Stabilizer Subgroups and the Simplicity of Reduced Crossed Products
Yair Hartman, Mehrdad Kalantar · 2026
Read Stabilizer Subgroups and the Simplicity of Reduced Crossed Products by Yair Hartman, Mehrdad Kalantar (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-2605-22430.txt.
Introduction
Hartman and Kalantar prove that a simple reduced crossed product forces a stabiliser with trivial amenable radical. The setting is a minimal group action on a compact space. From this, they characterise simplicity for reduced crossed products of minimal actions. Covered classes are countable linear groups, hyperbolic groups, and groups with countably many amenable subgroups. The result answers Ozawa's 2014 question for these classes. For infinite uniformly recurrent subgroups of C*-simple groups, almost every subgroup has trivial amenable radical. The measure is fully supported and atomless.
Catalog record
| CREATOR | Yair Hartman, Mehrdad Kalantar |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 15806 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.22430
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Stabilizer Subgroups and the Simplicity of Reduced Crossed Products Thanks: YH was supported by the funding from the European Research Council (ERC) under the European Union’s Seventh Framework Programme (FP7-2007-2013) (Grant agreement No. 101078193). MK was supported by the NSF grant DMS-2155162, and The Simons Foundation grant SFI-MPS-TSM-00014307. For the purpose of open access, the authors have applied a CC-BY license to any author accepted manuscript arising from this submission. Yair Hartman, Mehrdad Kalantar Abstract . Given a minimal action G ↷ X G\curvearrowright X of a countable group G G on a compact space X X , we prove that if the reduced crossed product G ⋉ r C ( X ) G\ltimes_{r}C(X) is simple, then there exists a point whose stabilizer subgroup has trivial amenable radical. As a consequence, we give a complete characterization of the simplicity of the reduced crossed product of minimal actions of countable linear groups, hyperbolic groups, and, more generally, for groups with countably many amenable subgroups. This answers a question of Ozawa (2014) for these classes of groups. Furthermore, in the case of an infinite uniformly recurrent subgroup of a C ∗ C^{*} -simple group, we prove that almost every subgroup has a trivial amenable radical, with respect to a fully supported, atomless probability measure. The crossed product C^{*} -algebras associated with dynamical systems were among the very first constructed examples of operator algebras. Since their introduction, understanding the relationship between the dynamical properties of an action and the algebraic structure of the resulting crossed product has remained a central problem in the field. In particular, an old problem in C^{*} -algebra theory, which is the main focus of this work, is to determine when the reduced crossed product of a minimal action is simple, i.e. has no non-zero proper closed ideals. More precisely, let G be a countable group, X a compact Hausdorff space, and G\curvearrowright X by homeomorphisms. We are concerned with the question of finding necessary and sufficient conditions, in terms of the stabilizer subgroups, for the reduced crossed product G\ltimes_{r}C(X) to be simple. In [ 9 ] Elliott proved that G\ltimes_{r}C(X) is simple if G\curvearrowright X is topologically free, that is, the set of points with trivial stabilizer is dense. For a minimal action, this is equivalent to the existence of one point with a trivial stabilizer. This improved earlier results of Effros–Hahn [ 8 ] . Furthermore, for amenable groups, topological freeness was proved to characterize simplicity of the reduced crossed product by Kawamura and Tomiyama [ 14 ] (see also Archbold–Spielberg [ 3 ] along with Anantharaman-Delaroche [ 2 ] for the setup of topological amenable actions). However, the above characterization fails miserably for general actions of non-amenable groups. For example, the trivial action of the free group \mathbb{F}_{2} , which is as far as can be from being free, yields the simple reduced C^{*} -algebra of \mathbb{F}_{2} [ 19 ] . In [ 5 ] , it was proven that for a boundary action G\curvearrowright X , if G_{x} is C^{*} -simple for some x\in X , then G\ltimes_{r}C(X) is simple. Ozawa noted in [ 18 ] that the same holds for any minimal action G\curvearrowright X , and then asked whether the converse is true: Does simplicity of G\ltimes_{r}C(X) imply C^{*} -simplicity of stabilizer subgroups? The affirmative answer to this question would consequently give a complete characterization of the simplicity of G\ltimes_{r}C(X) in terms of the point stabilizers. In this direction, it was proven in [ 18 ] and [ 13 ] that if G\ltimes_{r}C(X) is simple, then the neighborhood stabilizer subgroups G_{x}^{o} consisting of group elements that pointwise fix an open neighborhood of x , are either all trivial or all non-amenable. While the problem of characterizing simplicity in terms of point stabilizers is formulated in topological terms, our strategy relies on “state-theoretic” and probabilistic techniques. We use the non-commutative stationary theory developed by the authors in [ 11 ] , and its generalization by Amrutam-Ursu [ 1 ] , which provides a characterization of simplicity in terms of stationary states. Applying this stationary approach to the crossed product yields Theorem A , while its application to the reduced group C^{*} -algebra yields Theorem B . The following is our main result. Theorem A . Let G be a countable group, X a compact space, and G\curvearrowright X by homeomorphisms. If G\ltimes_{r}C(X) is simple, then there exists a point x_{0}\in X such that the stabilizer subgroup G_{x_{0}} has trivial amenable radical. Recall that every C^{*} -simple group has trivial amenable radical ( [ 12 , 5 ] ), but the converse is not true in general [ 17 ] . However, for large classes of groups, including all linear groups, hyperbolic groups, and, more generally, groups with countably many amenable subgroups, these properties are equivalent (see e.g. [ 5 , Theorem 6.12] ). In particular, Theorem A yields a complete characterization of simplicity of the reduced crossed product C^{*} -algebras associated to minimal actions of these classes of groups. This answers the question of Ozawa in these cases. Corollary A . Let G be a countable group that is either linear or has countably many amenable subgroups (e.g., a hyperbolic group), and let G\curvearrowright X be a minimal action. Then G\ltimes_{r}C(X) is simple if and only if there exists a point x_{0}\in X such that the stabilizer subgroup G_{x_{0}} is C^{*} -simple. Remark 1 . Note that in the statement of Corollary A , we may replace the stabilizers with open stabilizers. Indeed, note that if for some x\in X , the stabilizer G_{x} is C^{*} -simple, then G^{o}_{x} is also C^{*} -simple, since it is a normal subgroup of G_{x} . Conversely, it was proven in [ 13 ] that if there exists some point x\in X whose open stabilizer G^{o}_{x} is C^{*} -simple, then the reduced crossed product is simple. Before we get into the proofs of the above results, let us briefly recall some definitions and fix our notation. We denote by {\rm Sub}(G) the Chabauty space of G , that is, the set of all subgroups of G , equipped with the product topology. This is a compact G -space, where G\curvearrowright{\rm Sub}(G) by conjugations. We denote by {\rm Sub}_{am}(G) the subset of all amenable subgroups of G , which is closed in {\rm Sub}(G) . For H\leq G we denote by R_{a}(H) the amenable radical of H . We recall the notion of C(X) -valued probability measures as defined and studied by Amrutam–Ursu in [ 1 ] . Given a compact space X , denote by C(X)^{+} the set of all non-zero positive functions f\in C(X) . By a C(X) -valued probability measure on G we mean a map \mu:G\ni s\mapsto\mu(s)\subset C(X)^{+} such that \sum_{s\in G}\sum_{{\varphi}\in\mu(s)}{\varphi}(x)=1 for every x\in X . We may formally represent \mu=\sum_{s\in G}\sum_{i\in I_{s}}{\varphi}_{i}\delta_{s} , where I_{s} is the index set such that \mu(s)=\{{\varphi}_{i}:i\in I_{s}\} . If G\curvearrowright X , then for a state \eta\in\mathcal{S}(G\ltimes_{r}C(X)) , the convolution \mu*\eta\in\mathcal{S}(G\ltimes_{r}C(X)) is defined by \mu*\eta(a)=\eta\Big(\sum_{s\in G}\sum_{i\in I_{s}}\sqrt{{\varphi}_{i}}\,s\!\cdot\!a\,\sqrt{{\varphi}_{i}}\,\Big) for all a\in G\ltimes_{r}C(X) . The state \eta\in\mathcal{S}(G\ltimes_{r}C(X)) is said to be \mu -stationary if \mu*\eta=\eta . Note that in the case of the trivial G -action, the above