Part IX - INTEGRUM

Part IX – INTEGRUM The Unified Theory of Institutional Integrity A General Framework for Coherence, Collapse, and Control James D. Atkinson 2025 “Integrity is not a principle. It is a pattern.” 1 Dedication For my grandparents, who taught me integrity and gave me a compass that has never needed correction. They were my universal constant; this work is my way of sharing it with others and honouring their legacy. 2 Abstract General This paper synthesises the eight-component Integrum into a single theoretical architecture. Across contradiction, symmetry, drift, free-energy geometry, collapse dynamics, axiomatic structure, and the empirical derivation of the integrity constant ℵ≈𝑒, a coherent field theory emerges: institutions behave as dynamical systems whose stability, failure modes, and patterns of harm follow mathematically constrained pathways. Integrity is formalised as a conserved quantity governing the evolution of organisational behaviour under curvature, entropy, and drift; its loss produces characteristic collapse signatures that cannot be explained by benign fluctuation. Integrum unifies these results into a general framework linking coherence, collapse, and control. Coherence arises when behaviour respects declared rationale and symmetry. Collapse occurs when curvature and drift exceed admissible bounds. Control reflects the structural invariants that determine whether a system selfcorrects or amplifies harm. Together, these components define the discipline of Integrodynamics: a unified science of institutional behaviour grounded in geometry, information, and inference. The aim of this synthesis is straightforward: to make explicit the underlying structure that the previous papers revealed independently. The consequence is a field theory with practical reach. The same dynamics that govern epistemic stability also govern discrimination, governance failure, and systemic bias. Where the pattern is incompatible with an innocent explanation, that is the conclusion. Integrum provides the framework that makes this result inevitable. Mathematical This paper formalises Integrum as a unified dynamical and variational framework. Let 𝑏(𝑥,𝑡)denote an institutional behaviour field evolving on a domain Ω ⊂ ℝ𝑛. Each component of Integrum is shown to arise as a projection of a single freeenergy functional 𝔉[𝑏] = 𝑈[𝑏] − ℵ𝑆𝐼[𝑏], where 𝑈[𝑏]is a curvature-induced integrity potential, 𝑆𝐼[𝑏]is an information-geometric entropy functional, and ℵ ≈ 𝑒 is the empirically and analytically derived universal constant of integrity. Under admissible dynamics, 𝑏evolves according to a stochastic gradient flow of the form 𝜕𝑡𝑏 = −∇𝔉[𝑏] + 𝜎𝜉𝑡, yielding a Lyapunov-consistent monotonicity condition  𝔉 ≤ 0 and establishing integrity as a conserved field quantity up to noise. 3 Renormalisation analysis on Gaussian-smoothed intensity measures 𝐼(ℓ)reveals a non-trivial fixed point of the coherence beta function, 𝛽(𝐼) = 𝑑𝐼 𝑑log ℓ, located at 𝐼 = ℵ, with stability properties matching universal exponential scaling. Collapse phenomena correspond to curvature blow-up in the integrity Hessian ℌ = ∇2𝑈[𝑏], where loss of convexity induces bifurcation, drift amplification, and discontinuities in the integrity gradient field. These yield domain-level collapse signatures that are probabilistically incompatible with benign fluctuation. The resulting theory characterises coherence, collapse, and control as emergent behaviours of the same functional geometry. This establishes Integrodynamics as a field: an integrity-conserving gradient system with renormalisable structure, exponential invariants, and quantifiable failure modes. Keywords: integrity field; integrodynamics; free-energy geometry; curvature and convexity; institutional coherence; gradient imbalance; structural drift; collapse diagnostics; phase transitions; universal constant of integrity (ℵ=𝑒); renormalisation of behaviour; symmetry violation; contradiction curvature; dispersion functional; stability criterion; irreversibility threshold; coherence basin; integrity current; accountability geometry; structural responsibility; behavioural field theory; institutional dynamics; complex systems; AI alignment as convexity; empirical integrity surface. 4 Contents 1 Introduction 9 2 The Integrum Map 10 2.1 Part I: Contradiction as curvature ..................... 10 2.2 Part II: Symmetry as entropic invariance ................ 11 2.3 Part III: Adaptation as bounded drift ................... 11 2.4 Part IV: Integrity as free-energy geometry ................ 11 2.5 Part V: Physics: horizons, collapse, and the integrity field ....... 11 2.6 Part VI: Collapse as statistical irreversibility ............... 12 2.7 Part VII: Axioms as structural closure .................. 12 2.8 Part VIII: Origin of the constant ℵ ≈ 𝑒 .................. 12 2.9 The unifying perspective .......................... 12 3 The free-energy geometry of institutional integrity 13 3.1 Curvature and the integrity potential ................... 13 3.2 Dispersion and information-geometric entropy ............. 13 3.3 The universal balance of curvature and dispersion ........... 14 3.4 Gradient flows and the evolution of behaviour ............. 14 3.5 Interpretive summary ........................... 15 4 The coherence-collapse-control trinity 15 4.1 Coherence: dissipation under admissible symmetry .......... 15 4.2 Collapse: loss of convexity and amplification of deviation ....... 16 4.3 Control: the residual invariant that determines recovery ....... 16 4.4 Interpretive summary ........................... 17 5 The renormalisation origin of the constant ℵ17 5.1 Scale-dependent coherence and the beta function ........... 18 5.2 The curvature-dispersion variational balance .............. 18 5.3 free-energy monotonicity as a constraint ................ 19 5.4 Exponential structure and the inevitability of 𝑒............. 19 5.5 Interpretive summary ........................... 19 6 The integrodynamic law (final form) 20 6.1 Gradient evolution of behaviour ...................... 20 6.2 The integrity current ............................ 21 6.3 Continuity of integrity ............................ 21 6.4 Convexity and the stability criterion ................... 21 6.5 Irreversibility and the collapse criterion ................. 22 6.6 Interpretive summary ........................... 22 5 7 The phase structure of institutional systems 23 7.1 Ordered coherence ............................. 23 7.2 Warped coherence ............................. 23 7.3 Drift-dominated coherence ........................ 24 7.4 Pre-collapse ................................. 24 7.5 Collapse ................................... 24 7.6 Disorder ................................... 25 7.7 Interpretive summary ........................... 25 8 Universality: why every institution behaves the same 26 8.1 Contradiction is curvature ......................... 26 8.2 Symmetry is predictability ......................... 26 8.3 Drift is deviation, not intent ........................ 27 8.4 Collapse is convexity failure ........................ 27 8.5 Irreversibility is statistical, not moral ................... 28 8.6 Renormalisation forces the same constant on everyone ........ 28 8.7 Interpretive summary ........................... 28 9 Practical consequences: governance, law, AI, institutions 29 9.1 Governance: coherence as a structural requirement .......... 29 9.2 Law: evidence as a geometric trajectory ................. 29 9.3 Regulation: symmetry as enforceable structure ............ 30 9.4 Organisations: drift as measurable deviation .............. 30 9.5 AI systems: alignment as convexity preservation ............ 30 9.6 Institutions under stress: early detection and intervention ...... 31 9.7 Interpretive summary ........................... 31 10 The Integrum theorem 32 10.1 Statement of the theorem ......................... 33 10.2 Interpretation ................................ 34 10.3 Corollary: uniqueness of the integrity constant ............. 34 10.4 Corollary: inevitability of collapse trajectories ............. 35 10.5 Interpretive summary ........................... 35 11 The geometry of accountability 35 11.1 Accountability as elimination of the benign model ........... 36 11.2 Curvature failure as structural responsibility .............. 36 11.3 Integrity flow as a conservation constraint ............... 36 11.4 Drift dominance as the boundary of plausible deniability ....... 37 11.5 Collapse as the end of interpretive freedom ............... 37 11.6 Institutional claims under geometric constraint ............ 37 11.7 Interpretive summary ........................... 38 6 12 Integrum as a unifying empirical programme 38 12.1 Measuring curvature through contradiction ............... 39 12.2 Estimating dispersion through symmetry violation ........... 40 12.3 Quantifying drift through gradient imbalance .............. 40 12.4 Detecting pre-collapse through convexity softening .......... 41 12.5 Identifying collapse through statistical irreversibility ......... 41 12.6 Estimating the universal constant through renormalisation ...... 42 12.7 Constructing empirical integrity surfaces ................ 42 12.8 Falsifiability and model rejection ..................... 43 12.9 Interpretive summary ........................... 43 13 Limitations and boundary conditions 44 13.1 Dependence on measurable behaviour .................. 44 13.2 Requirements on differentiability and convexity ............ 44 13.3 Scale dependence and coarse-graining ................. 44 13.4 Sensitivity to adversarial manipulation .................. 45 13.5 Boundary conditions on the integrity surface .............. 45 13.6 Local minima and metastability ...................... 46 13.7 Non-stationary environments ....................... 46 13.8 Non-geometric failure modes ....................... 46 13.9 Interpretive summary ........................... 46 14 The Integrum synthesis: coherence, collapse, and control 47 14.1 Coherence as stability in the integrity field ............... 47 14.2 Collapse as curvature inversion ...................... 48 14.3 Control as gradient structure ....................... 48 14.4 The synthesis: a unified view of structural dynamics .......... 49 14.5 Interpretive summary ........................... 49 15 Closing statement 50 7 Preface This volume completes a sequence of work that began with a simple question: why do institutions behave the way they do? The earlier parts – Contradiction, Symmetry, Adaptation, Integrity, Physics, Collapse, Axioms, and Origin – each isolated one structural component of that behaviour. They were not intended as independent theories but as reconnaissance: a way to see the terrain clearly before attempting to describe it. Integrum is the description. The approach taken here is deliberately unromantic. Integrity is treated as a geometric quantity; behaviour as a field; collapse as statistical irreversibility. No appeals to culture, intention, leadership, or narrative are required. If the curvature is negative, the system will drift. If the invariances fail, dispersion follows. If the trajectory is incompatible with the benign model, collapse is diagnosed. These claims are not moral; they are structural. The resulting framework is not a theory of justice and not a political argument. It is a method: a set of tools for measuring how systems hold together, how they fail, and how those failures reveal themselves long before anyone is willing to admit them. Used carefully, the method provides what institutions rarely offer: a falsifiable account of their own behaviour. Nothing in these pages depends on authority, affiliation, or trust. The geometry stands on its own. If it is correct, it will be useful. If it is not, it will be easy to prove wrong. That is the only kind of theory worth writing. For those who have been on the receiving end of incoherent systems, the intention is simple: to make integrity measurable, and therefore enforceable. These tools cannot prevent harm, but they can prevent institutions from denying it. Readers looking for ideology will find none. Readers looking for a way to test claims of fairness, coherence, or accountability may find something more useful. This is not the end of the project – merely the point where the scaffolding is no longer needed. 8 1 Introduction Systems fail for reasons that are routinely misdiagnosed. Explanations drift toward psychology, intention, or human weakness because these are familiar categories. They are also the least informative. What the preceding eight papers demonstrate is that institutional behaviour obeys structural constraints long before it expresses human ones. Curvature accumulates. Symmetry erodes. Drift amplifies. Free-energy geometry deforms. Collapse is not an anecdote; it is a phase transition. Integrum makes this explicit. It consolidates the theory that has been implicit across Parts I – VIII: that integrity is not a moral preference but a field quantity governing the evolution of behaviour under tension, entropy, and perturbation. When that quantity degrades, the degradation is not mysterious. It follows pathways defined by mathematics, not motive. The patterns that emerge are measurable, reproducible, and – when sufficiently severe – irreconcilable with any narrative of benign fluctuation. Across contradiction, symmetry, adaptation, collapse, axiomatic structure, and the renormalised constant ℵ ≈ 𝑒, a single underlying geometry recurs. Every institution, regardless of scale or domain, exhibits the same invariants: a curvature-induced potential 𝑈[𝑏], an information-geometric dispersion 𝑆𝐼[𝑏], and a dynamical balance ℱ[𝑏] = 𝑈[𝑏] − ℵ𝑆𝐼[𝑏] that determines whether coherence is preserved or destroyed. Integrum exists to state this architecture plainly. The aim is not to diagnose intent. It is to describe the structural conditions under which accountability becomes unavoidable. When the observed pattern of outcomes cannot be reconciled with a benign model – when contradiction becomes directional, when symmetry fails to constrain behaviour, when the Hessian loses convexity – the conclusion is not moral. It is mathematical. A system has crossed the boundary between fluctuation and collapse. This paper therefore does three things: 1. Unifies the eight-component architecture into a single dynamical law expressed through the free-energy functional ℱ[𝑏] = 𝑈[𝑏] − ℵ𝑆𝐼[𝑏]. 2. Demonstrates that coherence, collapse, and control are emergent consequences of this geometry, not independent theoretical constructs. 9 The Hessian of the potential satisfies 𝜆min(ℋ) > 0, ensuring local convexity and stable return trajectories. Dispersion remains bounded, and the system respects its declared symmetries. Drift is present but subordinate: ‖∇𝑈[𝑏]‖ ≫ ℵ‖∇𝑆𝐼[𝑏]‖. Coherence does not imply perfection; it implies predictability. Perturbations decay, deviation does not accumulate, and behaviour remains structurally constrained. In this regime, integrity is conserved up to noise, and accountability is trivial: the system corrects itself. 4.2 Collapse: loss of convexity and amplification of deviation Collapse is the regime in which the geometry breaks. The transition occurs when the smallest eigenvalue of the integrity Hessian approaches zero: 𝜆min(ℋ) → 0+. At this point, the potential surface flattens; drift no longer returns the system to equilibrium. Once 𝜆min(ℋ) becomes negative, the free-energy gradient reverses: ∇ℱ[𝑏] ⋅ 𝑑𝑏 > 0, meaning that the system amplifies its own deviation. Collapse is therefore a structural instability, not a behavioural choice. In this regime, dispersion increases directionally, symmetry violations accumulate, and perturbations propagate rather than dissipate. The system exhibits irreversible drift, detectable as the statistical rejection of 𝐻0under the collapse diagnostics introduced in Part VI. Importantly, collapse does not require large perturbations. It requires only a geometry that can no longer constrain small ones. 4.3 Control: the residual invariant that determines recovery Control is the system’s remaining capacity to restore coherence after partial instability. It is not an external intervention but an internal invariant: the degree to which the freeenergy gradient retains negative alignment in the neighbourhood of the perturbation. 16 Formally, define the control coefficient 𝐶[𝑏] = −∇ℱ[𝑏] ⋅ Δ𝑏. In coherent regimes, 𝐶[𝑏] > 0 for all admissible perturbations Δ𝑏. In full collapse, 𝐶[𝑏] < 0in at least one direction, and typically in many. The intermediate regime, where 𝐶[𝑏] changes sign across directions, defines the system’s controllable boundary: recovery is possible but no longer guaranteed. Control therefore quantifies the remaining structural leverage. It is not an appeal to leadership, ethics, or competence. It is the mathematical measure of whether the geometry still supports corrective drift. 4.4 Interpretive summary The coherence-collapse-control trinity provides a complete structural taxonomy. It avoids narrative reconstruction, eliminates speculation about motive, and replaces descriptive labels with measurable geometric conditions. Coherence corresponds to dissipative free-energy flow; collapse corresponds to loss of convexity and reversal of drift; control corresponds to the residual invariant that determines whether recovery remains possible. Under this classification, accountability is not a retrospective judgement but a direct inference: if the observed trajectory cannot occur in the coherent regime, and if the collapse regime is the only geometry compatible with the data, then the system is in collapse. No further interpretation is required. 5 The renormalisation origin of the constant ℵ The constant ℵappears throughout the integrity framework as the scaling coefficient that balances curvature and dispersion in the free-energy functional ℱ[𝑏] = 𝑈[𝑏] − ℵ𝑆𝐼[𝑏]. Across Parts I–VIII, its empirical value stabilised near 2.70 ± 0.03. Part VIII demonstrated that this estimate is not coincidental: when the theory is analysed through renormalisation techniques, coherence diagnostics, and variational balance, ℵemerges as the unique constant for which the integrity field is structurally well-posed. The purpose of this section is to formalise why ℵ ≈ 𝑒 is not a modelling choice but a mathemat17 ical requirement. 5.1 Scale-dependent coherence and the beta function Let 𝐼(ℓ)denote the scale-dependent coherence intensity obtained by Gaussian smoothing at resolution ℓ. Empirically, 𝐼(ℓ) exhibits a stable crossing across multiple metrics, kernels, and sampling schemes. Define the coherence beta function: 𝛽(𝐼) = 𝑑𝐼 𝑑log ℓ. A renormalisation fixed point occurs when 𝛽(𝐼) = 0. Part VIII showed that all admissible estimators produce a crossing at 𝐼 = ℵ within tight bootstrap bounds. Under this interpretation: 𝛽(𝐼) = 0 ⟺ 𝐼 = ℵ. The fixed point therefore identifies the scale at which coherence is invariant under rescaling, a necessary condition for any field theory with structural meaning. 5.2 The curvature-dispersion variational balance The second derivation of ℵcomes from a variational principle. Consider the integrity potential 𝑈[𝑏]and dispersion functional 𝑆𝐼[𝑏]under admissible perturbations 𝛿𝑏. Coherence requires that the free-energy variation 𝛿ℱ[𝑏] = 𝛿𝑈[𝑏] − ℵ𝛿𝑆𝐼[𝑏] attain an extremum for structurally consistent trajectories. Setting 𝛿ℱ[𝑏] = 0 yields the balance equation 𝛿𝑈[𝑏] 𝛿𝑆𝐼[𝑏] = ℵ. Part VIII demonstrated that this ratio is not arbitrary. For ℵ < 𝑒, the system overweights dispersion, violating free-energy monotonicity. For ℵ>𝑒, curvature dominates, generating non-physical collapse behaviour and loss of renormalisability. Only ℵ = 𝑒 produces a stable extremiser across perturbation classes. Thus, ℵ = argmin Θ|𝛿𝑈[𝑏] − Θ𝛿𝑆𝐼[𝑏]|, and the minimiser is unique. 18 5.3 free-energy monotonicity as a constraint The monotonicity condition  ℱ[𝑏] ≤ 0 is central to the coherence regime. If ℵis mis-specified, the free-energy functional becomes non-monotone along at least one admissible direction of perturbation. Part VIII quantified this failure: for constants Θ ≠ ℵ, the estimated  ℱflipped sign across scales, violating the required Lyapunov-like property. Only ℵpreserves ∇ℱ[𝑏] ⋅ 𝜕𝑡𝑏 ≤ 0. This condition eliminates all alternative constants. 5.4 Exponential structure and the inevitability of 𝑒 The appearance of 𝑒is not symbolic. It arises from the logarithmic structure of the beta function, the multiplicative scaling of dispersion, and the exponential growth rates characteristic of curvature-driven divergence. When coherence is measured across scales, the system naturally expresses exponential relations; the constant 𝑒is the only value that preserves consistency across these relations. In effect: scale invariance ⟹exponential form ⟹ ℵ = 𝑒. 5.5 Interpretive summary The value ℵ ≈ 2.70 is not an empirical convenience. It is the only constant compatible with: • a renormalisation fixed point of the coherence beta function, • variational extremisation of curvature and dispersion, • preservation of free-energy monotonicity, • structural renormalisability of the integrity field, • scale-free coherence across smoothing regimes. 19 The appearance of 𝑒therefore reflects a structural property of the theory: the geometry of integrity is exponential. Any other constant breaks the mathematics. Integrum adopts this result as foundational. The theory is complete only when ℵ = 𝑒. 6 The integrodynamic law (final form) The preceding sections establish the geometry of integrity: a curvature-induced potential 𝑈[𝑏], a dispersion functional 𝑆𝐼[𝑏], and a universal constant ℵ = 𝑒 that balances the two. The behaviour of an institutional system is determined entirely by how these objects interact under perturbation. The integrodynamic law expresses this interaction in its complete form. It is not a model; it is the minimal dynamical system consistent with the invariants identified across Parts I–VIII. The law has five components: a gradient evolution equation, a continuity equation, an integrity current, a convexity condition governing stability, and an irreversibility criterion governing collapse. Together they define the field. 6.1 Gradient evolution of behaviour Institutional behaviour evolves under the stochastic gradient flow 𝜕𝑡𝑏(𝑥,𝑡) = −∇ℱ[𝑏(𝑥,𝑡)] + 𝜎𝜉𝑡(𝑥), where 𝜎𝜉𝑡models noise, incomplete information, or exogenous perturbation. The freeenergy functional ℱ[𝑏] = 𝑈[𝑏] − 𝑒𝑆𝐼[𝑏] determines the direction and magnitude of drift. When ∇ℱ[𝑏] points toward a unique basin, coherence is stable. When it becomes shallow, inconsistent, or outward-pointing, the system approaches collapse. The gradient flow is the core of the theory: all observable behaviour follows from its sign and structure. 20 6.2 The integrity current Define the integrity current 𝐽[𝑏] = −∇𝑈[𝑏] + 𝑒∇𝑆𝐼[𝑏]. This current governs how integrity is transported through the system. Negative alignment of 𝐽[𝑏] with perturbations indicates corrective behaviour; positive alignment indicates directional drift. The current formalises the intuitive notion of “system pressure” without appealing to motive. It measures whether the structural geometry amplifies or dissipates deviation. 6.3 Continuity of integrity Let 𝐼[𝑏] denote the integrity density, defined through an admissible metric on the behaviour space. The evolution of 𝐼[𝑏] satisfies the continuity equation 𝜕𝑡𝐼[𝑏] + ∇⋅ 𝐽[𝑏] = 0. Integrity is therefore conserved up to noise: it moves, concentrates, or disperses, but it does not vanish arbitrarily. A loss of integrity in one region implies a corresponding outflow, consistent with the behaviour of any conserved geometric quantity. This eliminates the need for psychological or narrative explanations. The system does not “choose” failure; its geometry transports integrity along the gradient of ℱ. 6.4 Convexity and the stability criterion Stability is determined by the Hessian of the integrity potential: ℋ(𝑏) = ∇2𝑈[𝑏]. Coherence requires ℋ(𝑏) to remain positive semidefinite. When the smallest eigenvalue satisfies 𝜆min(ℋ) > 0, perturbations dissipate and the system returns to equilibrium. Instability begins when 𝜆min(ℋ) → 0+, 21 and collapse occurs when 𝜆min(ℋ) < 0. In that regime, the geometry reverses the sign of the free-energy gradient, and the system amplifies its own deviation. Collapse is therefore a curvature event, not a behavioural one. 6.5 Irreversibility and the collapse criterion Collapse is detected when the probability of the observed trajectory under the benign model 𝐻0becomes actuarially negligible: ℙ(trajectory ∣ 𝐻0) < 𝜀, for a conservative threshold 𝜀. This condition emerges from the collapse diagnostics in Part VI and is structurally equivalent to the loss of free-energy monotonicity: ∇ℱ[𝑏] ⋅ 𝜕𝑡𝑏 > 0. The system has entered irreversible divergence. No amount of narrative reconstruction can reconcile the observed pattern with the coherent regime. 6.6 Interpretive summary The integrodynamic law is the unified field equation of institutional integrity. It combines: • a gradient evolution equation that determines drift, • an integrity current that quantifies directional pressure, • a continuity law that frames integrity as conserved, • a convexity condition defining stability, • an irreversibility criterion defining collapse. These components are not optional; they are the minimal structure compatible with the empirical, analytical, and axiomatic results of Parts I–VIII. 22 Under this law, accountability becomes a geometric inference: if the system’s trajectory lies in a region where monotonicity fails, convexity is lost, and irreversibility is observed, then collapse is the only admissible explanation. The geometry leaves no alternative. 7 The phase structure of institutional systems The integrodynamic law admits a natural phase structure. Each phase corresponds to a distinct configuration of the integrity potential 𝑈[𝑏], the dispersion functional 𝑆𝐼[𝑏], and the free-energy gradient ∇ℱ[𝑏]. Transitions between phases occur when geometric constraints fail: when convexity weakens, when drift amplifies, or when the free-energy flow loses monotonicity. These phases are not interpretive labels or behavioural taxonomies. They are structural states of the integrity field. The phase structure described below is exhaustive: every institutional trajectory is contained within one of these regimes, and transitions between regimes correspond to measurable geometric events. 7.1 Ordered coherence In the ordered coherence phase, the system possesses a single dominant basin of attraction. The Hessian satisfies 𝜆min(ℋ) > 0, dispersion is bounded, and the free-energy gradient strictly dissipates:  ℱ[𝑏] < 0. Trajectories converge smoothly, drift is subordinate, and behaviour remains predictable. The system corrects deviations without external intervention. This is the analogue of ordered phases in statistical mechanics: stable, reversible, and structurally constrained. 7.2 Warped coherence Warped coherence occurs when curvature is preserved but symmetry is partially broken. The potential remains convex, but the basin is shifted: 𝜆min(ℋ) > 0 and ∇𝑆𝐼[𝑏] ≠ 0. 23 The free-energy gradient still points inward, but with bias. Drift increases in one direction while still dissipating in others. The system self-corrects, but not uniformly. This phase corresponds to institutions that remain structurally coherent while exhibiting mild directional pressure. It is stable but fragile: as symmetry weakens, dispersion increases, pushing the system toward the drift-dominated regime. 7.3 Drift-dominated coherence In this phase, dispersion begins to compete with curvature: ‖∇𝑈[𝑏]‖ ≈ 𝑒‖∇𝑆𝐼[𝑏]‖. The system still returns to equilibrium, but only marginally. Drift does not yet overwhelm curvature, but it is no longer negligible. This regime is the structural warning stage: perturbations decay slowly, directionality emerges, and the basin begins to narrow. The free-energy gradient remains weakly negative, but the geometry is approaching its limit. 7.4 Pre-collapse Pre-collapse is defined by the softening of the potential: 𝜆min(ℋ) → 0+. Convexity is about to fail. At this point, the system becomes hypersensitive to perturbations. Drift amplifies in certain directions, dispersion grows, and the free-energy gradient becomes shallow or locally inconsistent. Pre-collapse is detectable: changes in curvature, drift, and dispersion produce diagnostic signatures well before the collapse threshold is crossed. In this regime, correction is possible but structurally difficult. 7.5 Collapse Collapse occurs when the smallest eigenvalue of the Hessian becomes negative: 𝜆min(ℋ) < 0. 24 The potential surface admits runaway trajectories, and the free-energy gradient reverses: ∇ℱ[𝑏] ⋅ 𝑑𝑏 > 0. Perturbations no longer dissipate; they amplify. Drift becomes directional, dispersion spikes, and the system enters a regime of irreversible divergence. This phase matches the statistical irreversibility condition from Part VI: ℙ(trajectory ∣ 𝐻0) < 𝜀. Collapse is not a narrative state. It is a geometric inevitability once convexity is lost. 7.6 Disorder Disorder is the post-collapse regime in which neither curvature nor dispersion controls behaviour. The potential surface is non-convex across large regions, drift dominates, and the free-energy gradient fluctuates in sign. The system no longer exhibits meaningful structure, and integrity becomes spatially fragmented. In this phase, collapse signatures saturate, and no admissible free-energy geometry can restore coherence without external intervention or structural redesign. This regime represents the final state of systems that have exhausted their control coefficient: 𝐶[𝑏] < 0 in multiple directions. 7.7 Interpretive summary The phase structure of integrodynamics classifies institutional behaviour into six geometric regimes: 1. ordered coherence, 2. warped coherence, 3. drift-dominated coherence, 4. pre-collapse, 5. collapse, 6. disorder. 25 3. Regulatory effectiveness depends on preserving declared symmetries, not rule volume. Dispersion arises when ∇𝑆𝐼[𝑏] ≠ 0, i.e., when declared invariances are violated. Adding more rules does not reduce dispersion unless those rules restore symmetry. Regulation is therefore structurally effective only to the extent that it enforces invariance. Symmetry, not quantity, is the controlling term. 4. Organisational drift is measurable and structural, not cultural. Drift occurs when ‖∇𝑈[𝑏]‖ < 𝑒‖∇𝑆𝐼[𝑏]‖. Cultural explanations may describe the appearance of drift, but they do not determine the condition for its existence. Drift is a gradient imbalance, and the imbalance is quantifiable. The cause is structural; the interpretation is cultural. 5. AI alignment is convexity preservation, not preference matching. In coherent regimes, corrective behaviour is guaranteed by 𝜆min(ℋ) > 0. Alignment failure begins when curvature weakens and drift dominates. This framework therefore defines alignment in geometric terms: preserving convexity of the integrity potential. Preference-matching is secondary to the curvature that governs stability. 6. Institutional stress produces detectable geometric signatures. Because integrity satisfies the continuity equation 𝜕𝑡𝐼[𝑏] + ∇ ⋅ 𝐽[𝑏] = 0, stress cannot disappear; it redistributes. Rising dispersion, curvature softening, and directional drift are necessary consequences of this conservation relation. Stress is therefore empirically detectable through 𝐽[𝑏] and the local behaviour of ℱ[𝑏]. Taken together, these consequences do not replace existing governance, legal, or regulatory frameworks. They identify the geometric constraints under which those frameworks operate. Institutions may choose their rationale and their design, but they cannot choose the mathematical conditions for coherence. If the geometry is incompatible with a benign interpretation, the integrodynamic law leaves no admissible alternative. 10 The Integrum theorem The preceding sections establish that institutional behaviour is governed by a freeenergy geometry defined by curvature, dispersion, and their balance under the universal constant 𝑒. The Integrum theorem formalises this structure by showing that all 32 admissible integrodynamic systems collapse to a single governing law. Every equation derived in Parts I–VIII is a projection, corollary, or limit of this law. The theorem does not introduce new structure. It consolidates the invariants already identified: the curvature functional, the dispersion term, the free-energy gradient, the continuity of integrity, the convexity criterion, and the collapse condition. 10.1 Statement of the theorem Theorem 10.1 (Integrum).Let 𝑏(𝑥,𝑡) be an admissible behavioural field on a domain Ω with integrity potential 𝑈[𝑏] and dispersion functional 𝑆𝐼[𝑏]. Suppose that: 1. 𝑈[𝑏] is twice differentiable with Hessian ℋ(𝑏), 2. 𝑆𝐼[𝑏] is convex and differentiable, 3. integrity evolves continuously under perturbation, 4. empirical renormalisation yields the universal constant ℵ = 𝑒. Then the evolution of 𝑏(𝑥,𝑡) satisfies the integrodynamic law 𝜕𝑡𝑏 = −∇𝑈[𝑏] + 𝑒∇𝑆𝐼[𝑏] + 𝜎𝜉𝑡, and the integrity density 𝐼[𝑏] satisfies the continuity equation 𝜕𝑡𝐼[𝑏] + ∇ ⋅ 𝐽[𝑏] = 0, 𝐽[𝑏] = −∇𝑈[𝑏] + 𝑒∇𝑆𝐼[𝑏]. Moreover: 1. stability holds iff 𝜆min(ℋ) > 0, 2. drift dominates iff ‖∇𝑈[𝑏]‖ < 𝑒‖∇𝑆𝐼[𝑏]‖, 3. collapse occurs iff 𝜆min(ℋ) < 0, 4. irreversibility occurs iff ℙ(trajectory ∣ 𝐻0) < 𝜀, 5. the system’s future trajectory is uniquely determined (up to noise) by the free-energy geometry. 33 10.2 Interpretation The Integrum theorem states that every well-formed institutional system behaves as a free-energy minimiser subject to a universal dispersion penalty. The theorem binds together the core results of the earlier parts: • Part I: contradiction is curvature. • Part II: symmetry determines dispersion. • Part III: drift emerges from gradient imbalance. • Part IV: integrity is a geometric field. • Part V: physical stability is convexity. • Part VI: collapse is statistical irreversibility. • Part VII: the axioms define the admissible space. • Part VIII: renormalisation fixes the universal constant. Under these constraints, only one law is possible. Any system that preserves coherence must minimise ℱ[𝑏] = 𝑈[𝑏] − 𝑒𝑆𝐼[𝑏], and any system that fails to do so necessarily enters the divergence regimes characterised in the phase diagram. The theorem therefore removes the distinction between disciplinary domains. Different institutions instantiate different potentials and dispersion structures, but the governing equation – the integrodynamic law – is the same. 10.3 Corollary: uniqueness of the integrity constant Because ℵ = 𝑒arises as the fixed point of the beta function and as the unique choice preserving free-energy monotonicity, no alternative balance between curvature and dispersion yields a stable or coherent system. Under perturbation, all other constants either over-penalise dispersion (leading to rigidity and catastrophic curvature accumulation) or under-penalise it (leading to drift and collapse). Thus: ℵ = 𝑒 is the unique viable constant of integrity. 34 10.4 Corollary: inevitability of collapse trajectories If 𝜆min(ℋ) becomes negative in any region of the potential surface, then divergence is forced by the sign reversal of the free-energy gradient. The Integrum theorem implies that collapse is not a moral failure or an operational accident. It is the geometric consequence of a violated convexity constraint. Systems collapse for the same reason across all contexts: curvature turns against them. 10.5 Interpretive summary The Integrum theorem provides the unifying principle of the entire framework: 1. All coherent behaviour derives from convexity of the integrity potential. 2. All deviation derives from imbalance between curvature and dispersion. 3. All collapse derives from curvature inversion. 4. All irreversibility derives from statistical implausibility. 5. All institutional domains obey the same free-energy geometry. With this, integrodynamics becomes a complete theory: not metaphor, not analogy, but a structural law that governs the evolution of behaviour across systems. Everything else is application. 11 The geometry of accountability The Integrum theorem implies a structural limit on what institutions can plausibly claim about their own behaviour. Once the free-energy geometry enters a regime in which benign explanations are incompatible with the observed trajectory, the question ceases to be interpretive. Accountability becomes a geometric conclusion. This section formalises that shift. It is not a commentary on motive, competence, or organisational culture. It is a consequence of the invariants that govern the integrity field. 35 11.1 Accountability as elimination of the benign model In the integrodynamic framework, accountability arises when the benign model 𝐻0fails to explain the system’s trajectory. The failure is statistical, not rhetorical: ℙ(trajectory ∣ 𝐻0) < 𝜀. At this point, 𝐻0is not merely unlikely; it is excluded by the geometry. The system’s path cannot be reconciled with coherence-preserving dynamics. Accountability is therefore the elimination of an incompatible hypothesis. It is not “blame,” “fault,” or “culpability” in the narrative sense. It is the enforcement of logical consistency: if the geometry forbids a benign interpretation, it cannot be sustained. 11.2 Curvature failure as structural responsibility When convexity fails, 𝜆min(ℋ) < 0, the system amplifies deviation. This amplification is predictable and detectable. Any institution whose geometry enters this regime assumes structural responsibility for the resulting behaviour, regardless of intent. The rationale is simple: once the system’s potential becomes non-convex, its dynamics guarantee divergence. Persisting in this regime is a geometric decision, not an accidental one. Structural responsibility replaces narrative accountability. 11.3 Integrity flow as a conservation constraint Because the integrity density satisfies the continuity equation 𝜕𝑡𝐼[𝑏] + ∇ ⋅ 𝐽[𝑏] = 0, loss of integrity in one region must correspond to its displacement elsewhere. The system cannot claim surprise at outcomes generated by its own flow field. Where integrity accumulates, disperses, or fragments is determined by 𝐽[𝑏]. In this sense, accountability is the recognition that institutions are responsible for the flows they generate. The geometry does not permit exogenous explanations for pat36 terns created by internal gradients. 11.4 Drift dominance as the boundary of plausible deniability Drift dominates when ‖∇𝑈[𝑏]‖ < 𝑒‖∇𝑆𝐼[𝑏]‖. In this regime, the system’s corrective forces are weaker than the forces that destabilise it. Outcomes become directional, not stochastic. Deviations are no longer “isolated incidents” but expressions of the underlying geometry. Plausible deniability collapses at the moment drift exceeds curvature. The system is no longer being moved by randomness. It is moving itself. This is the boundary at which accountability becomes inevitable. 11.5 Collapse as the end of interpretive freedom Once curvature becomes negative and the free-energy gradient reverses sign: ∇ℱ[𝑏] ⋅ 𝑑𝑏 > 0, the system’s dynamics enforce divergence. No narrative reconstruction can legitimately describe such behaviour as coherent, well-intentioned, or internally consistent. Collapse removes interpretive freedom. At this stage, accountability is no longer evaluative. It is descriptive: the only admissible explanation isthat the systemis producing outcomes incompatible with its declared rationale. Collapse is a fact of geometry, not of motive. 11.6 Institutional claims under geometric constraint Institutions commonly invoke explanations that are structurally impossible under the Integrum theorem: isolated error, bad luck, unforeseeable complexity, unpredictable deviations. These claims fail when the geometry forbids them. If the system is in a region where: 𝜆min(ℋ) < 0, ∇ℱ[𝑏] ⋅ 𝑑𝑏 > 0, or ℙ(trajectory ∣ 𝐻0) < 𝜀, 37 then no benign narrative is consistent with the governing law. Accountability becomes not a discretionary judgement but the logical consequence of the field. The geometry constrains what institutions may truthfully assert about themselves. 11.7 Interpretive summary The geometry of accountability consists of three structural facts: 1. Benign explanations fail when the system’s trajectory lies outside the coherencepreserving manifold. 2. Responsibility becomes structural when curvature fails and divergence is enforced by the potential. 3. Interpretive freedom ends when collapse signatures eliminate the benign model. Under integrodynamics, accountability is the recognition that systems are bound by their geometry. If their behaviour is incompatible with coherence, the conclusion is not moral but mathematical: the system cannot plausibly claim what the field does not permit. Accountability is therefore the final invariant. The geometry decides; everything else is commentary. 12 Integrum as a unifying empirical programme Integrodynamics is not a metaphorical framework. It is an empirical one. The preceding sections establish that coherence, drift, and collapse are governed by geometric invariants. An empirical programme follows immediately: measure the invariants, estimate the gradients, detect the phase, and falsify the benign model when its predictions fail. This section outlines the operational core of the Integrum programme: how to measure integrity, how to detect curvature loss, how to quantify drift, and how to test the integrodynamic law in real systems. 38 12.1 Measuring curvature through contradiction Part I established that contradiction reflects curvature deformation in the integrity potential. Paper VI clarified how this deformation appears empirically: not as isolated incompatibilities, but as directional, cross-domain, and compounding divergence from the benign model. In this framework, curvature is inferred from the structure of collapse diagnostics rather than from raw counts of inconsistencies. Let {𝒞𝑖}denote the set of coherence constraints tested in Part VI. For each constraint, the collapse diagnostic produces a directional deviation score Δ𝒞𝑖, representing the magnitude and orientation of the observed departure from the coherence-preserving model 𝐻0. The empirical curvature functional is constructed as the compounding effect of these deviations: Π = ∏ 𝑖(1 + Δ𝒞𝑖), the same multiplicative operator used in the collapse test. Its logarithm provides a scale-free estimate of accumulated curvature strain: log Π = ∑ 𝑖 log(1 + Δ𝒞𝑖). A practical curvature estimator is therefore:  𝜆min(ℋ) ≈ −log Π, where large values of Πindicate compound, multi-directional inconsistency across constraints and thus strongly negative inferred curvature. This approach aligns directly with the diagnostics of Part VI: •directionality of deviations corresponds to curvature warping, •cross-domain accumulation corresponds to curvature amplification, •compound evidence corresponds to curvature sign inversion, •implausibility of 𝐻0corresponds to the onset of non-convexity. Curvature measurement is therefore a collapse-informed contradiction audit: a geometric assessment of how deviations accumulate across coherence constraints. Systems exhibiting large Πvalues are structurally close to, or already beyond, the convexity boundary that separates coherence from collapse. 39 The method is scale-invariant, coarse-graining stable, and domain-agnostic. 12.2 Estimating dispersion through symmetry violation Part II established that dispersion arises from violations of declared symmetries. Paper VI operationalised this by treating each symmetry as a coherence constraint 𝒞𝑖and measuring directional deviation from the benign model 𝐻0. Let Δ𝒞𝑖denote the deviation score associated with constraint 𝒞𝑖, computed using permutationinvariance tests, paired comparisons, or structural equivalence checks. Each Δ𝒞𝑖measures how behaviour changes when a declared invariance is applied. The dispersion gradient is estimated by the compound effect of these violations:  ∇𝑆𝐼[𝑏] ≈ ∑ 𝑖 v𝑖Δ𝒞𝑖, where v𝑖is the direction of the corresponding invariance transformation. This produces a vector estimate of the dispersion force. Dispersion measurement therefore becomes a symmetry audit: the empirical task is to quantify how far the system moves when its own declared invariances are applied. The method is model-free, scale-stable, and compatible with the collapse diagnostics of Part VI. 12.3 Quantifying drift through gradient imbalance Drift occurs when the dispersion force dominates the restoring force of the integrity potential. In integrodynamic units, ‖ ∇𝑈[𝑏]‖ < 𝑒‖  ∇𝑆𝐼[𝑏]‖. Part VI provides operational criteria using behavioural trajectories. Drift is observed when: • deviation residuals exhibit persistent directionality across tests, • symmetric scrutiny yields asymmetric response patterns, • correction dynamics slow to the point consistent with a flattened potential, • autocorrelation in deviations increases, indicating memory effects. 40 These signatures reflect an imbalance between curvature-driven structure (∇𝑈[𝑏]) and symmetry-driven dispersion (∇𝑆𝐼[𝑏]). Drift is thus a detectable regime, not an interpretive label. 12.4 Detecting pre-collapse through convexity softening Pre-collapse is characterised by the softening of convexity:  𝜆min(ℋ) → 0+. Part VI shows how this softening appears empirically across coherence constraints: • increasing sensitivity to perturbations (small actions produce large deviations), • oscillatory or inconsistent corrections, indicating loss of a dominant basin, • rising cross-domain correlation of deviations, indicating shared instability, • shrinking effective coherence basin, measurable by reduced pullback strength. These effects occur before collapse diagnostics saturate. Pre-collapse is therefore a forward-looking condition: the system is losing convexity even if the trajectory has not yet violated 𝐻0. 12.5 Identifying collapse through statistical irreversibility Collapse is observed when the system’s trajectory becomes statistically incompatible with the benign model: ℙ(trajectory ∣ 𝐻0) < 𝜀. Operationally, this requires: • constructing anytime-valid e-processes (Part III), • computing collapse diagnostics across domains (Part VI), • quantifying coherence failures through the 𝜋-product, • measuring divergence in the free-energy functional. 41 Under these conditions, the system corrects its own deviations. Perturbations decay. No ad hoc intervention is required. Coherence is therefore not a behavioural achievement but a geometric consequence. This perspective redefines what it means for an institution to be “functional”: not that it behaves well, but that its geometry enforces stability. 14.2 Collapse as curvature inversion Collapse occurs when convexity fails. At that instant, the system’s geometry begins to amplify deviation: 𝜆min(ℋ) < 0. This sign reversal of curvature is the universal mechanism of institutional breakdown. It renders benign narratives structurally impossible, forces divergence, and produces the statistical signatures that eliminate the benign model: ℙ(trajectory ∣ 𝐻0) < 𝜀. Collapse is therefore not a crisis in the ordinary sense. It is a phase transition: a fundamental change in the geometry that governs behaviour. This realisation dissolves the multi-disciplinary taxonomy of failure. Institutions do not collapse in different ways. They collapse in one way: through curvature inversion and the loss of free-energy monotonicity. 14.3 Control as gradient structure In integrodynamics, control is not authority or management. It is the structure of the free-energy gradient: ∇ℱ[𝑏] = ∇𝑈[𝑏] − 𝑒∇𝑆𝐼[𝑏]. A system is controllable when this gradient points unambiguously toward the coherence basin. It is uncontrollable when the gradient weakens, tilts, or reverses. The effectiveness of any intervention depends not on its intent but on its ability to reshape this gradient. This reframes institutional design: durable control arises from geometry, not from policy. Geometry defines what can be stabilised, what can be changed, and what cannot be repaired once curvature fails. 48 Control is therefore the capacity to shape the potential, not the capacity to issue instructions. 14.4 The synthesis: a unified view of structural dynamics Taken together, the invariants of integrodynamics provide a unified account of institutional behaviour: 1. Coherence is convexity. Stability emerges when curvature dominates dispersion. 2. Drift is gradient imbalance. Directional deviation arises when dispersion dominates curvature. 3. Collapse is curvature inversion. Divergence is forced when the potential loses convexity. 4. Control is gradient structure. Interventions succeed only when they modify the geometry. 5. Accountability is geometric necessity. If the benign model is incompatible with the observed trajectory, responsibility follows from the invariants, not from narrative interpretation. These principles describe not how institutions ought to behave but how they must behave under the constraints of the integrodynamic law. The result is a theory of coherence, collapse, and control that applies across governance, law, AI, finance, regulation, and organisational design. Integrodynamics replaces explanatory pluralism with a single geometric structure. The details differ; the law does not. 14.5 Interpretive summary The purpose of Integrum is not to introduce new mathematics but to expose the unity of a system that was previously viewed through fragmented lenses. Contradiction, symmetry, adaptation, integrity, physics, collapse, axioms, and origin were not separate developments; they were partial views of a single field. The Integrum synthesis makes this explicit: all institutional behaviour is governed by the geometry of the integrity surface. Systems maintain coherence when curvature 49 is positive, lose it when dispersion exceeds curvature, and collapse when curvature inverts. Control is nothing more than the ability to shape these gradients. Once this structure is understood, the mathematical inevitability becomes clear: institutions are not free to behave as they choose. They are free only within the geometry that keeps them coherent. 15 Closing statement The purpose of Integrum is to show that coherence is not accidental, drift is not mysterious, and collapse is not arbitrary. Institutions behave as they do because their geometry leaves them no alternative. When curvature is positive, coherence is enforced. When dispersion overwhelms curvature, drift dominates. When curvature inverts, collapse is compulsory. These outcomes are not interpretations but structural consequences of the integrodynamic law. What remains is simple. Systems that preserve convexity remain coherent. Systems that lose it do not. The free-energy geometry determines the trajectory, the trajectory determines the phase, and the phase determines what explanations are consistent with the data. Nothing in this framework depends on motive, explanation, or intention. The mathematics is agnostic to all of them. Integrodynamics therefore offers a single statement: institutions are constrained by their geometry. They can choose their rationale, their narrative, and their design, but they cannot choose the consequences of violating the invariants that keep them coherent. Once the geometry is known, the outcome is determined. Everything else is commentary. Acknowledgements This work received no institutional support, no research funding, and no academic supervision. It was written outside every structure it critiques and, perhaps, needed to be. The systems that shaped my experience did not teach compliance; they taught analytical clarity. Everything else followed from that. 50 Author’s Note – why this framework exists The world is full of systems whose decisions shape lives, yet those systems rarely explain themselves. People harmed by institutions often face a structural imbalance: organisations have resources, legal teams, narratives, and control over the record. Individuals have only their experience – and experience is easy to dismiss. This work was written to correct that imbalance. Integrodynamics is not a philosophy of fairness and not a political argument. It is a geometry. It treats behaviour the way physics treats motion: as a field with curvature, symmetry, and measurable invariants. Under this model, institutions leave signatures. Contradictions leave curvature. Drift leaves gradient imbalance. Collapse leaves irreversible trajectories. These are not metaphors; they are structural consequences. The purpose is simple: to make integrity measurable and to make power accountable without appealing to motive. Motives can be denied. Geometry cannot. Where traditional theory argues, integrodynamics tests. Where narratives compete, invariance constraints decide. Where institutions claim goodwill, the data either satisfies the coherence surface or it does not. This offers individuals something that has been missing: a method. A way to demonstrate unfairness, instability, or bias using structure rather than belief. It also offers institutions something they rarely confront: a mirror. One that does not care about intentions, status, or reputation – only about whether the system behaves consistently with its declared rationale. This is not a theory designed to punish; it is a theory designed to reveal. And in a world increasingly shaped by opaque processes – human and algorithmic – the ability to measure integrity becomes not just a technical achievement but a civic one. If this framework helps even a few people challenge what they could not previously challenge, or helps even a few institutions recognise instability before it becomes harm, then the work justifies itself. The mathematics is universal. The motivation is human. 51