A finite Linear Dependence of Discrete Series Multiplicities

Kaustabh Mondal and Gunja Sachdeva · 2025

Read A finite Linear Dependence of Discrete Series Multiplicities by Kaustabh Mondal and Gunja Sachdeva (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-2506-00542.txt.

Introduction

Mondal and Sachdeva prove a finiteness result for discrete series multiplicities in twenty twenty-five. They work with a semisimple Lie group that has a compact Cartan subgroup. A uniform lattice in the group gives a quotient space for analysis. Discrete series representations are a distinguished class of unitary representations of the group.

For any finite set of such representations that meets a set condition, the authors study their multiplicities. Those few multiplicities determine all discrete series multiplicities in the quotient space. That yields a refinement of the strong multiplicity one result for this class. As an application, they turn to spaces of cusp forms at two given levels. Equality of dimensions over a finite set of weights implies equality for all weights.

Catalog record

CREATORKaustabh Mondal and Gunja Sachdeva
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE63 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2506-00542

Browse the full library · About the Tower of Babel

Full text

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

A finite Linear Dependence of Discrete Series Multiplicities Kaustabh Mondal and Gunja Sachdeva Abstract . Let G G be a connected semisimple simply connected Lie group with a compact Cartan subgroup and let Γ \Gamma be a uniform lattice in G G . Let G ^ d \widehat{G}_{d} denote the set of equivalence classes of unitary discrete series representations of G G . We prove that for any finite subset of G ^ d \widehat{G}_{d} satisfying a certain condition, the associated finite set of discrete series multiplicities in L 2 ​ ( Γ \ G ) L^{2}(\Gamma\backslash G) determines all discrete series multiplicities in L 2 ​ ( Γ \ G ) L^{2}(\Gamma\backslash G) . This allows us to obtain a refinement of the strong multiplicity one result for discrete series representations. As an application, we deduce that for two given levels, the equality of the dimensions of the spaces of cusp forms over a suitable finite set of weights implies the equality of the dimensions of the spaces of cusp forms for all weights. 2020 Mathematics Subject Classification 22E40, 22E46, 11F11 1. Introduction 1.1. Discrete Series Multiplicities Let G be a semisimple connected Lie group whose complexification G_{{\mathbb{C}}} is simply connected. We assume that G contains a compact Cartan subgroup, i.e. \text{rank}(G)=\text{rank}(K) , where K is a maximal compact subgroup in G . Consider \Gamma to be a lattice in G such that \text{vol}(\Gamma\backslash G)<\infty with respect to the unique G -invariant Haar measure on \Gamma\backslash G . We have the right regular representation R_{\Gamma} of G on (1.1) L^{2}(\Gamma\backslash G):=\{\phi:G\rightarrow{\mathbb{C}}\ |\ \phi(\gamma x)=\phi(x)\ \text{for all}\ \gamma\in\Gamma,x\in G\ \text{and}\ \int\limits_{\Gamma\backslash G}|\phi(x)|^{2}dx<\infty\}. Let L^{2}_{0}(\Gamma\backslash G) be the maximal semisimple subrepresentation of L^{2}(\Gamma\backslash G) . Then L^{2}_{0}(\Gamma\backslash G)\cong\widehat{\bigoplus\limits_{\pi\in\widehat{G}}}m(\pi,\Gamma)\pi, where \widehat{G} is the set of equivalence classes of unitary irreducible representations of G , and m(\pi,\Gamma)=\text{dim}\ {\rm Hom}_{G}(\pi,L^{2}_{0}(\Gamma\backslash G)). Since G has compact Cartan subgroups it possesses the discrete series representations. It is well-known that discrete series representations are parametrized by the Harish-Chandra parameters \lambda , and denoted by \pi_{\lambda} (see Subsec. 2.1 ). The purpose of this article is to present a linear dependence among the multiplicities m(\pi_{\lambda},\Gamma) of the discrete series representations \pi_{\lambda} of G . The multiplicity m(\pi_{\lambda},\Gamma) depends on \text{vol}(\Gamma\backslash G) . For example, if \Gamma is uniform torsion free lattice in G then the multiplicity m(\pi_{\lambda},\Gamma) is equal to d_{\pi}\text{vol}(\Gamma\backslash G) , where d_{\pi} is the formal degree of \pi (see [ KN01 ] for the definition). Moreover, if \Gamma contains (at most finitely many) nontrivial elliptic elements (i.e. torsion elements) then these elliptic elements also contribute to m(\pi_{\lambda},\Gamma) . Let us fix a maximal compact subgroup K in G . By the assumption on G , there exists a compact Cartan subgroup T\subset K . Let \mathfrak{g}_{0},\mathfrak{k}_{0},\mathfrak{t}_{0} be the Lie algebras of G,K,T respectively, and \mathfrak{g},\mathfrak{k},\mathfrak{t} be their complexifications respectively. Let \Phi=\Phi(\mathfrak{g},\mathfrak{t}) and \Phi_{c}=\Phi(\mathfrak{k},\mathfrak{t}) be the root systems of (G,T) and (K,T) respectively such that \Phi_{c}\subset\Phi . Let \Phi^{+} be a fixed but arbitrary choice of a positive system of roots in \Phi . A Harish-Chandra parameter \lambda is an integral regular linear form in the complex dual space \mathfrak{t}^{*} of \mathfrak{t} . Let \mathcal{D}(G) denote the set of all Harish-Chandra parameters in G (see Subsec. 2.1 for detailed definition). Every element \lambda\in\mathcal{D}(G) determines a unique positive system of roots \{\alpha\in\Phi:\langle\lambda,\alpha\rangle>0\} which is denoted by P^{\lambda} . Any discrete series representation is uniquely determined by such a Harish-Chandra parameter \lambda up to W_{c} -conjugate (see Thm. 2.1.1 for details). Here, W_{c} denotes the Weyl group of the root system \Phi_{c} . Definition 1.1.1 . Assume \lambda_{1},\lambda_{2}\in\mathcal{D}(G) are given such that P^{\lambda_{1}}=P^{\lambda_{2}} . The infinite set \mathcal{S}(\lambda_{1},\lambda_{2}):=\{\lambda_{1}+k\lambda_{2}\in\mathfrak{t}^{*}:k\in{\mathbb{N}}\cup\{0\}\} of integral linear forms is called a string with base \lambda_{1} and direction \lambda_{2} . For each k\in{\mathbb{N}}\cup\{0\} , we denote \lambda_{1}+k\lambda_{2} by \lambda_{k} . Note that every linear form \lambda_{k}\in\mathcal{S}(\lambda_{1},\lambda_{2}) is regular and therefore belongs to \mathcal{D}(G) . The discrete series representation determined by \lambda_{k}\in\mathcal{D}(G) is denoted by \pi_{\lambda_{k}} . Each \lambda_{k} determines the same positive system of roots i.e. P^{\lambda_{k}}=P^{\lambda_{1}}=P^{\lambda_{2}} . We need to impose a mild condition on \lambda\in\mathcal{D}(G) . We denote \mathcal{D}^{\star}(G):=\{\lambda\in\mathcal{D}(G):\langle\lambda-\delta^{\lambda},\alpha\rangle>0\ \text{for all}\ \alpha\in P^{\lambda}\cap\Phi_{n}\}, where \Phi_{n}=\Phi\setminus\Phi_{c} and \delta^{\lambda}:=\frac{1}{2}\sum\limits_{\alpha\in P^{\lambda}}\alpha . From Lem. 2.4.3 , it can be seen that if the base or the direction of a string belongs to \mathcal{D}^{\star}(G) then every linear form in the string belongs to \mathcal{D}^{\star}(G) . Now we are ready to display our first main result. Theorem 1.1.2 . Let G be a semisimple connected Lie group whose complexification G_{{\mathbb{C}}} is simply connected. Let K be a maximal compact subgroup in G . Assume that G contains a compact Cartan subgroup. Let \Gamma be a uniform lattice in G , and q=|\Gamma\cap K| . For two Harish-Chandra parameters \lambda_{1} and \lambda_{2} with P^{\lambda_{1}}=P^{\lambda_{2}} , consider the string \mathcal{S}(\lambda_{1},\lambda_{2}) such that the base or the direction belongs to \mathcal{D}^{\star}(G) . If a finite set \mathcal{A}\subset{\mathbb{N}}\cup\{0\} satisfies |\mathcal{A}\cap(j+q{\mathbb{Z}})|\geq|\Phi^{+}|+1 for all 0\leq j\leq q-1 , then for any \lambda_{k}\in\mathcal{S}(\lambda_{1},\lambda_{2}) , the discrete series multiplicity m(\pi_{\lambda_{k}},\Gamma) can be expressed linearly in terms of the finitely many discrete series multiplicities \{m(\pi_{\lambda_{\ell}},\Gamma):\ \ell\in\mathcal{A}\}. From the proof 4.1 of Thm. 1.1.2 , one can see that the coefficients in {\mathbb{Z}} of the above linear expression do not depend on \Gamma . Hence we have the following corollary: Corollary 1.1.3 . Assume the hypothesis on G and K as in Thm. 1.1.2 . Let \Gamma_{1} and \Gamma_{2} be two uniform lattices in G , and q=\text{l.c.m.}\ (|\Gamma_{1}\cap K|,|\Gamma_{2}\cap K|) . For two Harish-Chandra parameters \lambda_{1} and \lambda_{2} with P^{\lambda_{1}}=P^{\lambda_{2}} , consider the string \mathcal{S}(\lambda_{1},\lambda_{2}) such that the base or the direction belongs to \mathcal{D}^{\star}(G) . Let us choose a finite set \mathcal{A}\subset{\mathbb{N}}\cup\{0\} which satisfies |\mathcal{A}\cap(j+q{\mathbb{Z}})|\geq|\Phi^{+}|+1 for all 0\leq j\leq q-1 . If m(\pi_{\lambda_{k}},\Gamma_{1})=m(\pi_{\lambda_{k}},\Gamma_{2}) for all k\in\mathcal{A} , then m(\pi_{\lambda_{k}},\Gamma_{1})=m(\pi_{\lambda_{k}},\Gamma_{2}) for all \lambda_{k}\in\mathcal{S}(\lambda_{1},\lambda_{2}) . In Subsec. 4.3 , we obtain a generalised version of Thm. 1.1.2 for a multi-directed string (see Def. 4.3.1 ) of discrete series multiplicities. Even in this general case it is sufficient to assume that only one direction belongs to \mathcal{D}^{\star}(G) . Remark 1.1.4 . If G has one dimensional Cartan subgroup T i.e. \text{dim}\ \mathfrak{t}=1 (e.g. G=\mathrm{SL}(2,{\mathbb{R}}) ), then for any Harish-Chandra parameter \lambda\in\mathcal{D}^{\star}(G) the set \{k\lambda:k\in{\mathbb{Z}}\setminus\{0\}\} exhausts all but finitely many Harish-Chandra parameters in \mathcal{D}^{\star}(G) . Therefore Thm. 1.1.2 concludes that if G has one dimensional Cartan subgroup, then there exists a finite subset \mathcal{F}\subset\mathcal{D}^{\star}(G) such that any discrete series multiplicity m(\pi_{\mu},\Gamma) for some \mu\in\mathcal{D}^{\star}(G) is determined by (a linear expression of) the multiplicities \{m(\pi_{\lambda},\Gamma):\lambda\in\mathcal{F}\} . As an application of Thm. 1.1.2 , we obtain the following strong multiplicity one result for a string of discrete series representations. In fact, it holds for any infinite set of exceptions in the string with sufficiently small density. Corollary 1.1.5 . Assume the above hypothesis on G and K as in Thm. 1.1.2 . Let \Gamma_{1} and \Gamma_{2} be two uniform lattices in G , and q=\ \text{l.c.m.}\ (|\Gamma_{1}\cap K|,|\Gamma_{2}\cap K|) . For two Harish-Chandra parameters \lambda_{1} and \lambda_{2} with P^{\lambda_{1}}=P^{\lambda_{2}} , consider the string \mathcal{S}(\lambda_{1},\lambda_{2}) such that the base or the direction belongs to \mathcal{D}^{\star}(G) . If the following condition holds: \limsup\limits_{t\rightarrow\infty}\frac{|\left\{0\leq k\leq t:m(\pi_{\lambda_{k}},\Gamma_{1})\neq m(\pi_{\lambda_{k}},\Gamma_{2})\right\}|}{t}<\frac{1}{q}, then m(\pi_{\lambda_{k}},\Gamma_{1})=m(\pi_{\lambda_{k}},\Gamma_{2})~\text{for all}~\lambda_{k}\in\mathcal{S}(\lambda_{1},\lambda_{2}) . In particular, from Cor.

Read the complete text