Multiple rational normal forms in Lie theory
Dmitriy Voloshyn · 2025
Read Multiple rational normal forms in Lie theory by Dmitriy Voloshyn (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2506-01530.txt.
Introduction
Dmitriy Voloshyn studies how to split group elements in a solo 2025 mathematics paper. The setting is the algebra of connected reductive groups. Any typical element can be written as a product of structured parts. One part comes from a unipotent subgroup, another from a Borel subgroup.
A Weyl group representative sits between those parts in the formula. Voloshyn defines a class of rational versions of those representatives. Only some of them yield valid splits of this form. He then works out the properties of the good cases. The maps in the split are rational, meaning they can fail on thin exceptional sets. The paper adds tools to Lie theory's study of group structure.
Catalog record
| CREATOR | Dmitriy Voloshyn |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 75 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2506-01530
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Multiple rational normal forms in Lie theory Dmitriy Voloshyn Abstract We study the decomposition of a generic element g ∈ G g\in G of a connected reductive complex algebraic group G G in the form g = N ( g ) B ( g ) u ¯ N ( g ) − 1 g=N(g)B(g)\bar{u}N(g)^{-1} where N : G ⇢ 𝒩 − N:G\dashrightarrow\mathcal{N}_{-} and B : G ⇢ ℬ + B:G\dashrightarrow\mathcal{B}_{+} are rational maps onto a unipotent subgroup 𝒩 − \mathcal{N}_{-} and a Borel subgroup ℬ + \mathcal{B}_{+} opposite to 𝒩 − \mathcal{N}_{-} , and u ¯ \bar{u} is a representative of a Weyl group element u u . We introduce a class of rational Weyl group elements that give rise to such decompositions, and study their various properties. 0 0 footnotetext: 2010 Mathematics Subject Classification. Primary 20G07; Secondary 20F55, 13F60. 0 0 footnotetext: Key words and phrases. Normal form, rational Weyl group element, cluster algebra. 1 Introduction Let G be a connected complex reductive algebraic group, (\mathcal{B}_{+},\mathcal{B}_{-}) a pair of opposite Borel subgroups, \mathcal{N}_{\pm}\subset\mathcal{B}_{\pm} their unipotent radicals, \mathcal{H}:=\mathcal{B}_{+}\cap\mathcal{B}_{-} a Cartan subgroup, W:=N_{G}(\mathcal{H})/\mathcal{H} the corresponding Weyl group. The goal of the paper is to address the following problem: Problem 1.1 . For a given G , describe elements u of its Weyl group W for which there exist rational maps N:G\dashrightarrow\mathcal{N}_{-} and B:G\dashrightarrow\mathcal{B}_{+} such that for a generic g\in G , g=N(g)B(g)\overline{u}N(g)^{-1} (1.1) where \overline{u}\in N_{G}(\mathcal{H}) is a representative of u . Throughout the paper, we call an element u\in W a solution of Problem 1.1 if there exist rational maps B and N that give rise to the decomposition ( 1.1 ). In the next example, we illustrate the problem in \GL_{2}(\mathbb{C}) . Example 1.2 . For G:=\GL_{2}(\mathbb{C}) , identify the Weyl group with the group of permutation matrices: W=\left\{w_{0}:=\begin{bmatrix}0&1\\ 1&0\end{bmatrix},\ I:=\begin{bmatrix}1&0\\ 0&1\end{bmatrix}\right\}. (1.2) Set D:=\{X\in\GL_{2}(\mathbb{C})\ |\ x_{12}\neq 0\} . Then every X\in D has a decomposition of the form X=\begin{bmatrix}1&0\\ \frac{x_{22}}{x_{12}}&1\end{bmatrix}\begin{bmatrix}x_{12}&x_{11}+x_{22}\\ 0&-\frac{\det X}{x_{12}}\end{bmatrix}\begin{bmatrix}0&1\\ 1&0\end{bmatrix}\begin{bmatrix}1&0\\ -\frac{x_{22}}{x_{12}}&1\end{bmatrix}; (1.3) therefore, the element w_{0} is a solution of Problem 1.1 . On the other hand, let us set \Delta:=\Delta(X):=x_{11}^{2}-2x_{11}x_{12}+x_{22}^{2}+4x_{12}x_{21}. (1.4) Then for the identity permutation, X\in D decomposes as X=\begin{bmatrix}1&0\\ \frac{-x_{11}\pm\sqrt{\Delta}}{2x_{12}}&1\end{bmatrix}\begin{bmatrix}\frac{1}{2}(x_{11}+x_{22})\pm\frac{1}{2}\sqrt{\Delta}&x_{12}\\ 0&\frac{1}{2}(x_{11}+x_{22})\mp\frac{1}{2}\sqrt{\Delta}\end{bmatrix}\begin{bmatrix}1&0\\ -\frac{-x_{11}\pm\sqrt{\Delta}}{2x_{12}}&1\end{bmatrix} (1.5) Clearly, the identity permutation is not a solution of Problem 1.1 . Our motivation for studying Problem 1.1 originates in the theory of cluster algebras, as defined by S. Fomin and A. Zelevinsky in [ 6 ] , and in particular, in its interaction with Poisson geometry. For an overview of the subject and its applications, we refer to [ 5 , 8 , 9 , 14 , 16 ] . In [ 10 ] , M. Gekhtman, M. Shapiro, and A. Vainshtein constructed a generalized cluster structure in the ring of invariants 1 1 1 We use the notation \Ad\mathcal{N}_{-} to distinguish the invariant ring from the one where \mathcal{N}_{-} acts by left or right translations. So, p\in\mathbb{C}[\GL_{n}]^{\Ad\mathcal{N}_{-}} if and only if p(ngn^{-1})=p(g) for all g\in G , n\in\mathcal{N}_{-} . \mathbb{C}[\GL_{n}]^{\Ad\mathcal{N}-} , where \mathcal{N}_{-} acts on \GL_{n} by conjugation. The initial extended cluster of \mathbb{C}[\GL_{n}]^{\Ad\mathcal{N}_{-}} is a part of the initial extended clusters in \mathbb{C}[\GL_{n}] and \mathbb{C}[\GL_{n}\times\GL_{n}] . In the first case, the obtained cluster structure is compatible 2 2 2 A cluster structure is called compatible with a Poisson bracket if in any given extended cluster, the Poisson bracket is quadratic. with the Poisson bracket induced from the Poisson dual \GL_{n}^{*} of \GL_{n} , and in the second case, the Poisson structure on \GL_{n}\times\GL_{n} is that of the Drinfeld double of \GL_{n} . As we showed in [ 11 , 15 ] , the same initial extended cluster of \mathbb{C}[\GL_{n}]^{\Ad\mathcal{N}_{-}} is a part of the initial extended clusters in \mathbb{C}[\GL_{n}] and \mathbb{C}[\GL_{n}\times\GL_{n}] compatible with the Poisson brackets from the Belavin-Drinfeld class. We refer to [ 3 , 4 ] for details on Poisson geometry. The construction of the initial extended cluster in \mathbb{C}[\GL_{n}]^{\Ad\mathcal{N}_{-}} utilizes a solution of Problem 1.1 . Let C:=s_{1}\cdots s_{n-1} be a special Coxeter element in type A_{n-1} . By [ 10 , Lemma 8.2] , there exist rational maps B:\GL_{n}(\mathbb{C})\dashrightarrow\mathcal{B}_{+} and N:\GL_{n}(\mathbb{C})\dashrightarrow\mathcal{N}_{-} such that for a generic X\in\GL_{n}(\mathbb{C}) , X=N(X)B(X)\overline{C}N(X)^{-1}. (1.6) The initial extended cluster comprises n-1 isolated frozen variables, which are the characters \chi^{\omega_{i}} corresponding to the fundamental weights \omega_{1},\ldots,\omega_{n-1} , and the rest of the variables is obtained from the initial extended cluster of \mathbb{C}[\GL_{n-1}]^{\mathcal{N}_{-}} , where \mathcal{N}_{-} acts by multiplication on the right, via the pullback by the map \GL_{n}\ni X\mapsto B(X)^{[2,n]}_{[2,n]}\in\GL_{n-1}. (1.7) It is not known how to extend the above construction to other Lie types. In order to construct a generalized cluster structure in \mathbb{C}[G]^{\Ad\mathcal{N}_{-}} (and in related Poisson varieties), we are thus led to the study of Problem 1.1 , which we also find to be interesting in its own right. Let us briefly explain our results. We call an element u\in W of a Weyl group W rational if a certain graph \Gamma(u) associated with u is acyclic (see Definition 2.3 ). In [ 10 , Lemma 8.3] , it was shown that the longest Weyl group element w_{0}\in W is a solution of Problem 1.1 in any Lie type. We develop a recursive procedure that constructs a pair of maps (B,N) for u from the pair of maps for w_{0} (or, more generally, from any other solution of Problem 1.1 ). If u is rational, then the recursive procedure ends in a finite number of steps, and thus one obtains the decomposition ( 1.1 ) for u . We also prove a partial converse: if u is a sufficiently long element and is a solution of Problem 1.1 , then u is rational (see Theorem 2.4 ). We conjecture that the class of solutions of Problem 1.1 is precisely the class of rational Weyl group elements. We then define a rationality graph \Gamma(W) whose vertices are the rational Weyl group elements of W , and two vertices u,v\in W are connected by an edge iff there is a simple reflection s such that u=sv . Surprisingly, the graph \Gamma(W) is connected, and it contains more than one vertex if and only if w_{0} induces a nontrivial automorphism of the Dynkin diagram; that is, only in type A_{r} , r\geq 2 , D_{r} for r odd, and E_{6} (see Theorem 2.11 ; for examples of \Gamma(W) , see Appendix A ). In the context of constructing a generalized cluster structure in \mathbb{C}[\SL_{n}]^{\Ad\mathcal{N}_{-}} , the special Coxeter element C and its inverse C^{-1} exhibit the following properties: 1) The two elements are the only rational Coxeter elements; 2) The two elements are the only vertices of valency 1 in \Gamma(W) . However, no Coxeter element in any other Lie type is rational; nevertheless, in type D_{r} for r odd, the rationality graph \Gamma(W) also contains two elements of valency 1 , and this similarity leads us to believe that the two elements play a role in the construction of a generalized cluster structure in \mathbb{C}[\Spin(2r)]^{\Ad\mathcal{N}_{-}} . We will study this question in our next paper. We give precise definitions and statements of our main results in the next section. In sections 3 and 4 we provide the proofs. Acknowledgements. At various