K K -type multiplicities in degenerate principal series via Howe duality

Mark Colarusso, William Q. Erickson, Andrew Frohmader, Jeb F. Willenbring · 2025

Read K K -type multiplicities in degenerate principal series via Howe duality by Mark Colarusso, William Q. Erickson, Andrew Frohmader, Jeb F. Willenbring (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-2502-19505.txt.

Introduction

Mark Colarusso and co-authors count how representations split when a big symmetry group shrinks to a block-built subgroup. Their 2025 paper covers the complex classical groups and their degenerate principal series, a family of representations of real groups built by extension from a smaller piece. The counts, called K-type multiplicities, are basic data for harmonic analysis.

Using Howe duality and a seesaw trick as one frame, the authors prove a formula for these counts as a sum of generalized Littlewood-Richardson numbers, valid in a stable range. When the subgroup shrinks to single blocks, the formula turns fully combinatorial: tableaux, the little diagrams of the field, read off the answer. The work ties representation theory to concrete counting rules.

Catalog record

CREATORMark Colarusso, William Q. Erickson, Andrew Frohmader, Jeb F. Willenbring
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE116 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2502-19505

Browse the full library · About the Tower of Babel

Full text

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

K K -type multiplicities in degenerate principal series via Howe duality Mark Colarusso, William Q. Erickson, Andrew Frohmader, Jeb F. Willenbring Abstract . Let K K be one of the complex classical groups O k {\rm O}_{k} , GL k {\rm GL}_{k} , or Sp 2 ​ k {\rm Sp}_{2k} . Let M ⊆ K M\subseteq K be the block diagonal embedding O k 1 × ⋯ × O k r {\rm O}_{k_{1}}\times\cdots\times{\rm O}_{k_{r}} or GL k 1 × ⋯ × GL k r {\rm GL}_{k_{1}}\times\cdots\times{\rm GL}_{k_{r}} or Sp 2 ​ k 1 × ⋯ × Sp 2 ​ k r {\rm Sp}_{2k_{1}}\times\cdots\times{\rm Sp}_{2k_{r}} , respectively. By using Howe duality and seesaw reciprocity as a unified conceptual framework, we prove a formula for the branching multiplicities from K K to M M which is expressed as a sum of generalized Littlewood–Richardson coefficients, valid within a certain stable range. By viewing K K as the complexification of the maximal compact subgroup K ℝ K_{\mathbb{R}} of the real group G ℝ = GL ⁡ ( k , ℝ ) G_{\mathbb{R}}={\rm GL}(k,\mathbb{R}) , GL ⁡ ( k , ℂ ) {\rm GL}(k,\mathbb{C}) , or GL ⁡ ( k , ℍ ) {\rm GL}(k,\mathbb{H}) , respectively, one can interpret our branching multiplicities as K ℝ K_{\mathbb{R}} -type multiplicities in degenerate principal series representations of G ℝ G_{\mathbb{R}} . Upon specializing to the minimal M M , where k 1 = ⋯ = k r = 1 k_{1}=\cdots=k_{r}=1 , we establish a fully general tableau-theoretic interpretation of the branching multiplicities, corresponding to the K ℝ K_{\mathbb{R}} -type multiplicities in the principal series. 2020 Mathematics Subject Classification Primary 20G05; Secondary 05E10; 17B10 1. Introduction Motivation from real groups The principal series plays a fundamental role in the representation theory of real groups. Casselman [ 3 ] established that every irreducible admissible representation of a real reductive group G_{\mathbb{R}} can be realized as a submodule of some principal series representation of G_{\mathbb{R}} . (See also [ 5 ] *Cor. 1.18.) Principal series representations arise via induction from a finite-dimensional irreducible representation of a minimal parabolic subgroup P_{\mathbb{R}} . More generally, if P_{\mathbb{R}} is not necessarily minimal, then the parabolically induced representations are said to be degenerate principal series representations. These representations are infinite-dimensional. A key technique in infinite-dimensional representation theory is to decompose admissible representations of G_{\mathbb{R}} under restriction to the action of a maximal compact subgroup K_{\mathbb{R}} . This allows for the possibility of a combinatorial description of the structure of the representation by determining a formula for the multiplicities of the irreducible representations of K_{\mathbb{R}} (called K_{\mathbb{R}} -types ) in the decomposition. Although the main result in this paper is a finite-dimensional branching rule from K_{\mathbb{R}} down to a certain subgroup M_{\mathbb{R}} , this rule can be viewed as a K_{\mathbb{R}} -type multiplicity formula for degenerate principal series of G_{\mathbb{R}} . In more detail, let G_{\mathbb{R}} be a real reductive group with maximal compact subgroup K_{\mathbb{R}} . Let P_{\mathbb{R}} be a (not necessarily minimal) parabolic subgroup of G_{\mathbb{R}} . Let \bm{\mu} be an irreducible finite-dimensional representation of P_{\mathbb{R}} on which its nilradical acts trivially. By inducing \bm{\mu} from P_{\mathbb{R}} to G_{\mathbb{R}} , in the sense of [ 39 ] *§1.5, we obtain a degenerate principal series representation of G_{\mathbb{R}} , whose underlying Harish-Chandra module we denote by \mathcal{P}(\bm{\mu}) : \mathcal{P}(\bm{\mu})\coloneqq\text{the underlying Harish-Chandra module of }{\rm Ind}^{G_{\mathbb{R}}}_{P_{\mathbb{R}}}\bm{\mu}. From our perspective, the goal is to understand the K_{\mathbb{R}} -structure of \mathcal{P}(\bm{\mu}) . To this end, let M_{\mathbb{R}} denote the intersection of K_{\mathbb{R}} with the reductive Levi factor of P_{\mathbb{R}} . In the context of this paper, we may suppose that \bm{\mu} remains irreducible upon restriction from P_{\mathbb{R}} to M_{\mathbb{R}} . As representations of K_{\mathbb{R}} , (1.1) \mathcal{P}(\bm{\mu})\cong{\rm Ind}^{K_{\mathbb{R}}}_{M_{\mathbb{R}}}\bm{\mu}, where the right-hand side denotes algebraic induction from M_{\mathbb{R}} to K_{\mathbb{R}} . Now we wish to describe the multiplicity of a given K_{\mathbb{R}} -type, say \lambda , inside \mathcal{P}(\bm{\mu}) . By applying Frobenius reciprocity to ( 1.1 ), we compute this multiplicity as \displaystyle\dim{\rm Hom}_{K_{\mathbb{R}}}(\lambda,\mathcal{P}(\bm{\mu})) \displaystyle=\dim{\rm Hom}_{K_{\mathbb{R}}}(\lambda,{\rm Ind}^{K_{\mathbb{R}}}_{M_{\mathbb{R}}}\bm{\mu}) \displaystyle=\dim{\rm Hom}_{M_{\mathbb{R}}}(\bm{\mu},{\rm Res}^{K_{\mathbb{R}}}_{M_{\mathbb{R}}}\lambda)\eqqcolon b^{\lambda}_{\bm{\mu}}, where b^{\lambda}_{\bm{\mu}} denotes the multiplicity of the M_{\mathbb{R}} -type \bm{\mu} inside the K_{\mathbb{R}} -type \lambda . In this way, the problem of determining K_{\mathbb{R}} -type multiplicities in degenerate principal series is equivalent to the restriction problem (for finite -dimensional representations) from K_{\mathbb{R}} to M_{\mathbb{R}} . In the case where P_{\mathbb{R}} is a minimal parabolic, the numbers b^{\lambda}_{\bm{\mu}} are the K_{\mathbb{R}} -type multiplicities in the principal series representation induced from \bm{\mu} . The branching multiplicities b^{\lambda}_{\bm{\mu}} are the focus of this paper. Overview of results Recall that the finite-dimensional real associative division algebras are the real numbers \mathbb{R} , the complex numbers \mathbb{C} , and the quaternions \mathbb{H} . In this paper, we let G_{\mathbb{R}} be one of the real groups {\rm GL}(k,\mathbb{R}) , {\rm GL}(k,\mathbb{C}) , or {\rm GL}(k,\mathbb{H}) . We choose K_{\mathbb{R}} to be the subgroup of G_{\mathbb{R}} preserving the norm induced by conjugation on \mathbb{R} , \mathbb{C} , or \mathbb{H} , respectively. We take the subgroup M_{\mathbb{R}}\subseteq K_{\mathbb{R}} to be the block-diagonally embedded direct sum of lower-rank groups of the same Lie type as K_{\mathbb{R}} ; thus each possible M_{\mathbb{R}} is given by a choice of positive integers k_{1},\ldots,k_{r} summing to k . We let K and M denote the complexifications of K_{\mathbb{R}} and M_{\mathbb{R}} . Concrete details are given below: (1.2) \begin{array}[]{c|c|c|c|c}G_{\mathbb{R}}&K_{\mathbb{R}}&M_{\mathbb{R}}&K&M\\ \hline\cr{\rm GL}(k,\mathbb{R})&{\rm O}(k)&{\rm O}(k_{1})\times\cdots\times{\rm O}(k_{r})&{\rm O}_{k}&{\rm O}_{k_{1}}\times\cdots\times{\rm O}_{k_{r}}\\ {\rm GL}(k,\mathbb{C})&{\rm U}(k)&{\rm U}(k_{1})\times\cdots\times{\rm U}(k_{r})&{\rm GL}_{k}&{\rm GL}_{k_{1}}\times\cdots\times{\rm GL}_{k_{r}}\\ {\rm GL}(k,\mathbb{H})&{\rm Sp}(k)&{\rm Sp}(k_{1})\times\cdots\times{\rm Sp}(k_{r})&{\rm Sp}_{2k}&{\rm Sp}_{2k_{1}}\times\cdots\times{\rm Sp}_{2k_{r}}\end{array} Note that the groups K are the complex classical groups (orthogonal, general linear, and symplectic). For each of these groups K , the finite-dimensional irreducible complex representations U^{\lambda} (with rational matrix coefficients) are indexed by a certain set of partitions \lambda , or pairs \lambda=(\lambda^{+},\lambda^{-}) of partitions for K={\rm GL}_{k} . For M , then, the finite-dimensional irreducible rational representations W^{\bm{\mu}} are labeled by r -tuples \bm{\mu}=(\mu_{1},\ldots,\mu_{r}) of partitions, or (when K={\rm GL}_{k} ) by pairs \bm{\mu}=(\bm{\mu}^{+},\bm{\mu}^{-}) of such r -tuples. As described above, our main interest in this paper is the restriction problem from K_{\mathbb{R}} to M_{\mathbb{R}} , or equivalently from K to M . That is, our main results are combinatorial formulas for the branching multiplicities (1.3) b^{\lambda}_{\bm{\mu}}\coloneqq\dim{\rm Hom}_{M}(W^{\bm{\mu}},{\rm Res}^{K}_{M}U^{\lambda}). The Littlewood–Richardson coefficients are ubiquitous in algebraic combinatorics [ 11 , 16 ] ; given partitions \lambda,\mu,\nu , the Littlewood–Richardson coefficient c^{\lambda}_{\mu\nu} gives the coefficient of the Schur function s_{\lambda} in the product s_{\mu}s_{\nu} . One can generalize these coefficients for products of arbitrarily many Schur functions, so that, for example, c^{\lambda}_{\bm{\mu}\bm{\nu}} gives the coefficient of s_{\lambda} in the product s_{\mu_{1}}\cdots s_{\mu_{r}}s_{\nu_{1}}\cdots s_{\nu_{t}} . Recall that a partition is said to have even rows (resp., columns) if every row (resp., column) in its Young diagram has even length. Our first main result is the following formula (see Theorem 3.1 ) for the branching multiplicity b^{\lambda}_{\bm{\mu}} expressed as a sum of generalized Littlewood–Richardson coefficients. Theorem 1.1 (Stable branching rule from K to M ) . Let b^{\lambda}_{\bm{\mu}} denote the branching multiplicity ( 1.3 ) from K to M . Let k_{1},\ldots,k_{r} be positive integers summing to k . If \lambda and \bm{\mu} lie within a certain stable range (to be specified below) depending on \min_{i}\{k_{i}\} , then

Read the complete text