A complex-analytic characterization of Lagrangian immersions in $\mathbb C^n$ with transverse double points

Basic Information

• Paper ID: 2511.15306
• Title: A complex-analytic characterization of Lagrangian immersions in $\mathbb C^n$ with transverse double points
• Authors: Purvi Gupta, Rudranil Sahu (Indian Institute of Science, Bangalore)
• Classification: math.CV (Complex Variables), math.SG (Symplectic Geometry)
• Publication Date: November 20, 2025
• Paper Link: https://arxiv.org/abs/2511.15306

Abstract

This paper studies compact smooth totally real immersed $n$-dimensional submanifolds $M\subset\mathbb C^n$ with finitely many transverse double points. It is known that if $M$ is Lagrangian with respect to some Kähler form on $\mathbb C^n$, then $M$ is rationally convex in $\mathbb C^n$ (Gayet, 2000), but the converse does not hold (Mitrea, 2020). This paper proves that $M$ is Lagrangian with respect to some Kähler form on $\mathbb C^n$ if and only if $M$ is rationally convex and at each double point, the pair of transverse tangent planes satisfies a diagonalizability condition: there exists a complex linear transformation on $\mathbb C^n$ mapping this plane pair to $\left(\mathbb R^n,(D+i)\mathbb R^n\right)$, where $D$ is some real diagonal $n\times n$ matrix.

Research Background and Motivation

Problem Background

This paper addresses a fundamental question at the intersection of complex geometry and symplectic geometry: characterizing Lagrangian submanifolds with self-intersections.

Core Concepts:

1. Rational Convexity: A compact set $K\subset\mathbb C^n$ is called rationally convex if its complement is a union of complex hypersurfaces in $\mathbb C^n$. This endows $K$ with important approximation-theoretic properties: holomorphic functions on neighborhoods of $K$ can be uniformly approximated by rational functions.
2. Lagrangian Submanifolds: An $n$-dimensional real submanifold $M\subset\mathbb C^n$ is called Lagrangian with respect to a Kähler form $\omega$ if $\iota^*\omega=0$, where $\iota:M\to\mathbb C^n$ is the immersion map.
3. Totally Real Submanifolds: Submanifolds whose real subspaces contain no complex lines.

Existing Results

The classical Duval-Sibony theorem (1991-1995) establishes a perfect correspondence in the case without self-intersections:

• Embedding Case: The image of a compact smooth totally real embedding $\iota:M\to\mathbb C^n$ is rationally convex if and only if it is Lagrangian with respect to some Kähler form.

For the case with self-intersections:

• Gayet (2000): If an immersion $\iota:M\to\mathbb C^n$ has only finitely many transverse double points and is Lagrangian, then $\iota(M)$ is rationally convex (one-way implication).
• Mitrea (2020): Constructs counterexamples showing the converse fails, and proves that if $\iota(M)$ is rationally convex, then there exists a non-negative $(1,1)$-form $\omega$ that is strictly positive outside self-intersection points and satisfies $\iota^*\omega=0$.

Research Motivation

Core Problem: In the case with self-intersection points, what additional conditions make rational convexity imply Lagrangian property?

Key Insight: The crux of the problem lies in the local geometric structure at double points. Mitrea's counterexample shows that certain unions of totally real planes, though rationally convex, are not Lagrangian. Therefore, additional geometric constraints must be imposed at double points.

Core Contributions

1. Main Theorem (Theorem 1.1): Provides a complete characterization of immersed Lagrangian submanifolds with transverse double points: $M\text{ is Lagrangian} \Longleftrightarrow M\text{ is rationally convex} + \text{all double points are diagonalizable}$
2. Characterization of Diagonalizability (Proposition 3.2): Proves that the union of two totally real planes $\mathbb R^n$ and $S(A)=(A+i)\mathbb R^n$ is Lagrangian if and only if the matrix $A$ is diagonalizable over the reals.
3. Construction of Local Kähler Forms (Lemma 4.1): Explicitly constructs a local Kähler form $dd^cf$ near each diagonalizable double point satisfying $\iota^*(d^cf)=0$, which is key to subsequent global gluing.
4. Global Gluing Technique (Section 5): Cleverly glues local Kähler forms with Mitrea's degenerate Kähler form to obtain a global Kähler form.

Detailed Methodology

Problem Setup

Input: A compact smooth $n$-dimensional manifold $M$ and a totally real immersion $\iota:M\to\mathbb C^n$ with only finitely many transverse double points $p_1,\ldots,p_m$.

Output: Determine whether there exists a Kähler form $\omega$ on $\mathbb C^n$ such that $\iota^*\omega=0$.

Key Conditions:

• Rational convexity (necessary but not sufficient)
• Double point diagonalizability (supplementary condition)

Core Technical Framework

1. Linear Model Analysis (Section 3)

Problem Reduction: Through complex linear transformation, the local geometry of any transverse double point can be reduced to the union of two totally real planes $\mathbb R^n$ and $S(A)=(A+i)\mathbb R^n$ at the origin, where $A\in M_{n\times n}(\mathbb R)$.

Key Lemma (Lemma 3.3): If $P\in M_{n\times n}(\mathbb R)$ is nonsingular, then the complex linear transformation $P$ maps $\mathbb R^n$ to $\mathbb R^n$ and $S(A)$ to $S(PAP^{-1})$.

Main Result (Proposition 3.2): $\mathbb R^n\cup S(A)$ is Lagrangian with respect to some Kähler form if and only if $A$ is diagonalizable.

Proof Strategy:

• Consider constant-coefficient Kähler forms $\omega=\sum_{j,k}h_{jk}dz_j\wedge d\bar{z}_k$, where $h=(h_{jk})$ is a positive definite Hermitian matrix.
• Parametrization: $\iota(t)=(t_1,\ldots,t_n)$, $\iota_A(t)=(\sum a_{1k}t_k+it_1,\ldots,\sum a_{nk}t_k+it_n)$
• Compute pullbacks: $\iota^*\omega=2i\sum_{1\leq j<k\leq n}(\text{Im}\,h_{jk})dt_j\wedge dt_k$$\iota_A^*\omega=2i\sum_{1\leq j<k\leq n}c_{jk}(A)dt_j\wedge dt_k$ where $c_{jk}(A)=\sum_r((\text{Re}\,h_{jr})a_{rk}-(\text{Re}\,h_{kr})a_{rj})$
• Sufficiency: If $A$ is diagonalizable, by Lemma 3.3 we can assume $A$ is diagonal, and then the standard Kähler form satisfies the condition.
• Necessity: If $A$ is not diagonalizable, its real Jordan form $J(A)$ must contain nontrivial Jordan blocks. Consider two cases:
  • Real Eigenvalue Case: $J(A)$ contains $\begin{pmatrix}\lambda & 1 \\ & \lambda & \ddots \\ && \ddots & 1 \\ &&& \lambda\end{pmatrix}_{r\times r}$, $r\geq 2$. Computing $c_{12}(J(A))=h_{11}=0$, contradicting positive definiteness of $h$.
  • Complex Eigenvalue Case: $J(A)$ contains $\begin{pmatrix}C & I_2 \\ & C & \ddots \\ && \ddots & I_2 \\ &&& C\end{pmatrix}$, where $C=\begin{pmatrix}s & -t \\ t & s\end{pmatrix}$. Computing $c_{12}(J(A))=-t(h_{11}+h_{22})=0$, contradiction.

2. Local Kähler Form Construction (Section 4)

Goal: Construct a function $f$ near a diagonalizable double point $p$ such that:

• $\iota^*(d^cf)=0$ (stronger than just $\iota^*(dd^cf)=0$)
• $dd^cf>0$

Coordinate Setup: Through complex affine transformation, assume $p=0$, $T_0R=\mathbb R^n$, $T_0S=(A+i)\mathbb R^n$, where $A=\text{diag}(\lambda_1,\ldots,\lambda_n)$.

Local Representation: By the implicit function theorem, there exist smooth functions $\varphi,\psi:\mathbb R^n\to\mathbb R^n$ satisfying $\varphi(0)=\psi(0)=D\varphi(0)=D\psi(0)=0$ such that: $R=\{t+i\varphi(t)\}, \quad S=\{(A+i)t+\psi(t)\}$

Construction Strategy: Set $f(x,y)=\|x\|^2+\|y\|^2+r(x,y)$, where $r$ must satisfy: $(V r)(t,\varphi(t))=p(t)$$(W r)(A\cdot t+\psi(t),t)=q(t)$ where:

• $V=\nabla_x\cdot(D\varphi\circ\pi_x)-\nabla_y$
• $W=\nabla_x-\nabla_y\cdot(A+D\psi\circ\pi_y)$
• $p(t)=2\varphi(t)^T-2t^T\cdot(D\varphi)(t)$
• $q(t)=2t^T\cdot(D\psi)(t)-2\psi(t)^T$

Coordinate Change: Introduce the transformation $\Theta(u,v)=(x(u,v),y(u,v))$: $x(u,v)=u+A\cdot v+\psi(v), \quad y(u,v)=v-(A+(D\psi)(v))\cdot u$

This transformation straightens $S$ to $\{u=0\}$ and pulls back $W$ along $S$ to $\nabla_u$ along $\{u=0\}$.

Explicit Construction: Define $\tilde{r}(u,v)=Q(v)\cdot u+\alpha(u)\cdot(v-\sigma(u))$ where $\sigma$ is determined by the implicit function theorem such that $\Theta(\{v=\sigma(u)\})=R$, and $\alpha(u)=(P(u)-Q(\sigma(u))\cdot B(u)-u^T\cdot(DQ)(\sigma(u))\cdot C(u))\cdot(C(u)-(D\sigma)(u)\cdot B(u))^{-1}$

Through careful computation, verify that $\tilde{r}$ satisfies all requirements, then set $r=\tilde{r}\circ\Theta^{-1}$.

3. Global Gluing (Section 5)

Gluing Strategy:

1. For each double point $p_j$, Lemma 4.1 yields a local function $f_j$ on the ball $B^n(p_j,\varepsilon)$ satisfying $dd^cf_j>0$ and $\iota^*(d^cf_j)=0$.
2. Choose $\delta\in(0,\varepsilon)$ and a cutoff function $\chi:[0,\infty)\to[0,1]$ satisfying $\chi\equiv 1$ on $[0,\delta]$ and $\chi\equiv 0$ on $[\varepsilon,\infty)$.
3. By Mitrea's result, there exists \phi:\mathbb C^n\to\mathbb R} such that:
   • $\phi$ is plurisubharmonic
   • $\phi$ is strictly plurisubharmonic on $\mathbb C^n\setminus\bigcup_j B^n(p_j,\delta/2)$
   • $\iota^*dd^c\phi=0$
4. Choose sufficiently large $C>0$ such that $Cdd^c\phi+\sum_{j=1}^m d\chi_j\wedge d^cf_j>0 \text{ on } \mathbb C^n\setminus\bigcup_j B^n(p_j,\delta/2)$
5. Define the global Kähler form: $\omega=Cdd^c\phi+\sum_{j=1}^m d(\chi_j d^cf_j)$

Verification:

• On $\bigcup_j B^n(p_j,\delta)$: $dd^cf_j>0$ and $d\chi_j\equiv 0$ ensure $\omega>0$
• On $\mathbb C^n\setminus\bigcup_j B^n(p_j,\delta)$: Step 4 and $\chi_j dd^cf_j\geq 0$ ensure $\omega>0$
• Lagrangian property: $\iota^*\omega=C\iota^*dd^c\phi+\sum_j\iota^*d\chi_j\wedge\iota^*d^cf_j+\sum_j\chi_j d(\iota^*d^cf_j)=0$

Technical Innovations

1. Precise Characterization of Diagonalizability: First explicitly identifies diagonalizability as the essential obstruction to Lagrangian property, providing complete proof through Jordan normal form analysis.
2. Stronger Local Condition: Constructs local Kähler potentials satisfying $\iota^*(d^cf)=0$ rather than merely $\iota^*(dd^cf)=0$, which is key to the gluing technique.
3. Creative Application of Implicit Function Theorem: Through carefully designed coordinate transformation $\Theta$, converts nonlinear problems into solvable linear equation systems.
4. Explicit Construction: Provides completely explicit expressions for local Kähler potentials, avoiding the abstraction of existence proofs.

Experimental Setup

This is pure mathematical theoretical research with no numerical experiments or datasets. All results are obtained through rigorous mathematical proofs.

Experimental Results

Main Theoretical Results

Theorem 1.1 (Main Theorem): Let $M$ be a compact smooth $n$-dimensional manifold and $\iota:M\to\mathbb C^n$ a totally real immersion with finitely many transverse double points $p_1,\ldots,p_m$, such that $\iota(M)$ is a smooth submanifold except at these points. Then: $\iota(M)\text{ is Lagrangian} \Longleftrightarrow \iota(M)\text{ is rationally convex and } p_1,\ldots,p_m\text{ are diagonalizable}$

Proposition 3.2 (Linear Model): $\mathbb R^n\cup S(A)$ is Lagrangian $\Longleftrightarrow$ $A$ is diagonalizable over the reals.

Lemma 4.1 (Local Construction): Near a diagonalizable double point, there exists a local Kähler potential $f$ satisfying $\iota^*(d^cf)=0$ and $dd^cf>0$.

Theoretical Significance

1. Completeness: Provides the precise converse of Gayet's theorem in the transverse double point case, completely characterizing Lagrangian property.
2. Geometric Intuition: The diagonalizability condition has clear geometric meaning—tangent plane pairs can be simultaneously diagonalized, avoiding "twisting" caused by complex eigenvalues.
3. Technical Breakthrough: Explicit construction of local Kähler potentials may become a standard tool for similar problems.

Related Work

Classical Results

1. Duval (1991, 1994) and Duval-Sibony (1995): Establish foundational theory for the embedding case, proving equivalence of rational convexity and Lagrangian property without self-intersections.
2. Gayet (2000): Extends results to immersions with transverse double points, proving one-way implication (Lagrangian $\Rightarrow$ rationally convex).
3. Duval-Gayet (2008): Further extends to cases with second-order tangencies.
4. Weinstock (1988): Studies polynomial convexity of unions of two totally real planes, giving necessary and sufficient conditions for local rational convexity (matrix $A$ has no purely imaginary eigenvalues with modulus greater than 1).
5. Mitrea (2020): Constructs counterexamples showing rational convexity does not imply Lagrangian property, and proves existence of degenerate Kähler forms.

Positioning of This Paper

This paper completes a full cycle in this research direction:

• Compared to Gayet: provides precise conditions for the converse theorem
• Compared to Mitrea: finds positive characterization starting from counterexamples
• Compared to Weinstock: elevates from rational convexity to Lagrangian property

Conclusions and Discussion

Main Conclusions

This paper completely resolves when a totally real immersion with transverse double points is Lagrangian: it must and only needs to satisfy rational convexity plus diagonalizability at double points. This result elegantly unifies two perspectives: complex analysis (rational convexity) and symplectic geometry (Lagrangian property).

Limitations

1. Restriction to Self-Intersection Type: Only considers transverse double points, excluding more complex self-intersections (higher multiplicity points, tangency points, etc.).
2. Verification of Diagonalizability: In practical applications, determining whether a given double point is diagonalizable requires computing tangent spaces and matrix eigenvalues.
3. Higher-Dimensional Generalization: The method heavily depends on special structures of $\mathbb C^n$; generalizing to more general Kähler manifolds requires new techniques.
4. Non-Uniqueness of Construction: The Kähler form construction involves multiple arbitrary choices (cutoff functions, constant $C$), lacking canonicality.

Future Directions

The paper does not explicitly propose future directions, but the following are worth exploring:

1. Generalization to Other Self-Intersection Types: Study more complex cases such as quartic tangencies and triple points.
2. Quantitative Theory: Study curvature bounds of Kähler forms and geometric invariants of immersions.
3. Dynamical Applications: Lagrangian submanifolds have important applications in Hamiltonian systems; results here may have dynamical significance.
4. Algorithmic Implementation: Develop effective algorithms to determine whether a given immersion satisfies the conditions.

In-Depth Evaluation

Strengths

1. Problem Importance: Resolves a core problem unsolved for over 20 years in this field, perfecting the Duval-Sibony theoretical framework.
2. Completeness of Results: Provides necessary and sufficient conditions with no ambiguity. The diagonalizability condition is both simple and verifiable.
3. Technical Depth of Proofs:
   • Jordan normal form analysis (Proposition 3.2) cleverly exploits positive definiteness contradictions
   • Local Kähler potential construction (Lemma 4.1) involves complex coordinate transformations and differential equation solving
   • Global gluing (Section 5) ingeniously uses Mitrea's degenerate forms
4. Clarity of Exposition:
   • Well-structured: from linear models → local construction → global gluing, progressing systematically
   • Comprehensive notation system (Section 2 explains notation in detail)
   • Key steps are sufficiently explained
5. Universality of Methods: The local-to-global gluing technique may apply to other geometric problems.

Weaknesses

1. Computational Complexity: Section 4 computations are extremely lengthy (particularly around equations (15)-(24)), though rigorous, readability is affected. Could be simplified through more abstract frameworks.
2. Insufficient Geometric Intuition: The geometric meaning of the diagonalizability condition (why it is precisely the obstruction) lacks intuitive explanation. Deeper understanding from symplectic geometry perspective would be valuable.
3. Missing Counterexamples: While citing Mitrea's counterexamples, the paper provides no concrete examples of non-diagonalizable double points, weakening intuition for necessity of the condition.
4. Insufficient Discussion of Generalizations: Lacks discussion of implications for more general cases (non-compact manifolds, higher-multiplicity self-intersections).
5. Weak Application Background: Does not discuss potential applications in symplectic topology, mirror symmetry, and other fields.

Impact

Theoretical Contributions:

• Completes the final piece of the Duval-Gayet-Mitrea research program
• Reveals subtle differences between rational convexity and Lagrangian property
• Provides a paradigm for studying Lagrangian submanifolds with self-intersections

Technical Contributions:

• Explicit construction of local Kähler potentials may become standard tool
• Application of Jordan normal forms in symplectic geometry demonstrates power of algebraic methods

Limitations:

• As pure theoretical result, limited immediate applications
• High technical threshold requiring dual expertise in complex and symplectic geometry

Applicable Scenarios

1. Theoretical Research: Provides tools for studying topology and geometry of Lagrangian submanifolds.
2. Mirror Symmetry: Lagrangian submanifolds are central to mirror symmetry; results may help understand singular cases.
3. Symplectic Topology: When studying Lagrangian skeleta of symplectic manifolds, need to handle self-intersecting submanifolds.
4. Complex Dynamics: Totally real submanifolds relate to invariant sets in certain dynamical systems; conditions here may help characterize invariant set properties.

References (Key References in the Paper)

1 J. Duval. Convexité rationnelle des surfaces lagrangiennes. Invent. Math., 104(1):581–599, 1991.

2 J. Duval. Une caractérisation kählérienne des surfaces rationnellement convexes. Acta Math., 172:77–89, 1994.

4 J. Duval and N. Sibony. Polynomial convexity, rational convexity, and currents. Duke Math J., 97(2):487–513, 1995.

5 D. Gayet. Convexité rationnelle des sous-variétés immergées lagrangiennes. Ann. Scient. Éc. Norm. Sup., 33(2):291–300, 2000.

6 O. Mitrea. A characterization of rationally convex immersions. J. Geom. Anal., 30(1):968–986, 2020.

7 B. M. Weinstock. On the polynomial convexity of the union of two maximal totally real subspaces of $\mathbb C^n$. Math. Ann., 282(1):131–138, 1988.

Summary

This paper is an important theoretical contribution at the intersection of complex geometry and symplectic geometry. By introducing the diagonalizability condition, it completely characterizes when totally real immersions with transverse double points are Lagrangian. The proof combines algebraic (Jordan normal forms), analytic (implicit function theorem, differential equations), and geometric (coordinate transformations, gluing techniques) tools, demonstrating sophisticated technical mastery. While as pure theoretical work its applications are limited, its completeness and elegance make it a landmark result in the field. For researchers studying singularity theory of Lagrangian submanifolds, this is essential reading.