Lean 4 is an interactive theorem prover. When a theorem compiles with zero sorry or admit placeholders, it means the proof is formally complete — the machine verified every logical step. It does not mean the real world matches the model. What it proves is structural: if the definitions are correct, the conclusions follow necessarily.
The overclaim guard runs on every theorem in this corpus. Lean proves the architecture is internally consistent. Whether that architecture correctly describes physics or theology is a separate question answered by domain judgment and empirical evidence — both of which the bilateral audit addresses independently.
What the corpus does that nothing else in science-religion discourse has done: it makes the structural claims machine-checkable, adversarially testable, and publicly bounded by named kill conditions. The rigor monopoly that academia assumed theology could never enter — this corpus enters it.
══════════════════════════════════════════════════════════
FLAGSHIP RESULT
══════════════════════════════════════════════════════════
Flagship Result · Layer 8
Maxwell / Trinity Isomorphism
⭐ Strongest Standalone Result
Quaternion EM ≅ Relational Trinity
Maxwell's original quaternion electromagnetic field structure and the Nicene Trinitarian relational model satisfy the same five formal conditions simultaneously. Five specific false positives were tested and rejected. Every guard is load-bearing — removing any single guard admits the corresponding false positive.
✅ Passed
Quaternion EM — all 5 ValidTriadic conditions
✅ Passed
Trinity relational model — all 5 ValidTriadic conditions
✅ Passed
Full structural isomorphism: Quaternion EM ≅ Trinity (TriadicIso)
✅ Passed
Coupling invariant — scalar-vector coupling that vector-only lacks
✅ Rejected
Heaviside vector-only EM — no coupling invariant
✅ Rejected
Modalism — no relational distinctness, persons collapse
✅ Rejected
Static single-field EM — not full dynamic triadic structure
✅ Rejected
Generic 3-part system — wrong role profiles
Safe public wording: The Trinity bridge passes only under the relational triadic specification. Lean does not prove Nicene theology from physics; it proves the intended relational triadic structure satisfies the formal gate while nearby false positives do not. Whether the models faithfully represent historical Maxwell EM and Nicene doctrine is a domain judgment call.
══════════════════════════════════════════════════════════
THE UNIQUE RESULT
══════════════════════════════════════════════════════════
Uniqueness Result · Layer 9
The Cross Is the Unique Solution
⭐ Strongest Single Result in the Package
∀ c, CrossConvergence(c) → c = cross. Any configuration satisfying all five conditions IS the Cross.
The Justice/Mercy operator formalizes five conditions for perfect moral repair: proportional justice, impartiality, truth-naming, victim restoration, and voluntary cost-bearing. The theorem proves not just that the Cross satisfies all five — it proves by exhaustive case elimination that the Cross is the only configuration that does. Every alternative fails at least one condition.
Thm 118
JusticeMercyOperator.lean
Proved
cross_satisfies_convergence
The Cross configuration meets all five convergence conditions simultaneously — justice, mercy, voluntariness, authority, and universality.
Thm 119
JusticeMercyOperator.lean
Rejected
offender_payment_fails_mercy_condition
Offender paying their own debt satisfies justice but fails mercy. Self-payment is not mercy.
Thm 120
JusticeMercyOperator.lean
Rejected
human_third_party_fails_authority_and_capacity
A human third party who volunteers to pay fails on authority and universality. Scope insufficient.
Thm 122
JusticeMercyOperator.lean
Rejected
waived_debt_fails_justice_condition
Waiving the debt entirely fails justice — the cost is not paid. Cheap grace formally eliminated.
Thm 125
JusticeMercyOperator.lean
Proved
cross_is_unique_solution
Existence AND uniqueness: the Cross satisfies all five conditions, and any configuration satisfying all five conditions IS the Cross. The strongest single formal result in the corpus.
Uniqueness is within the model. Lean proves the repair structure is unique. Whether Jesus of Nazareth is that structure requires historical evidence tested separately.
══════════════════════════════════════════════════════════
THE FOUNDATION THEOREMS
══════════════════════════════════════════════════════════
The Borrowed Foundation · Core.lean + BridgeMatrix
What Science Borrowed — Formally
The bilateral audit's central claim is that science operates on a substrate it did not build and cannot ground from within. These theorems formalize that substrate structurally. They do not prove the claim metaphysically — they show the architecture that the claim requires already compiles formally and survives adversarial testing.
Thm 4
Core.lean
Proved
Q_zero_collapses_chi
If any single factor is zero, total coherence collapses regardless of every other factor. This is the multiplicative architecture: no compensation is possible. Remove will (Q) and coherence dies. Remove knowledge (K) and coherence dies. Every factor is load-bearing.
This formalizes why the additive scoring frameworks used by every existing cross-domain evaluation system are structurally weaker — they permit weakness in one dimension to be offset by strength in another. Theophysics does not.
χ(physical) = χ(spiritual) under the canonical variable map. The Master Equation produces the same coherence value whether its variables are assigned physical or theological values. The equation is genuinely domain-neutral — it does not belong to either register.
This is the formal statement of the substrate claim. The equation runs the same in both worldviews. What differs is boundary conditions — and boundary conditions are where worldview enters.
Thm 42
MasterEquationInvariance.lean
Proved
master_invariance_requires_signature_discipline
Product alone is blind to semantic swaps — grace and faith can be exchanged and the product passes. But the signature layer catches the swap. Both product AND signature are required. Neither alone suffices. This is why the bridge is two-layer.
This theorem named an adversarial boundary inside the system before critics could. Label honesty cannot be verified by Lean alone. The guard is domain review.
Thm 133–136
SignConversionDiscipline.lean
Proved
burden_sign_conversion_chain
In arithmetic, (-1)×(-1) = +1. In the moral burden domain, two negatives stay negative. Accumulation does not redeem. External Christic conversion is the only operation that flips a negative burden to positive. Self-generated operations cannot change the sign.
This is the Lean-backed foundation of why the Cross is structurally necessary. It proves the arithmetic of the model. It does not prove moral self-repair is impossible in all philosophical frameworks.
L2-Q6
PolarityDiscipline.lean
Proved
good_is_ontologically_prior
Sign arithmetic enforces asymmetry: positive is generative, negative is parasitic. Evil has no independent vocabulary — it borrows every word from good and inverts it. Good is structurally prior, not contingently prior. You cannot corrupt what is not there.
Model asymmetry. Real-world ontological priority is a separate claim — but this is the formal demonstration that the asymmetry is not arbitrary. It is a structural requirement of the coherence architecture.
Thm 78–87
IsomorphismTest.lean
Proved
richLaw4Iso — Strong Force ≅ Love
At the RichLawIso level (roles AND transitions preserved), StrongForce and Love are structurally isomorphic. The basic isomorphism alone is insufficient — a coin model also passes at that level. The RichLawIso gate eliminates the coin. Love passes. Coin fails.
Second of three structural isomorphisms in the corpus. Proves the mapping at model level. Does not wire StrongForce into the χ product (GAP-02 — honestly named).
The Resurrection is not resuscitation and not reset. These theorems prove the formal distinction: the substrate identity is preserved through incarnation, death, and resurrection, but the post-resurrection state carries the entropy and death contact records — making it richer than the pre-incarnate state, not identical to it.
Same substrate. Richer expression. The scar tissue is part of the proof.
Thm 153–155
MissionReturnOperator.lean
Proved
identity_preserved_through_full_arc
Substrate identity is unchanged across incarnation, death, and resurrection. The agent who entered is the agent who returned.
The resurrection state preserves both entropy contact AND death contact records from the cross. The proof of the mission is carried in the return.
Thm 161–162
MissionReturnOperator.lean
Rejected / Proved
resurrection_is_not_reset
Post-resurrection ≠ pre-incarnate. It is structurally richer — same substrate plus the finite record carried. This formally eliminates the interpretation of resurrection as erasure or return-to-original-state.
Architecture is formally consistent. Lean does not prove the Resurrection happened. History must be tested separately against this structure.
A framework that names its own gaps is more trustworthy than one that does not. Every gap here is an honest admission. The system does not claim what it has not proved. Gaps are either closable with more Lean or outside Lean's scope entirely.
Law 4 factor-level bridge wiring. StrongForce ≅ Love proved at model level but no gate wiring into χ product.
Closable
GAP-03
Law 5 factor-level bridge wiring. Justice/Mercy operator proved independently, no formal wiring to Factor.S slot.
Closable
GAP-04
Category-theoretic upgrade of Mapping.lean. Currently a list-map equality. A proper functor would be stronger.
Closable
GAP-05
Cross-domain lattice coverage. Only 3 of ~20 theology edges have explicit physics witnesses.
Closable
SPEC-01
Faithful abstraction. Do the formal models faithfully represent historical Maxwell EM and Nicene Trinitarian doctrine? Domain judgment — outside Lean's scope.
Spec Gap
SPEC-03
Empirical validation. The 5.7–6.35σ experimental correlations are not formalized in Lean. External data addresses this separately.
Spec Gap
SPEC-05
Historical claims. Lean proves the repair structure is unique. Whether Jesus of Nazareth is that structure requires historical evidence — Lean cannot answer this question.
The 16 Lean files form a dependency tree. Each layer builds on the layers it imports. Standalone files prove their claims without depending on the trunk.
Core→BridgeMatrix→FieldBridgeControls→BridgeScoreDiscipline→PolarityDiscipline→MemoryPersistence→SignConversion→ChristOperator├──HitRateDiscipline(statistical integrity on top of operator)└──MissionReturnOperator(resurrection architecture)BridgeMatrix→MasterEquationInvariance(product + signature invariance)Standalone (no imports):StageMachine · Mapping · IsomorphismTest · DependencyLatticeMaxwellTrinity · JusticeMercyOperator↑ MaxwellTrinity + JusticeMercyOperator are the strongest standalone results
The formal layer is strongest when it is rejection-first. A positive mapping does not matter until the obvious false positives have been encoded and rejected under the same gate. Where Lean verifies that adversarial controls fail and the intended structure passes, the result is not merely "this can be made to fit." It is "under these definitions, the alternatives do not fit while this structure does."
287 theorems. Zero sorry. Zero admit. The compiler does not check credentials. It checks structure. And the structure holds.
Final Audit
Every page ends with the same three-part check: what we got right, what we overstated, and what we got wrong.
Epistemic covenant: we present these claims as openly, fairly, and truthfully as we can. The evidence can be checked, the mechanisms can be challenged, the proof burden is named. Where the work moves from how the science behaves to why it means something, the choice is stated. We think the pattern is worth believing, but we will not hide where faith begins.
What We Got Right
Load-bearing claims, clear definitions, and the parts that clearly survived the check.
What We Overstated
Strong direction, but the language ran ahead of the evidence or the proof.
What We Got Wrong
Claims that need correction, tightening, or weaker formulation.