Gao (Math. Res. Lett. 2025) generalized Huang's Lemma — the central rigidity tool for proper holomorphic mappings between balls and hyperquadrics — by proving that solutions of with and have a rank-one multiplier and partially linear structure, and correspondingly that local orthogonal pairs from into are quasi-standard or null. A generalized Whitney map shows that both conclusions fail at the borderline , where the multiplier can have Hermitian rank greater than one, and no structural description is known there; likewise the classification of local orthogonal pairs between and indefinite targets is available only in restricted ranges. The problem is to classify the local structure of orthogonal pairs (equivalently, of solutions of the generalized Huang equation) at this borderline codimension — whether every pair is null, quasi-standard, or of Whitney type — and to complete the classification for indefinite targets in the ranges not covered by existing rigidity theorems, which is exactly the input the transfer principle for Levi-degenerate sources consumes as a hypothesis.
Background
The rigidity of holomorphic mappings between bounded domains has been a central theme of several complex variables since Poincaré. A benchmark is proper holomorphic maps between unit balls (maps whose preimage of every compact set is compact; between bounded domains this forces the boundary to be mapped to the boundary): Huang proved that a map which is -smooth up to the boundary with is a totally geodesic (linear) embedding. The key tool, now called Huang's Lemma, states that if holomorphic functions () vanish at the origin and satisfy with real-analytic, then . The invariant controlling such identities is the Hermitian rank: a real-analytic function has Hermitian rank at most when it can be written as with holomorphic , and the minimal such is its rank (a notion originating in Hilbert's 17th problem and its sums-of-squares analogues). Huang's Lemma says a multiplier of is forced to vanish in the low-term range; Xiao proved the first non-degenerate version: at term count , either or has rank one and the tuples are linear up to holomorphic factors.
The paper under consideration (Gao) works in a coordinate-free geometric framework. Let denote with the standard, possibly indefinite and degenerate Hermitian form (the last coordinates enter trivially), and let be its projectivization. The generalized ball is its set of positive points; the boundary, consisting of null points, is a CR hypersurface (a real hypersurface of complex space whose tangent spaces carry a maximal complex subspace — the natural setting for boundary rigidity questions) which is Levi-degenerate when (the Levi form, the intrinsic pseudoconvexity curvature of such a hypersurface, is here the Hermitian form itself, which has a -dimensional kernel along the boundary). A local orthogonal pair is a pair of holomorphic maps from an open subset of to such that with respect to the target form whenever with respect to the source form; this is the projective incarnation of a pair of Segre maps (the Segre variety of a boundary point is the set of points joined to it by complex lines lying in the ambient hypersurface — a standard local invariant that linearizes CR mapping questions), and the diagonal case is a local orthogonal map, the coordinate-free model of a local CR map between hyperquadrics. A pair is null if for all (a trivial solution), standard if it is induced by a linear conformal pair of the underlying Hermitian spaces, and quasi-standard if it splits as a direct sum of a standard part and a null part. In homogeneous coordinates an orthogonal pair satisfies a generalized Huang equation , and the multiplier — determined up to a product of two holomorphic functions — completely characterizes the two rigid alternatives: the pair is null if and only if , and quasi-standard if and only if has Hermitian rank one.
Gao's main theorem extends Huang's Lemma to the sharp term range for arbitrary signature: if and , then and the tuples are partially linear, and , with constant matrices satisfying and . A companion transfer theorem deduces rigidity for Levi-degenerate sources from the non-degenerate case: if all local orthogonal pairs from to are quasi-standard or null, then the same holds from for every . The numerical bound is sharp: at the generalized Whitney map of Huang–Xiao satisfies the equation with multiplier of Hermitian rank three (throughout, denotes the non-degenerate source dimension), so the rank-one conclusion, and with it quasi-standard rigidity, fails. The known rigidity patches are restricted: Gao's earlier theorem covers pairs from into positive-definite targets with ; Gao–Ng cover local orthogonal maps when (with ), when (with ), and when (with ); Huang–Xiao independently proved the orthogonal-map transfer theorem for degenerate sources. No structural description exists at the borderline or for general indefinite targets outside these ranges — precisely the input that further applications of the transfer principle require as a hypothesis.
Problem Statement
Let with , and set . Classify the local orthogonal pairs (and, as the diagonal case, the local orthogonal maps) from into the positive-definite target , up to the natural equivalence generated by linear conformal changes of source and target (that is, by linear conformal pairs in the sense of Definition 2.2 of the source paper: pairs of linear maps satisfying for some ) and by the direct-sum splitting into standard and null parts. Concretely, determine whether every such pair is null, quasi-standard, or equivalent to a generalized Whitney-type pair as in Example 1.2 of the source paper (the pair whose multiplier is the Hermitian-rank-three polynomial described in the background), or whether further local structures occur at this borderline codimension; in the non-degenerate case this amounts to classifying the solutions of the generalized Huang equation at term count , including the possible Hermitian ranks of the multiplier, the uniqueness and moduli of the holomorphic factors, and the structure of the complementary components . As the linked second component of the same objective — the missing hypothesis for the transfer theorem (Theorem 1.3 of the source) — determine for which signature pairs and which target dimensions beyond every local orthogonal pair from the non-degenerate source to is quasi-standard or null, and classify the exceptions when this dichotomy fails.
References
- Yun Gao, Hermitian rank and rigidity of holomorphic mappings, Mathematical Research Letters 32(3), 771-788 (2025). DOI: 10.4310/MRL.250729002731, https://arxiv.org/abs/2402.07126
- Yun Gao, Orthogonal Pair and a Rigidity Problem for Segre Maps Associated to Real Hyperquadrics, The Journal of Geometric Analysis 32, Article 211 (2022). DOI: 10.1007/s12220-022-00949-5, https://arxiv.org/abs/2109.09340
- Yun Gao, Sui-Chung Ng, Local orthogonal maps and rigidity of holomorphic mappings between real hyperquadrics, Journal de Mathématiques Pures et Appliquées 179, 219-231 (2023). DOI: 10.1016/j.matpur.2023.09.005, https://arxiv.org/abs/2110.04046
- Xiaojun Huang, Ming Xiao, Rigidity of Mappings Between Degenerate and Indefinite Hyperbolic Spaces, The Journal of Geometric Analysis 33, Article 23 (2023). DOI: 10.1007/s12220-022-01080-1
- Ming Xiao, A theorem on Hermitian rank and mapping problems, Mathematical Research Letters 30(3), 945-968 (2023). DOI: 10.4310/mrl.2023.v30.n3.a12
- Xiaojun Huang, Ming Xiao, Holomorphic mappings between hyperquadrics with positive signature, Pure and Applied Mathematics Quarterly 18(2), 599-616 (2022). DOI: 10.4310/pamq.2022.v18.n2.a11
- Xiaojun Huang, On a linearity problem for proper holomorphic maps between balls in complex spaces of different dimensions, Journal of Differential Geometry 51, 13-33 (1999). DOI: 10.4310/jdg/1214425024
- Abdullah Al Helal, Jiří Lebl, Hermitian rank in ideal powers, arXiv:2503.05976 (2025), https://arxiv.org/abs/2503.05976
- Yun Gao, Sui-Chung Ng, On the Rank of Hermitian Polynomials and the SOS Conjecture, International Mathematics Research Notices 2023(13), 11276-11290 (2023). DOI: 10.1093/imrn/rnac145
- Yun Gao, Rigidity of holomorphic maps between Shilov boundaries of type-I bounded symmetric domains, arXiv:2508.05980 (2025), https://arxiv.org/abs/2508.05980
- Yun Gao, Gap phenomenon of holomorphic maps between Shilov boundaries of type-I bounded symmetric domains, arXiv:2608.01259 (2026), https://arxiv.org/abs/2608.01259
- Yun Gao, Some results related to Macaulay's Theorem about Hilbert functions and applications, arXiv:2512.21590 (2025), https://arxiv.org/abs/2512.21590
- Mingyue Chi, Sui-Chung Ng, On the Restrictions of Rational Maps on Linear Subspaces and Some Applications in CR Geometry, Acta Mathematica Sinica, English Series 42, 1274-1282 (2024). DOI: 10.1007/s10114-024-3437-y
- Michael Reiter, Duong Ngoc Son, On CR maps between hyperquadrics and Winkelmann hypersurfaces, arXiv:2404.16543 (2024), https://arxiv.org/abs/2404.16543
- Nguyen Gia Hien, Michael Reiter, Duong Ngoc Son, The classification of CR maps from hyperquadrics into tubes over null cones of symmetric forms, arXiv:2510.13252 (2025), https://arxiv.org/abs/2510.13252