On an algebraic approach to the Zelevinsky classification for classical \(p\)-adic groups (Q958531)

From MaRDI portal
scientific article
Language Label Description Also known as
English
On an algebraic approach to the Zelevinsky classification for classical \(p\)-adic groups
scientific article

    Statements

    On an algebraic approach to the Zelevinsky classification for classical \(p\)-adic groups (English)
    0 references
    0 references
    0 references
    5 December 2008
    0 references
    The Langlands classification for a connected reductive \(p\)-adic group \(G\) attaches to each irreducible admissible representation \(\pi\) of \(G\) a triple \((P,\tau, \chi)\) consisting of a standard parabolic subgroup \(P\), a tempered representation \(\tau\) of \(P\) and a strongly positive character \(\chi\). The representation \(\pi\) then appears as the unique quotient (the Langlands quotient) in the induced representation \(\text{Ind}_P^G(\chi \otimes \tau)\). If we apply the duality operator (as defined by Aubert, Schneider-Stuhler and Zelevinsky), we obtain a dual classification. The inducing data in the dual Langlands classification contain the dual of a tempered representation. Such a representation is called a negative representation. In the case of the general linear groups, the dual Langlands classification is the well-known Zelevinsky classification [\textit{A. V. Zelevinsky}, Ann. Sci. Éc. Norm. Supér. (4) 13, 165--210 (1980; Zbl 0441.22014)]. More recently, the dual Langlands classification for connected reductive \(p\)-adic groups has been proved in [\textit{D. Ban} and \textit{C. Jantzen}, Mich. Math. J. 56, No.~3, 637--653 (2008; Zbl 1165.22014)]. In the paper under review, the authors describe the dual Langlands classification, or Zelevinsky classification, for classical \(p\)-adic groups, in an algebraic way, in terms of segments. The groups \(G_n\) considered are the split groups \(Sp(2n,F), SO(2n+1,F)\) and \(O(2n,F)\), and the quasi-split even orthogonal groups, where \(F\) is a non-Archimedean field of characteristic different from 2. Let \(\nu=|\det|\). Let \(\rho\) be an irreducible supercuspidal unitary representation of \(GL(\ell,F)\). A segment \(\Delta\) is a set of the form \( [\nu^\alpha \rho, \nu^\beta \rho] = \{\nu^\alpha \rho, \nu^{\alpha+1} \rho, \dots, \nu^\beta \rho \}, \) where \(\alpha, \beta \in \mathbb{R}\), \(\beta - \alpha \in \mathbb{Z}_{\geq 0}\). The induced representation \(\nu^\alpha \rho \times \dots \times \nu^\beta \rho\) contains the unique irreducible subrepresentation denoted by \(\langle \Delta \rangle\). Define \(e(\Delta) = (\alpha + \beta)/2\). The main result is the following. {\parindent5mm \begin{itemize}\item[1.] Suppose that \(\Delta_1, \dots, \Delta_k\) is a sequence of segments satisfying \(e(\Delta_1)\geq \cdots \geq e(\Delta_k) > 0\). Let \(\sigma_{\text{neg}}\) be a negative representation of \(G_m\). The induced representation \( \langle \Delta_1 \rangle \times \langle \Delta_2 \rangle \cdots \times \langle \Delta_k \rangle \rtimes \sigma_{\text{neg}} \) has a unique irreducible subrepresentation; denote it by \(\langle \Delta_1, \dots, \Delta_k; \sigma_{\text{neg}} \rangle\). \item[2.] If \(\sigma\) is an irreducible admissible representation of \(G_n\), there exist \(\Delta_1, \dots, \Delta_k\) and \(\sigma_{\text{neg}}\) as above such that \(\sigma \cong \langle \Delta_1, \dots, \Delta_k; \sigma_{\text{neg}} \rangle\). \end{itemize}}
    0 references
    0 references
    \(p\)-adic groups
    0 references
    Zelevinsky classification
    0 references
    dual Langlands classification
    0 references
    0 references
    0 references
    0 references
    0 references

    Identifiers