Extending a nilpotent action by a normalizing element #
Let M be a Lie module over L, let H be a Lie subalgebra of L acting nilpotently on M, and
let y : L normalize H and act nilpotently on M. This file proves that the Lie subalgebra
LieSubalgebra.lieSpan R L (insert y ↑H) spanned by y and H again acts nilpotently on M, so
that in particular every element t • y + h acts nilpotently. This is the step by which
Hochschild's proof of Ado's theorem enlarges a nilpotently-acting subalgebra one element at a time,
and it is applied there to two different modules, so it is stated once for a general module.
Because y normalizes H, that Lie span is already spanned by y and H as a module: its
underlying submodule is R ∙ y ⊔ H.toSubmodule. Its elements are therefore exactly the t • y + h
with t : R and h ∈ H, which is the form in which the extension lemma is used.
No Noetherian or finiteness hypothesis is needed for the subalgebra statement. Engel's theorem
enters only in the variants whose hypothesis on H is the pointwise one,
∀ x ∈ H, IsNilpotent (toEnd x), and those carry [IsNoetherian R M].
Main statements #
LieSubalgebra.lieSpan_insert_toSubmodule: the Lie span ofinsert y ↑Hhas underlying submoduleR ∙ y ⊔ H.toSubmodule, whenynormalizesH.LieSubalgebra.lieModule_isNilpotent_lieSpan_insert: the nilpotent-extension lemma. IfHacts nilpotently onMand a normalizing elementyacts nilpotently onM, then the Lie span ofyandHacts nilpotently onM.LieSubalgebra.mem_lieSpan_insert_iff: the elements of that Lie span are exactly thet • y + hwithh ∈ H, with intro formLieSubalgebra.smul_add_mem_lieSpan_insert.LieSubalgebra.lt_lieSpan_insert: that Lie span lies strictly aboveHwheny ∉ H.LieSubalgebra.isNilpotent_toEnd_of_mem_lieSpan_insertandLieSubalgebra.isNilpotent_toEnd_of_mem_lieSpan_insert_of_forall: the pointwise readings, the second one taking the pointwise hypothesis onHas well, together with thet • y + hreadingLieSubalgebra.isNilpotent_toEnd_smul_add_of_mem.LieIdeal.isNilpotent_toEnd_of_mem_span_singleton_sup: the special case of an ideal, where the normalizing hypothesis is automatic and the conclusion can be read off the submoduleR ∙ y ⊔ I.toSubmoduledirectly, together with itst • y + xreadingLieIdeal.isNilpotent_toEnd_smul_add_of_mem.Ado.LieAlgebra.isNilpotent_ad_of_mem_span_singleton_sup_nilradicalandAdo.LieAlgebra.isNilpotent_ad_smul_add_of_mem_nilradical: the adjoint specialization, thatt • y + nisad-nilpotent fornin the nilradical andyad-nilpotent.
References #
- G. Hochschild, An addition to Ado's theorem, Proc. Amer. Math. Soc. 17 (1966), 531-533.
- [W. Fulton and J. Harris, Representation Theory: A First Course][fulton-harris1991], Appendix E, §E.2.
- Mathlib's
LieAlgebra.exists_engelian_lieSubalgebra_of_lt_normalizerinMathlib/Algebra/Lie/Engel.lean, the source this file follows.
When y normalizes a Lie subalgebra H, the Lie subalgebra generated by H together with y
is already spanned by them as a module: its underlying submodule is R ∙ y ⊔ H.toSubmodule.
When y normalizes a Lie subalgebra H, the elements of the Lie subalgebra spanned by H
together with y are exactly the t • y + h with t : R and h ∈ H.
When y normalizes a Lie subalgebra H, every t • y + h with h ∈ H lies in the Lie
subalgebra spanned by H together with y.
A Lie subalgebra H lies strictly below the Lie span of H together with an element y ∉ H.
No normalizing hypothesis is needed.
The nilpotent-extension lemma. If a Lie subalgebra H acts nilpotently on M, and an
element y normalizing H acts nilpotently on M, then the Lie subalgebra spanned by y and H
acts nilpotently on M.
Every element of the Lie subalgebra spanned by a normalizing element y and H acts
nilpotently on M, as soon as H does and y does.
The t • y + h reading of LieSubalgebra.isNilpotent_toEnd_of_mem_lieSpan_insert: if H acts
nilpotently on M and a normalizing element y acts nilpotently on M, then so does every
t • y + h with h ∈ H.
The pointwise form of the nilpotent-extension lemma: if every element of H acts nilpotently on
a Noetherian module M, and so does a normalizing element y, then so does every element of the
Lie subalgebra spanned by y and H.
The nilpotent-extension lemma for an ideal I of L: the normalizing hypothesis is
automatic, and the elements covered by the conclusion are exactly those of the submodule
R ∙ y ⊔ I.toSubmodule, that is, the elements t • y + x with x ∈ I.
The t • y + x reading of LieIdeal.isNilpotent_toEnd_of_mem_span_singleton_sup: for an ideal
I all of whose elements act nilpotently on M, and y acting nilpotently on M, every
t • y + x with x ∈ I acts nilpotently on M.
The adjoint specialization of the nilpotent-extension lemma: if y is ad-nilpotent, then so
is every element of R ∙ y ⊔ nilradical R L. Every element of the nilradical is ad-nilpotent, so
no nilpotency hypothesis on the ideal is needed.
The t • y + n reading of isNilpotent_ad_of_mem_span_singleton_sup_nilradical: if y is
ad-nilpotent, then so is t • y + n for every n in the nilradical.