Ordinary Wall Specialization Forgets the Unique Resonant Direction
Question
Entry 1767 identifies one relative-sign Hom resonance over each exceptional divisor above
Can the existing ordinary finite-field specialization determine whether the source extension occupies that slot?
Labelled quotient coordinates
Use the frozen order
Entry 873 gives the generic (D)-wall fixed mask
Thus (q_{\rm wall1}) is not fixed across the primitive exact-lift nullspace. Entry 1767 identifies precisely (q_{\rm wall1}) as the unique (-\tfrac12) source direction resonant with the relative-sign algebraic line over (P_6\cap D).
Likewise, the generic (H)-wall fixed mask is
Thus (q_{\rm wall2}) is not fixed, and Entry 1767 identifies precisely (q_{\rm wall2}) as the resonant direction over (P_6\cap H).
Result
Therefore evaluating a chosen primitive witness on either wall cannot decide the exceptional extension class. The ambiguity is not incidental workflow noise; it occurs in the exact coordinate required by the indicial theorem.
Required next object
The finite falsifier must retain both labelled normals before specialization:
where (C_{132}) denotes the complete source reduction system. It must:
- preserve the labelled top, wall-1, and wall-2 quotient coordinates;
- retain the entire primitive exact-lift module;
- derive the exceptional transition rather than choose a witness;
- project the exceptional associated grade to the unique Hom kernel of Entry 1767.
A nonzero projection establishes a supported relative-sign coefficient extension over the existing carrier. A zero projection closes the route.
Durable artifacts
- checker:
research/benincasa/checkers/p6_wall_rees_necessity.rs; - result:
research/benincasa/results/p6-wall-rees-necessity.json; - convention note:
research/benincasa/p6-wall-rees-necessity.md; - allocator claim:
seqclaim-401295932a05aee682c3e176.