Harish-Chandra's Principle in Representation Theory

• Harish-Chandra's Principle is a foundational concept in representation theory that describes the discrete decomposition of irreducible unitary representations with finite multiplicities, especially for real reductive Lie groups.
• It employs spherical functions, the Harish-Chandra c-function, and Paley–Wiener results to ensure analytic control and prevent continuous spectrum accumulation in subgroup restrictions.
• The principle underpins spectral analysis on locally symmetric spaces by enabling eigenfunction expansions and uniform multiplicity bounds, with applications to both Riemannian and pseudo-Riemannian geometries.

Harish-Chandra's Principle concerns the structural behavior of irreducible unitary representations of real reductive Lie groups, their restrictions to maximal compact and non-compact reductive subgroups, and the resulting decomposition properties. It is foundational in representation theory, harmonic analysis, and the spectral theory of locally symmetric spaces.

1. Admissibility Theorem for Restrictions to Maximal Compact Subgroups

Let $G$ be a real reductive Lie group with Lie algebra $\mathfrak{g}$, and $K \subset G$ a maximal compact subgroup. Denote by $\mathrm{Irr}(G)$ the set of irreducible unitary representations of $G$ and by $\mathrm{Irr}(K)$ the set of finite-dimensional irreducible representations of $K$. A continuous representation $\pi$ of $G$ on a Banach or Fréchet space $V$ is called admissible (or $\mathfrak{g}$0-admissible) if for every $\mathfrak{g}$1, the isotypic subspace

$\mathfrak{g}$2

is finite dimensional.

Passing to the algebraic framework, a ($\mathfrak{g}$3)-module $\mathfrak{g}$4 of finite length is admissible if each $\mathfrak{g}$5-type $\mathfrak{g}$6 occurs with finite multiplicity: $\mathfrak{g}$7

Theorem 1.1 (Harish-Chandra):

For $\mathfrak{g}$8, for every $\mathfrak{g}$9,

$K \subset G$0

That is, $K \subset G$1 decomposes discretely with finite multiplicities.

The proof uses the theory of spherical functions and the Harish-Chandra $K \subset G$2-function, a Paley–Wiener theorem for $K \subset G$3-finite matrix coefficients, and classification of irreducible ($K \subset G$4)-modules via infinitesimal character and highest-weight theory. The analytic continuation, control of singularities, and highest weight behavior prevent the accumulation of poles, ensuring only finitely many $K \subset G$5-types for each irreducible $K \subset G$6 (Kobayashi, 2024).

2. Generalizations: Non-compact Reductive Subgroups

Harish-Chandra's principle has been extended to restrictions to non-compact reductive subgroups.

2.1 Discrete Decomposability with Finite Multiplicities

Given a closed reductive subgroup $K \subset G$7, irreducible restrictions to $K \subset G$8 generally exhibit continuous spectrum. A representation $K \subset G$9 is said to be $\mathrm{Irr}(G)$0-admissible if

$\mathrm{Irr}(G)$1

with no continuous part—i.e., discretely decomposable with finite multiplicities.

This is equivalent, at the ($\mathrm{Irr}(G)$2)-module level, to the absence of continuous families in the restriction to ($\mathrm{Irr}(G)$3) together with finite multiplicity for each irreducible constituent.

Theorem 2.3 (Kobayashi--Criterion for $\mathrm{Irr}(G)$4-admissibility):

Let $\mathrm{Irr}(G)$5, and let $\mathrm{Irr}(G)$6 be a maximal compact subgroup of $\mathrm{Irr}(G)$7. The following are equivalent:

• (i) $\mathrm{Irr}(G)$8 is $\mathrm{Irr}(G)$9-admissible: $G$0 for all $G$1.
• (ii) A transversality condition: $G$2 where $G$3 is the asymptotic $G$4-support of $G$5, and $G$6 is the momentum cone for the cotangent bundle $G$7.

If (ii) holds, then $G$8 is $G$9-admissible, i.e., discretely decomposable with finite multiplicities (Kobayashi, 2024).

Significant examples include theta correspondences and tensor products of holomorphic discrete series.

2.2 Finite and Uniformly Bounded Multiplicity

If discrete decomposability fails, one may relax to finite (and even uniformly bounded) multiplicity. Two crucial structures arise:

• A homogeneous $\mathrm{Irr}(K)$0-space $\mathrm{Irr}(K)$1 is real-spherical if a minimal parabolic $\mathrm{Irr}(K)$2 has an open orbit on $\mathrm{Irr}(K)$3;
• Its complexification $\mathrm{Irr}(K)$4 is spherical if a Borel subgroup of $\mathrm{Irr}(K)$5 has an open orbit on $\mathrm{Irr}(K)$6.

Two key theorems:

• Finite-Multiplicity Pairs: For a pair $\mathrm{Irr}(K)$7 of real reductive groups, the following are equivalent:
  1. $\mathrm{Irr}(K)$8, $\mathrm{Irr}(K)$9, $K$0
  2. The double coset space $K$1 is real-spherical.

• Uniformly Bounded Multiplicity Pairs: Uniform boundedness of the form

$K$2

is equivalent to sphericity of $K$3 (Kobayashi, 2024).

Algebraically, these multiplicity conditions correspond to polynomiality or commutativity properties of invariant rings, such as $K$4 or algebras of invariant differential operators.

3. Bounded-Multiplicity Triples and Distinguished Representations

Further generalization concerns bounded-multiplicity triples $K$5. For a class $K$6, typically the $K$7-distinguished representations (those $K$8 such that $K$9), say that the triple has bounded multiplicity if

$\pi$0

Theorem 4.6 (Kobayashi):

If $\pi$1 is a reductive symmetric pair and $\pi$2 is a reductive subgroup, then $\pi$3 is a bounded-multiplicity triple for $\pi$4-distinguished $\pi$5 if and only if the flag variety $\pi$6 (for a relative Borel) is $\pi$7-spherical.

In particular, when $\pi$8 acts spherically on the complex flag variety attached to $\pi$9, one recovers uniform bounds for multiplicities in restriction (Kobayashi, 2024).

4. Spectral Theory of Locally Symmetric Spaces

Harish-Chandra’s principle supports new developments in the spectral analysis of locally symmetric spaces, including in pseudo-Riemannian and indefinite settings.

Let $G$0 be a reductive symmetric space and let $G$1 be a discrete group acting properly discontinuously on $G$2. The quotient $G$3 is equipped with a Laplacian $G$4 and an algebra $G$5 of invariant differential operators.

Theorem 5.6 (Kassel–Kobayashi): If $G$6 acts spherically on $G$7, then every compactly supported $G$8 admits an expansion

$G$9

into joint eigenfunctions of $V$0. The pseudo-Riemannian Laplacian is essentially self-adjoint, and the $V$1-admissibility (discrete decomposability plus uniform multiplicity) of the $V$2-action on $V$3 is crucial for establishing the full Plancherel-type decomposition (Kobayashi, 2024).

Examples in which this applies include odd-dimensional anti-de Sitter geometries, indefinite Kähler spaces, and space-forms of other signatures, with implications far beyond the classical Riemannian field.

5. Analytic and Harmonic Analysis Underpinnings

The principle is fundamentally analytic: it relies on the expansion of matrix coefficients, the analytic properties of characters, and the use of spherical functions and the $V$4-function. The Paley–Wiener theorem for $V$5-finite functions, analytic continuation, and the structure theory of ($V$6)-modules guarantee the restriction decomposes discretely with finite multiplicities.

Geometric and algebraic characterizations of multiplicity (via sphericity, momentum cones, and transversality) are crucial for admissibility criteria in the non-compact subgroup case. The machinery developed extends to a range of settings, enabling precise control of restrictions in branching problems, applications to automorphic forms, and spectral expansions in harmonic analysis.

6. Impact and Developments

Harish-Chandra’s principle, and its modern extensions, provide foundational tools across representation theory, automorphic forms, and non-Riemannian geometry. They enable:

• Systematic analysis of branching laws for representations under subgroup restriction;
• The reduction of complex spectral problems (in both Riemannian and pseudo-Riemannian cases) to questions about sphericity and multiplicity;
• New analytic frameworks for the spectral theory of locally symmetric spaces beyond the classical positive-definite context;
• Control over the Plancherel decomposition, eigenfunction expansions, and the behavior of invariant differential operators, crucial for understanding automorphic spectra and harmonic analysis on homogeneous spaces.

These results form the core of an ongoing program in the study of unitary representation restriction phenomena, intertwining geometric, analytic, and algebraic structures, with applications extending to mathematical physics and number theory (Kobayashi, 2024).