Kill Switches

Each kill switch in Claim / Test / Status / Recovery form. KS-35, KS-36, KS-37 per the Master Kill Switch Registry. No renaming.

KS-35 — Self-consistency multiplicity

Claim. The fixed-point equation ε = f(ε), once formalised from the axioms alone, admits a unique solution. The Unique Residual Conjecture (§5.1) depends on uniqueness.

Test. Formalise f(ε) from {S, B, R, C} (debt D1) and solve. If more than one solution exists, KS-35 fires.

Status. LIVE — HARD. The fixed-point equation is defined but not explicitly constructed. Until f(ε) is written from the axioms, multiplicity cannot be ruled out.

Recovery. If KS-35 fires, the six-face identification survives but the Unique Residual Conjecture does not. α_em is determined to one of a finite set of values rather than uniquely; further structure is required to select among them. The corpus is unaffected (AP24 is non-load-bearing).

KS-36 — Self-consistency wrong value

Claim. The fixed-point equation, when solved, gives α_em ≈ 1/137.036. Any other value falsifies the six-face identification.

Test. Solve ε = f(ε) from the axioms (debt D1) and compute the value. Compare to the measured α_em ≈ 1/137.035999… Disagreement beyond computational uncertainty fires KS-36.

Status. LIVE — HARD. Same dependency as KS-35: the equation is defined but not solved. The toy model (§5.6) demonstrates the mechanism but uses SM input.

Recovery. If KS-36 fires, the six-face identification is wrong: either the faces are not all readings of one object, or the relations between them are not the correct consistency conditions. The Unique Residual Conjecture fails. The corpus is unaffected.

KS-37 — Independent variation of constants

Claim. The six faces cannot vary independently. They are readings of one object; varying one forces co-variation of the others.

Test. Empirical. Observations of constants varying in different epochs, different regions of spacetime, or under different conditions. If G is shown to drift without c or α_em co-varying — or if any one face varies independently of the others — KS-37 fires.

Status. LIVE — EMPIRICAL. Currently the best constraints on time variation of α_em (quasar absorption spectra, atomic clocks) are consistent with no variation at the 10⁻¹⁷/yr level. Constraints on G variation are weaker but consistent. AP24 predicts: any future drift in one face appears in all faces together. Pure single-face drift is forbidden.

Recovery. If KS-37 fires, the six faces are independent quantities. The identity is false. The six-face identification collapses; the corpus is unaffected (no upstream paper depends on the identification).