Bourbaki

Lemma 9

Tag 0117, permanent. Cite it as [Bourbaki, Tag 0117].

Lie Groups and Lie Algebras, chapter VIII, § 4. Split semi-simple Lie algebra defined by a reduced root system

Let μ,ν∈Q\mu, \nu \in Q and w∈W(R)w\in W(R) be such that wμ=νw\mu=\nu. There exists an automorphism of g\mathfrak{g} that takes gμ\mathfrak{g}^\mu to gν\mathfrak{g}^{\nu}.

For all α∈B\alpha \in B, let sαs_{\alpha} be the reflection in V defined by α\alpha. Since W(R) is generated by the sαs_{\alpha} (Chap. VI, §1, no. 5, Remark 1), it suffices to prove the lemma when w=sαw=s_{\alpha}. In view of Lemma 8, we can define

θα=eadXαeadX−αeadXα \theta_{\alpha}=e^{adX_{\alpha}}e^{adX_{-\alpha}}e^{adX_{\alpha}}

It is verified as in Chap. I, §6, no. 8, that θα\theta_{\alpha} is an automorphism of g\mathfrak{g}. We have

1^1 An endomorphism uu of a vector space V is called locally nilpotent (or almost

nilpotent) if, for every v∈Vv\in V, there exists a positive integer nn such that un(v)=0u^n(v) = 0

(cf. Commutative Algebra, Chap. IV, §1, no. 4, Def. 2). Then exp(uu), or eue^u, is

defined by the formula eu(v)=∑n≥0(1/ne^u(v) =\sum_{n\geq 0}(1/n!)un(v)u^n(v) for all v∈Vv\in V. θα(Hα)=eadXαeadX−α(Hβ−n(α,β)Xα)\theta_{\alpha}(H_{\alpha}) =e^{adX_{\alpha}}e^{adX_{-\alpha}}(H_{\beta}-n(\alpha , \beta )X_{\alpha})

=eadXα(Hβ−n(α,β)Xα+n(α,β)X−α−n(α,β)Hα−n(α,β)22X−α) =e^{adX_{\alpha}}(H_{\beta}-n(\alpha , \beta )X_{\alpha}+n(\alpha , \beta )X_{-\alpha}-n(\alpha , \beta )H_{\alpha}-\frac{n(\alpha , \beta)}{2}2X_{-\alpha})
=eadXα(Hβ−n(α,β)Hα−n(α,β)Xα) =e^{adX_{\alpha}}(H_{\beta}-n(\alpha , \beta )H_{\alpha}-n(\alpha , \beta )X_{\alpha})
=Hβ−n(α,β)Hα−n(α,β)Xα−n(α,β)Xα−n(α,β)(−2Xα) =H_{\beta}-n(\alpha , \beta )H_{\alpha}-n(\alpha , \beta )X_{\alpha}-n(\alpha , \beta )X_{\alpha}-n(\alpha , \beta )(-2X_{\alpha})
=Hβ−n(α,β)Hα =H_{\beta}-n(\alpha , \beta )H_{\alpha}

If z∈gμz\in \mathfrak{g}^\mu,

[Hβ,θα−1z]=θα−1[Hβ−n(α,β)Hα,z] [H_{\beta}, \theta_{\alpha}^{-1}z] =\theta_{\alpha}^{-1}[H_{\beta}-n(\alpha , \beta )H_{\alpha}, z]
=θα−1(⟨μ,β∨⟩z−n(α,β)⟨μ,α∨⟩z) =\theta_{\alpha}^{-1}(\langle \mu, \beta^{\vee}\rangle z-n(\alpha , \beta )\langle \mu, \alpha^{\vee}\rangle z)
=⟨μ−⟨α∨,μ⟩α,β∨⟩θα−1z=⟨sαμ,β∨⟩θα−1z =\langle \mu- \langle \alpha^{\vee}, \mu\rangle \alpha , \beta^{\vee}\rangle \theta^{-1}_{\alpha}z=\langle s_{\alpha}\mu, \beta^{\vee}\rangle \theta^{-1}_{\alpha}z

so θα−1z∈gsαμ\theta_{\alpha}^{-1}z\in \mathfrak{g}^{s_{\alpha}\mu}. This shows that θα−1gμ⊂gsαμ\theta^{-1}_{\alpha}\mathfrak{g}^\mu\subset \mathfrak{g}^{s_{\alpha}\mu}. Since θα\theta_{\alpha} is an automorphism and since this inclusion holds for all μ∈Q\mu\in Q, we see that θα−1gμ=gsαμ\theta_{\alpha}^{-1}\mathfrak{g}^\mu=\mathfrak{g}^{s_{\alpha}\mu}, which proves the lemma.

Bourbaki sets the proof directly after the statement, and the corpus does not separate them, so what is above is the statement and its proof together.

Cited by

Cites

  • Commutative Algebra, Chap. IV, §1, no. 4, Def. 2 not in this corpus: AC IV
  • Chap. I, §6, no. 8
  • Chap. VI, §1, no. 5, Remark 1 not in this corpus
  • Lemma 8
Transcribed from the printed volume. Not translated.