🔢 mathematics

Borel subalgebras of Lie algebras of vector fields

This paper introduces the concept of integrable Borel subalgebras as the tangent algebras of Borel subgroups in the automorphism groups of affine varieties and provides a classification of these subalgebras for toric affine surfaces, including the affine plane and its cyclic quotients.

Original authors: Ivan Arzhantsev, Mikhail Zaidenberg

Published 2026-05-15

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Ivan Arzhantsev, Mikhail Zaidenberg

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

Imagine you are exploring a vast, infinite city called Aut(X). This city isn't made of buildings and streets, but of symmetries—all the possible ways you can twist, stretch, or rearrange a geometric shape (called a variety $X$ ) without tearing it apart.

In mathematics, every city has a "skeleton" or a "map" that describes how things move. For our city of symmetries, this map is a Lie algebra. Think of the Lie algebra as the collection of all possible "instantaneous moves" or "vector fields" you can make in the city.

The paper by Arzhantsev and Zidenberg is about finding the most important, "maximal" structures within this map. Specifically, they are hunting for Borel subalgebras.

The Core Concept: The "Maximal Solvable" Team

To understand a Borel subalgebra, let's use a team analogy.

The City (Aut(X)): The group of all symmetries.
The Map (Lie Algebra): The list of all possible moves.
Solvable: Imagine a team of workers who can get a job done by breaking it down into simple, step-by-step tasks. They don't create chaotic loops; they are orderly.
Borel Subalgebra: This is the largest possible orderly team you can form. You can't add any more workers to this team without making the whole group chaotic (unsolvable).

In the world of finite, simple shapes (like a standard sphere), there is a perfect one-to-one match: every "maximal orderly team" of workers corresponds to a "maximal orderly team" of symmetries. It's like saying, "If you have the best team of movers, you automatically have the best moving company."

The Problem: The City is Too Big

The authors point out that for more complex, infinite cities (like the affine plane or surfaces with special singularities), this perfect match breaks down.

The Issue: You can have a "maximal orderly team" of moves (a Borel subalgebra) that does not correspond to any actual "maximal orderly moving company" (a Borel subgroup) in the city.
The Metaphor: Imagine you have a list of instructions that is perfectly orderly and can't be improved. However, there is no actual company that follows exactly those instructions as a whole. The instructions exist, but the company doesn't.

The Solution: "Locally Integrable"

The authors introduce a new filter called "locally integrable."

Integrable: A set of moves is integrable if you can actually build a real, connected "moving company" (a subgroup) that follows those moves.
Locally Integrable: This is a slightly looser version. It means the moves can be built up in small, manageable chunks, each of which corresponds to a real company.

The Main Discovery (The "Aha!" Moment):
The paper proves a beautiful correspondence:

The "maximal orderly teams" of moves that are "locally integrable" are exactly the same as the "maximal orderly moving companies" in the city.

If you take a Borel subalgebra and check if it's "locally integrable," you are guaranteed that it comes from a real Borel subgroup. If it's not locally integrable, it's a "ghost" team that exists on paper but has no real-world counterpart.

The Case Studies: The Plane and the Torus

The authors test this theory on specific shapes:

The Affine Plane ( $A^2$ ):
- Here, the "triangular" teams (moves that only affect variables in a specific order, like $x$ then $y$ ) are the kings.
- They found that for the 2D plane, the triangular team is indeed the maximal orderly team, and it corresponds perfectly to a real subgroup.
- However, they also found "ghost" teams (Borel subalgebras) that are maximal but not locally integrable. These are the ones that don't correspond to real subgroups.
The 3D Space ( $A^3$ ):
- Things get messy in 3D. The "triangular" team, which was the king in 2D, is no longer the maximal orderly team in 3D.
- There are bigger, more complex orderly teams that the triangular team is a part of. This means the triangular team is not a Borel subalgebra in 3D, even though it feels like it should be.
Toric Surfaces (The "Cyclic Quotients"):
- These are surfaces made by folding the plane in specific ways (like taking a piece of paper and gluing edges with a twist).
- The authors classified the "maximal orderly teams" for these shapes.
- The Result: Depending on the "twist" (mathematically, a number $e$ and $d$ ), there is either one type of maximal team or two distinct types.
- If the twist is "symmetric" ( $e^2 \equiv 1$ ), there's only one type of team. If it's "asymmetric," there are two distinct types that cannot be transformed into each other.

A Surprising Twist: Simplicity

Finally, the authors looked at a specific type of move: zero divergence (moves that preserve volume, like squeezing a balloon without changing its size).

In the standard plane, the team of volume-preserving moves is "simple" (it has no smaller, independent sub-teams).
The Finding: On these folded, singular surfaces ( $X_{d,e}$ ), the team of volume-preserving moves is not simple. It has hidden sub-teams inside it. This is a significant difference between smooth shapes and these singular, folded shapes.

Summary

In simple terms, this paper is a guidebook for navigating the "symmetry cities" of mathematics.

It warns us that not every "perfectly ordered list of moves" corresponds to a real "moving company."
It introduces a test ("locally integrable") to tell the real companies from the ghost lists.
It proves that once you filter for the real companies, the map is perfect: every maximal orderly company has a unique, matching list of moves.
It maps out exactly what these companies look like for specific shapes, revealing that in 3D, the rules change, and on folded surfaces, the structure of the teams depends on how the surface is twisted.

Technical Summary: Borel Subalgebras of Lie Algebras of Vector Fields

Problem Statement
The paper addresses the structural relationship between the automorphism group $\text{Aut}(X)$ of an affine variety $X$ and its associated Lie algebra $\text{Lie}(\text{Aut}(X))$ . In the context of finite-dimensional algebraic groups, there is a well-established bijection between Borel subgroups and Borel subalgebras (maximal solvable subalgebras). However, for infinite-dimensional ind-groups like $\text{Aut}(X)$ , this correspondence breaks down. Specifically, not every Borel subalgebra of $\text{Lie}(\text{Aut}(X))$ corresponds to the tangent algebra of a Borel subgroup of $\text{Aut}(X)$ . The authors aim to characterize which subalgebras do correspond to Borel subgroups and to classify these "locally integrable Borel subalgebras" for specific classes of varieties, notably affine spaces and toric surfaces.

Methodology
The authors employ a combination of ind-group theory, the theory of derivations, and the structure of solvable subgroups in infinite-dimensional settings.

Definitions and Correspondence: They introduce the notion of a locally integrable Lie subalgebra, defined as one filtered by a sequence of locally finite (and thus finite-dimensional) Lie subalgebras. They establish that a subalgebra is locally integrable if and only if it is contained in the Lie algebra of a connected nested ind-subgroup.
Borel Correspondence: The core theoretical tool is the proof that the map $G \mapsto \text{Lie}(G)$ induces a bijection between the conjugacy classes of Borel subgroups of $\text{Aut}(X)$ and the $\text{Ad}$ -conjugacy classes of locally integrable Borel subalgebras of $\text{Lie}(\text{Aut}(X))$ . This relies on recent results regarding the structure of solvable and unipotent subgroups of $\text{Aut}(X)$ (citing works by Kraft, Arzhantsev, and others).
Case Studies: The methodology is applied to specific varieties:
- Affine Space ( $A^n$ ): The authors analyze the triangular de Jonquières subgroups and their Lie algebras. They utilize the bigrading of derivations on $k[x_1, \dots, x_n]$ to test maximality among solvable subalgebras.
- Toric Surfaces ( $X_{d,e}$ ): For normal affine toric surfaces defined as quotients $A^2/G_{d,e}$ , the authors utilize the amalgamated product structure of $\text{Aut}(X_{d,e})$ and the fixed-point subalgebras under the group action $G_{d,e}$ to describe the Lie algebra structure.

Key Contributions and Results

Characterization of Borel Subalgebras:
The paper proves that the Borel subgroups of $\text{Aut}(X)$ correspond precisely to the locally integrable Borel subalgebras of $\text{Lie}(\text{Aut}(X))$ (Theorem 1.3). This clarifies that while Borel subalgebras exist in the Lie algebra, only those satisfying the local integrability condition arise from algebraic Borel subgroups.
Classification for Affine Spaces ( $A^n$ ):
- $n=2$ : The triangular Lie subalgebra $j_2 = \text{Lie}(\text{JONQ}(A^2))$ is shown to be a Borel subalgebra (maximal among solvable subalgebras). All Borel subgroups of $\text{Aut}(A^2)$ are conjugate, and thus all locally integrable Borel subalgebras are $\text{Ad}$ -conjugate to $j_2$ .
- $n=3$ : The triangular subalgebra $j_3 = \text{Lie}(\text{JONQ}(A^3))$ is not a Borel subalgebra; it is not maximal among solvable subalgebras of $\text{Lie}(\text{Aut}(A^3))$ . The authors construct a strictly larger solvable subalgebra containing $j_3$ . Consequently, $\text{Aut}(A^3)$ possesses non-conjugate Borel subgroups and non-conjugate locally integrable Borel subalgebras.
Classification for Toric Surfaces ( $X_{d,e}$ ):
The authors classify locally integrable Borel subalgebras for the quotient surfaces $X_{d,e} = A^2/G_{d,e}$ .
- The number of conjugacy classes depends on the arithmetic condition $e^2 \equiv 1 \pmod d$ .
- If $e^2 \equiv 1 \pmod d$ , there is exactly one conjugacy class of locally integrable Borel subalgebras.
- If $e^2 \not\equiv 1 \pmod d$ , there are exactly two conjugacy classes (Corollary 1.4).
Structure of Lie Algebras on Toric Surfaces:
- The Lie algebra $\text{Lie}(\text{Aut}(X_{d,e}))$ is identified with a subalgebra of $\text{Lie}(\text{Aut}(A^2))$ fixed by the $\text{Ad}$ -action of $G_{d,e}$ .
- Under the condition $e \mid (d+1)$ , the authors prove that $\text{Lie}(\text{Aut}(X_{d,e}))$ coincides exactly with the fixed-point subalgebra $\text{Lie}(\text{Aut}(A^2))^{\text{Ad}(G_{d,e})}$ .
- Crucially, the subalgebra of zero divergence vector fields within $\text{Lie}(\text{Aut}(X_{d,e}))$ is not simple for singular toric surfaces ( $d > 1$ ). This contrasts with the smooth case ( $A^2$ ), where the zero divergence subalgebra is simple. This result provides a Lie-algebraic characterization of the smoothness of the toric surface.

Significance and Claims
The paper claims to resolve the ambiguity regarding the correspondence between Borel subgroups and Borel subalgebras in the infinite-dimensional setting of automorphism groups. By introducing "local integrability" as the necessary and sufficient condition, the authors provide a rigorous framework to distinguish between algebraic Borel subgroups and arbitrary maximal solvable subalgebras.

The work highlights a fundamental divergence between low and high dimensions in affine spaces: while the triangular subalgebra is maximal (and thus Borel) in dimension 2, it fails to be maximal in dimension 3. Furthermore, the paper extends the classification of Borel structures to singular toric surfaces, linking the number of Borel conjugacy classes to the arithmetic properties of the quotient group. Finally, the non-simplicity of the zero divergence subalgebra on singular toric surfaces is presented as a distinguishing feature of the singular case, offering a potential Lie-theoretic criterion for detecting singularities in affine toric varieties.