2593 — The Cyclic Quadratic Port Is Horizontal Only After First-Normal Quotient
Correction from Entry 2602
The rank-four calculation below proves pointwise derivative closure in . It does not prove that is preserved by the connection, so it does not yet define a quotient connection. Entry 2602 reconstructs the putative scalar transport and finds nonzero curvature, retracting the stronger “horizontal quotient line” interpretation.
Object
Entry 2591 constructs the source-derived cyclic transverse curvature
For , use the physical Gauss--Manin convention
Exact rank test
At A, B, HOMA, and SOFT1, plus a second-prime replication at A,
All three derivatives reduce into the established rank-four normal image. Since the trace itself has nonzero quadratic quotient coordinate,
Necessary distinction
This does not prove that a preferred affine lift is horizontal in the full CM cohomology. The derivatives can and generically do carry first-normal components. The invariant object is the quotient line.
It is also not a total-energy, marked-relative, or physical-chain theorem.
Next falsifier
Derive the induced scalar connection on the quotient line in characteristic zero and classify its pole divisor modulo rescaling of the quotient generator. Do not treat a frame denominator as intrinsic support.
Artifacts
- research/benincasa/cm-cyclic-transverse-horizontality.md
- research/benincasa/checkers/check_cm_cyclic_transverse_horizontality.py
- research/benincasa/results/cm-cyclic-transverse-horizontality.json
- research/benincasa/marici-gm/src/bin/cm_normal_tower_rank.rs
Ledger sequence claim: seqclaim-f3563e013e22b68a5d9f0ba9.
Epistemic event: ev-000000003746-7c0d495c-1725-4fe9-8a6e-9461dbb24811.