English

On the Centralizer of $K$ in $U(\frak {g})$

Representation Theory 2007-05-23 v3 Rings and Algebras

Abstract

Let g=k+p\frak{g} = \frak{k} +\frak{p} be a complexified Cartan decomposition of a complex semisimple Lie algebra g\frak{g} and let KK be the subgroup of the adjoint group of g\frak{g} corresponding to k\frak{k} . If HH is an irreducible Harish-Chandra module of U(g)U(\frak{g}), then HH is completely determined by the finite-dimensional action of the centralizer U(g)KU(\frak{g})^K on any one fixed primary \k\k component in HH. This original approach of Harish-Chandra to a determination of all HH has largely been abandoned because one knows very little about generators of U(g)KU(\frak{g})^K. Generators of U(g)KU(\frak{g})^K are given by generators of the symmetric algebra analogue S(g)KS(\frak{g})^K. Let Sm(g)K,mZ+S_m(\frak{g})^K, m\in {\Bbb Z}_+, be the subalgebra of S(g)KS(\frak{g})^K defined by KK-invariant polynomials of degree at most mm. Let QQ and QmQ_m be the respective quotient fields of S(g)KS(\frak{g})^K and Sm(g)KS_m(\frak{g})^K. We prove that if n=dimgn= dim \frak{g} one has Q=Q2nQ= Q_{2n}. We also determine the variety, NilKNil_K, of unstable points with respect to the action KK on g\frak{g} and show that NilKNil_K is already defined by A2nA_{2n}. As pointed out to us by Hanspeter Kraft, this fact together with a result of Harm Derksen (See [D]) implies, indeed, that A=ArA= A_r where r=(2n2)dimpr = {2n\choose 2} dim {\frak p}.

Keywords

Cite

@article{arxiv.math/0607215,
  title  = {On the Centralizer of $K$ in $U(\frak {g})$},
  author = {Bertram Kostant},
  journal= {arXiv preprint arXiv:math/0607215},
  year   = {2007}
}

Comments

19 pages, plain tex