Wednesday, 27 December 2006

rt.representation theory - Innocent question on tensor products of modular representations

Jim has already given the correct answer "no".



Here is a hopefully instructive example. Let kk be an alg. closed field of pos. char pp, and let G=SL2(k)G=SL2(k). Write V=k2V=k2 for the "natural" 2-dimensional representation of GG say with basis e1,e2e1,e2. Let W=SpVW=SpV be the pp-th symmetric power of VV. Then WW contains a 2 dimensional submodule AA spanned by the pp-th powers ep1ep1 and ep2ep2; the module AA is isom. to the "first Frobenius twist" of VV.



It is an exercise to check that there is no GG-stable complement to AA in WW; i.e. the SES
0toAtoWtoW/Ato00toAtoWtoW/Ato0
is not split.
Thus
WW is not completely reducible. Evidently there is a surjective mapping
VotimesptoWVotimesptoW, thus also the pp-th tensor power VotimespVotimesp is not completely
reducible.



But VV is a simple (hence completely reducible) GG-module; thus tensor powers of a completely reducible module are not in general completely reducible. In fact,
the (p1)(p1)-th tensor power Votimesp1Votimesp1 is completely reducible;
arguing as before, one sees that Votimes(Votimesp1)Votimes(Votimesp1) is not completely
reducible; thus in general the tensor product of two completely reducible modules
is not completely reducible.



I gave some further remarks about semisimplicity of tensor products in an answer to
this question.

No comments:

Post a Comment