Remarks on the intersection of two quadrics

Nigel Hitchin

REVIEW 3 major objections 3 minor 1 cited by

Remarks on the intersection of two quadrics

T0 review · 3 major / 3 minor · reviewed 2026-08-06 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read For the intersection of two quadrics, a point is very stable exactly when one natural polynomial has distinct roots.

desk verdict A solid research note with a genuinely new very-stable criterion and a clean parabolic Higgs reinterpretation; the stress-test counterexample is wrong, but two or three spots need tightening. read the letter →

arxiv 2506.23671 v1 pith:VZH54RJ3 submitted 2025-06-30 math.AG

Nigel Hitchin This is my paper

classification math.AG MSC 14H6014H7014N05

keywords intersectionoftwoquadricsintegrablesystemsHiggsbundlesquasi-parabolicverystablepointswobblylocusgeometricLanglandshyperellipticcurves

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper takes the explicit formula for the integrable system on the cotangent bundle of an intersection $Q\cap Q_1$ of two quadrics and reinterprets it as the quasi-parabolic rank-2 Higgs bundle system on the projective line. In this language a point of the quadric intersection is very stable, meaning no nonzero cotangent vector makes all conserved quantities vanish, precisely when a single polynomial $p(z)=\sum_i x_i^2\prod_{j\neq i}(z-\mu_j)$ has distinct roots, counting $z=\infty$ as a possible root. The payoff is a concrete description of the wobbly locus, the complement of the very stable points, as a discriminant hypersurface, together with a bridge from the classical geometry of quadrics to the geometric and analytic Langlands programmes.

What carries the argument

The central object is the quasi-parabolic rank-2 Higgs bundle on $\mathbb{P}^1$ with Higgs field $\Phi=\sum_{i=1}^N v_i\otimes v_i/(z-\mu_i)\,dz$, where the $v_i$ live in a two-dimensional symplectic space and the $\mu_i$ are the parameters of the pencil of quadrics. From it one forms the polynomial $p(z)=\sum_i x_i^2\prod_{j\neq i}(z-\mu_j)$, whose degree $n=N-3$ is the dimension of $X$; its roots $a_k$ are the points at which $\Phi$ preserves the trivial subbundle. The Poisson-commuting functions of the system become $\lambda_k^2=\left(\sum_i x_i y_i/(a_k-\mu_i)\right)^2$, so the integrable-system map factors as a linear isomorphism followed by a sum of squares. The determinant of the matrix linking these coordinates is a product of $x_i$, values $p(\mu_i)$, and Vandermonde factors, so it is nonzero when all $x_i\neq 0$ and the $\mu_i$ and $a_k$ are distinct; the cases with $x_i=0$ are folded in by induction on $n$, and nilpotence is detected by the matrix form of $\Phi$.

What would settle it

On a smooth intersection with $N=5$ and distinct $\mu_i$, take a point where $p$ has a double root, compute the resultant $\operatorname{Res}(p,p')$, and solve the linear system $\ell_i(y)=0$ for a nonzero cotangent vector $y$; if the resultant vanishes but no such $y$ exists, or vice versa, Proposition 1's criterion fails.

Watch

Extended reading notes

Core claim

The paper's central claim, Proposition 1, is that for a smooth intersection $X=Q\cap Q_1$ of two quadrics, a point $(x_1,\dots,x_N)$ of $X$ is very stable with respect to the integrable system if and only if the polynomial $p(z)=\sum_{i=1}^N x_i^2\prod_{j\neq i}(z-\mu_j)$ has distinct roots, with $z=\infty$ counted as a root. The proof translates the system into a meromorphic Higgs field $\Phi=\sum_i v_i\otimes v_i/(z-\mu_i)\,dz$ on the rank-2 bundle $O\oplus O(-1)$ over $\mathbb{P}^1$, whose degree-$n=N-3$ polynomial $p$ records the points where $\Phi$ preserves the distinguished trivial subbundle. When all roots of $p$ are distinct, the map from the cotangent bundle to the base of the integrable system is a linear isomorphism followed by a sum of squares, hence proper, so the point is very stable; when $p$ has a multiple zero, a nonzero nilpotent Higgs field exists, so the point is wobbly. The paper then identifies the wobbly locus with the discriminant of $p$ and connects this rank-2 picture to $\tau$-invariant $\operatorname{Spin}(2g)$ bundles on a hyperelliptic curve.

Load-bearing premise

The whole construction presupposes that the two quadrics meet smoothly, so the parameters $\mu_1,\dots,\mu_N$ are mutually distinct; the proof does not extend to singular intersections in which two of them coincide.

Editorial extensions

If this is right

The wobbly locus is the inverse image of the discriminant hypersurface defined by the resultant of $p$ and $p'$ under the map $X\to\mathbb{P}^n$ sending $x$ to the squares $x_i^2$.
For $n=3$, the criterion recovers the known description of wobbly bundles on a genus-2 curve as a discriminant, matching the twistor Hecke eigensheaf picture in that case.
The same data define commuting second-order differential operators on a square root of the canonical bundle of $X$, giving a concrete model for the analytic Langlands correspondence.
Through the identification with $\tau$-invariant $\operatorname{Spin}(2g)$ bundles on a hyperelliptic curve, the rank-2 description provides a moduli space in which the nilpotent cone of the Higgs-bundle integrable system is compactified.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

The resultant criterion gives a direct computational test for very stability: evaluate $\operatorname{Res}(p,p')$ at any point of $X$ with all $x_i\neq 0$, and the point is wobbly exactly when the resultant vanishes, without needing to solve for cotangent vectors.
The roots $a_1,\dots,a_n$ of $p$ provide separation-of-variables coordinates, so on the open set where they are distinct they should give action-angle coordinates for the integrable system and simplify the integration of its flows.
The formula for $p$ is algebraic in the coordinates $x_i$ and the parameters $\mu_i$, so one could test numerically whether the discriminant locus varies continuously as the quadrics degenerate and whether the limiting wobbly locus matches the singular intersection case.

Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

The short version: this is a solid research note, and the headline result, Proposition 1, looks right. The stress-test counterexample does not survive contact with the paper: the parametrized y is a linear combination of x and μx, so it represents the zero vector in the cotangent fiber. Evaluating s_i on it proves nothing. You need a y not in span{x, μx} to test wobbliness, and the paper's criterion handles the rest.

What's genuinely new: the very-stable criterion for arbitrary n (previously only n=3 in [5]), the clean reinterpretation of the Beauville–Höring–Liu–Voisin system as a rank-2 quasi-parabolic Higgs system on P^1, and the differential-operator perspective in Section 6. The link to Benedetti et al. in Section 7 is mostly expository but useful — it makes the rank-2 nature of the Spin(2g) Higgs field visible.

Soft spots, in proportion:

1. Section 4's determinant evaluation is asserted without proof and has a typo ('μ_n' should be 'μ_m'). It's a standard Cauchy-like computation, so this is a cosmetic flaw, not a substantive one.

2. Section 6 silently assumes the canonical bundle has a square root. The formula k = -(N-4)/2 gives an integer only for N even. For odd N, O(k) is not a line bundle on P^{N-1}, so the analytic-Langlands analogy only applies to even N (odd-dimensional X). That should be stated.

3. The proof of Proposition 1 leans on a PGL(2) invariance claim that is asserted, not proved. I believe it's true — reparametrizing the pencil does not change X or the integrable system up to isomorphism — but a reader has to fill that in. The 'including z=∞' clause is handled by moving infinity, and the stress-test worry about that step is unfounded, as I said.

The paper is for people who work on integrable systems or geometric Langlands and want a concrete computable example. It deserves a serious referee. I'd want the typo fixed and the parity/PGL(2) issues clarified before publication.

Referee Report

3 major / 3 minor

Summary. The paper interprets the Beauville–Hörng–Liu–Voisin integrable system on the cotangent bundle of the intersection X of two quadrics in terms of quasi-parabolic rank-2 Higgs bundles on P^1. It characterizes very stable points by the roots of the polynomial p(z) (Proposition 1), constructs commuting second-order differential operators on a square root of the canonical bundle, and identifies X with a moduli space of invariant Spin(2g) bundles following Ramanan and Benedetti–Hörng–Liu. The central claim is that a point x in Q∩Q1 is very stable if and only if p has distinct roots, including a possible root at infinity.

Significance. The paper is clearly written and valuable for its explicit formulas and for the reinterpretation in terms of parabolic Higgs bundles and Spin(2g) bundles; if the very-stability criterion were correct, it would give a concrete model of the wobbly locus as a discriminant, useful in geometric Langlands. The paper also makes a genuine link between the quadric intersection and recent work of Benedetti, Hörng, and Liu. However, the central criterion in Proposition 1 is false as stated; the explicit counterexample given in the report shows that a point with only a root at infinity is not necessarily very stable. This affects the main theorem and the proof's PGL(2)-invariance step. The remaining sections may be salvageable, but the central characterization requires substantial revision.

major comments (3)

[Section 5, Proposition 1] The criterion is false as stated. Take N=4, μ=(0,1,2,3), and x=(i,√3,i√3,1). Then Σ x_i^2=0 and Σ μ_i x_i^2=0, so x lies on Q∩Q1, and p(z)=6 is constant, i.e. p has only the simple root at infinity in the sense of the statement. The proposition therefore declares x very stable. But the cotangent fiber is one-dimensional and is covered by y(β)=(1, -i√3/2+β√3, 2i√3β, i/2+3β), β∈C; one checks x·y(β)=μx·y(β)=0 and substitution into Eq. (3) gives s_i(y(β))=0 for i=1,2,3,4 and every β. Hence every cotangent vector at x lies in the zero fiber of the integrable system, so x is maximally wobbly, not very stable. The flaw is in the PGL(2)-reduction in the proof: a Möbius transformation changes the marked points μ_i and the Hecke modification at infinity, and no isomorphism preserving very-stability is exhibited. The 'including z=∞' clause is exactly the failure point and needs a separate argument.
[Section 4, Eq. (8)] The determinant evaluation is asserted without proof and is load-bearing for the linear independence of the ℓ_i used in Proposition 1. The displayed formula also has an index typo: the last product should involve μ_ℓ−μ_m rather than μ_ℓ−μ_n. Since the argument must handle degeneracies such as x_i=0 or a_i=μ_j, this computation, with its precise hypotheses, needs to be supplied or replaced by a reference.
[Section 5, converse in Proposition 1] The proof for a repeated root is only sketched. After choosing the basis of H^0(P^1,O(n−1)) by evaluation at a1,a3,...,an and derivative at a1, the text asserts that trΦ^2=(z−a1)^2(d0+…) makes all functions of the integrable system vanish; this needs an explicit statement of how the corresponding section of O(n−1) is zero, and the modification for several multiple zeros is not given. The argument should be completed, since this is the key step proving that points with multiple finite roots are not very stable.

minor comments (3)

[Section 4, determinant formula] The product in the determinant evaluation is written as ∏_{ℓ<m}(μ_ℓ−μ_n); the index should be m, not n, and the range of the product should be stated consistently.
[Section 6] The claim that O(k) with k=−(N−4)/2 is a square root of the canonical bundle requires the existence of such a line bundle on X. For a smooth intersection of two quadrics with n≥3, Pic(X)=Z, and when N−4 is odd no such square root exists; for example, N=7 gives K_X=O(−3), which is not divisible by two in Pic(X). The differential-operator construction therefore needs an additional hypothesis or a spin-structure discussion.
[Section 5, Remark 2] The statement that the complement of the very stable points is the inverse image of the discriminant hypersurface should be revisited after the correction to Proposition 1, since points with a root at infinity are not captured by the resultant of p and p′ in the usual affine sense.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the very-stability criterion is proved by explicit determinant and polynomial computations, not by presupposing the conclusion.

full rationale

The paper's main claim, Proposition 1, identifies very stable points with those for which the polynomial p(z) has distinct roots. This is not circular: the proof derives the equivalence from the explicit structure of the integrable system. The functions of the system are the coefficients of tr(Phi^2); after moving to the parabolic Higgs bundle description, they coincide with squares of linear forms ell_i(y) evaluated at the roots a_i of p. The key step is the determinant computation of the matrix (8), which is shown to be nonzero exactly when the a_i are distinct and the x_i are nonzero. Very stability is then equivalent to the linear independence of the ell_i, which is the same determinant condition. This is a genuine mathematical equivalence rather than a restatement of the definition. The formula for the integrable system is taken from Beauville et al. [2] as an input, and the paper reinterprets it; it does not rename a known result as a prediction. The self-citation [9] is used to identify the commuting functions with the parabolic Hitchin system, but that is a standard external framework and is not load-bearing in the sense of assuming Proposition 1. There are no fitted parameters, no prediction renamed from a fit, and no uniqueness theorem imported from the authors' prior work to force the conclusion. The skeptical counterexample concerning a root at infinity, if valid, would show a mathematical error in the proof's PGL(2) reduction or in the treatment of the point z = infinity; it would be a correctness flaw, not circularity, because the contested assertion is not an input of the derivation. Accordingly, the circularity score is low.

Assumptions & free parameters 0 free parameters · 7 assumptions · 0 invented entities

The paper relies on standard theorems in symplectic geometry, Higgs bundle theory, and orthogonal bundle theory, cited from [2], [4], [9], [13], [15]. All are external, previously established results; none are introduced ad hoc for this paper. No free parameters are fitted. No new physical or mathematical entities are postulated beyond standard constructions (the polynomial p(z), the bundle W, etc., which are derived, not postulated).

assumptions (7)

domain assumption The symplectic quotient m^{-1}(0)/SL(2,C) is the cotangent bundle of the quadric Q (as a coadjoint orbit of SO(N)).
Used in Section 2 to pass from the moment-map equations (1) to T*Q. This is a standard fact in symplectic geometry, but stated without proof.
domain assumption The integrable system for parabolic Higgs bundles has Poisson-commuting functions given by the coefficients of tr Φ^2.
Taken from the author's earlier paper [9] and used in Section 3 to identify the Beauville et al. formula with the parabolic Hitchin system.
domain assumption The formula for commuting functions on X given by Beauville et al. [2, Prop. 7.4] is correct and is the starting point of the paper.
Section 3 explicitly says 'this is the formula in [2] (Proposition 7.4)'.
domain assumption Ramanan's theorem identifies X (odd dimension) with the moduli space of τ-invariant Spin(2g) bundles on a hyperelliptic curve C.
Used in Section 7, cited from [15].
domain assumption Bhosle's correspondence gives an equivalence between τ-invariant orthogonal bundles on C and degenerate orthogonal bundles on P^1.
Used in Section 7 to translate the Spin(2g) Higgs field to a meromorphic endomorphism of W; cited from [4].
standard math For the intersection of two quadrics to be smooth, the parameters μ_i must be distinct.
Stated in Section 4 ('the μi are distinct for a smooth intersection'); it underlies the degree and residue computations.
standard math The operators Σ_i Ω_{ij}/(μ_i-μ_j) commute when the Ω_{ij} satisfy the Kohno-Drinfeld relations.
Used in Section 6 to show [Δ_i, Δ_j]=0; cited as 'exactly as in the KZ-equation'.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Remarks on the intersection of two quadrics." pith.science (2026). https://pith.science/paper/VZH54RJ3

@misc{pith2026250623671,
  author       = {Pith},
  title        = {Pith review of: Remarks on the intersection of two quadrics},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VZH54RJ3}},
  note         = {Machine review of arXiv:2506.23671}
}

read the original abstract

The article takes the formula for the integrable system defined by Beauville et al on the cotangent bundle of the intersection of two quadrics X, and interprets it in terms of rank 2 quasi parabolic Higgs bundles on the projective line. We then discuss aspects related to the geometric Langlands programme in this simple concrete context. We conclude with a description of the link with the recent paper of Benedetti et al identifying X and its integrable system in terms of invariant Spin(2g) bundles on a hyperelliptic curve of genus g.

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

A universal Higgs bundle moduli space
math.DG 2026-03 conditional novelty 7.0 of 10

A symplectic connection over Teichmüller space, derived from the harmonic-map energy, carries an integrable complex structure that realizes the universal Higgs bundle moduli space and satisfies Higgs-bundle-like curva...

Reference graph

Works this paper leans on

16 extracted references · 16 canonical work pages · cited by 1 Pith paper

[5]

R.Donagi, T.Pantev & C.Simpson, Twistor Hecke eigensheaves in genus 2, arXiv 2403.17045

work page Pith review arXiv
[1]

M.F.Atiyah, Complex fibre bundles and ruled surfaces , Proc. Lond. Math. Soc. 5 (1955) 40– 434

work page 1955
[2]

doi:10.1112/mod.2024.3

A.Beauville, A.H¨ oring, J.Liu & C.Voisin, Symmetric tensors on the in- tersection of two quadrics and Lagrangian fibration , Moduli (2024);1:e4. doi:10.1112/mod.2024.3

work page doi:10.1112/mod.2024.3 2024
[3]

V.Benedetti, A.H¨ oring & J.Liu,Intersection of two quadrics: modular interpreta- tion and Hitchin morphism , arXiv 2506.04707

work page Pith review arXiv
[4]

Bhosle-Desale, Degenerate symplectic and orthogonal bundles on P1, Math

U. Bhosle-Desale, Degenerate symplectic and orthogonal bundles on P1, Math. Annalen 267 (1984) 347–364

work page 1984
[6]

P.Etingof, E.Frenkel & D.Kazhdan, A general framework for the analytic Lang- lands correspondence, Pure and Applied Mathematics Quarterly 20 (2024) 307– 426

work page 2024
[7]

P.Etingof & H.Liu, Hitchin systems and their quantization , arXiv 2409.09505

work page arXiv
[8]

Hausel, Enhanced mirror symmetry for Langlands dual Hitchin systems , Pro- ceedings of ICM 2022, 2228–2249

T. Hausel, Enhanced mirror symmetry for Langlands dual Hitchin systems , Pro- ceedings of ICM 2022, 2228–2249

work page 2022

Show all 16 references

[9]

N.J.Hitchin, Stable bundles and integrable systems, Duke Math. J. 57 (1988), 91–114

1988
[10]

N.J.Hitchin, Multiplicity algebras for rank 2 bundles on curves of small genus , International Journal of Mathematics 35 (2024)

2024
[11]

J.Hurtubise, Separation of variables and the geometry of Jacobians , Sigma 3 (2007) 017

2007
[12]

G.Laumon, Un analogue global du ˆ cone nilpotent, Duke Math. J. 54 (1987), 647– 671

1987
[13]

P.Newstead, Stable bundles of rank 2 and odd degree over a curve of genus 2, Topology 7 (1968) 205-215

1968
[14]

C.Pauly & A.Pe´ on-Nieto,Very stable bundles and properness of the Hitchin map , Geometriae Dedicata 198 (2019) 143–148. 12

2019
[15]

Indian Acad

S.Ramanan, Orthogonal and spin bundles over hyperelliptic curves , Proc. Indian Acad. Sci. Math. Sci., 90 (1981) 151— 166

1981
[16]

Mathematical Institute University of Oxford Woodstock Road Oxford OX2 6GG UK 13

E.Sklyanin, Separation of variables in the Gaudin model, Journal of Soviet Math- ematics 47 (1989) 2473–2488. Mathematical Institute University of Oxford Woodstock Road Oxford OX2 6GG UK 13

1989

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.

[1] [5]

R.Donagi, T.Pantev & C.Simpson, Twistor Hecke eigensheaves in genus 2, arXiv 2403.17045

work page Pith review arXiv

[2] [1]

M.F.Atiyah, Complex fibre bundles and ruled surfaces , Proc. Lond. Math. Soc. 5 (1955) 40– 434

work page 1955

[3] [2]

doi:10.1112/mod.2024.3

A.Beauville, A.H¨ oring, J.Liu & C.Voisin, Symmetric tensors on the in- tersection of two quadrics and Lagrangian fibration , Moduli (2024);1:e4. doi:10.1112/mod.2024.3

work page doi:10.1112/mod.2024.3 2024

[4] [3]

V.Benedetti, A.H¨ oring & J.Liu,Intersection of two quadrics: modular interpreta- tion and Hitchin morphism , arXiv 2506.04707

work page Pith review arXiv

[5] [4]

Bhosle-Desale, Degenerate symplectic and orthogonal bundles on P1, Math

U. Bhosle-Desale, Degenerate symplectic and orthogonal bundles on P1, Math. Annalen 267 (1984) 347–364

work page 1984

[6] [6]

P.Etingof, E.Frenkel & D.Kazhdan, A general framework for the analytic Lang- lands correspondence, Pure and Applied Mathematics Quarterly 20 (2024) 307– 426

work page 2024

[7] [7]

P.Etingof & H.Liu, Hitchin systems and their quantization , arXiv 2409.09505

work page arXiv

[8] [8]

Hausel, Enhanced mirror symmetry for Langlands dual Hitchin systems , Pro- ceedings of ICM 2022, 2228–2249

T. Hausel, Enhanced mirror symmetry for Langlands dual Hitchin systems , Pro- ceedings of ICM 2022, 2228–2249

work page 2022