Bourbaki

Theorem 2

Tag 00YW, permanent. Cite it as [Bourbaki, Tag 00YW].

Lie Groups and Lie Algebras, chapter VIII, § 2. Root system of a split semi-simple Lie algebra

(i) The set R=R(g,h)R = R(\mathfrak{g},\mathfrak{h}) is a reduced root system in h∗\mathfrak{h}^*.

(ii) Let α∈R\alpha \in R. The map sα,Hα:λ→λ−λ(Hα)αs_{\alpha ,H_{\alpha}}:\lambda \rightarrow \lambda -\lambda (H_{\alpha})\alpha from h∗\mathfrak{h}^* to h∗\mathfrak{h}^* is the unique reflection ss of h∗\mathfrak{h}^* such that s(α)=−αs(\alpha ) =-\alpha and s(R)=Rs(R) = R. For all t∈k∗t\in k^*, ss is the transpose of θα(t)∣h\theta_{\alpha}(t)|\mathfrak{h}.

First, R generates h∗\mathfrak{h}^*, for if h∈hh\in \mathfrak{h} is such that α(h)=0\alpha (h) = 0 for all α∈R\alpha \in R, then ad h=0h= 0 and hence h=0h= 0 since the centre of g\mathfrak{g} is zero. By definition, 0∉R0\notin R. Let α∈R\alpha \in R. Since α(Hα)=2,s=sα,Hα\alpha (H_{\alpha}) = 2,s=s_{\alpha ,H_{\alpha}} is a reflection such that s(α)=−αs(\alpha ) =-\alpha. Then s(R)=Rs(R) = R by Lemma 1 (iii), and β(Hα)∈Z\beta (H_{\alpha})\in \mathbf{Z} for all β∈R\beta \in R (Prop. 2 (i)). This shows that R is a root system in h∗\mathfrak{h}^*. For all h∈hh\in \mathfrak{h} and λ∈h∗\lambda \in \mathfrak{h}^*,

⟨s(λ),h⟩=⟨λ−λ(Hα)α,h⟩=⟨λ,h−α(h)Hα⟩=⟨λ,θα(t)h⟩ \langle s(\lambda ), h\rangle =\langle \lambda -\lambda (H_{\alpha})\alpha , h\rangle =\langle \lambda , h-\alpha (h)H_{\alpha}\rangle =\langle \lambda , \theta_{\alpha}(t)h\rangle

so ss is the transpose of θα(t)∣h\theta_{\alpha}(t)|\mathfrak{h}. Finally, we show that the root system R is reduced. Let α∈R\alpha \in R and y∈g2αy\in \mathfrak{g}^{2\alpha}. Since 3α∉R3\alpha \notin R (Chap. VI, §1, no. 3, Prop. 88),[Xα,y]=0, [X_{\alpha}, y] = 0; on the other hand, [X−α,y]∈g−α+2α=gα=kXα[X_{-\alpha}, y]\in \mathfrak{g}^{-\alpha+2\alpha}=\mathfrak{g}^{\alpha}=kX_{\alpha}, so [Xα,[X−α,y]]=0[X_{\alpha},[X_{-\alpha}, y]] = 0; thus

4y=2α(Hα)y=[Hα,y]=−[[Xα,X−α],y]=0 4y= 2\alpha (H_{\alpha})y= [H_{\alpha}, y] =-[[X_{\alpha}, X_{-\alpha}], y] = 0

so y=0y= 0 and g2α=0\mathfrak{g}^{2\alpha}= 0. In other words, 2α2\alpha is not a root. Q.E.D.

Identify h\mathfrak{h} canonically with h∗∗\mathfrak{h}^{**}. With the notations of Chap. VI, §1, no. 1, we then have, by Th. 2 (ii),

Hα=α∨H_{\alpha}=\alpha^{\vee} for all α∈R\alpha \in R. (2)

The HαH_{\alpha} thus form the root system R∨R^{\vee} in h\mathfrak{h} inverse to R.

We shall call R(g,h)R(\mathfrak{g},\mathfrak{h}) the root system of (g,h)(\mathfrak{g},\mathfrak{h}). The reflections sα,Hαs_{\alpha ,H_{\alpha}} will be denoted simply by sαs_{\alpha}. The Weyl group, group of weights, Coxeter number ... .. of R(g,h)R(\mathfrak{g},\mathfrak{h}) are called the Weyl group, group of weights, Coxeter number ... .. of (g,h)(\mathfrak{g},\mathfrak{h}). As in Chap. VI, §1, no. 1, we consider the Weyl group as operating not only on h∗\mathfrak{h}^*, but also on h\mathfrak{h} by transport of structure, so that sα=θα(t)∣hs_{\alpha}=\theta_{\alpha}(t)|\mathfrak{h}. Since the θα(t)\theta_{\alpha}(t) are elementary automorphisms of g\mathfrak{g}, we have:

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

Transcribed from the printed volume. Not translated.