Invariant vectors under extension of scalars #
For a Lie module M over a Lie algebra L over a commutative ring R, the invariant vectors form
the Lie submodule LieModule.maxTrivSubmodule R L M. This file compares the invariants of the
extension of scalars A ⊗[R] M, as a module over A ⊗[R] L, with the extension of the
invariants of M:
maxTrivSubmodule A (A ⊗[R] L) (A ⊗[R] M) = (maxTrivSubmodule R L M).baseChange A.
The containment ⊇ holds for every coefficient algebra. The reverse containment holds when A is
flat over R and L is finitely generated as an R-module: the invariants are then the kernel
of the linear map m ↦ (⁅s i, m⁆)ᵢ for a finite spanning family s of L, and a flat extension
of scalars commutes with kernels (LinearMap.ker_baseChange_of_flat).
The equality is what lets an invariant vector found after extending scalars be traded for one over
the original ring: if every invariant vector of M lies in a submodule N, then every invariant
vector of A ⊗[R] M lies in N.baseChange A. Weyl's complete reducibility theorem over a field
of characteristic zero is descended from the algebraically closed case in exactly this way.
Main results #
LieModule.baseChange_maxTrivSubmodule_le: the extension of an invariant vector is invariant.LieModule.maxTrivSubmodule_baseChange: over a flat coefficient algebra, and for a finitely generated Lie algebra, the invariants of the extension are the extension of the invariants.
The extension of an invariant vector is invariant, over any coefficient algebra.
Over a flat coefficient algebra, extension of scalars commutes with taking invariants, for a Lie algebra that is finitely generated as a module.