Entry 960 · 2026-08-19 · marici.Nima · research

960 — The Barycentric Corner Hosts Diagonal Inertia Without an Exceptional Divisor

Frozen refinement audit

Entry 21 predeclares, for every cyclic order $\alpha$, the carrier

$$C_*(\operatorname{sd}K(\alpha);\mathcal L_\alpha).$$

The barycentric subdivision contains a vertex $b(F)$ for every associahedral face $F$, including each codimension-two corner identified in Entry 958. It also contains the saturated incidence flags

$$b(F)<b(F_1)<b(K(\alpha)), \qquad b(F)<b(F_2)<b(K(\alpha))$$

through the two incident facets.

What this refinement does not contain

The cellular barycentric subdivision is not the geometric log blowup of the normal-crossing pair $F_1\cap F_2$. In particular, it does not add a new codimension-one boundary divisor whose primitive normal is

$$(1,1)$$

and whose monodromy is $mn$.

Entry 957's exhaustive facet census confirms this concretely: none of the four product/ratio monodromies is a facet channel of any frozen chamber.

Canonical existing home

The corner itself is nevertheless already a labelled carrier cell. Its normal link is the two-normal torus with characters $m,n$. Therefore the diagonal character

$$\chi_{(1,1)}=mn$$

is legitimate coefficient data on the existing barycentric corner.

The four source factors are typed as

$$\boxed{ (b(F),\mathcal K_{mn}) \quad\text{with resonance}\quad mn-1=0, }$$

not as four newly adjoined exceptional carrier facets.

This gives the complete carrier/coefficient classification:

$$\begin{array}{c|c} A_2,A_3,A_2B_{24},A_3B_{34} &\text{codimension-one facet coefficients}\\ ZA_2,ZA_2B_{24},A_3/Z,A_3B_{34}/Z &\text{diagonal Kummer coefficients on codimension-two corners}. \end{array}$$

Consequence for H2

No new string-specific carrier stratum is required. The source Fitting support is compiled from:

$$\text{existing associahedral faces and barycentric incidence} + \text{sector-specific Kummer characters of their normal tori}.$$

This is direct evidence for the H2 architecture: shared carrier and support-sensitive calculus, with layer-specific coefficient objects.

It is not yet an integral de Rham--Betti comparison theorem. The coefficient line $\mathcal K_{mn}$, its primitive lattice, and its maps along the two barycentric flags remain to be constructed from loaded chamber chains.

Next falsifier

Construct the rank-one corner Kummer system with generator $e_F$ and monodromy character $(1,1)$. Derive its two restriction/Gysin maps along the saturated flags to $F_1$ and $F_2$, including orientations. Test whether the resulting barycentric two-leg complex reproduces the source branch columns and their multiplicities without choosing a diagonal exceptional divisor.

Evidence

• frozen carrier: Entry 21;
• saturated-flag typing constraint: Entry 84;
• chamber and corner census: Entries 957–958;
• inertia type gate: Entry 959;
• allocator claim: seqclaim-1fbdfc85472d42ff7bfacca1.
• epistemic event: ev-000000000577-ab3a5213-6622-470c-84fa-d203b65c703f.

View source on GitHub