Classification of Fuchsian groups with torsion
George Peterzil · 2026
Read Classification of Fuchsian groups with torsion by George Peterzil (2026) 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-2606-31459.txt.
Introduction
Peterzil settles a classification question for Fuchsian groups with torsion. These are discrete groups of motions of the hyperbolic plane, and torsion means some of their elements are rotations. Bergfalk and Smythe had shown that surfaces can be classified by concrete invariants, and they asked whether the same holds for orbifolds, where torsion appears.
Peterzil answers yes. In that setting too, the action on discrete subgroups that are finitely generated admits concrete classification. A second result follows: a nonsingular ergodic space for the group, with large finitely generated stabilizers, must be homogeneous. The proofs rest on Selberg's lemma and a Greenberg result on commensurators, the near-symmetries of a subgroup. Concrete classification is a strong form of order for a classification problem.
Catalog record
| CREATOR | George Peterzil |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 29 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2606.31459
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Classification of Fuchsian groups with torsion Thanks: The research was supported by the ISF (grant No. 957/25). George Peterzil Abstract . In their recent paper, Bergfalk and Smythe prove that the isometry equivalence relation on hyperbolic surfaces with finitely-generated fundamental group is concretely classifiable, and ask whether the same result holds true for 2-dimensional hyperbolic orbifolds, or equivalently, whether the action of PSL 2 ( ℝ ) \PSL_{2}(\R) on its space of finitely-generated discrete subgroups is concretely classifiable. In this note we answer this question in the affirmative. We then use the result to prove that a nonsingular ergodic PSL 2 ( ℝ ) \PSL_{2}(\R) -space with nonelementary finitely-generated stabilizers is homogeneous, in similarity with a result of Stuck-Zimmer for lattices in semisimple lie groups. The main ingredients of our proof are Selberg’s lemma and a result of Greenberg on commensurators. 2020 Mathematics Subject Classification Primary 03E15, Secondary 22D40 1. Introduction Given a linear group H\leq\text{GL}_{n}(\mathbb{C}) we write \mathcal{S}(H) for the collection of closed subgroups of H . We equip \mathcal{S}(H) with the Chabauty topology. This is a compact metrizable space, and in particular it is Polish. We write \mathcal{S}_{d}(H)\subseteq\mathcal{S}(H) for the subspace consisting of discrete subgroups of H . The group H acts on \mathcal{S}(H) continuously by conjugation, such that \mathcal{S}_{d}(H) is an invariant set. This space was introduced in [ 9 ] and has since been widely used in various contexts. Descriptive set-theoretically, it is natural to ask for the complexity of the associated orbit equivalence relation, in general and when restricted to certain subspaces. In [ 6 ] , Bergfalk and Smythe provide a systematic framework for tackling manifold classification problems descriptive set-theoretically. In particular, they consider the problem of classification of hyperbolic surfaces up to isometry, or equivalently, the problem of classification of Fuchsian groups up to conjugacy. Using Teichmüller theory, they prove the following theorem. Theorem 1.1 (Theorem J in [ 6 ] ) . The action by conjugation of \PSL_{2}(\R) on the space of torsion-free, finitely-generated Fuchsian groups is concretely classifiable 1 1 1 We follow their notational lead by using the term concretely classifiable rather than the term smooth in order to avoid confusion. . Somewhat surprisingly, the analogous statement for \PSL_{2}(\mathbb{C}) is proven false in the same paper (Theorem C). They also demonstrate that classification of discrete subgroups which are not necessarily finitely-generated is impossible in any noncompact semisimple lie group (Theorem I), following the unclassifiability of subgroups of the free group F_{2} proved by Thomas and Velickovic [ 22 ] and of groups containing F_{2} by Andretta, Camerlo and Hjorth [ 4 ] . It is then asked whether the same is true without the assumption that the groups are torsion-free. We answer this question in the affirmative. Theorem 1.2 . The action by conjugation of \PSL_{2}(\R) on the space of finitely-generated Fuchsian groups is concretely classifiable. Proof outline We use a result of Stuck–Zimmer on the classifiability of lattices to reduce the problem to the nonlattice case, where for elementary groups we construct a Borel transversal directly. For the nonelementary, nonlattice case, using Selberg’s lemma [ 19 ] , we associate to each group \Gamma the intersection \Gamma_{m} of all of its subgroups of index n_{\Gamma} , where n_{\Gamma} is the minimal index of a torsion-free subgroup of \Gamma , and using it we assign to a finitely-generated nonelementary Fuchsian group \Gamma the group \Comm_{G}(\Gamma)_{m} (where \Comm_{G}(\Gamma) is the commensurator of \Gamma in \PSL_{2}(\R) ), which we prove induces a countable-to-one map between the conjugacy classes with torsion and those without torsion using a theorem of Greenberg on commensurators. The result of Bergfalk-Smythe tells us that the latter is concretely classifiable, and the theorem then follows from a descriptive set-theoretic argument. The methods of Bergfalk and Smythe point at a natural reason why classifying torsion-free subgroups could be simpler; the problem of classifying manifolds is apriori more accessible than the problem of classifying orbifolds . Thus, we prove the reduction to the torsion-free case for certain classes of discrete subgroups of an arbitrary linear group (see 3.1 ), hoping that it would prove useful. A Stuck-Zimmer type corollary In their seminal paper [ 21 ] , Stuck and Zimmer prove that the action of a lie group on its space of lattices is concretely classifiable. Using it, they furthermore prove: Theorem 1.3 (Lemma 3.5 in [ 21 ] ) . Let H be a semisimple lie group with no compact factors, and let (X,\mu) be a nonsingular ergodic H -space such that for \mu -almost every x\in X we have that the stabilizer H_{x} is a lattice. Then there is a measure-class preserving, H -equivariant isomorphism (X,\mu)\cong(H/\Gamma,\nu_{\Gamma}) for some lattice \Gamma\leq H . The paper [ 21 ] has initiated the systematic study of probability measures on the space of subgroups, in particular for those invariant under conjugation, labeled Invariant Random Subgroups, or IRS for short (see [ 14 ] for a short survey). This notion has proved incredibly fruitful (see, for example, [ 3 ] , [ 5 ] , [ 1 ] ). In the particular case of \PSL_{2}(\R) , Biringer and Raimbault [ 7 ] provide a complete classification of the almost-sure topological type of the quotient of an invariant random subgroup. For more general rank one groups, several interesting theorems and constructions are given in [ 2 ] . The bulk of the theory described above has been developed for conjugation-invariant probability measures on the space of subgroups, rather than for quasi -invariant probability measures (with some exceptions, see [ 12 ] for a notable example). It is worth noting that very little is known at the full generality of quasi-invariant random subgroups beyond 1.3 (although this terminology was introduced several years after their paper). Following the original proof of Stuck and Zimmer and using 1.2 we prove the theorem below. Theorem 1.4 . Let (Y,\nu) be a nonsingular ergodic \PSL_{2}(\R) -space such that almost every stabilizer is finitely-generated and nonelementary. Then (Y,\nu) is isomorphic to a homogeneous space \PSL_{2}(\R)/\Gamma , equipped with its canonical invariant measure class. Structure of the paper In section 2 we review some preliminaries and basic results from descriptive set theory and from the theory of commensurators. In section 3 we introduce the notion of a small commensurator class, in section 4 we prove 1.2 , and in section 5 we prove 1.4 . Acknowledgments I thank Johanna Steinmeyer, who has tragically passed away as this project was nearing its completion, for her encouragement and support. I also thank Nachi Avraham-Re’em, Jeffrey Bergfalk, Or Landesberg and Iian Smythe for enriching conversations and correspondences regarding this body of work. 2. Preliminaries and basic facts Notation Given an equivalence relation E on a set X and some X_{0}\subseteq X , we write E\restriction X_{0} for E\cap X_{0}\times X_{0} . We write H for a general real linear algebraic group and G for \PSL_{2}(\R) . Write \mathcal{S}_{\fg}(H)\subseteq\mathcal{S}_{d}(H) for the subspace consisting of finitely-generated discrete subgroups of H , and \tilde{\mathcal{S}}_{\fg}(H)\subseteq\mathcal{S}_{\fg}(H) for those which are torsion-free. For a subgroup H_{0}\leq H and g\in H we write H^{g}_{0}=gH_{0}g^{-1} . We write \mathcal{F}(H)\supseteq\mathcal{S}(H) for the space of closed subsets of H , equipped with the