On the classification of unitary representations of reductive Lie groups

From MaRDI portal



Abstract: Suppose G is a real reductive Lie group in Harish-Chandra's class. We propose here a structure for the set Pi_u(G) of equivalence classes of irreducible unitary representations of G. (The subscript u will be used throughout to indicate structures related to unitary representations.) We decompose Pi_{u}(G) into disjoint subsets with a (very explicit) discrete parameter set Lambda_u: Pi_u(G) = �igcup_{lambda_u in Lambda_u} Pi_u^{lambda_u}(G). Each subset is identified conjecturally with a collection of unitary representations of a certain subgroup G(lambda_u) of G. (We will give strong evidence and partial results for this conjecture.) In this way the problem of classifying Pi_u(G) would be reduced (by induction on the dimension of G) to the case G(lambda_u) = G.


Let \(G\) be a real reductive Lie group in Harish-Chandra's class. This paper presents a program for classifying unitary irreducible representations along the lines of the Vogan-Zuckerman classification of general irreducible representations. More precisely, the set \(\Pi_u(G)\) of (equivalence classes of) irreducible unitary representations of \(G\) is partitioned into disjoint subsets \(\Pi_u^{\lambda_u}(G)\), where the parameter \(\lambda_u\) runs over a very explicit set \(\Lambda_u\). The representations in each subset \(\Pi_u^{\lambda_u}\) are defined by a certain condition on their lowest \(K\)-types, but it is conjectured that they are all obtained by cohomological parabolic induction from unitary representations of a certain reductive subgroup \(G(\lambda_u)\) of \(G\). If the conjecture is true, then it yields an inductive framework for studying unitary representations of \(G\) which is in particular powerful enough to classify those of sufficiently regular infinitesimal character. The authors reduce the proof of the conjecture to showing that certain Hermitian representations with large infinitesimal character cannot be unitary. They then present some supporting evidence for this assertion, using Parthasarathy's Dirac operator inequality, and show how a sharpening of that inequality would yield the desired conjecture. In the last section they show that their partition of the unitary dual behaves well with respect to the Fell topology and observe that, even if not all the representations in each piece are cohomologically induced from unitary representations (as required by the conjecture), then some closed subsets of them are so induced. The techniques are largely those of \textit{D. A. Vogan}'s book [Representations of real reductive Lie groups, Prog. Math. 15 (Boston-Basel-Stuttgart 1981; Zbl 0469.22012)] as refined and extended in \textit{A. W. Knapp} and \textit{D. A. Vogan}'s more recent book [Cohomological induction and unitary representations, Princeton Math. Ser. 45 (Princeton 1995; Zbl 0863.22017)].




Cited in
(44)








This page was built for publication: On the classification of unitary representations of reductive Lie groups

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q1281445)