BG Theorem 3.10(a): module-level core
DefinitionOddOrder_BG_Ch1_Preliminary_S03g_Thm310Corefeit-thompsonfinite-groupsgroup-theoryishida-odd-order
Builds the representation-theoretic core of Theorem 3.10(a) of Bender–Glauberman, Local Analysis for the Odd Order Theorem on top of the free block permutation keystone and the weight-space machinery of §3E. The spine is prime_card_of_freeBlock_cond3: a finite subgroup acting on and permuting the blocks freely, with a primality condition, has prime order.
The main declaration: mulActionSubgroupOfLe (The restriction of a ↥R-action on ι to a subgroup ↥S (S ≤ R), via the inclusion ↥S →* ↥R).
The bundle also proves 6 supporting lemmas used by later statements.
Definition code
/-
Copyright (c) 2026 Yawara Ishida. All rights reserved.
Released under Apache 2.0 license as described in the file LICENSE.
Authors: Yawara Ishida
-/
-- From github.com/yawara/odd-order @ 20a6fccbb63c (Apache-2.0; LICENSE at that commit):
-- OddOrder/BG/Ch1_Preliminary/S03g_Thm310Core.lean
-- Generated for Prove2Me by deleting declarations; the remaining commands are verbatim except:
-- `private` dropped where another file uses the declaration, and a solution's main theorem
-- renamed `_root_.solution` (platform convention).
import Definitions.Def_OddOrder_GroupTheory_RepresentationTheory_FreeBlockPermutation
import Definitions.Def_OddOrder_BG_Ch1_Preliminary_S03e_Thm35
open Module LinearMap
namespace OddOrder.BG.Ch1.S03
variable {F : Type*} [Field F] {G : Type*} [Group G]
variable {V : Type*} [AddCommGroup V] [Module F V]
/-- The restriction of a `↥R`-action on `ι` to a subgroup `↥S` (`S ≤ R`), via the inclusion
`↥S →* ↥R`. Used to apply the free block keystone to `R` and to `⟨x⟩ ≤ R` with the same blocks. -/
@[reducible]
noncomputable def mulActionSubgroupOfLe {R S : Subgroup G} (hSR : S ≤ R)
{ι : Type*} [MulAction ↥R ι] : MulAction ↥S ι :=
MulAction.compHom ι (Subgroup.inclusion hSR)
/-- **BG Theorem 3.10(a), block form.** Let a finite subgroup `R ≤ G` act via `ρ` on a
finite-dimensional nontrivial `V = ⨁ᵢ Wᵢ` (internal direct sum), permuting the blocks **freely**
(`ρ h` maps `Wᵢ` onto `W (h • i)`, and `h • i = i ⟹ h = 1`). If for every `x ∈ R^#` the fixed
subspaces satisfy `finrank V^⟨x⟩ = finrank V^R` (the "prime action" condition (3)), then `|R|` is
prime.
This is the heart of Theorem 3.10(a) Case 2: the keystone `finrank V = |H|·finrank V^H` applied to
`H = R` and to `H = ⟨x⟩` for an `x` of prime order `p` yields `|R|·finrank V^R = p·finrank V^⟨x⟩`,
and the condition collapses this to `|R| = p`. -/
theorem prime_card_of_freeBlock_cond3
[FiniteDimensional F V] [Nontrivial V] (ρ : Representation F G V)
{R : Subgroup G} [Finite ↥R] (hRne : R ≠ ⊥)
{ι : Type*} [Finite ι] [DecidableEq ι] [MulAction ↥R ι]
{W : ι → Submodule F V} (hW : DirectSum.IsInternal W)
(hfree : ∀ (h : ↥R) (i : ι), h • i = i → h = 1)
(hperm : ∀ (h : ↥R) (i : ι), (W i).map (ρ (h : G)) = W (h • i))
(hcond3 : ∀ x : G, x ∈ R → x ≠ 1 →
finrank F (Representation.invariants (ρ.comp (Subgroup.zpowers x).subtype))
= finrank F (Representation.invariants (ρ.comp R.subtype))) :
∃ p : ℕ, p.Prime ∧ Nat.card ↥R = p := by
classical
have : Fintype ι := Fintype.ofFinite ι
-- Cauchy: pick `x ∈ R` of prime order `p`.
obtain ⟨p, hp, hpdvd⟩ := (Nat.card ↥R).exists_prime_and_dvd
(fun h1 => hRne (Subgroup.card_eq_one.mp h1))
have : Fact p.Prime := ⟨hp⟩
obtain ⟨x₀, hx₀⟩ := exists_prime_orderOf_dvd_card' (G := ↥R) p hpdvd
have hxR : (x₀ : G) ∈ R := x₀.2
have hxord : orderOf (x₀ : G) = p := by
have h := orderOf_injective R.subtype R.subtype_injective x₀
rw [hx₀] at h; exact h
have hxne : (x₀ : G) ≠ 1 := by
intro h; rw [h, orderOf_one] at hxord; exact hp.ne_one hxord.symm
have hPR : Subgroup.zpowers (x₀ : G) ≤ R := Subgroup.zpowers_le.mpr hxR
have hPcard : Nat.card ↥(Subgroup.zpowers (x₀ : G)) = p := by rw [Nat.card_zpowers, hxord]
have : Finite ↥(Subgroup.zpowers (x₀ : G)) :=
Finite.of_injective _ (Subgroup.inclusion_injective hPR)
-- Keystone applied to `R`.
have hR := ρ.finrank_eq_card_mul_finrank_invariants_of_freeBlock R hW hfree hperm
-- Keystone applied to `⟨x⟩` (same blocks, restricted action).
let : MulAction ↥(Subgroup.zpowers (x₀ : G)) ι := mulActionSubgroupOfLe hPR
have hfreeP : ∀ (h : ↥(Subgroup.zpowers (x₀ : G))) (i : ι), h • i = i → h = 1 := by
intro h i hi
have hi' : (Subgroup.inclusion hPR h) • i = i := hi
exact Subgroup.inclusion_injective hPR
((hfree (Subgroup.inclusion hPR h) i hi').trans (map_one (Subgroup.inclusion hPR)).symm)
have hpermP : ∀ (h : ↥(Subgroup.zpowers (x₀ : G))) (i : ι),
(W i).map (ρ (h : G)) = W (h • i) := by
intro h i
have h1 : ((h : G)) = ((Subgroup.inclusion hPR h : ↥R) : G) := by
rw [Subgroup.coe_inclusion]
rw [h1]
exact hperm (Subgroup.inclusion hPR h) i
have hP := ρ.finrank_eq_card_mul_finrank_invariants_of_freeBlock
(Subgroup.zpowers (x₀ : G)) hW hfreeP hpermP
rw [hPcard] at hP
-- `finrank V^⟨x⟩ = finrank V^R` (the prime-action condition (3)).
have hcond := hcond3 (x₀ : G) hxR hxne
rw [hcond] at hP
-- `finrank V ≥ 1`, hence `finrank V^R ≥ 1`.
have hdRpos : 0 < finrank F (Representation.invariants (ρ.comp R.subtype)) := by
rcases Nat.eq_zero_or_pos (finrank F (Representation.invariants (ρ.comp R.subtype)))
with h0 | h
· rw [h0, Nat.mul_zero] at hR
exact absurd hR Module.finrank_pos.ne'
· exact h
refine ⟨p, hp, ?_⟩
-- `|R| · dR = finrank V = p · dR`, cancel `dR`.
exact Nat.eq_of_mul_eq_mul_right hdRpos (hR.symm.trans hP)
section WeightFreeness
open OddOrder.BG.Ch1.S03e OddOrder.RepresentationTheory
variable (ρ : Representation F G V) {K : Subgroup G}
/-- A nonzero weight space forces `χ k ≠ 0` (`ρ k` is invertible). -/
theorem weightChar_ne_zero {χ : ↥K → F} (hχ : weightSpace ρ K χ ≠ ⊥) (k : ↥K) : χ k ≠ 0 := by
obtain ⟨v, hv, hv0⟩ := Submodule.exists_mem_ne_zero_of_ne_bot hχ
rw [mem_weightSpace] at hv
intro h0
have hzero : ρ (k : G) v = 0 := by rw [hv k, h0, zero_smul]
exact hv0 ((ρ.apply_bijective (k : G)).injective (hzero.trans (map_zero _).symm))
/-- A nonzero weight space forces `χ` to be multiplicative. -/
theorem weightChar_mul {χ : ↥K → F} (hχ : weightSpace ρ K χ ≠ ⊥) (k k' : ↥K) :
χ (k * k') = χ k * χ k' := by
obtain ⟨v, hv, hv0⟩ := Submodule.exists_mem_ne_zero_of_ne_bot hχ
rw [mem_weightSpace] at hv
have h1 : ρ ((k * k' : ↥K) : G) v = χ (k * k') • v := hv (k * k')
have h2 : ρ ((k * k' : ↥K) : G) v = (χ k * χ k') • v := by
rw [Subgroup.coe_mul, map_mul, Module.End.mul_apply, hv k', map_smul, hv k, smul_smul,
mul_comm (χ k') (χ k)]
have heqv : χ (k * k') • v = (χ k * χ k') • v := h1.symm.trans h2
have h3 : (χ (k * k') - χ k * χ k') • v = 0 := by rw [sub_smul, heqv, sub_self]
rcases smul_eq_zero.mp h3 with h | h
· exact sub_eq_zero.mp h
· exact absurd h hv0
/-- A nonzero weight space gives `χ 1 = 1`. -/
theorem weightChar_one {χ : ↥K → F} (hχ : weightSpace ρ K χ ≠ ⊥) : χ 1 = 1 := by
have h := weightChar_mul ρ hχ 1 1
rw [mul_one] at h
exact (mul_left_cancel₀ (weightChar_ne_zero ρ hχ 1) (by rw [mul_one, ← h])).symm
/-- A nonzero weight space gives `χ a⁻¹ = (χ a)⁻¹`. -/
theorem weightChar_inv {χ : ↥K → F} (hχ : weightSpace ρ K χ ≠ ⊥) (a : ↥K) :
χ a⁻¹ = (χ a)⁻¹ := by
have h : χ a * χ a⁻¹ = 1 := by rw [← weightChar_mul ρ hχ, mul_inv_cancel, weightChar_one ρ hχ]
exact (inv_eq_of_mul_eq_one_right h).symm
/-- **Freeness (L4).** If conjugation by `c⁻¹` is fixed-point-free on the abelian normal kernel `K`
(equivalently `C_K(c) = 1`), then a `conjChar c`-fixed character with nonzero weight space is
trivial. The Frobenius hypothesis on the complement supplies the fixed-point-freeness. -/
theorem weightChar_eq_one_of_conjChar_fixed [Finite G] [K.Normal]
{c : G} (hFPF : MonoidHom.FixedPointFree (conjNormalMulAut K c⁻¹))
{χ : ↥K → F} (hχ : weightSpace ρ K χ ≠ ⊥) (heq : conjChar K c χ = χ) :
χ = fun _ => 1 := by
funext j
obtain ⟨k, hk⟩ := hFPF.commutatorMap_surjective j
rw [MonoidHom.commutatorMap_apply] at hk
have hfix : χ (conjNormalMulAut K c⁻¹ k) = χ k := congrFun heq k
calc χ j = χ (k / conjNormalMulAut K c⁻¹ k) := by rw [hk]
_ = χ k * (χ (conjNormalMulAut K c⁻¹ k))⁻¹ := by
rw [div_eq_mul_inv, weightChar_mul ρ hχ, weightChar_inv ρ hχ]
_ = χ k * (χ k)⁻¹ := by rw [hfix]
_ = 1 := mul_inv_cancel₀ (weightChar_ne_zero ρ hχ k)
end WeightFreeness
end OddOrder.BG.Ch1.S03
Source
Yawara Ishida, Lean 4 formalization of the Feit–Thompson (odd order) theorem, github.com/yawara/odd-order @ 20a6fccbb63c (Apache-2.0), https://github.com/yawara/odd-order/blob/20a6fccbb63c0b138acda9f7fafd798c42f9441b/OddOrder/BG/Ch1_Preliminary/S03g_Thm310Core.lean