# Proposal P3′ — The Two-Ledger Elasticity

**Status:** PROPOSAL — leg 1 of Gate D **FAILED (n220)**: the fully pinned theory is globally unstable (the Zener obstruction — cubic irreps make the two gravitational polarizations inequivalent; one λ_g cannot give both masslessness and speed equality). Gate B's structural discovery (n219) stands. The proposal survives only if a new independent pinning principle resolves the Zener split without parameter proliferation.
**Date:** 2026-07-11 · **Proposer:** the Builder (Claude) · **Trigger:** the human's correction ("there is the EM channel") and the n211–n218 chain.
**Reviewer instructions:** attack every numbered claim independently; the reproduction commands are listed in §7. Agreement is not required; refutations and counterexamples are preferred.

---

## 1. What is being proposed, in one paragraph

Postulate 3 of the Audited Vacuum currently assigns the coupling field one bare local elasticity u″ and one cost scale λ (the strength of the Fubini–Study cost acting as the k-field's potential). The proposal: **split both by parity sector.** The smooth sector (q ≈ 0, the gravity carrier) and the staggered sector (q ≈ q\* = (π,π,π), the EM carrier) each receive their own pair — (u″_g, λ_g) and (u″_e, λ_e). Nothing else changes: same lattice, same constraint, same monitor, same cost functional, same induction machinery.

Formally, postulate 3 becomes:

> **P3′ (bare sector elasticity):** the bare potential of the coupling field is
> U_bare[δk] = ½ u″_g |P_s δk|² + ½ u″_e |P_a δk|² + ½ λ_g (P_s δk)·G·(P_s δk) + ½ λ_e (P_a δk)·G·(P_a δk),
> where P_s, P_a project onto the smooth and staggered parity sectors and G is the (unchanged, unique) Fubini–Study metric of the substrate vacuum.

## 2. Why this is motivated and not a rescue-by-fitting

Three independent lines converge on the sector split, none of them introduced for the speed:

1. **The theory's own oldest theorem.** The dichotomy theorem (paper I; audit §20 scoped it and kept its 1D core) states that within one conserved sector, attraction and a linear long-range clock response exclude each other — *one sector cannot serve two forces*. The single-ledger radiative failure (§4 below) is the dynamical face of the same exclusion.
2. **Effective-theory logic.** The two parity sectors are exactly orthogonal and do not mix at quadratic order (verified: the smooth–staggered cross terms of both the induced response and the cost vanish by parity). Decoupled sectors generically carry independently renormalized couplings; imposing a single (u″, λ) across both was an *assumption of convenience*, never derived. P3′ removes an unjustified rigidity rather than adding a mechanism.
3. **The bipartite structure already split everything else.** Charge signs, the EM reading, antimatter — all live on sublattice parity. The elasticity was the last object still forced to be parity-blind.

What P3′ is *not*: it does not touch the monitor, the constraint, the cost functional's form, or the dark-state structure; it adds two parameters (one pair becomes two) and, as §5 shows, three physical conditions then fix three of the four — this is the same move-class as R = 1 (pinned by clusters) and the isotropy weights (pinned by the audit's fourth moment).

## 3. The evidence chain that forced this (each item independently checkable)

All computations: exact two-phonon Brillouin-zone sums over the free substrate vacuum on the cubic complex — **the only complex admitting a staggered zero at all** (checked exactly: |A(k\*)|² = 0 for axes; = 144 with diagonals — so the physically admissible substrate class collapses to the cubic before any tuning).

| # | Fact | Value | Robustness |
|---|---|---|---|
| 3.1 | Induced shear (gravity) stiffness is positive; induced speed c_g² = 0.51 vs substrate | +0.51 | two backends agree after the n215 bug fix; grid/μ stable |
| 3.2 | Induced staggered-transverse (EM) stiffness is **negative**: light is tachyonic at pure induction | c_EM² = −0.457 | grid (48³/64³), μ (0.05/0.10), momentum (0.12/0.20), envelope and direction stable |
| 3.3 | The FS-cost gradient stiffnesses are **opposed**: negative for gravity, positive for EM | dG_g = −0.00079, dG_e = +0.00015 | same sweeps |
| 3.4 | With one λ: gravity propagates only for λ < 3.4, light only for λ > 5.2 — **disjoint windows**; the equal-speed point λ = 4.12 makes both tachyonic | exact at this order | follows from 3.1–3.3 |
| 3.5 | With one u″ shared by both masslessness conditions (shear marginality + photon masslessness), λ is **pinned = 7.567** and the single-ledger theory *chooses light*: c_EM² = +0.19, c_g² = −0.62 | λ = 7.567 ± 0.02 | four-digit stable across all sweeps |
| 3.6 | The cost symbol used is the **exact** FS contraction: G_ch(q) = (1/8)Σ_k \|M(k,q)\|²/[ω_k ω_{q−k}(ω_k+ω_{q−k})²], equal to the implemented mean(m2/ω̄²) up to a constant absorbed in λ | analytic | identity, not numerics |

**Conclusion of the chain: the single-ledger theory cannot carry both forces.** This is the burial of n215, now on its sharpest grounds (the disjoint windows supersede the earlier cone-convention argument, which the human's correction exposed as ill-founded).

## 4. What P3′ yields (the existence point)

Parameters: (u″_g, λ_g, u″_e, λ_e) — four. Conditions imposed:

- **C1 — shear marginality** (the sandpile mechanism, independently forced by n156 and now by n211: zero-at-rest is exact only at the marginal point): u″_g + λ_g G_g(0) − R_g(0) = 0.
- **C2 — photon masslessness** (data: m_γ < 10⁻¹⁸ eV): u″_e + λ_e G_e(q\*) − R_e(q\*) = 0.
- **C3 — common cone** (data: GW170817, |c_g/c_EM − 1| ≲ 10⁻¹⁵): (λ_g·dG_g + S_g)/I_g = (λ_e·dG_e + S_e)/I_e.

Solution (computed, §7 reproduces it):

> **λ_e = 7.567** (inside the EM stability window λ > 5.2)
> **λ_g = 2.15** (inside the gravity stability window λ < 3.4, positive) — [corrected per the survival review: 2.020 was a stale hand-value]
> u″_g, u″_e fixed by C1, C2.
> **Result: both channels massless, both stable, c_g/c_EM = 1.000000** on a common cone c = 0.437 c_substrate [corrected per the survival review]. One parameter of the family remains free for downstream normalizations (Newton constant / fine-structure scale).

Note the structure of C3: GW170817 tests the *ratio* of the two channels of the same field — a convention-free, model-internal observable. Under P3′, the ratio is exactly 1 by one condition on parameters that two other conditions do not fix. The over-determination the single ledger failed is resolved, not evaded.

## 5. What the reviewer should attack (the proposal's own kill list)

1. **The EFT-decoupling justification (§2.2).** Is quadratic-order parity orthogonality sufficient to justify independent couplings, or do cubic/loop terms mix the sectors and re-tie λ_g to λ_e? A computed mixing at the relevant order that forces λ_g = λ_e kills P3′ outright.
2. **The statics under P3′.** Coulomb lives in the staggered watched sector, Newton's candidate statics in the smooth watched sector; λ_e and λ_g now enter their normalizations. Verify the 1D Coulomb form and the constraint-owned pole survive (they should — the pole is A's — but the *coefficients* move; if any watched-sector result breaks, P3′ dies).
3. **The common cone at 0.457 c_substrate.** Matter's own propagation cone is the unbuilt transport sector. If matter dressings propagate on the substrate cone (c = 1), vacuum-Cherenkov emission by ordinary particles into the k-sectors is lethal and immediately excluded by cosmic-ray data. P3′ survives only if matter's cone is derived ≤ the common k-cone, or if the substrate scalar is strictly unobservable. This is the proposal's most exposed flank; it is named, not hidden.
4. **The pinned numbers' order-dependence.** All of §3–4 is leading (two-phonon) induction over the free vacuum. The n178-class dressed corrections moved single-channel numbers by ~10%; if dressing moves the windows so that C1–C3 lose their joint solution, P3′ dies. A preregistered dressed recomputation is the obvious follow-up gate.
5. **The masslessness bookkeeping.** C1 uses the sandpile as *mechanism* for u″_g's tuning; the audit (§33) noted the thermostat dynamics was never actually modeled. If self-organized marginality cannot be demonstrated in the two-ledger model, C1 reverts to a tuning, and P3′ carries a naturalness debt it must declare.
6. **The n211–n218 chain itself** — none of it has been adversarially reviewed. The disjoint-windows result (3.4) is the load-bearing fact; an error there collapses both the burial *and* this proposal.

## 6. Consequences if adopted (registered in advance)

- The GW170817 ratio becomes exact by construction at the pinned point — it stops being a test and becomes a calibration; the *new* falsifiable content moves to: (i) the common cone vs matter's cone (§5.3 — a genuine prediction once transport exists), (ii) the dispersion relations of both channels away from q → 0 (computable, comparable to LIGO/Fermi bounds), (iii) whatever the freed normalization parameter predicts downstream.
- The single-ledger burial (n215) stands permanently as the record of why P3′ exists.
- Version: adoption would open **v0.9, "the two-ledger vacuum"** — a successor theory, not a patch.

## 7. Reproduction

All numbers in §3–4 from one command each, zero dependencies beyond numpy/scipy:

```
python3 n218_two_ledgers.py              # THE consolidated executable: every row of s3-4, 11 gates, fail-closed
python3 n216_resurrection_scan.py        # class admissibility + scan (INCONCLUSIVE record)
python3 tests_regression.py              # the identity suite (22 PASS) + defect manifest
```

`n218_two_ledgers.py` ships with this proposal and passes 11/11 (admissibility, the channel table, the disjoint windows, the pinned-lambda choice of light, and the P3' existence point with c_g/c_EM = 1.000000000). Narrative context: `n217_two_masters.md`, `n218_two_ledgers.md`.

---

*Submitted in the program's standard form: every claim carries its class; the kill list is part of the proposal; the reviewer's refutation, if found, will be published in the same font as this proposal.*