The spectral space of conjugacy classes of subgroups of a compact Lie group

J.P.C.Greenlees · 2025

Read The spectral space of conjugacy classes of subgroups of a compact Lie group by J.P.C.Greenlees (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-2509-19409.txt.

Introduction

J. P. C. Greenlees, writing alone, describes the space of conjugacy classes of subgroups of a compact Lie group, under the Zariski topology. That space splits into separate blocks that do not touch. Each block is governed by one subgroup whose Weyl group is finite. Inside a block, an integral representation of the governing subgroup's component group controls the structure. The abstract states this decomposition in a single dense sentence, and this summary stays close to it.

Catalog record

CREATORJ.P.C.Greenlees
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE66 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2509-19409

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

The spectral space of conjugacy classes of subgroups of a compact Lie group Thanks: The work is partially supported by EPSRC Grant EP/W036320/1. The author would also like to thank the Isaac Newton Institute for Mathematical Sciences, Cambridge, for support and hospitality during the programme Equivariant homotopy theory in context, where later parts of work on this paper was undertaken. This work was supported by EPSRC grant EP/Z000580/1. J.P.C.Greenlees Abstract . The space of conjugacy classes of subgroups of a compact Lie groups with its Zariski topology splits as a disjoint union of clopen blocks, each dominated by a subgroup H H with finite Weyl group, and the structure of each block is controlled by an integral representation of π 0 ​ ( H ) \pi_{0}(H) . 1. Introduction 1.A. Overview The space \mathfrak{X}_{G}=\mathrm{Sub}(G)/G of conjugacy classes of closed subgroups of G has two topologies of interest. First of all, it has the h-topology, which is the quotient topology of the Hausdorff metric topology on \mathrm{Sub}(G) . This is a Stone space (compact, Hausdorff, totally disconnected). The space also has a Zariski topology whose closed sets are the h-closed sets which are also closed under passage to cotoral subgroups 1 1 1 L is cotoral in K if L is normal in K with quotient being a torus. For conjugacy classes (L) is cotoral in (K) if the relation holds for some representative subgroups. . Our focus is on the Zariski topology. In fact \mathfrak{X}_{G} (with the Zariski topology) is the Balmer spectrum of finite rational G -spectra [ 8 , 1 ] making it a spectral space. This also makes it the basis of an algebraic model for rational G -spectra. We will briefly describe the intended applications to G -equivariant cohomology theories in Subsection 1.D , but this paper is purely about the space \mathfrak{X}_{G} itself. It was shown in [ 1 ] that the space \mathfrak{X}_{G} has finite Cantor-Bendixson rank (equal to the rank of G ) and that the Cantor-Bendixson and Thomason filtrations agree. The results of the present paper refine this numerical information to give a rather detailed description of the space. We will show (Theorem 2.6 ) that the space \mathfrak{X}_{G} decomposes as a finite union of Zariski clopen subspaces (which we refer to as ‘blocks’) \mathfrak{X}_{G}=\mathrm{Sub}(G)/G=\mathcal{V}^{G}_{H_{1}}\amalg\cdots\amalg\mathcal{V}^{G}_{H_{n}}, where H_{1},H_{2},\cdots,H_{n} are subgroups with finite Weyl group. The block \mathcal{V}^{G}_{H} is dominated by a subgroup H in the sense that all conjugacy classes are represented by subgroups of H , and the structure of \mathcal{V}^{G}_{H} is controlled by the subgroup H . 1.B. Single blocks Here we give an outline of the structure of a single block \mathcal{V}^{G}_{H} in the sense that we describe the block ‘up to finite error’ to give a crude guide to the appearance. The body of the paper will make this more precise. 1.B.1. Shape We now describe the structure of an individual block \mathcal{V}^{G}_{H} : up to finite error we have \mathcal{V}^{G}_{H}\sim\mathfrak{P}^{G}_{H}\times\mathcal{N}^{G}_{H} where \mathfrak{P}^{G}_{H} is the poset of subgroups of a torus with the cotoral ordering and \mathcal{N}^{G}_{H} is a Stone space which, again up to finite error, has a product decomposition \mathcal{N}^{G}_{H}\sim\mathcal{N}^{G}_{H,2}\times\cdots\times\mathcal{N}^{G}_{H,t} as a product of standard Stone spaces (‘integral Grassmannians’). We outline the shape of these below. Altogether \mathcal{V}^{G}_{H}\sim\mathfrak{P}^{G}_{H}\times\mathcal{N}^{G}_{H,2}\times\cdots\times\mathcal{N}^{G}_{H,t},; its Cantor-Bendixson rank is finite and equal to the sum of those of the factors. 1.B.2. Fusion First of all, the block \mathcal{V}^{G}_{H} for G is obtained from the block \mathcal{V}^{H}_{H} by passing to G -conjugacy (the map \mathcal{V}^{H}_{H}\longrightarrow\mathcal{V}^{G}_{H} is a quotient map with finite fibres of bounded size). We therefore get the general form by looking at all top blocks \mathcal{V}^{G}_{G} . We first describe the two simplest types of block and then explain how to mix them. 1.B.3. Weyl-finite blocks The simplest case is when the block \mathcal{V}^{G}_{G} is Weyl-finite in the sense that all elements of \mathcal{V}^{G}_{G} have finite Weyl group. The reason for the simplicity is that in this case \mathcal{V}^{G}_{G} is itself a Stone space (with trivial cotoral ordering). An example to bear in mind is G=H=O(2) , then \mathcal{V}^{G}_{G}=\{(D_{2n})\;|\;n\geq 1\}\cup\{O(2)\}, topologised as the one point compactification of the positive integers. 1.B.4. Noetherian blocks The next case is when the subgroup G is a finite central extension of a torus. In this case the topology on \mathcal{V}^{G}_{G} is determined by the the partially ordered set of subgroups of the identity component G_{e} . The first example is when G=H is itself a circle, and \mathcal{V}^{G}_{G}=\{(C_{n})\;|\;n\geq 1\}\cup\{SO(2)\}, topologised like \mathrm{Spec}(\mathbb{Z}) : the proper closed subspaces are precisely the finite sets of finite subgroups C_{n} . 1.B.5. Mixed blocks For a general block, we mix the behaviour of Weyl-finite and Noetherian blocks. If G_{e}=\Sigma\times_{Z}T where \Sigma is semisimple, T is a torus and Z is a finite central subgroup then T is the identity component of the centre of the identity component, so we call it the central torus of G_{e} and write T=T_{z}G . We can identify the underlying spectral space \mathcal{V}^{G}_{G} in terms of the integral representation of the component group G_{d} on \Lambda_{0}=H_{1}(T) , together with the extension class \epsilon(G)\in H^{2}(G_{d};T)\cong H^{3}(G_{d};\Lambda_{0}) . Up to finite multiplicity, \mathcal{V}^{G}_{G} is a product of a Stone space \mathcal{N}^{G}_{G} space and a poset \mathfrak{P}^{G}_{G} . The Stone space \mathcal{N}^{G}_{G} is a neighbourhood of G in the space \Phi G of conjugacy classes of subgroups with finite Weyl groups. The poset \mathfrak{P}^{G}_{G} is obtained from \mathrm{Sub}(T_{z}^{1}G) by passing to G -conjugacy classes, where T_{z}^{1}G is the central torus of G itself (i.e., the subtorus of T_{z}G on which G_{d} acts trivially). In fact there is a continuous map \lambda:\mathcal{V}^{G}_{G}\longrightarrow\mathfrak{P}^{G}_{G}\times\mathcal{N}^{G}_{G} which is an isomorphism up to finite indeterminacy. We take \lambda((K))=(K\cap T_{z}^{1}G,\omega(K)) , where \omega(K)/K is the conjugacy class of the maximal torus of W_{G}(K) . The map \lambda is almost surjective in a sense we will explain, and the fibres over each pair are finite and of bounded cardinality. 1.C. The isotypical decomposition of the Stone space We may decompose further. Suppose \Lambda_{0}\otimes\mathbb{Q}=H_{1}(T;\mathbb{Q})=S_{1}^{\oplus m_{1}}\oplus S_{2}^{\oplus m_{2}}\oplus\cdots\oplus S_{t}^{\oplus m_{t}} where S_{1},S_{2},\ldots,S_{t} are distinct simple modules with S_{1} having trivial action. The factor S_{1}^{\oplus m_{1}} gives rise to the Noetherian factor \mathfrak{P}^{G}_{G} above. The factor \mathcal{N}^{G}_{G,S_{i}} corresponds to S_{i}^{\oplus m_{i}} , and is a Stone space of Cantor-Bendixson rank m_{i} , which is, up to finite error, an m_{i} -stage compactification of the spaces of full rank W -sublattices of \mathbb{Z}^{\oplus m_{i}} . 1.D. Application to G -spectra This subsection explains the motivation for this paper. It will play no mathematical role, so can be ignored by incurious readers. If we decompose the space \mathrm{Sub}(G)/G as a union of Zariski clopen subsets \mathrm{Sub}(G)/G=\mathfrak{X}_{1}\amalg\cdots\amalg\mathfrak{X}_{n}, then since the rational Burnside ring is [S^{0},S^{0}]^{G}=A(G)=C(\mathfrak{X}_{G},\mathbb{Q}) , idempotents in the Burnside ring show that the category of rational G -spectra decomposes into blocks \mbox{$G$-{\bf spectra}}\simeq\mbox{$G$-{\bf spectra}}\langle\mathfrak{X}_{1}\rangle\times\cdots\times\mbox{$G$-{\bf spectra}}\langle\mathfrak{X}_{n}\rangle where \mbox{$G$-{\bf spectra}}\langle\mathfrak{X}\rangle is the category of G -spectra with geometric isotropy in \mathfrak{X} . Thus we may conduct the analysis one block at a time. It is conjectured [ 6 ] that there is a small and calculable abelian model \mathcal{A}(G|\mathfrak{X}) so that the category DG- \mathcal{A}(G|\mathfrak{X}) of differential graded objects in \mathcal{A}(G|\mathfrak{X}) is a model for \mbox{$G$-{\bf spectra}}\langle\mathfrak{X}\rangle in the sense that there is a

Read the complete text