overwhelm and shutdown become more likely
• relational misattunement becomes more threatening
• sensory irregularities become disproportionately costly
• trust requires longer periods of stable entrainment

AToM therefore predicts that neurodivergent trauma is not qualitatively different, but geometrically deeper: curvature increases more sharply, dimensionality collapses more quickly, and boundaries become more brittle.

8.6 Neurodivergence as a Coherence-Diagnostic Asset

Neurotypical cognition excels at smoothing. It keeps groups cohesive by down-weighting inconsistencies. However, smoothing blinds systems to early coherence fractures. Neurodivergent systems, by contrast, detect micro-fractures long before neurotypical systems recognize them.

This makes them powerful coherence sensors.

They often detect:

• logical inconsistencies
• narrative incoherence
• emotional misalignment
• ethical contradictions
• structural unfairness
• policy misconfigurations
• institutional bottlenecks

What society often labels as “complaining,” “being too literal,” “being too sensitive,” or “fixating on details” is actually early detection of coherence breakdown. AToM reframes these traits as signals—not noise.

This sensitivity becomes especially valuable in:

• scientific research (precision pattern detection)
• technical work (low-error tolerance)
• ethical reasoning (sensitivity to norm contradiction)
• policy evaluation (detection of structural incoherence)
• organizational analysis (identification of misalignment)
• interpersonal dynamics (early detection of rupture)

Neurodivergence therefore becomes central—not peripheral—to the maintenance of coherence in complex systems.

8.7 Complementary Coherence Architectures

AToM emphasizes that human groups function best when different coherence phenotypes work together. Neurotypical smoothing and neurodivergent precision form a complementary pair:

• Neurotypical systems stabilize large-scale group coherence by smoothing irregularities.
• Neurodivergent systems protect structural integrity by detecting irregularities early.

One provides coherence breadth; the other coherence depth.

When both are present, a group can maintain stability while also adapting to subtle disruptions.

This complementarity explains:

• why neurodivergent individuals excel in high-variance creative fields
• why they contribute essential error detection in engineering or science
• why they often serve as the ethical or structural conscience of groups
• why mixed neurotypes produce more innovative and robust teams

AToM frames neurodiversity as an evolutionarily advantageous coherence architecture.

8.8 Neurodivergence as a Different Meaning Geometry

Neurodivergent cognition experiences meaning differently because meaning is coherence felt from the inside. A system with high precision, narrow smoothing, and altered curvature sensitivity will experience coherence—and thus meaning—with different thresholds and constraints. This explains:

• why autistic meaning often centers around pattern, structure, and truth
• why ADHD meaning often centers around novelty and momentum
• why neurodivergent individuals gravitate toward high-coherence domains (math, coding, systems, music)
• why ambiguous or contradictory environments feel disorienting

These are not quirks—they are properties of the underlying coherence geometry.

Meaning feels different because coherence is different.

8.9 Conclusion: Neurodivergence Is Not a Deficit—It Is a Geometry

In AToM, neurodivergence is not pathology. It is a difference in how coherence is constructed, preserved, and restored. Neurodivergent systems are precision sensors, early detectors of coherence fractures, and essential contributors to complex human systems. They reveal the geometry of meaning itself by making visible the coherence conditions that neurotypical smoothing conceals.

Neurodivergent cognition is not an exception to coherence.

It is a different form of coherence—one that society deeply needs.

9. The Coherence Sensor Stack: Contemporary Measurement Infrastructure 

AToM’s claim that coherence is a structural, measurable property is not speculative.

It is empirically actionable.

Modern technology now provides a multilayered “coherence sensor stack” capable of capturing coherence geometry across linguistic, physiological, behavioral, and topological modalities. These tools—most of which have emerged only in the last decade—allow AToM to transition from theoretical framework to empirical science. They enable the mapping, quantification, and prediction of coherence across individuals, relationships, organizations, and cultures.

The coherence sensor stack consists of four main layers:

1. linguistic coherence sensing (LLMs, embeddings, narrative geometry)
2. physiological coherence sensing (HRV, EDA, interoception, autonomic coupling)
3. topological coherence sensing (persistent homology, attractor analysis, bottleneck detection)
4. multimodal fusion architectures (systems that integrate all the above into unified coherence manifolds)

Each layer captures a different aspect of coherence geometry. Together, they render meaning measurable.

9.1 Linguistic Coherence Sensing: Embeddings, Entropy, and Narrative Geometry

Modern large language models (LLMs) operate in high-dimensional embedding spaces that encode relationships between words, sentences, and entire discourses. These models are not merely predictive; they are geometric instruments. They track curvature, integrability, and divergence across linguistic manifolds. This allows coherence to be quantified directly.

Several linguistic indicators serve as coherence measures:

Entropy Gradients

LLMs compute token-by-token uncertainty. Sharp increases in entropy correspond to coherence fractures—points where the discourse becomes disorganized, contradictory, or unstable.

Embedding Smoothness

When sentences or paragraphs move smoothly across embedding space, coherence is high. When they zigzag or jump unpredictably, coherence is low.

Narrative Curvature

Narrative arcs can be modeled as trajectories. High curvature indicates unexpected turns, emotional whiplash, or incoherent plot logic. Low curvature indicates stable, meaningful progression.

Cross-Document Coherence

Comparing embeddings across documents reveals alignment within organizations, institutions, or cultural groups. Divergence indicates fragmentation.

These methods allow AToM to operationalize coherence in linguistic systems—therapy transcripts, conversation logs, essays, organizational documents, policymaking discussions, or cultural narratives.

Linguistic coherence becomes a measurable geometric object.

9.2 Physiological Coherence Sensing: HRV, Vagal Tone, and Autonomic Geometry

Meaning collapses psychologically because coherence collapses biologically. Physiological coherence is the foundation of experiential coherence. Modern wearables allow continuous measurement of several physiological variables that index coherence geometry.

Heart-Rate Variability (HRV)

HRV reflects the flexibility of the autonomic nervous system. High HRV = high coherence; low HRV = reduced dimensionality and increased curvature.

Electrodermal Activity (EDA)

EDA measures sympathetic arousal. Sudden spikes signal prediction-error surges or coherence rupture.

Respiration Rhythms

Respiration patterns synchronize with affect and cognition. Smooth sinusoidal rhythms correspond to low curvature; erratic rhythms indicate instability.

Vagal Tone

Vagal regulation serves as the physiological backbone of coherence. Strong vagal tone correlates with stable manifold geometry.

Physiological Synchrony

Between individuals, coherence emerges as coupled HRV, synchronized breathing, shared autonomic rhythms. These are the biological signatures of trust, attunement, and relational safety.

Physiological coherence signals are thus real-time indicators of meaning, stability, and cross-scale alignment.

9.3 Topological Coherence Sensing: Persistent Homology and Attractor Mapping

Linguistic and physiological signals capture local coherence. Topological data analysis (TDA) captures global coherence: the structure of the manifold itself. TDA tools such as persistent homology, mapper, and Vietoris–Rips filtrations reveal the shape of high-dimensional data.

AToM uses these tools to model coherence and incoherence in structural terms:

H_0: Connectedness

Fragmented H_0 structure indicates isolated states or subgraphs—analogous to dissociation, siloing, or institutional fragmentation.

H_1: Loops and Attractor Cycles

Highly persistent loops indicate stable—but sometimes pathological—attractors.

Examples:

• rumination loops
• addictive cycles
• organizational dysfunction patterns
• cultural polarization cycles

H_2 and Higher: Cavities and Voids

Voids represent regions the system cannot access—states or experiences that are structurally blocked.

Example:

• emotional experiences that cannot be integrated
• institutional procedures that never connect
• cultural topics that remain unprocessed

Topological measures reveal the architecture of coherence. They show whether the system’s geometry is flexible, fractured, bottlenecked, or trapped in attractors.

TDA transforms coherence from intuition into measurable shape.

9.4 Multimodal Fusion: Unified Coherence Manifolds

The final layer of the coherence sensor stack integrates all preceding layers into a single manifold. Modern machine learning architectures—such as Perceiver, Perceiver IO, multimodal transformers, and diffusion-based fusion networks—can take inputs across modalities and align them into a unified latent space.

This makes cross-scale coherence measurable.

Multimodal fusion provides:

• alignment of linguistic coherence with physiological coherence
• integration of interpersonal synchrony with behavioral patterns
• mapping of organizational communication with cultural narratives
• identification of global coherence collapse before visible failure

For example:

• A conversation may appear verbally coherent but physiologically incoherent.
• A couple may exhibit synchronized speech patterns but desynchronized HRV.
• A company may display polished messaging but fractured topological communication structure.
• A culture may generate unified narratives but diverge drastically in physiological stress markers or social network topology.

Multimodal manifolds reveal these cross-modal coherence mismatches.

AToM predicts that systems fail when coherence collapses in one modality faster than others can compensate. Multimodal sensing detects these early-warning signs.

9.5 Coherence Reconstruction Through Sensor Feedback

The coherence sensor stack is not only descriptive—it is prescriptive.

It enables real-time coherence modulation.

Examples:

• HRV feedback helps individuals reduce curvature.
• Conversation analytics guide therapists toward coherence-restoring interventions.
• Organizational communication dashboards highlight misalignment before crisis.
• Multimodal cultural sensing detects early polarization and recommends stabilizing interventions.

The stack thus becomes not just a measurement system but a coherence-regulation infrastructure.

It gives systems the tools to detect drift, reduce curvature, restore dimensionality, reopen bottlenecks, and re-establish cross-scale entrainment.

Meaning becomes instrumentable.

9.6 The Coherence Stack as the Empirical Foundation of AToM

AToM asserts that meaning is coherence under constraint. The coherence sensor stack proves that coherence can be observed, quantified, and modified. It makes the key constructs of AToM—curvature, dimensionality, persistence, reversibility, entrainment—empirically tractable.

These tools allow researchers to:

• test AToM’s predictions
• falsify or refine its hypotheses
• build coherence datasets
• map coherence trajectories across therapy, learning, group dynamics, or cultural evolution
• integrate neurodivergent coherence profiles into systems-level models

Most importantly, the coherence sensor stack closes the theoretical loop. Meaning ceases to be a philosophical abstraction. It becomes a measurable structural invariant—and therefore a candidate for scientific unification.

The stack gives AToM the empirical backbone it needs to move from conceptual synthesis to a mature science of coherence.

If you want, I can now continue with Section 10 (Empirical Tests and Falsifiability) or begin assembling the entire manuscript into one continuous, publication-ready text.

9. Empirical Tests and Falsifiability

AToM makes multiple empirically testable predictions that integrate cleanly with FEP.

9.1 Neural Tests

Prediction:

Coherence breakdown should appear as increased cross-frequency curvature (greater error in phase coupling).

Method:

EEG or MEG coherence spectra.

9.2 Physiological Synchrony

Prediction:

Dyads approaching rupture show early HRV decoupling before subjective awareness.

Method:

Wearables measuring HRV, EDA, respiration.

9.3 Linguistic Coherence

Prediction:

Linguistic curvature spikes (entropy gradients) precede therapeutic breakthroughs or relational rupture.

Method:

LLM embedding analysis.

9.4 Organizational Breakdown

Prediction:

Topological bottlenecks in communication networks predict institutional failure.

Method:

TDA (persistent homology) on email/chat graphs.

9.5 Cultural Fragmentation

Prediction:

Divergent narrative attractors (memetic clusters) predict polarization.

Method:

Topic modeling + diffusion maps.

Each test directly corresponds to AToM’s coherence geometry and FEP’s expected surprise minimisation.

10. Empirical Tests and Falsifiability

AToM is not a metaphorical framework or speculative synthesis. It is a theory that makes specific, testable, falsifiable predictions. Because AToM defines meaning as coherence under constraint, and because coherence is a measurable geometric property of dynamical systems, the theory generates empirical claims across multiple scales. These claims can be evaluated using existing sensing technologies, data structures, and analytic methods—linguistic, physiological, topological, multimodal, and behavioral.

For a theory of meaning to be scientifically viable, it must satisfy three criteria:

1. It must define measurable variables.
2. It must make predictions that could be shown wrong.
3. It must integrate results across modalities and scales.

AToM meets all three.

This section outlines empirically falsifiable predictions derived from the coherence-curvature-inference structure at the heart of AToM. These predictions are not rhetorical; they map to measurable signals that contemporary tools can detect.

10.1 Prediction: Coherence Metrics Outperform Psychometric Self-Reports

AToM predicts that coherence metrics derived from physiological and linguistic signals will correlate with well-being, stability, and therapeutic progress more strongly than traditional psychometric self-report instruments.

Test:

Collect HRV, respiration, EDA, and linguistic curvature data during therapy sessions, daily check-ins, or ecological momentary assessment. Compare coherence composite scores against standard symptom scales (PHQ-9, GAD-7, PCL-5, etc.).

Falsification condition:

If psychometric instruments consistently outperform coherence metrics in predicting future stability or crisis, the theory’s claims about coherence as a superior invariant weaken.

Rationale:

HRV variance, respiratory smoothness, and linguistic curvature track integrative function directly; self-report measures track narrative interpretation, which may lag behind structural changes.

10.2 Prediction: Linguistic Coherence Shifts Precede Therapeutic Breakthroughs

AToM asserts that changes in linguistic coherence occur before patients consciously report insight or change. This reflects shifts in the underlying coherence geometry prior to narrative awareness.

Test:

Analyze therapy transcripts using embedding smoothness, entropy gradients, and narrative curvature. Look for systematic coherence increases preceding breakthrough sessions.

Falsification condition:

If narrative coherence increases only after subjective reports of change, then AToM’s claim that geometry precedes narrative is incorrect.

Rationale:

Generative models update at the structural level before they update at the narrative level; this is a direct implication of both AToM and active inference.

10.3 Prediction: Autistic + LLM Systems Detect Incoherence Earlier Than Neurotypicals

AToM argues that autistic cognition is a high-resolution coherence sensor. When paired with the precision of LLMs, these hybrid teams should detect inconsistencies—logical, relational, organizational, narrative—earlier and more reliably than neurotypical raters.

Test:

Provide mixed groups (autistic, neurotypical, autistic+LLM, neurotypical+LLM) with transcripts, organizational protocols, or conflict scenarios. Measure accuracy and speed in detecting inconsistencies, contradictions, or incoherent structures.

Falsification condition:

If autistic+LLM teams do not outperform neurotypical teams, the coherence-sensing hypothesis weakens.

Rationale:

Neurotypical smoothing masks incoherence; autistic precision reveals it.

10.4 Prediction: Dyadic Entrainment Breakdowns Forecast Relationship Rupture

AToM predicts that loss of physiological synchrony—especially HRV coupling—will precede visible interpersonal conflict or relational rupture.

Test:

Measure HRV synchrony in couples, therapeutic dyads, parent–child pairs, or team partnerships. Track synchrony over weeks or months alongside self-reported relational quality.

Falsification condition:

If relational rupture or distress occurs without preceding decoupling of physiological synchrony, AToM’s entrainment model is incomplete.

Rationale:

Before a relationship “breaks,” cross-blanket prediction error rises and coupling collapses. The body detects incoherence before awareness does.

10.5 Prediction: Organizational Collapse Is Preceded by Topological Bottlenecks

AToM predicts that organizations approaching functional or ethical breakdown exhibit characteristic topological deformation: communication networks collapse into bottlenecks, disjoint clusters intensify, and high-persistence cycles form around dysfunctional patterns.

Test:

Apply persistent homology (H_0, H_1) to email metadata, messaging graphs, code repositories, decision trees, or workflow logs in organizations undergoing stress.

Falsification condition:

If organizational failure occurs without preceding bottlenecks or fragmentation, the theory’s organizational coherence claims weaken.

Rationale:

Institutional meaning collapses when coherence geometry collapses. Bottlenecking is the structural signature of this collapse.

10.6 Prediction: Cultural Polarization Exhibits Diverging Narrative Curvature

AToM predicts that polarization is not simply ideological difference but a divergence in coherence geometry. Competing subcultures develop different narrative attractors that diverge in curvature over time, reducing integrability across groups.

Test:

Use embedding trajectories to analyze narrative curvature in social media, news sources, or cultural texts across different demographic groups.

Falsification condition:

If polarization occurs without measurable divergence in narrative curvature, AToM’s cultural coherence model would be incomplete.

Rationale:

Group-level coherence requires shared narrative scaffolds. Divergence in narrative curvature signals breakdown of cross-group integrability.

10.7 Prediction: Trauma-Relapse Risk Is Detectable Through Topological Narrowing

Trauma is modeled as collapse in manifold dimensionality and topological connectivity. AToM predicts that trauma relapse will occur when bottlenecks reappear in linguistic, physiological, or behavioral TDA signatures.

Test:

Track a trauma survivor’s coherence manifold over time; measure H_0 and H_1 structure under stress. Look for narrowing prior to relapse events.

Falsification condition:

If relapse occurs without preceding geometric narrowing, AToM’s trauma model fails.

Rationale:

Trauma re-emerges when coherence geometry contracts back toward pathological attractors.

10.8 Prediction: Coherence Across Modalities Converges in Healthy Systems and Diverges in Fragmented Systems

A healthy system exhibits cross-modal coherence: linguistic, physiological, behavioral, and relational signals align. A fragmented system exhibits divergence: each modality reflects a different geometry.

Test:

Feed multimodal data (physiology, language, behavior, relationship patterns) into a multimodal embedding model and track alignment.

Falsification condition:

If individuals in good psychological health exhibit strong modal divergence, or if individuals in crisis exhibit strong modal convergence, AToM’s coherence-integration hypothesis would be contradicted.

Rationale:

Cross-scale coupling is the core of coherence. Divergence signals fragmentation.

10.9 Prediction: AI-Assisted Coherence Reconstruction Enables Faster Recovery

AToM predicts that tools capable of monitoring and modulating coherence geometry—LLM-based linguistic coaching, HRV biofeedback, multimodal pattern detection—will accelerate therapeutic, relational, and organizational recovery.

Test:

Randomize participants into coherence-tracking interventions versus standard treatment; measure speed and durability of recovery.

Falsification condition:

If coherence-informed interventions offer no improvement over traditional approaches, AToM’s applied predictions weaken.

Rationale:

If coherence is the invariant, then direct coherence instrumentation should out-perform approaches that address content alone.

10.10 Conditions Under Which AToM Could Be Refuted

To make AToM genuinely falsifiable, we specify clear refutation conditions:

• coherence metrics fail to predict breakdown more accurately than traditional tools
• linguistic curvature shifts do not precede change
• physiological synchrony does not predict relational stability
• topological bottlenecks do not forecast trauma or organizational collapse
• narrative curvature does not correlate with cultural fragmentation
• neurodivergent precision does not detect coherence fractures
• multimodal fusion fails to integrate coherence across modalities
• coherence-informed interventions show no measurable advantage

Failure in any of these domains would weaken or refute core claims.

10.11 Summary: Coherence as a Scientific Construct

AToM satisfies the criteria of scientific theory:

• It defines measurable variables (curvature, dimensionality, persistence, synchrony).
• It produces non-trivial, falsifiable predictions.
• It integrates neuroscience, psychology, physiology, linguistics, organizational theory, and cultural analysis.
• It specifies clear failure points.

Meaning becomes an empirical question—coherence can be measured, tested, predicted, and restored.

AToM thus stands as a domain-appropriate, mathematically grounded, scientifically falsifiable framework for understanding coherence across human systems. It moves meaning from metaphor to measurement.

11. Conclusion: Meaning as Coherence

AToM advances a unified claim: meaning is coherence under constraint. Across neurological, physiological, relational, organizational, and cultural scales, systems flourish when they maintain integrated, low-entropy trajectories, and fracture when coherence collapses. Meaning is not an abstraction hovering above these processes; it is their felt signature. Meaning is what coherence feels like from the inside. Coherence is what meaning looks like from the outside.

This manuscript has shown that the Free-Energy Principle already contains the mathematical skeleton of coherence: living systems must minimize long-run expected surprise to maintain their form. AToM takes this skeleton and gives it the musculature, connective tissue, and multi-scale architecture required to explain human meaning. It provides the structural language that allows expected-surprise minimisation to be understood simultaneously as neural prediction, physiological regulation, interpersonal attunement, group alignment, institutional stability, and cultural continuity.

Meaning emerges only when coherence propagates across these scales. When neural predictions align with sensory signals, when bodily rhythms synchronize with context, when interpersonal attunement supports emotional integration, when organizational processes match collective purpose, and when cultural narratives provide stable scaffolding for interpretation—meaning is high. When coherence fails at any of these layers, meaning collapses.

AToM integrates diverse fields by identifying coherence as their common structural invariant. Predictive processing highlights prediction error minimisation; enactivism highlights sensorimotor coupling; attachment theory highlights relational co-regulation; trauma science highlights coherence collapse; information geometry highlights curvature and dimensionality; and cultural evolution highlights the transmission of stabilizing patterns. AToM weaves these into a coherent geometry.

This geometry has six recurring invariants that define coherent systems:

1. Curvature: Coherent systems exhibit low curvature—small perturbations produce manageable changes. Incoherent systems have high curvature—tiny deviations lead to large destabilizations.
2. Dimensionality: Coherent systems maintain flexible degrees of freedom. Trauma collapses dimensionality, narrowing the system’s ability to adapt.
3. Persistence: Coherent systems preserve stable structural patterns. Incoherent systems exhibit pathological attractors and bottlenecks that trap them in maladaptive cycles.
4. Coupling: Coherent systems synchronize across levels. Incoherent systems lose alignment between neural, physiological, relational, organizational, or cultural layers.
5. Boundaries: Coherent systems maintain flexible, permeable Markov blankets. Incoherent systems experience ruptures—too rigid, too porous, or unstable boundaries.
6. Reversibility: Coherent systems return to stability after perturbation. Trauma introduces hysteresis—making return to prior states costly or impossible.

These invariants appear everywhere in human systems. They describe the neural dynamics of emotional regulation, the physiological rhythms of safety, the interpersonal timing of attunement, the organizational patterns of alignment, and the cultural structures of shared meaning. They show that meaning is not primarily symbolic but structural.

AToM’s definition—meaning as coherence under constraint—therefore provides a single, domain-appropriate explanation capable of integrating vast areas of the cognitive, social, and behavioral sciences. It avoids reduction (e.g., turning culture into neurons) and avoids metaphorical overreach (e.g., treating physics as a universal model for meaning). It preserves domain-specific complexity while identifying cross-scale invariants. It explains not only what meaning is, but how meaning fails and how it can be restored.

Meaning fails when coherence collapses. Trauma collapses coherence geometry: it increases curvature, reduces dimensionality, introduces bottlenecks, imposes hysteresis, and fractures boundaries. Interpersonal breakdowns—misattunement, conflict, betrayal—collapse dyadic coherence. Organizational dysfunction collapses group coherence. Cultural polarization collapses collective coherence. The same signatures appear across all layers, because coherence is a cross-domain invariant.

Meaning is restored through entrainment. Neural entrainment integrates oscillatory patterns; physiological entrainment stabilizes autonomic rhythms; interpersonal entrainment re-establishes mutual prediction; group entrainment aligns shared goals; cultural entrainment synchronizes narratives across populations. Entrainment is coherence reconstruction, the inverse of trauma. It smooths curvature, expands dimensionality, opens topological pathways, reduces hysteresis, and reintegrates boundaries.

AToM’s information-geometric tools—curvature, KL gradients, persistent homology, dimensionality analysis, reversibility metrics—provide a rigorous vocabulary for measuring coherence. These tools transform coherence from an intuitive idea into an empirical construct. The coherence sensor stack then provides the measurement infrastructure—linguistic embeddings, physiological telemetry, topological data analysis, multimodal fusion architectures—to test, refine, and falsify the theory.

AToM therefore offers a fully testable research program. It predicts that coherence metrics outperform traditional psychometric tools; that linguistic curvature shifts precede therapeutic breakthroughs; that physiological synchrony predicts relational stability; that topological bottlenecks precede organizational collapse; that narrative divergence predicts cultural fragmentation; and that coherence-informed interventions improve outcomes. Each prediction can be empirically supported or contradicted, giving the theory the falsifiability required for scientific legitimacy.

Crucially, AToM establishes a domain boundary. It positions physics as the high-symmetry, low-variability pole where coherence is rigidly enforced by invariance. Meaning systems inhabit the opposite pole: high variability, soft constraint, developmental plasticity, and multiple viable coherence geometries. AToM does not attempt to derive meaning from physics; it uses physics as a boundary case to clarify where AToM properly applies. It is a theory of coherence for systems whose stability depends on adaptive negotiation rather than rigid symmetry.

The theory also reframes neurodivergence. Autistic cognition becomes not a deficit but a high-resolution coherence-sensing system. ADHD becomes a tempo-mismatch architecture. Dyslexia becomes an alternative mapping between linguistic compression and perceptual structure. These coherence geometries are not errors; they are variations in how systems detect, maintain, or restore coherence. Neurodivergence becomes central to understanding human meaning, not peripheral.

Ultimately, AToM offers a way to unify the sciences of mind, behavior, organization, and culture without collapsing them into one another. It connects predictive processing, enactivism, trauma science, organizational theory, cultural evolution, and AI research under a single structural insight: meaning is coherence maintained through constraint, and coherence is the geometry of expected-surprise minimisation across nested systems.

Meaning is not a story we tell about ourselves—it is the structural condition that makes any story possible. It is the stability that allows prediction, action, identity, relationship, and culture to function. When coherence holds, meaning flows. When coherence breaks, meaning dissolves.

AToM provides the geometry, the mathematics, and the measurement framework for understanding that process across all human scales. It offers not a metaphor, but a science of meaning—empirically anchored, mathematically grounded, and aligned with the fundamental invariants that define human experience.

Coherence is the central variable.

Meaning is the felt sense of coherence.

AToM is the science of both.