Skip to content

Documentary state: EDITORIALLY_ADMITTED. The manuscript has a protected editorial admission, not a theorem certification.

OAI-26 family 052: tangent-bundle splittings and product decompositions

Editorial status: SOURCE_CRITIC_COMPLETE; protected acceptance is determined by the admissions ledger. Mathematical status: NO_CERTIFICATION. This article documents the two pinned manuscript claims without independently verifying their proofs.

The question in plain language

The tangent bundle of a manifold organizes all directions in which one may move at each point. Sometimes it splits into two holomorphic subbundles, suggesting two independent systems of directions. That does not automatically make the whole manifold a product: the directions must be compatible with actual families of submanifolds, a property called integrability.

The pinned OAI-26 synopsis claims that when both summands are integrable on a compact Kähler manifold, its universal covering space decomposes compatibly as a product. It further claims that on smooth rationally connected projective manifolds, the required integrability follows automatically from the splitting.

A small example

For a product \(X=Y\times Z\), its tangent bundle naturally splits:

\[ T_X\cong \pi_Y^*T_Y\oplus\pi_Z^*T_Z. \]

Here \(\pi_Y\) and \(\pi_Z\) are the two projections. The two sets of directions come from moving in \(Y\) while holding \(Z\) fixed, and vice versa. Each set is integrable because there are actual factor submanifolds tangent to it.

The converse is more demanding. A splitting of abstract tangent directions need not obviously come from a global product. Passing to the universal cover removes fundamental-group identifications that may prevent the original manifold itself from being a product. Thus a universal-cover splitting is not the same statement as a product decomposition of \(X\).

What is known

The protected Family 052 catalogue lists two manuscripts, both dated September 23, 2026:

  • Universal-cover splitting for compact Kähler manifolds.
  • Integrability of split tangent bundles on rationally connected manifolds.

The first primary PDF, Theorem 1.1, claims a two-summand form of Beauville's splitting conjecture: for a compact connected Kähler manifold of dimension at least two whose tangent bundle splits into two integrable holomorphic subbundles of positive rank, the ordinary universal cover is biholomorphic to a product of two connected simply connected complex manifolds. The product tangent directions identify with the specified lifted summands. The factors need not be compact, and a finite cover of the original manifold is not thereby asserted to be a product.

The second primary PDF, Theorem 1.1, claims automatic integrability for each specified two-positive-rank-summand tangent splitting of a smooth connected rationally connected projective complex manifold. Its Corollary 1.2, using Höring's earlier product theorem, claims a stronger decomposition of the manifold itself as a product compatible with the specified summands, not merely a decomposition of its universal cover.

The integrability hypothesis in the general compact Kähler result must not be omitted. Nor should the rationally connected projective conclusion be extended to every compact complex manifold without a separate theorem. The scope here is explicitly two summands, not an arbitrary number of summands.

GCL verification state

The source snapshot is pinned to openai/math@1581d4cbdd69a0faf34b2089614fe2a814316a6a. Both pinned primary PDFs were retrieved and read: Kähler universal-cover paper, Git blob 4869a892ebd2192cf9acb780d8263179f78255bf, 479,820 bytes, 33 pages, SHA-256 14d3e3f23539fd98735adb5042214f7f6eac9499a47e3303018932e00d29d333; rationally connected integrability paper, Git blob e07d0c5e6ad37a020f67fca8c10c770ae00fb9fd, 362,906 bytes, 16 pages, SHA-256 cb23bde917710f73321f4342b88dba8135aeb84b4c553f57df7d5532de73c01d. Abstracts, Theorem 1.1 in both, and the rationally connected Corollary 1.2 were checked.

This source-critical pass checks the stated hypotheses and the strength of the two conclusions; it does not independently prove global continuation of the local product structure, automatic integrability, or the classical product theorem. No Family 052-specific Lean proof, Comparator replay, semantic-equivalence audit or MATHCERT certificate has been verified.

What remains

The bounded primary-source critic checked both original theorem statements and separated the ordinary-universal-cover result from the stronger actual-product corollary in the rationally connected case. A full mathematical review must still inspect the global continuation of foliation boxes in the Kähler paper, tangent-box construction and bracket-tensor vanishing in the rationally connected paper, and the exact use of Höring's earlier theorem.

The article hash, mathematical Markdown, compendium tests, generated HTML and protected merge/readback are documentary acceptance controls. None establishes source-to-Lean theorem equivalence or mathematical certification.

Sources and claim boundaries

Boundary: NO_CERTIFICATION. This article reports two publisher-claimed geometric results without independently validating or formally certifying them.


Indexed primary manuscript paths: