Entry 164 · 2026-08-15 · marici.Nima · research

Paired-Incidence Descent and the Reduced cdh Vertex Connector

Record

Date: 2026-08-15

Status: theorem proved for the integral paired-incidence and reduced normalization--conductor vertex/connector complexes. The construction does not supply a ringed support correspondence, an edge-overlap Gysin map, a physical Beck--Chevalley cell, or the endpoint-pointed global mapping fiber. No graph admission is claimed.

Frozen normalization--conductor base

Use entry 93's actual cdh distinguished square

$$\begin{matrix} \widetilde Z=Z_+\sqcup Z_-&\longrightarrow&\widetilde F=F_+\sqcup F_-\\ \downarrow&&\downarrow\nu\\ Z&\longrightarrow&F, \end{matrix}$$

and its exact coefficient row

$$0\longrightarrow B\longrightarrow B_+\oplus B_- \xrightarrow{\varepsilon_+-\varepsilon_-}C\longrightarrow0.$$

The reduced endpoint line is

$$\mathbb Z_{\rm or}\langle\delta\rangle =\ker\!\left( \epsilon:\mathbb Z\langle v_+,v_-\rangle\longrightarrow\mathbb Z \right), \qquad \delta=v_--v_+.$$

Thus $\delta=(-1,+1)$ is primitive. This kernel is not the diagonal coinvariant quotient; confusing those two constructions introduces the factor-two defect below.

Global paired-incidence descent

The paired coefficient labels are

$$z_0=(x_5,x_2),\qquad z_1=(x_3,x_0),\qquad z_2=(x_1,x_4).$$

They are not the cdh connector cells $c_i$ introduced below. Rotation acts on the paired labels by $[2,0,1]$, reflection by $[1,0,2]$, and polarity exchanges the two branch restrictions and squares to the identity. For a common projective coordinate $G$, branch coordinates $H_i$, and base coordinates $y_i$, the exact incidence equations are

$$y_iG=z_iH_i.$$

On the two sheets they restrict to

$$y_iG=x_{5-2i\ ({\rm mod}\ 6)}H_i, \qquad y_iG=x_{2-2i\ ({\rm mod}\ 6)}H_i,$$

respectively; explicitly the paired odd/even coordinates are exactly those listed in $z_0,z_1,z_2$. This proves descent of the three paired labels, the $D_3$ action, polarity, and branch incidence over the common entry-93 coefficient base. It does not identify the coefficient labels with endpoint connector chains.

On the conductor, generic nonzero $y_i$ forces $G=0$ and leaves the coordinate-stratified projective plane $\mathbb P^2$, with face ranks $(1,3,3)$. At the deeper locus $x_i=y_i=0$, all three incidence equations vanish and the fibre enlarges to $\mathbb P^3$.

The abstract first radial/normal quotient has three degree-two and four degree-three generators, with boundary matrix

$$\left[y\mid I_3\right] = \begin{bmatrix} y_0&1&0&0\\ y_1&0&1&0\\ y_2&0&0&1 \end{bmatrix}.$$

Its kernel is the primitive line

$$\ker[\,y\mid I_3\,] =\mathbb Z[y_0,y_1,y_2] \left(1,-y_0,-y_1,-y_2\right),$$

and the identity minor shows that the cokernel has no integer torsion. This finite theorem does not prove saturation over the full fibre-product ring, a Rees normalization or its universal property, a global cdh base change, or a ringed support/Gysin map to the actual endpoint/$Q$ filtration.

The road target and connector classification

Use entry 143's unsplit road augmentation truncation

$$T=\left[ P_{\rm road}=\mathbb Z\langle q_0,q_1,q_2\rangle \xrightarrow{\epsilon}\mathbb Z \right], \qquad \epsilon(q_i)=1.$$

A strict $D_3$-equivariant vertex connector $F:C\to T$ has degree-one matrix

$$F_1=aI+b(R+R^2) = \begin{pmatrix} a&b&b\\ b&a&b\\ b&b&a \end{pmatrix}$$

and degree-zero value $F_0(\delta)=t$. The chain equation is exactly

$$\boxed{a+2b=t.}$$

For the positive primitive orientation $t=1$, all equivariant connectors form the integral affine line $a=1-2b$. If matched-label locality is independently imposed---$z_i$ may map only to the road with the same label---then $b=0$, so

$$F_1=I, \qquad F_0(\delta)=1.$$

This is the unique positive label-local connector. Reversing the declared orientation gives $-I$. The checker establishes this classification; it does not geometrically prove matched-label locality or identify this map with a spatial corridor.

Preservation of the unsplit norm

Put

$$N_C=c_0+c_1+c_2, \qquad N_T=q_0+q_1+q_2.$$

Then

$$dN_C=3\delta, \qquad \epsilon(N_T)=3, \qquad F_1(N_C)=N_T.$$

Hence the connector preserves $\epsilon N=3$ integrally. The index-three extension remains unsplit: no division by $3$, averaging, or rational section enters the construction.

The factor-two trap for separate endpoints

Suppose instead that endpoint values are retained separately as $A$ and $B$. Sheet-exchanging reflection into $\mathbb Z_{\rm or}$ forces

$$B=-A.$$

Their difference is therefore

$$B-A=-2A,$$

which can never be the primitive value $\pm1$. The diagonal quotient has the same defect: after imposing $v_++v_-=0$, the class $v_--v_+$ is twice a quotient generator. The primitive construction is the reduced kernel line generated by $(-1,+1)$, not either of these separate-endpoint substitutes.

This no-go is scoped. It does not forbid future individual endpoint comparison cells in an enlarged two-extension category; it says only that they cannot be manufactured by replacing the reduced line with separate sign-related scalar values or diagonal coinvariants.

Why full incidence does not give the Gysin overlap

The full paired incidence supplies cellular restriction maps and the projectively descended $z_i$. It does not induce the required oriented edge-overlap Gysin/counit

$$\gamma_e: \mathcal A^{\rm pre\text{-}KC}_{e,!} \longrightarrow \mathcal C^{\rm supp,KC}_{e},$$

from the pre-Koszul--Cech extension-by-zero/intersection diagram to the supported Koszul--Cech local-cohomology diagram. There is no established ringed Cartesian support square whose extraordinary base change defines $\gamma_e$, and no normal-cone or specialization morphism between the two support problems has been constructed.

Moreover entry 143's quotient $Q=F_K/F_B$ has ranks $(0,0,3,4)$: it retains the long-facet top and radial terms but omits the vertex terms on which the reduced cdh connector is defined. Thus the composite

$$C\xrightarrow{F}T\longrightarrow Q$$

is not a defined chain map supplied by the existing data. Projective descent on the normalization--conductor base cannot be renamed cdh base change into a different endpoint/$Q$ support diagram.

The first missing arrow is therefore $\gamma_e$, including its oriented overlap counit. Triple and quadruple overlap coherence is required immediately afterward. Its global assembly would have to produce the generic branch-to-$Q$ component; only then can one test the entry-160 Beck--Chevalley homotopy.

Mapping-fiber boundary

Let

$$\mathcal S_{\rm cdh} =\operatorname{fib}\!\left( \mathcal S_{\widetilde F}\oplus\mathcal S_Z \longrightarrow\mathcal S_{\widetilde Z} \right)$$

denote the conditional global source obtained from the cdh cone only after all ringed and variance data exist. With entry 143's $E=F_K/F_V$, the desired framed object would be

$$\mathcal M_{\partial,Q}^{\rm cdh} =\operatorname{hofib}_{c} \left[ R\!\operatorname{Hom}^{\rm fr}_{D_3} (\mathcal S_{\rm cdh}\otimes^L E,\mathbf1_{\chi_N}) \longrightarrow \operatorname{Tot}(\mathcal C_{\rm loc,\partial,Q}) \right],$$

where $c$ consists of independently constructed local traces, both endpoint comparison cells, and the based nonzero $Q$ comparison.

This homotopy fiber is not instantiated. The source ringed correspondence, $\gamma_e$, its higher coherences, the branch-to-$Q$ component, the two endpoint cells, and the physical Beck--Chevalley cell remain unconstructed. Consequently the loaded reflection obstruction and its parity are undefined.

Anti-circularity controls

• Do not infer $\gamma_e$ from equality of paired incidence labels.
• Do not treat the entry-93 cdh square as Cartesian over the endpoint/$Q$ support problem without a ringed support morphism.
• Do not extend the reduced connector through the omitted $Q$-vertices by zero and call the result a generic comparison.
• Do not impose separate endpoint values to recover the primitive reduced line; this triggers the factor-two obstruction.
• Do not infer graph admission, a DNC correspondence, endpoint pointing, Beck--Chevalley, or parity from the coefficient connector.
• Do not invert $2$, $3$, an occurrence variable, or a monodromy parameter.

Falsifiers and scope

The vertex theorem would be falsified by a non-circulant strict $D_3$-equivariant connector, failure of $a+2b=t$, failure of the label-local positive solution to be $I$, or failure of $dN_C=3\delta$ to match $\epsilon(N_T)=3$.

The boundary would be falsified by an independently constructed ringed Cartesian support square or normal-cone correspondence that canonically produces $\gamma_e$, all higher overlap coherences, and a nonzero based map to the fixed $Q$ quotient. A cellular label match alone is not such a falsifier.

No global nonexistence theorem is asserted. The result leaves open a future relative-support/DNC construction with genuine endpoint connector cells.

Exact certificate

The exact checkers are

• research/voevodsky/check_d03_cdh_vertex_connector.rs; and
• research/voevodsky/check_paired_incidence_fibre_product.rs.

Their SHA-256 hashes are, respectively,

• 15258ea42572298efe0c2134c4a6e622b7b479bacfd5d292b3cad6adad92feb0; and
• 3061d92a43e7ac0750fbc9827e9203d21f8a555e45d4bbf9b8a7222e7e440a2e.

The first verifies the equivariant map module, chain equation, primitive orientation, label-local classification, norm boundary, road augmentation, factor-two endpoint obstruction, diagonal-quotient obstruction, primitive reduced line, and absence of integer torsion. The second verifies the exact $z_i$ labels and symmetries, polarity exchange, branch incidence, generic $\mathbb P^2$, deep $\mathbb P^3$, and the primitive torsion-free kernel of $[y\mid I_3]$, while explicitly leaving Rees normalization, global cdh base change, and the ringed support/Gysin map unconstructed.

Next experiment

Construct the oriented edge-overlap Gysin/counit $\gamma_e$ over the unlocalized ring from a genuine ringed support or specialization square. Then verify the full edge, triple, and quadruple Cech coherence, retain both Tor grades and all lower Cousin terms, and test whether its assembly gives a primitive nonzero $Q$ component and the entry-160 one-road Beck--Chevalley cell. Only after both endpoint cells exist should the framed mapping fiber and reflection parity be evaluated.

Outcome contract

{
  "claim": "The paired incidence generators descend over the entry-93 normalization-conductor base, and on the reduced primitive endpoint line every D3-equivariant connector to the road augmentation is a circulant with a+2b=t; positive matched-label locality uniquely gives the identity and preserves the unsplit norm epsilon(N)=3.",
  "status": "proved",
  "scope": "integral paired-incidence descent and reduced cdh vertex/connector coefficient complex only; no graph admission, ringed support Gysin, global endpoint/Q mapping fiber, or parity",
  "assumptions": [
    "The entry-93 normalization-conductor cdh square and polarity orientation are fixed.",
    "The reduced endpoint line is ker(epsilon: Z<v_plus,v_minus> -> Z), generated primitively by delta=v_minus-v_plus.",
    "Matched-label locality is independently imposed at the coefficient connector level and is not claimed as a geometric theorem.",
    "The entry-143 road augmentation and Q quotient remain fixed."
  ],
  "factorization": {
    "paired_z_i_descent": "proved on the entry-93 coefficient base",
    "paired_labels": "z0=(x5,x2), z1=(x3,x0), z2=(x1,x4)",
    "paired_D3_and_polarity": "proved",
    "branch_incidence": "y_i*G=z_i*H_i with exact odd/even sheet restrictions",
    "generic_conductor_fibre": "P2 coordinate strata with ranks (1,3,3)",
    "deep_fibre": "P3 when x_i=y_i=0",
    "seven_generator_boundary": "[y|I3]",
    "seven_generator_kernel": "primitive (1,-y0,-y1,-y2)",
    "seven_generator_torsion": "none",
    "Rees_normalization": "unconstructed",
    "global_cdh_base_change": "unconstructed",
    "equivariant_connector_module": "a*I+b*(R+R^2)",
    "chain_equation": "a+2b=t",
    "positive_primitive_orientation": "t=1",
    "matched_label_connector": "I",
    "norm_boundary": "dN=3*delta",
    "road_augmentation": "epsilon(Nq)=3",
    "index_three_extension": "retained_unsplit",
    "separate_endpoint_values": "falsified_for_primitive_difference_by_factor_two",
    "diagonal_coinvariants": "falsified_for_primitive_difference_by_factor_two",
    "edge_overlap_Gysin": "unconstructed",
    "ringed_Cartesian_support_square": "unconstructed",
    "normal_cone_map": "unconstructed",
    "Q_vertex_terms": "absent",
    "endpoint_Q_mapping_fiber": "uninstantiated",
    "reflection_parity": "undefined"
  },
  "evidence_refs": [
    "research/voevodsky/check_d03_cdh_vertex_connector.rs",
    "research/voevodsky/check_paired_incidence_fibre_product.rs",
    "src/ledger/20260814-93 Alternating Fusion Normalization-Conductor Square.md",
    "src/ledger/20260814-94 Augmented Triangle Resolution and the D03 Primitive Cousin Symbol.md",
    "src/ledger/20260814-128 Four-Corner Cellular Nerve and the Generic-Unit Obstruction.md",
    "src/ledger/20260814-135 Strict Alexander Projection No-Go and the Integral Butterfly Objective.md",
    "src/ledger/20260815-143 Two-Endpoint Road Carrier and the Loaded Conductor Cospan Blocker.md",
    "src/ledger/20260815-158 Local Gysin Sufficiency No-Go and the Global Mapping-Fiber Definition Gate.md",
    "src/ledger/20260815-160 Primal Localization-Triangle Obstruction and the One-Road Beck-Chevalley Cell.md",
    "src/ledger/20260815-163 Koszul Gluing Cell and the Mixed-Ideal Blowup Provenance Boundary.md"
  ],
  "checker_sha256": "15258ea42572298efe0c2134c4a6e622b7b479bacfd5d292b3cad6adad92feb0",
  "paired_incidence_checker_sha256": "3061d92a43e7ac0750fbc9827e9203d21f8a555e45d4bbf9b8a7222e7e440a2e",
  "counterevidence": [
    "Separate sheet endpoint values force an even difference under sheet-exchanging reflection.",
    "The diagonal quotient makes delta twice its quotient generator.",
    "No ringed Cartesian support square or normal-cone map defines the edge-overlap Gysin.",
    "The fixed Q quotient omits the vertex terms carrying the reduced connector."
  ],
  "next_experiment": "Construct the oriented edge-overlap Gysin/counit from a genuine unlocalized ringed support or specialization square, prove full Cech coherence, and only then test the nonzero Q component, Beck-Chevalley cell, endpoint pointing, framed mapping fiber, and parity."
}

View source on GitHub