I am interested in an analogue of Proposition 14.31 in Fulton and Harris' book 'Representation Theory: A First Course' for positive characteristic. Let $G$ be a semisimple algebraic group of adjoint type defined over an algebraically closed field $K$ of characteristic 2, and $\mathfrak{g} = Lie(G)$. Suppose that $\mathfrak{g}$ is a Lie algebra of classical type $B, C$ or $D$. Let $T$ a maximal torus of $G$, and $W = N(T)/T$ the Weyl group of $T$. Set $\mathfrak{h} = Lie(T)$, a Cartan subalgebra of $\mathfrak{g}$, and let $\mathfrak{h}^*$ denote the dual vector space of $\mathfrak{h}$. Since I am working in bad characteristic, I don't have a $G$-equivariant isomorphism $\mathfrak{h} \to \mathfrak{h}^*$.

Now Fulton and Harris state that, in characteristic 0, the Weyl group $W$ acts irreducibly on $\mathfrak{h}^*$. Their proof requires that $\mathfrak{g}$ is a simple Lie algebra, which is not true in general over fields of characteristic 2, and that such a $G$-equivariant isomorphism $\mathfrak{h} \to \mathfrak{h}^*$ exists. It's easy enough to see that their proof works for fields of positive characteristic $p > 2$, since in this case the characteristic of $K$ is very good for $G$ and one can just apply the given proof in this situation.

Is it true that, when the characteristic of $K$ is 2, $W$ has an irreducible action on $\mathfrak{h}^*$, or does this property fail? Any reference or counterexample would be appreciated.


It fails very often! For instance, the Weyl group $S_3$ of type $A_2$ in characteristic $3$ has only two irreducible representations, both of dimension $1$. So there is no chance for it to act irreducibly on a two-dimensional vector space. (That is, it's not true that $W$ acts irreducibly on the dual Cartan in characteristic $p>2$).

You can see explicitly that the reflection representation of $S_3$ is reducible, as it contains the $S_3$-fixed element


Its dual then has the trivial module as a quotient. The easiest characteristic two example is probably the reflection representation of $W(A_3)=S_4$ or that of $W(B_2)$, both of which are reducible in characteristic $2$.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.