Part VII - AXIOMS: the axioms of integrodynamics

Part VII: AXIOMS The axioms of integrodynamics Foundations of the Physics of Integrity James D. Atkinson 2025 Abstract This paper formalises the axiomatic foundations of Integrodynamics: the unified dynamical theory governing contradiction, symmetry, adaptation, and integrity across epistemic, procedural, and institutional systems. Building on the six prior papers of the Integrity series, we consolidate their implicit assumptions into a minimal and complete axiom set from which the full integrity field theory follows. Integrity is treated not as a moral property but as a conserved structural quantity governed by free-energy dissipation, symmetry constraints, and evidential collapse under stochastic contradictions. Eight axioms are stated defining contradiction observability, free-energy monotonicity, conservation of evidential structure, symmetry-stabilised dynamics, bounded adaptation, and probabilistic irreversibility. From these axioms, the core integrodynamic law is derived, unifying contradiction, symmetry, adaptation, implausibility, and integrity into a single field-theoretic architecture. This paper establishes Integrodynamics as a complete structural theory of integrity, with testable predictions, conservation laws, and a well-defined phase structure. Keywords: integrodynamics; axiomatic systems; integrity theory; free-energy formalism; structural integrity; contradiction curvature; symmetry constraints; procedural symmetry; bounded adaptation; integrity entropy; evidential collapse; statistical irreversibility; legitimacy invariance; Lyapunov structure; phase structure; field-theoretic integrity; structural falsifiability; integrity diagnostics; epistemic entropy; institutional dynamics. 1 Contents 1 Purpose of the axiomatisation 3 2 Primitive Objects of Integrodynamics 3 2.1 Structural Assumptions on Admissible Symmetries . . . . . . . . . . 4 3 The Axioms of Integrodynamics 4 4 The Integrodynamic Law 6 5 Theorems and Derived Results 6 5.1 Lyapunov Structure and Monotonicity . . . . . . . . . . . . . . . . . . 7 5.2 Symmetry and Stability . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 5.3 Legitimacy and Invariance . . . . . . . . . . . . . . . . . . . . . . . . . 9 5.4 Statistical Irreversibility and Evidential Collapse . . . . . . . . . . . . 9 6 Derived Phase Structure 10 7 Position Within the Integrity Series 11 8 Falsifiability and Structural Failure Conditions 11 8.1 Failure of Free-Energy Monotonicity . . . . . . . . . . . . . . . . . . . 12 8.2 Failure of Symmetry–Stability Correspondence . . . . . . . . . . . . . 12 8.3 Failure of Legitimacy–Invariance Equivalence . . . . . . . . . . . . . . 12 8.4 Failure of Bounded Adaptation . . . . . . . . . . . . . . . . . . . . . . 13 8.5 Failure of Statistical Irreversibility . . . . . . . . . . . . . . . . . . . . 13 8.6 Failure of Deferred Intent . . . . . . . . . . . . . . . . . . . . . . . . . 13 8.7 Summary of Falsification Modes . . . . . . . . . . . . . . . . . . . . . 14 9 A Minimal Integrodynamic Toy Model 14 9.1 Definition of the model . . . . . . . . . . . . . . . . . . . . . . . . . . . 14 9.2 Gradient-flow dynamics . . . . . . . . . . . . . . . . . . . . . . . . . . 15 9.3 Free-energy monotonicity . . . . . . . . . . . . . . . . . . . . . . . . . 16 9.4 Equilibria and symmetry . . . . . . . . . . . . . . . . . . . . . . . . . . 16 9.5 Interpretation................................ 17 10 Conclusion 17 2 1 Purpose of the axiomatisation The preceding six papers establish the operational machinery of integrity diagnostics across epistemic, procedural, statistical, and institutional domains. What remains absent is a formal axiomatic closure: a minimal set of primitive assumptions from which all prior constructions follow as consequences rather than independent inventions. This paper supplies that closure. The objective is not philosophical justification, but structural completion. The axioms stated herein render the theory: • closed under derivation, • resistant to interpretive drift, • invariant under domain translation, • and fully falsifiable as a single integrated system. 2 Primitive Objects of Integrodynamics Let an integrodynamic system consist of the following primitive elements: • A behaviour field: 𝑏(𝑥,𝑡), • An integrity potential functional: 𝑈[𝑏], • An integrity entropy functional: 𝑆𝐼[𝑏], • An integrity free-energy functional: ℱ[𝑏]=𝑈[𝑏]−Θ𝑆𝐼[𝑏], • A declared admissible symmetry family: 𝒢, • A coherence current: 𝐽𝐼, • A bounded adaptive update operator: 𝒫′. All derived quantities, diagnostics, and institutional observables arise from these primitives. This paper introduces no new empirical claims. 3 2.1 Structural Assumptions on Admissible Symmetries The declared admissible symmetry family 𝒢acts on the configuration space 𝑋and satisfies the following algebraic properties: 1. Identity. There exists an element 𝑒∈𝒢 such that 𝑒(𝑥)=𝑥 for all 𝑥∈𝑋. 2. Closure. If 𝑔1,𝑔2∈𝒢, then their composition 𝑔2∘𝑔1∈𝒢. 3. Optional Invertibility (Group Case). If each 𝑔∈𝒢 is bijective and 𝑔−1 ∈𝒢, then 𝒢forms a group. In general, 𝒢need only be a monoid, allowing integrodynamic systems whose admissible processes are irreversible. These structural assumptions ensure that symmetry-based invariance (Axiom V), contradiction observability under admissible transformations (Axiom I), and symmetry-stabilised low-energy states (Axiom IV) are all well-defined. They also guarantee the compositional testability of integrity configurations under sequences of admissible procedures. 3 The Axioms of Integrodynamics Axiom I — Observability of contradiction If a system’s declared rationale and its observable behaviour cannot be jointly satisfied under admissible transformations, the resulting inconsistency manifests as measurable structural curvature. Contradiction is therefore an observable quantity, not a logical failure. Axiom II — Integrity as Free Energy The integrity state of a system is represented by a free-energy functional: ℱ[𝑏]=𝑈[𝑏]−Θ𝑆𝐼[𝑏], 4 whose evolution under admissible dynamics satisfies:  ℱ ≤0. Integrity cannot increase without compensating entropy export. Axiom III — Conservation of Evidential Structure In closed systems, the sum of operational integrity and epistemic entropy is approximately conserved: 𝐼∗+𝐻≈const. Integrity loss therefore reappears as disorder elsewhere in the informational environment. Axiom IV — Symmetry-Stabilised Low-Energy States Procedural symmetry corresponds to local minima of the integrity free-energy landscape. Persistent asymmetry corresponds to curvature away from stable basins. Symmetry without intervention is the unique passive equilibrium of integrity. Axiom V — Legitimacy as Invariance An institution is legitimate if and only if its observable behaviour is invariant under its declared admissible symmetry group. Legitimacy is therefore a demonstrated invariance, not a narrative claim. Axiom VI — Bounded Adaptation All adaptive updates must remain within the symmetry-preserving feasible region defined by declared commitments. Unbounded adaptation constitutes constitutional phase transition, not learning. 5 Axiom VII — Statistical Irreversibility When the probability of an observed multi-domain outcome under the innocent stochastic null model collapses below a defined threshold, directional structure is established as positive evidence. Null collapse is an irreversible evidential transition. Axiom VIII — Deferred Intent Intent is never assumed. It becomes the least-complex explanatory hypothesis only after symmetry violation, procedural deviation, and stochastic innocence have all failed. 4 The Integrodynamic Law From Axioms I–VIII, the governing law of integrodynamics follows: Integrity evolves as a conserved free-energy field whose stable states correspond to symmetry-preserving dynamics, whose drift generates evidential curvature, and whose stochastic irreversibility establishes structural directionality. Formally, the law may be expressed as:  ℱ =−Φcontradiction −Φasymmetry −Φevidential, with each dissipation term non-negative under admissible dynamics. 5 Theorems and Derived Results In this section we record several basic consequences of the axioms. The proofs are structural: they establish properties that hold in any domain where the axioms are satisfied, independently of the specific interpretation of 𝑏,𝑈,𝑆𝐼, and 𝒢. 6 5.1 Lyapunov Structure and Monotonicity Definition 5.1 (Admissible dynamics).A trajectory 𝑡 ↦ 𝑏(⋅,𝑡)is said to follow admissible dynamics if it respects all declared institutional and procedural constraints and satisfies the integrodynamic law. Theorem 5.2 (Second Law of Integrity).Under Axioms II and III, the free-energy functional ℱis a Lyapunov function for any closed integrodynamic system evolving under admissible dynamics. In particular,  ℱ(𝑡)≤0 for all 𝑡, with equality if and only if the system is in an integrity-equilibrium state. Proof. Axiom II states that for admissible dynamics the integrity state is represented by a free-energy functional ℱ[𝑏]whose evolution satisfies  ℱ ≤ 0. This is precisely the Lyapunov condition: ℱis non-increasing along trajectories. In a closed system, Axiom III implies that 𝐼∗+𝐻is approximately constant, so any strict reduction in ℱ corresponds to a reallocation between ordered integrity and epistemic entropy rather than an external injection of order. If  ℱ <0on a time interval, the system is moving down the free-energy landscape and is therefore out of equilibrium. Conversely, if  ℱ =0along a trajectory, then by Axiom II no further integrity dissipation is possible under admissible dynamics and the system resides at an equilibrium point of ℱ. Hence ℱis a Lyapunov function and the claim follows. Corollary 1 (Irreversibility of Integrity Dissipation).Under the assumptions of Theorem 5.2, there exists no admissible trajectory that returns the system from a state 𝑏2 to a state 𝑏1with ℱ[𝑏2]<ℱ[𝑏1]while preserving the closed-system condition. Proof. Suppose for contradiction that there exists an admissible trajectory from 𝑏1to 𝑏2 with ℱ[𝑏2]<ℱ[𝑏1], and another admissible trajectory returning from 𝑏2to 𝑏1in a closed system. Along the first trajectory, Theorem 5.2 implies ℱis non-increasing, so the inequality ℱ[𝑏2]<ℱ[𝑏1]is compatible with admissibility. Along the reverse trajectory, ℱwould have to increase from ℱ[𝑏2]back to ℱ[𝑏1], contradicting the monotonicity condition  ℱ ≤0for admissible dynamics. Hence no such reversible cycle exists, and integrity dissipation is irreversible in closed systems. 7 5.2 Symmetry and Stability Definition 5.3 (Symmetric configuration).A configuration 𝑏⋆is said to be symmetric if it is invariant under the declared admissible symmetry group 𝒢, i.e. 𝑔⋅𝑏⋆=𝑏⋆for all 𝑔∈𝒢. Theorem 5.4 (Symmetry Stabilisation Theorem).Assume Axioms II and IV. Let ℱbe invariant under the action of 𝒢and let 𝑏⋆be a symmetric configuration. If 𝑏⋆is a local minimiser of ℱwithin the space of admissible configurations, then 𝑏⋆is Lyapunov-stable under admissible dynamics. In particular, any sufficiently small admissible perturbation that preserves the symmetry constraints cannot decrease ℱ. Proof. By Axiom IV, procedural symmetry corresponds to local minima of the free-energy landscape and persistent asymmetry corresponds to curvature away from those minima. Let 𝑏⋆be a symmetric configuration that is a local minimiser of ℱ. Consider any admissible perturbation 𝑏⋆+𝛿𝑏that preserves the symmetry constraints; by local minimality we have ℱ[𝑏⋆+𝛿𝑏]≥ℱ[𝑏⋆] for all sufficiently small such perturbations. Since ℱis 𝒢-invariant, group actions do not change the value of ℱ, so the minimum is attained on the entire orbit of 𝑏⋆under 𝒢. Under Axiom II, admissible dynamics can only move the system in directions that weakly decrease ℱ. In a neighbourhood of 𝑏⋆, any direction that would strictly decrease ℱ corresponds to a departure from the symmetric basin and is therefore forbidden by the symmetry constraints. Hence trajectories starting sufficiently close to 𝑏⋆cannot move to states of strictly lower ℱwhile respecting admissibility, and Theorem 5.2 implies that they cannot move to states of higher ℱeither. This establishes Lyapunov stability of 𝑏⋆. Corollary 2 (Uniqueness of Passive Equilibrium).Under the hypotheses of Theorem 5.4, any passive equilibrium attainable under admissible dynamics is symmetric. Equivalently, symmetry without intervention is the unique passive equilibrium of integrity. Proof. Let 𝑏†be a passive equilibrium attainable under admissible dynamics, i.e. a state with  ℱ =0. If 𝑏†were not symmetric, Axiom IV would classify it as lying in a region of curvature away from any symmetric basin. In such a region, the free-energy gradient is non-zero and admissible dynamics must move the system towards lower values of ℱ, implying  ℱ <0and contradicting the assumption of passivity. Therefore every passive equilibrium must be symmetric, and symmetry without intervention is the unique form of passive equilibrium. 8 5.3 Legitimacy and Invariance Definition 5.5 (Legitimacy).An institution is legitimate if its observable behaviour is invariant under its declared admissible symmetry group 𝒢. Theorem 5.6 (Legitimacy–Invariance Equivalence).Assume Axiom V. Then legitimacy is equivalent to invariance: an institution is legitimate if and only if its observable behaviour is invariant under its declared admissible symmetry group 𝒢. Proof. The “only if” direction is immediate from Axiom V, which states that an institution is legitimate if and only if its observable behaviour is invariant under 𝒢. For the “if” direction, suppose that the institution’s behaviour is invariant under 𝒢. Then, again by Axiom V, this invariance is both necessary and sufficient for legitimacy as defined within the integrodynamic framework. No additional narrative or justificatory layer is required; invariance itself is the certificate of legitimacy. Remark 1. Theorem 5.6 shows that within Integrodynamics legitimacy is a derived property rather than a primitive assumption. Once the symmetry group and observables are specified, legitimacy is determined by invariance alone. 5.4 Statistical Irreversibility and Evidential Collapse Definition 5.7 (Evidential threshold).Let 𝑝0denote the probability, under the innocent stochastic null model, of observing a given multi-domain outcome. A pre-specified threshold 𝜀>0is called an evidential threshold if outcomes with 𝑝0≤𝜀are treated as statistically irreconcilable with the null. Theorem 5.8 (Evidential Irreversibility).Assume Axiom VII and fix an evidential threshold 𝜀. Let ℋ0denote the innocent null model and ℋ1a directional alternative. If an observed outcome 𝐷satisfies ℙ(𝐷∣ℋ0)≤𝜀and ℙ(𝐷∣ℋ1)≫ℙ(𝐷∣ℋ0), then: (i) the Bayes factor 𝐵(𝐷)=ℙ(𝐷∣ℋ1) ℙ(𝐷∣ℋ0)is large, and (ii) under any prior that assigns non-zero probability to ℋ1, the posterior odds in favour of ℋ1are strictly increased by observing 𝐷. Consequently, the evidential update favouring directionality is irreversible in the sense that no further observation of 𝐷can restore the original posterior odds. 9 9.3 Free-energy monotonicity We now verify that ℱis strictly non-increasing along trajectories of (9.2). By the chain rule,  ℱ(𝑡)=dℱ d𝑏􏿴𝑏(𝑡)􏿷⋅𝑏(𝑡). Using (9.1) we obtain  ℱ(𝑡)=dℱ d𝑏􏿴𝑏(𝑡)􏿷⋅􏿶−𝜅dℱ d𝑏􏿴𝑏(𝑡)􏿷􏿹=−𝜅􏿶dℱ d𝑏􏿴𝑏(𝑡)􏿷􏿹2. Since 𝜅>0and the square term is non-negative, we have  ℱ(𝑡)≤0 for all 𝑡, with equality if and only if dℱ d𝑏(𝑏(𝑡))=0. Thus, in this toy model, ℱis a Lyapunov function in the strict sense of Theorem 5.2: the integrity free energy cannot increase under the admissible dynamics (9.2). 9.4 Equilibria and symmetry Equilibria of the toy dynamics satisfy 𝑏=0, which from (9.2) occurs when either 𝑏=0 or 𝜆− 2Θ 1+𝑏2=0. The point 𝑏=0is always an equilibrium. Its symmetry is immediate: for every 𝑔∈𝒢, 𝑔⋅0=0, so the configuration is invariant under the declared symmetry group. A short calculation shows that near 𝑏=0, d2ℱ d𝑏2(0)=𝜆−2Θ, so for parameter regimes with 𝜆 > 2Θthe origin is a strict local minimum of ℱand therefore a stable symmetric equilibrium. Additional equilibria may exist if the algebraic condition 𝜆−2Θ/(1+𝑏2)=0admits nonzero solutions. In such cases one obtains symmetric pairs ±𝑏⋆with equal free energy, again reflecting the underlying ℤ2symmetry. Their stability depends on the sign of the second derivative d2ℱ/d𝑏2at those points; for typical parameter values, the symmetric configuration at 𝑏=0remains the global minimum. 16 9.5 Interpretation This toy model exhibits, in the simplest possible form, the structural features required by the axioms: • Axiom II (integrity as free energy) is realised via ℱ(𝑏)and the gradient-flow dynamics (9.1), with  ℱ ≤0holding identically. • Axiom IV (symmetry-stabilised low-energy states) is realised by the ℤ2symmetry and the symmetric equilibrium at 𝑏=0, which is a local (and for suitable parameters global) minimum of ℱ. • Theorem 5.4 is concretely instantiated: the symmetric equilibrium at 𝑏 = 0is Lyapunov-stable whenever 𝜆>2Θ. • Deviations 𝑏≠0generate curvature in the free-energy landscape and hence nonzero dissipation  ℱ <0until the system relaxes back towards a symmetric basin. Although highly simplified, the model shows that the axioms of Integrodynamics can be realised in a concrete dynamical system with explicit equations of motion, rather than remaining purely conceptual. 10 Conclusion With the axioms stated and their basic consequences derived, Integrodynamics becomes a complete dynamical theory of integrity rather than a collection of compatible methods. Contradiction becomes curvature, symmetry becomes stability, adaptation becomes bounded flow, and statistical collapse becomes irreversibility. All subsequent applications inherit their validity solely through these foundations. 17