# n216 — The resurrection scan: INCONCLUSIVE as frozen — and the autopsy finds the question itself is ill-posed

**Date:** 2026-07-11 · **Gate:** `gates/resurrection-scan.json` (sha `11a9f30d…`, frozen before execution). **Script:** `n216_resurrection_scan.py`. Directive honored: all parameters freed — weights of every direction family, dressing of both signs, worst-direction metric.

## What the scan returned

The optimizer converged to degenerate corners (near-zero axis and body weights) where per-direction ratios turn negative and the 20³/40³ grids disagree by 9.1% — **INCONCLUSIVE by the frozen clause.** No admissible member approached the PASS band anywhere the numerics were trustworthy.

## The autopsy — the real result of "trying everything"

The pathology traces to the gate's own light-cone metric, and the defect is not numerical. The staggered zero — the model's only candidate for a massless EM sector — requires k·d ≡ π (mod 2π) for **every** direction simultaneously. Checked exactly: the cubic complex has the zero (|A(k\*)|² = 0); **the 9- and 13-complexes have none** (|A(k\*)|² = 144 — no massless EM mode exists on any complex with diagonals). Two consequences, each heavier than the scan itself:

1. **Physical admissibility collapses the class before the optimizer ever runs.** A substrate must carry a massless EM sector (m_γ < 10⁻¹⁸ eV is data). Only pure-axis complexes qualify — and the cubic's shear ratio is 0.51 with no dressing that raises it. There were never any parameters to tune: **the class's one physically admissible member is the one that fails.**
2. **Deeper: the theory has no derived light cone at all** — this is audit §51 resurfacing at the exact load-bearing point. The photon here is a static kernel fact with no dynamics; every "c" used in the speed verdicts (n202's smooth-scalar speed, this gate's staggered curvature) is an internal convention, not light. The question "does c_g equal c?" **cannot even be posed inside the model** until an EM dynamical sector exists. The burial's cone identification inherited this hole — and so would any resurrection.

## The standing state after the best effort

The directive was to make it work by looking at all parameters. The honest finding: **the parameters were never the obstruction — the missing sector is.** The burial stands and deepens: the radiative program lacked not one ingredient (the graviton's speed) but two (the photon's existence as a wave). The only resurrection path is now explicitly sequenced and it does not run through parameter space: build the EM dynamical sector first (audit gates 56.2–56.3: canonical A_μ, Maxwell action, a real transverse propagator with a pole), *derive* the light cone from it, and only then re-pose the gravitational speed question as a new preregistered gate. Anything less is comparing one convention to another.

**Class: scan [A-model, frozen-gate execution, INCONCLUSIVE]; the admissibility collapse [A — exact check].**

**RULE-5 CORRECTION (n217, same day, credited to the human):** the 'no derived light cone / question cannot be posed' diagnosis was WRONG — the staggered transverse channel is dynamical on the same footing as the graviton channel, and its speed is computable with identical machinery. Computing it superseded this note's conclusion with a sharper one: see n217 (one field cannot carry both forces — the disjoint stability windows).