Bernstein–Zelevinsky

Bernstein–Zelevinsky
Name	Bernstein–Zelevinsky
Known for	Representation theory of p-adic groups

Bernstein–Zelevinsky describe a foundational framework in the representation theory of p-adic groups developed by Joseph Bernstein and Andrei Zelevinsky, which systematizes the classification of irreducible smooth representations of GL(n) over nonarchimedean local fields such as Q_p and finite extensions, and connects to structures studied by Harish-Chandra, Ilya Piatetski-Shapiro, Stephen Gelbart, and Frenkel-related programs. The work interfaces with harmonic analysis on p-adic groups, the theory of automorphic forms, and the local aspects of the Langlands program, influencing results by Jacquet, Shalika, Tadic, and Bushnell and Kutzko.

Introduction

The Bernstein–Zelevinsky theory arose from efforts by Joseph Bernstein and Andrei Zelevinsky to categorize smooth complex representations of GL(n) over local nonarchimedean fields such as Q_p, F_q((t)), and their finite extensions, building on earlier techniques of Harish-Chandra and Frobenius-type induction. Their approach synthesizes methods from the works of Pierre Deligne, Robert Langlands, Harris, and Taylor, and informs modern treatments by Bernstein's school, Bushnell and Kutzko, and expositions in the context of the Local Langlands correspondence and the Jacquet–Langlands correspondence. The framework uses parabolic induction, cuspidal representations classified following Bernstein's decomposition of the category of smooth representations, and segment combinatorics later exploited by Mœglin and Waldspurger.

Bernstein–Zelevinsky classification

The Bernstein–Zelevinsky classification expresses irreducible representations of GL(n) over a nonarchimedean local field as unique Langlands quotients of standard representations induced from essentially square-integrable constituents, a perspective that complements the Langlands classification for real groups by researchers such as Knapp and Zuckerman. They introduced multisegment parametrizations that relate to the work of Zelevinsky on derivatives and to the structure theory used by Tadic and Mœglin in the study of unitary duals, while interacting with techniques from Bernstein decomposition of categories, Kirillov models, and the Mackey theory of induced representations. This classification underpins later proofs of the Local Langlands conjectures for GL(n) achieved by Harris–Taylor and Henniart, and it is instrumental in the formulation of correspondences addressed by Bushnell–Henniart and computations by Paskunas.

Representations of GL(n) over local fields

Bernstein–Zelevinsky theory treats smooth representations of GL(n) over local fields like Q_p and F_q((t)) by decomposing categories into Bernstein components associated to cuspidal data analogous to constructions by Roche and Bernstein–Deligne. The methods use parabolic induction from Levi subgroups isomorphic to products of smaller GL(k) groups, echoing induction techniques in the works of Frobenius and Mackey, and rely on cuspidal building blocks studied by Jacquet and Shalika through their integrals and functionals. This perspective connects to harmonic analysis themes pursued by Gelbart and Piatetski-Shapiro and to computations in the representation theory of finite groups of Lie type as in Deligne–Lusztig theory.

Derivatives and Zelevinsky involution

Zelevinsky introduced functors of derivatives that assign to a representation of GL(n) a sequence of representations of smaller GL(k) groups, paralleling ideas in the works of Bernstein on category decompositions and resonating with the theory of Jacquet modules developed by Jacquet and Langlands. The Zelevinsky involution on the Grothendieck group interchanges standard modules and dual standard modules, influencing involutive symmetries explored by Aubert and further studied by Mœglin and Waldspurger; these operations relate to dualities appearing in the study of the Plancherel measure by Harish-Chandra and the trace formula techniques of Arthur. Derivative functors are computational tools in the analysis of reducibility points investigated by Tadic and in the construction of epsilon factors appearing in the works of Deligne and Tate.

Applications and connections

Bernstein–Zelevinsky ideas permeate the Local Langlands correspondence for GL(n) as established by Harris, Taylor, and Henniart and influence computational approaches in the Langlands program pursued by Scholze and Emerton. Their multisegment language is used in the classification of unitary duals by Tadic and in the study of distinction problems considered by Flicker and Kable, as well as in branching laws analyzed by Prasad and Gao. Connections extend to the theory of automorphic representations appearing in the work of Langlands, Gelbart, Jacquet–Langlands, and applications to the cohomology of Shimura varieties studied by Kottwitz and Harris–Taylor. Interactions with categorical representation theory appear in research by Bernstein and Riche and in geometric approaches inspired by Beilinson–Bernstein and Ginzburg.

Examples and explicit constructions

Explicit instances of the Bernstein–Zelevinsky classification include Speh representations studied by Speh and their role in the discrete spectrum considerations by Mœglin and Waldspurger, and principal series representations induced from characters of maximal tori related to classical computations by Casselman and Shalika. Other constructions demonstrate reducibility points for induced representations analyzed by Tadic and parabolic induction phenomena illustrated in the works of Bernstein–Zelevinsky themselves, with concrete combinatorial multisegment descriptions used in calculations by Zelevinsky, Moeglin, Vignéras, and Bushnell. These examples are central to explicit local-global compatibility checks in proofs by Harris–Taylor and computational verifications by Henniart.

Category:Representation theory