BG §15: Theorem 15.2 step 5, Corollary 15.5 (continuation)
DefinitionOddOrder_BG_Ch4_FamilyOfMaximal_S15_MF_Corollary155feit-thompsonfinite-groupsgroup-theoryishida-odd-order
A continuation of the Theorem 15.2 / Corollary 15.5 part of §15 of Bender–Glauberman, Local Analysis for the Odd Order Theorem: step 5 ( centralizes for narrow ) and conjunct 3 () as gated-endpoint skeletons.
Main declarations: D_centralizes_Q_of_narrow; derivedDerived_le_fittingInAmbient_of_inputs.
The bundle also proves 18 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/Ch4_FamilyOfMaximal/S15_MF/Corollary155.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_BG_Ch4_FamilyOfMaximal_S15_MF_Theorem152Assembly
namespace OddOrder.BG.Ch4.S15
open OddOrder.GroupTheory
open OddOrder.Isaacs
open OddOrder.BG.Ch3.S12
open OddOrder.BG.Ch4.S14
open scoped Pointwise
open scoped IsMulCommutative
open scoped commutatorElement
variable {G : Type*} [Group G]
/-- **`D ⋊ K` is a Frobenius group from the prime-manner action** (BG Theorem 15.2, mmd L4196-4200,
BG Theorem 3.10(b)(c) input): for the `q'`-Hall complement `D` of `Q` in `M_σ` and the `κ`-Hall
complement `K`, the group `D ⊔ K` is Frobenius with kernel `D` and complement `K`.
The Frobenius (fixed-point-free) condition is exactly the prime-manner action: a `k ∈ K#` fixing
`n ∈ D#` would centralize it, so `n ∈ C_{M_σ}(k) = K* ⊆ Q` (`hprime`, `hKstarQ`; `D ≤ M_σ`), while
`n ∈ D` and `D ∩ Q = 1` (`hDQ`), forcing `n = 1`. The remaining structure is bookkeeping:
`D ◁ D⊔K` (`K ≤ N_G(D)`, `hKnormD`), `D, K` complements (`D ∩ K = 1`, `hDK`), both nontrivial.
Discharges the `hfrob` hypothesis of `chiefFactor_card_and_commutator_of_inputs`. -/
theorem isFrobeniusGroup_DK_of_primeManner
{M K D Kstar Q : Subgroup G}
(hprime : ∀ x ∈ K, x ≠ 1 →
Subgroup.centralizer ({x} : Set G) ⊓ OddOrder.BG.Ch3.S10.Msigma M = Kstar)
(hDMσ : D ≤ OddOrder.BG.Ch3.S10.Msigma M) (hKstarQ : Kstar ≤ Q) (hDQ : Disjoint D Q)
(hKnormD : K ≤ Subgroup.normalizer (D : Set G)) (hDK : Disjoint D K)
(hDne : D ≠ ⊥) (hKne : K ≠ ⊥) :
OddOrder.Isaacs.Ch06.IsFrobeniusGroup ↥(D ⊔ K)
(D.subgroupOf (D ⊔ K)) (K.subgroupOf (D ⊔ K)) := by
have hDL : D ≤ D ⊔ K := le_sup_left
have hKL : K ≤ D ⊔ K := le_sup_right
-- `D ◁ D⊔K` from `D ≤ N(D)` and `K ≤ N(D)`.
have hDnormD : (D : Subgroup G) ≤ Subgroup.normalizer (D : Set G) := Subgroup.le_normalizer
have hDLnormal : (D.subgroupOf (D ⊔ K)).Normal :=
(Subgroup.normal_subgroupOf_iff_le_normalizer hDL).mpr (sup_le hDnormD hKnormD)
refine
{ isNormal := hDLnormal
isComplement := ?_
ne_bot_kernel := ?_
ne_bot_complement := ?_
conj_frobenius := ?_ }
· -- `D` and `K` are complements in `D ⊔ K`.
apply Subgroup.isComplement'_of_disjoint_and_mul_eq_univ
· rw [Subgroup.disjoint_def]
intro x hxD hxK
rw [Subgroup.mem_subgroupOf] at hxD hxK
exact Subtype.ext (Subgroup.disjoint_def.mp hDK hxD hxK)
· have hsup : D.subgroupOf (D ⊔ K) ⊔ K.subgroupOf (D ⊔ K) = ⊤ := by
rw [← Subgroup.subgroupOf_sup hDL hKL, Subgroup.subgroupOf_self]
have := Subgroup.normal_mul (D.subgroupOf (D ⊔ K)) (K.subgroupOf (D ⊔ K))
rw [hsup, Subgroup.coe_top] at this
exact this.symm
· -- kernel nontrivial.
intro hbot
exact hDne (by
have := Subgroup.map_mono (f := (D ⊔ K).subtype) (le_of_eq hbot)
rwa [Subgroup.map_subgroupOf_eq_of_le hDL, Subgroup.map_bot, le_bot_iff] at this)
· -- complement nontrivial.
intro hbot
exact hKne (by
have := Subgroup.map_mono (f := (D ⊔ K).subtype) (le_of_eq hbot)
rwa [Subgroup.map_subgroupOf_eq_of_le hKL, Subgroup.map_bot, le_bot_iff] at this)
· -- Frobenius condition = fixed-point-free = prime manner.
rintro a ha ha1 n hn hn1 hfix
rw [Subgroup.mem_subgroupOf] at ha hn
have haK : (a : G) ∈ K := ha
have ha1G : (a : G) ≠ 1 := fun h => ha1 (Subtype.ext h)
have hnG : (n : G) ≠ 1 := fun h => hn1 (Subtype.ext h)
have hfixG : (a : G) * (n : G) * (a : G)⁻¹ = (n : G) := Subtype.ext_iff.mp hfix
-- `n ∈ C_G(a)`: `a n a⁻¹ = n` ⟹ `a n = n a`.
have han : (a : G) * (n : G) = (n : G) * (a : G) := by
rw [mul_inv_eq_iff_eq_mul] at hfixG; exact hfixG
have hnCent : (n : G) ∈ Subgroup.centralizer ({(a : G)} : Set G) := by
rw [Subgroup.mem_centralizer_iff]
rintro g hg
rw [Set.mem_singleton_iff] at hg; subst hg
exact han
-- `n ∈ C_{M_σ}(a) = K* ⊆ Q`, while `n ∈ D` and `D ∩ Q = 1`.
have hnMσ : (n : G) ∈ OddOrder.BG.Ch3.S10.Msigma M := hDMσ hn
have hnKstar : (n : G) ∈ Kstar := by
rw [← hprime (a : G) haK ha1G]; exact Subgroup.mem_inf.mpr ⟨hnCent, hnMσ⟩
have hnQ : (n : G) ∈ Q := hKstarQ hnKstar
exact hnG (Subgroup.disjoint_def.mp hDQ hn hnQ)
/-- **BG Theorem 15.2 step 3-4, the chief-factor engine wiring** (mmd L4194-4196): given the
type-`P₁` data with `Q = O_q(M)`, the `K`-invariant complement `D` of `Q` in `M_σ`, and the
*output of `chiefFactor_Q0_normal_minimal_of_inputs`* (the normal `Q₀ = C_Q(D) ⊴ M`, `¬ K* ≤ Q₀`,
`Q₀ < Q`, and the lattice-minimality), it runs Theorem 3.10 on the Frobenius group `KD` and yields
the chief-factor index `[Q : Q₀] = q^{|K|}` with `|K|` prime, the commutator constraint
`D' ⊆ C_D(Q̄)`, and the elementary abelian section `Q̄ = Q/Q₀`.
Chains the chief-factor producers: `isElementaryAbelian_chiefFactor_of_minimalNormal`
(`hEA`/`hNT`), `card_centralizer_quotient_eq_of_kstar` (`hCfix`/`hCcard`),
`isFrobeniusGroup_DK_of_primeManner` (`hfrob`), `mem_centralizer_of_centralizes_quotient`
(`hFPF`), `actsPrimeManner_quotient_of_inputs` (`hcond3`), and the Theorem 3.10 engine
`chiefFactor_card_and_commutator_of_inputs`. The coprimality `gcd(|D ⊔ K|, |Q|) = 1` uses
`|D ⊔ K| = |K|·|D|` (the Frobenius semidirect structure). -/
theorem chiefFactor_engine_of_inputs [Finite G]
(hG : OddOrder.BG.IsMinimalSimpleOdd G) {M K Kstar Q D Q0 : Subgroup G} {q : ℕ} [Fact q.Prime]
[(Q0.subgroupOf Q).Normal]
(hM : M ∈ maximalSubgroups G) (hP1 : S14.IsTypeP1 M) (hKM : K ≤ M)
(hK : Ch03.IsHallSubgroup (S14.kappa M) (K.subgroupOf M))
(hKstar : Kstar = OddOrder.BG.Ch3.S10.Msigma M ⊓ Subgroup.centralizer (K : Set G))
(hQ : Q = opiCoreInG ({q} : Set ℕ) M)
(hQMσ : Q ≤ OddOrder.BG.Ch3.S10.Msigma M) (hMnormQ : M ≤ Subgroup.normalizer (Q : Set G))
(hKstarQ : Kstar ≤ Q) (hKne : K ≠ ⊥)
(hKMσdisj : Disjoint K (OddOrder.BG.Ch3.S10.Msigma M))
(hcop : Nat.Coprime (Nat.card ↥K) (Nat.card ↥(OddOrder.BG.Ch3.S10.Msigma M)))
(hDq' : q ∉ (Nat.card ↥D).primeFactors)
(hDMσ : D ≤ OddOrder.BG.Ch3.S10.Msigma M) (hKnormD : K ≤ Subgroup.normalizer (D : Set G))
(hQDdisj : Disjoint Q D) (hDne : D ≠ ⊥)
(hQ0def : Q0 = Q ⊓ Subgroup.centralizer (D : Set G))
(hMNQ0 : M ≤ Subgroup.normalizer (Q0 : Set G)) (hKstarNotQ0 : ¬ Kstar ≤ Q0)
(hQ0ltQ : Q0 < Q)
(hmin : ∀ H : Subgroup G, Q0 < H → H ≤ Q → (H.subgroupOf M).Normal → Q ≤ H) :
(Nat.card ↥K).Prime ∧
(Q0.subgroupOf Q).index = q ^ Nat.card ↥K ∧
(∀ g ∈ ⁅D, D⁆, ∀ x ∈ Q, ⁅g, x⁆ ∈ Q0) ∧
OddOrder.GroupTheory.IsElementaryAbelian q (↥Q ⧸ Q0.subgroupOf Q) := by
classical
have hMσM : OddOrder.BG.Ch3.S10.Msigma M ≤ M := OddOrder.BG.Ch3.S10.Msigma_le M
have hQM : Q ≤ M := hQMσ.trans hMσM
have hDM : D ≤ M := hDMσ.trans hMσM
have : Group.IsSolvable ↥M := hG.isSolvable_of_mem_maximalSubgroups hM
have hQpg : IsPGroup q ↥Q := by rw [hQ]; exact isPGroup_opiCoreInG_singleton M
have hQ0Q : Q0 ≤ Q := hQ0ltQ.le
have hKQ : K ≤ Subgroup.normalizer (Q : Set G) := hKM.trans hMnormQ
have hDNQ : D ≤ Subgroup.normalizer (Q : Set G) := hDM.trans hMnormQ
have hQQ0 : Q ≤ Subgroup.normalizer (Q0 : Set G) := hQM.trans hMNQ0
have hKQ0 : K ≤ Subgroup.normalizer (Q0 : Set G) := hKM.trans hMNQ0
have hDNQ0 : D ≤ Subgroup.normalizer (Q0 : Set G) := hDM.trans hMNQ0
have hKstarN : Kstar ≤ Subgroup.normalizer (Q0 : Set G) := hKstarQ.trans hQQ0
have hSolvQ : Group.IsSolvable ↥Q :=
Group.isSolvable_of_isSolvable_injective (Subgroup.inclusion_injective hQM)
have hsolvDK : Group.IsSolvable ↥(D ⊔ K) :=
Group.isSolvable_of_isSolvable_injective (Subgroup.inclusion_injective (sup_le hDM hKM))
have hKstarP : (Nat.card ↥Kstar).Prime := kstar_card_prime_of_inputs hG hM hP1 hKM hK hKstar
have hKstarEqQ : Nat.card ↥Kstar = q := by
obtain ⟨n, hn⟩ := hQpg.exists_card_eq
have hdvd : Nat.card ↥Kstar ∣ Nat.card ↥Q := Subgroup.card_dvd_of_le hKstarQ
rw [hn] at hdvd
exact (Nat.prime_dvd_prime_iff_eq hKstarP Fact.out).mp (hKstarP.dvd_of_dvd_pow hdvd)
have hqD : ¬ q ∣ Nat.card ↥D := fun hdvd =>
hDq' (Nat.mem_primeFactors.mpr ⟨Fact.out, hdvd, Nat.card_pos.ne'⟩)
have hcopDQ : Nat.Coprime (Nat.card ↥D) (Nat.card ↥Q) := by
obtain ⟨n, hn⟩ := hQpg.exists_card_eq
rw [hn]
rcases Nat.eq_zero_or_pos n with h0 | h0
· subst h0; simp
· rw [Nat.coprime_pow_right_iff h0]
exact ((Fact.out : q.Prime).coprime_iff_not_dvd.mpr hqD).symm
have hcopKQ : Nat.Coprime (Nat.card ↥K) (Nat.card ↥Q) :=
hcop.coprime_dvd_right (Subgroup.card_dvd_of_le hQMσ)
have hDKdisj : Disjoint D K := hKMσdisj.symm.mono_left hDMσ
have hcopDKQ : Nat.Coprime (Nat.card ↥(D ⊔ K)) (Nat.card ↥Q) := by
have hcardsup : Nat.card ↥(D ⊔ K) = Nat.card ↥K * Nat.card ↥D := by
rw [sup_comm]
exact card_sup_eq_mul_of_le_normalizer_of_disjoint hKnormD (disjoint_iff.mp hDKdisj.symm)
rw [hcardsup]; exact Nat.coprime_mul_iff_left.mpr ⟨hcopKQ, hcopDQ⟩
have hprime := actsPrimeManner_of_typeP hG hM hP1.1 hKM hK hKstar
obtain ⟨hEA, hNT⟩ :=
isElementaryAbelian_chiefFactor_of_minimalNormal hQ0ltQ hQM hQpg hMnormQ hMNQ0 hmin
obtain ⟨C, hQ0C, hCQ, hCfix, hCcard⟩ :=
card_centralizer_quotient_eq_of_kstar hKstar hQMσ hKstarQ hKstarEqQ hQ0Q hKstarNotQ0 hKQ hQQ0
hKQ0 hKstarN hcopKQ (Or.inr hSolvQ)
have hfrob := isFrobeniusGroup_DK_of_primeManner (M := M) hprime hDMσ hKstarQ hQDdisj.symm hKnormD
hDKdisj hDne hKne
have hFPF : ∀ x ∈ Q, (∀ d ∈ D, ⁅d, x⁆ ∈ Q0) → x ∈ Q0 :=
fun x hxQ hfix => mem_centralizer_of_centralizes_quotient hQ0def hDNQ hQQ0 hDNQ0 hcopDQ
(Or.inr hSolvQ) hxQ hfix
have hcond3 := actsPrimeManner_quotient_of_inputs hKstar hprime hQMσ hKstarQ hQ0Q hKQ hQQ0 hKQ0
hcopKQ (Or.inr hSolvQ)
obtain ⟨hKprime, hindex, hDcomm⟩ :=
chiefFactor_card_and_commutator_of_inputs hQ0Q hQ0C hCQ hDne hEA hNT hDNQ hKQ hDNQ0 hKQ0 hQQ0
hsolvDK hfrob hcopDKQ hFPF hcond3 hCfix hCcard
exact ⟨hKprime, hindex, hDcomm, hEA⟩
/-- **Theorem 15.2(g) reverse inclusion, reduced to `C_M(Q) ⊆ M_σ`** (mmd L4196-4198): if the
centralizer `C_M(Q)` of the normal `q`-subgroup `Q` lies in `M_σ` (the genuinely BG-specific input,
from `σ`-uniqueness — it does *not* follow from local structure, cf. the ChatGPT-verified counter-
example `M = (C₇⋊C₃)×(C₃₁⋊C₅)`), then `C_M(Q) ⊆ F(M)`. `C_M(Q)` is nilpotent
(`isNilpotent_of_centralizes_normal_of_quotient_isNilpotent`: it centralizes `Q ◁ M_σ` and
`M_σ/Q` is nilpotent) and normal in `M` (`M ≤ N_G(Q) ≤ N_G(C_G(Q))`), so a nilpotent normal
subgroup of `M` lands in `F(M)` (`nilpotent_normal_le_fitting`). Discharges the `hcent` input of
`fittingInAmbient_eq_sup_centralizer_inf_of_inputs`. -/
theorem centralizer_inf_le_fittingInAmbient_of_le_Msigma [Finite G]
(_hG : OddOrder.BG.IsMinimalSimpleOdd G) {M Q : Subgroup G} (_hM : M ∈ maximalSubgroups G)
(hMnormQ : M ≤ Subgroup.normalizer (Q : Set G))
[(Q.subgroupOf (OddOrder.BG.Ch3.S10.Msigma M)).Normal]
[Group.IsNilpotent (↥(OddOrder.BG.Ch3.S10.Msigma M) ⧸
Q.subgroupOf (OddOrder.BG.Ch3.S10.Msigma M))]
(hCle : Subgroup.centralizer (Q : Set G) ⊓ M ≤ OddOrder.BG.Ch3.S10.Msigma M) :
Subgroup.centralizer (Q : Set G) ⊓ M ≤ fittingInAmbient M := by
have : Group.IsSolvable ↥M := _hG.isSolvable_of_mem_maximalSubgroups _hM
-- `C_M(Q)` is nilpotent (central extension over the nilpotent `M_σ/Q`).
have : Group.IsNilpotent ↥(Subgroup.centralizer (Q : Set G) ⊓ M) :=
isNilpotent_of_centralizes_normal_of_quotient_isNilpotent hCle inf_le_left
-- `C_M(Q) ◁ M` (`M` normalizes `Q`, hence `C_G(Q)`, hence `C_G(Q) ⊓ M`).
have hMnormC : M ≤ Subgroup.normalizer
((Subgroup.centralizer (Q : Set G) ⊓ M : Subgroup G) : Set G) :=
le_normalizer_inf (hMnormQ.trans (normalizer_le_normalizer_centralizer Q))
Subgroup.le_normalizer
have : ((Subgroup.centralizer (Q : Set G) ⊓ M).subgroupOf M).Normal :=
(Subgroup.normal_subgroupOf_iff_le_normalizer inf_le_right).mpr hMnormC
have : Group.IsNilpotent ↥((Subgroup.centralizer (Q : Set G) ⊓ M).subgroupOf M) :=
Group.nilpotent_of_mulEquiv (Subgroup.subgroupOfEquivOfLe (inf_le_right :
Subgroup.centralizer (Q : Set G) ⊓ M ≤ M)).symm
-- Nilpotent normal subgroup of `M` lands in `F(M)`.
calc Subgroup.centralizer (Q : Set G) ⊓ M
= ((Subgroup.centralizer (Q : Set G) ⊓ M).subgroupOf M).map M.subtype :=
(Subgroup.map_subgroupOf_eq_of_le inf_le_right).symm
_ ≤ (OddOrder.Isaacs.Ch01.fitting ↥M).map M.subtype :=
Subgroup.map_mono OddOrder.Isaacs.Ch01.nilpotent_normal_le_fitting
_ = fittingInAmbient M := rfl
/-- **BG Theorem 15.2(g), assembled from the single `σ`-gap `C_M(Q) ⊆ M_σ`** (mmd L4196-4198):
chains the equality skeleton (`fittingInAmbient_eq_sup_centralizer_inf_of_inputs`) with the reduced
reverse inclusion (`centralizer_inf_le_fittingInAmbient_of_le_Msigma`). For `Q = O_q(M)` with
`M_σ/Q` nilpotent, the conjunct `F(M) = Q ⊔ (C_G(Q) ⊓ M)` follows from `C_M(Q) ⊆ M_σ` alone. This
is the wrapper-facing form: the only outstanding input is the BG-specific `C_M(Q) ⊆ M_σ` (a
`σ`-uniqueness fact to be supplied from the global analysis / forward input). -/
theorem fittingInAmbient_eq_sup_centralizer_inf_of_le_Msigma [Finite G]
(hG : OddOrder.BG.IsMinimalSimpleOdd G) {M Q : Subgroup G} {q : ℕ} [Fact q.Prime]
(hM : M ∈ maximalSubgroups G) (hMnormQ : M ≤ Subgroup.normalizer (Q : Set G))
(hQ : Q = opiCoreInG ({q} : Set ℕ) M)
[(Q.subgroupOf (OddOrder.BG.Ch3.S10.Msigma M)).Normal]
[Group.IsNilpotent (↥(OddOrder.BG.Ch3.S10.Msigma M) ⧸
Q.subgroupOf (OddOrder.BG.Ch3.S10.Msigma M))]
(hCle : Subgroup.centralizer (Q : Set G) ⊓ M ≤ OddOrder.BG.Ch3.S10.Msigma M) :
fittingInAmbient M = Q ⊔ (Subgroup.centralizer (Q : Set G) ⊓ M) :=
fittingInAmbient_eq_sup_centralizer_inf_of_inputs hQ
(centralizer_inf_le_fittingInAmbient_of_le_Msigma hG hM hMnormQ hCle)
/-- **`D = ⁅D, K⁆` from the Frobenius (fixed-point-free) action** (BG Theorem 15.2, mmd L4202,
BG Lemma 6.3(a) flavour but via the coprime decomposition): if `D ⊔ K` is a Frobenius group with
kernel `D` and complement `K` (so `K` acts fixed-point-freely on `D`), `K ≤ N_G(D)`, and the orders
of `K` and `D` are coprime with `D`/`K` one-sided solvable, then `⁅D, K⁆ = D`.
`Proposition 1.6(d)` (`subgroup_coprime_decomposition`) gives `D = C_D(K) ⊔ ⁅D, K⁆`; the Frobenius
condition forces `C_D(K) = C_G(K) ⊓ D = ⊥` (a nontrivial `d ∈ D` centralizing a nontrivial `k ∈ K`
would be fixed by conjugation, contradicting `conj_frobenius`), so the decomposition collapses to
`D = ⁅D, K⁆`. -/
theorem commutator_eq_self_of_frobenius_DK [Finite G] {D K : Subgroup G}
(hKne : K ≠ ⊥)
(hFrobFPF : ∀ a ∈ K, a ≠ 1 → ∀ n ∈ D, n ≠ 1 → a * n * a⁻¹ ≠ n)
(hKnormD : K ≤ Subgroup.normalizer (D : Set G))
(hcop : Nat.Coprime (Nat.card ↥K) (Nat.card ↥D))
(hSolv : Group.IsSolvable ↥K ∨ Group.IsSolvable ↥D) :
⁅D, K⁆ = D := by
-- `C_G(K) ⊓ D = ⊥`: a nontrivial common element contradicts the Frobenius condition.
have hCDK : (Subgroup.centralizer (K : Set G) ⊓ D : Subgroup G) = ⊥ := by
rw [eq_bot_iff]
intro d hd
rw [Subgroup.mem_inf] at hd
obtain ⟨hdcent, hdD⟩ := hd
by_contra hdne
rw [Subgroup.mem_bot] at hdne
-- Pick a nontrivial `k ∈ K`.
have : Nontrivial ↥K := (Subgroup.nontrivial_iff_ne_bot K).mpr hKne
obtain ⟨k, hkK, hkne⟩ := (Subgroup.nontrivial_iff_exists_ne_one K).mp inferInstance
-- `k` and `d` commute (from `d ∈ C_G(K)`), so `k * d * k⁻¹ = d`, contradicting Frobenius.
have hcomm : k * d = d * k := (Subgroup.mem_centralizer_iff.mp hdcent) k hkK
have hfix : k * d * k⁻¹ = d := by rw [hcomm]; group
exact hFrobFPF k hkK hkne d hdD hdne hfix
-- Proposition 1.6(d): `D = (C_G(K) ⊓ D) ⊔ ⁅D, K⁆`.
have hdecomp := OddOrder.BG.Ch3.S13.subgroup_coprime_decomposition hKnormD hcop hSolv
rw [hCDK, bot_sup_eq] at hdecomp
exact hdecomp.symm
set_option synthInstance.maxHeartbeats 80000 in
-- raised heartbeat budget for the heavy elaboration below
/-- **Theorem 15.2 step 5 — `D` centralizes `Q` for narrow `Q`** (mmd L4202, BG Theorem 5.5(a)):
if `Q` is a narrow `q`-group (`q` odd), `D ⊔ K` is a Frobenius group with kernel `D` and complement
`K` acting in a prime manner, `D ⊔ K ≤ N_G(Q)`, `D` is a `q'`-group (`q ∤ |D|`), and the orders of
`K`, `D` are coprime, then `D ⊆ C_G(Q)`.
`N := N_G(Q)` is proper (`Q ≠ 1, G` in the simple `G`), hence solvable; the conjugation action
`ψ : N → MulAut Q` has kernel `C_G(Q) ⊓ N`. Theorem 5.5(a) (`solvableAut_of_narrow`, applied to the
faithful action of `N/ker`) gives that `(N/ker)'` is a `q`-group. By the Frobenius condition
`⁅D, K⁆ = D` (`commutator_eq_self_of_frobenius_DK`), so `D ⊆ ⁅N, N⁆ = N'`; the image of `D` in
`N/ker` therefore lies in `(N/ker)'` (a `q`-group) yet is a `q'`-group (`q ∤ |D|`), hence trivial.
Trivial image means `D ⊆ ker ψ = C_G(Q) ⊓ N`, i.e. `D ⊆ C_G(Q)`. -/
theorem D_centralizes_Q_of_narrow [Finite G] (hG : OddOrder.BG.IsMinimalSimpleOdd G)
{Q D K : Subgroup G} {q : ℕ} [Fact q.Prime]
(hq_odd : Odd q) (hQpg : IsPGroup q ↥Q) (hQnarrow : OddOrder.GroupTheory.IsNarrow q ↥Q)
(hQne : Q ≠ ⊥)
(hKne : K ≠ ⊥)
(hFrobFPF : ∀ a ∈ K, a ≠ 1 → ∀ n ∈ D, n ≠ 1 → a * n * a⁻¹ ≠ n)
(hKnormD : K ≤ Subgroup.normalizer (D : Set G))
(hcop : Nat.Coprime (Nat.card ↥K) (Nat.card ↥D))
(hKsolv : Group.IsSolvable ↥K)
(hDKN : D ⊔ K ≤ Subgroup.normalizer (Q : Set G))
(hqD : ¬ q ∣ Nat.card ↥D) :
D ≤ Subgroup.centralizer (Q : Set G) := by
classical
have hp_prime : q.Prime := Fact.out
-- `N := N_G(Q)` is a proper (hence solvable) subgroup of the simple `G`.
set N : Subgroup G := Subgroup.normalizer (Q : Set G) with hN_def
have hNlt : N < ⊤ := by
rw [lt_top_iff_ne_top]
intro htop
have hQnorm : Q.Normal := by rw [← Subgroup.normalizer_eq_top_iff]; exact htop
rcases hG.simple.eq_bot_or_eq_top_of_normal Q hQnorm with h | h
· exact hQne h
· have hGpg : IsPGroup q G := (h ▸ hQpg : IsPGroup q ↥(⊤ : Subgroup G)).of_surjective
(Subgroup.topEquiv : (⊤ : Subgroup G) ≃* G).toMonoidHom Subgroup.topEquiv.surjective
have : Group.IsNilpotent G := hGpg.isNilpotent
exact hG.notSolvable inferInstance
have hNsolv : Group.IsSolvable ↥N := hG.isSolvable_of_lt_top N hNlt
-- The conjugation action `ψ : N → Aut Q` with kernel `C_G(Q) ∩ N`.
set ψ : ↥N →* MulAut ↥Q := Q.normalizerMonoidHom with hψ_def
have hψker : ψ.ker = (Subgroup.centralizer (Q : Set G)).subgroupOf N :=
Q.normalizerMonoidHom_ker
-- `A := N / ker ψ` acts faithfully, is solvable and odd.
have hA_odd : Odd (Nat.card (↥N ⧸ ψ.ker)) := by
refine hG.odd.of_dvd_nat (dvd_trans ?_ (Subgroup.card_subgroup_dvd_card N))
simpa [Subgroup.index] using Subgroup.index_dvd_card ψ.ker
-- Theorem 5.5(a): `(N / ker)'` is a `q`-group.
obtain ⟨hcomm, -, -, -⟩ := Ch1.S05.solvableAut_of_narrow hq_odd hQpg hQnarrow
(QuotientGroup.kerLift ψ) (QuotientGroup.kerLift_injective ψ) hA_odd
have hA' : IsPGroup q (_root_.commutator (↥N ⧸ ψ.ker)) := by
have hle : _root_.commutator (↥N ⧸ ψ.ker) ≤ Ch01.opCore q (↥N ⧸ ψ.ker) := by
rw [_root_.commutator, Subgroup.commutator_le]
intro x _ y _
have h1 : QuotientGroup.mk' (Ch01.opCore q (↥N ⧸ ψ.ker)) ⁅x, y⁆ = 1 := by
rw [map_commutatorElement, commutatorElement_eq_one_iff_mul_comm]
exact hcomm _ _
exact (QuotientGroup.eq_one_iff _).mp h1
exact (Ch01.opCore_isPGroup q _).to_le hle
-- `D ⊔ K ≤ N`, so `D ≤ N`; and `D = ⁅D, K⁆ ≤ N'`.
have hDN : (D : Subgroup G) ≤ N := le_sup_left.trans hDKN
have hDcommDK : ⁅D, K⁆ = D :=
commutator_eq_self_of_frobenius_DK hKne hFrobFPF hKnormD hcop (Or.inl hKsolv)
have hDcomm : (D : Subgroup G).subgroupOf N ≤ _root_.commutator ↥N := by
have hDder : (D : Subgroup G) ≤ derivedInG N := by
rw [← hDcommDK]
calc ⁅D, K⁆ ≤ ⁅N, N⁆ := Subgroup.commutator_mono hDN (le_sup_right.trans hDKN)
_ = derivedInG N := (Subgroup.map_subtype_commutator N).symm
have key : ((_root_.commutator ↥N).map N.subtype).comap N.subtype = _root_.commutator ↥N :=
Subgroup.comap_map_eq_self_of_injective N.subtype_injective (_root_.commutator ↥N)
calc (D : Subgroup G).subgroupOf N
≤ (derivedInG N).comap N.subtype := Subgroup.comap_mono hDder
_ = _root_.commutator ↥N := key
-- The image of `D` in `A` is `≤ (N/ker)'` (a `q`-group) and is a `q'`-group: hence trivial.
set DA : Subgroup (↥N ⧸ ψ.ker) :=
((D : Subgroup G).subgroupOf N).map (QuotientGroup.mk' ψ.ker) with hDA_def
have hDA_q : IsPGroup q ↥DA := by
refine hA'.to_le ?_
calc DA ≤ (_root_.commutator ↥N).map (QuotientGroup.mk' ψ.ker) := Subgroup.map_mono hDcomm
_ ≤ _root_.commutator (↥N ⧸ ψ.ker) := by
rw [_root_.commutator, _root_.commutator, Subgroup.map_commutator]
exact Subgroup.commutator_mono le_top le_top
-- `q ∤ |DA|`: `|DA|` divides `|D|` (surjective images), and `q ∤ |D|`.
have hDA_card_dvd : Nat.card ↥DA ∣ Nat.card ↥D := by
have h1 : Nat.card ↥DA ∣ Nat.card ↥((D : Subgroup G).subgroupOf N) :=
Subgroup.card_map_dvd (H := (D : Subgroup G).subgroupOf N) (QuotientGroup.mk' ψ.ker)
have h2 : Nat.card ↥((D : Subgroup G).subgroupOf N) = Nat.card ↥D :=
Nat.card_congr (Subgroup.subgroupOfEquivOfLe hDN).toEquiv
rw [h2] at h1; exact h1
have hqDA : ¬ q ∣ Nat.card ↥DA := fun h => hqD (h.trans hDA_card_dvd)
-- A `q`-group with `q ∤ |DA|` is trivial.
have hDA_bot : DA = ⊥ := by
obtain ⟨k, hk⟩ := hDA_q.exists_card_eq
rcases Nat.eq_zero_or_pos k with hk0 | hkpos
· rw [hk0, pow_zero] at hk; exact Subgroup.card_eq_one.mp hk
· exact absurd (hk ▸ dvd_pow_self q hkpos.ne') hqDA
-- Trivial image means `D ≤ ker ψ = C_G(Q) ∩ N`, hence `D ≤ C_G(Q)`.
have hDker : (D : Subgroup G).subgroupOf N ≤ ψ.ker := by
rw [hDA_def, Subgroup.map_eq_bot_iff, QuotientGroup.ker_mk'] at hDA_bot
exact hDA_bot
rw [hψker] at hDker
intro x hx
have hxN : x ∈ N := hDN hx
have : (⟨x, hxN⟩ : ↥N) ∈ (Subgroup.centralizer (Q : Set G)).subgroupOf N :=
hDker (by rw [Subgroup.mem_subgroupOf]; exact hx)
rw [Subgroup.mem_subgroupOf] at this
exact this
/-- **Theorem 15.2 step 5 — `D` centralizes `Q` from `q ∉ β(M)`** (mmd L4202): the `hDcent` input of
`mem_beta_of_inputs`, with the narrowness of `Q` discharged from `q ∉ β(M)`.
When `Q = O_q(M)` is (the image in `G` of) a Sylow `q`-subgroup `P` of `M` — which holds in the
type-P1 setting, since `M_σ/Q` is a `q'`-group, so the normal Sylow `q` of `M_σ` is a Sylow `q` of
`M` — narrowness of `↥Q ≅ ↥P` follows from `q ∉ β(M)` (`isNarrow_sylow_of_not_mem_beta`, BG Lemma
10.8 setup). Chaining with `D_centralizes_Q_of_narrow` (the Theorem 5.5(a) gate) gives
`q ∉ β(M) → D ⊆ C_G(Q)`, exactly the `hDcent` hypothesis of `mem_beta_of_inputs`. -/
theorem D_centralizes_Q_of_not_mem_beta [Finite G] (hG : OddOrder.BG.IsMinimalSimpleOdd G)
{M Q D K : Subgroup G} {q : ℕ} [Fact q.Prime] (hM : M ∈ maximalSubgroups G)
(hq_odd : Odd q) (hQpg : IsPGroup q ↥Q) (hQne : Q ≠ ⊥)
(hqπ : q ∈ (Nat.card ↥M).primeFactors)
(P : Sylow q ↥M) (hQP : Q = (P : Subgroup ↥M).map M.subtype)
(hKne : K ≠ ⊥)
(hFrobFPF : ∀ a ∈ K, a ≠ 1 → ∀ n ∈ D, n ≠ 1 → a * n * a⁻¹ ≠ n)
(hKnormD : K ≤ Subgroup.normalizer (D : Set G))
(hcop : Nat.Coprime (Nat.card ↥K) (Nat.card ↥D))
(hKsolv : Group.IsSolvable ↥K)
(hDKN : D ⊔ K ≤ Subgroup.normalizer (Q : Set G))
(hqD : ¬ q ∣ Nat.card ↥D) :
q ∉ OddOrder.BG.Ch3.S10.beta M → D ≤ Subgroup.centralizer (Q : Set G) := by
intro hqβ
-- Narrowness of the Sylow `P`, transferred along `↥Q ≅ ↥P`.
have hPnarrow : OddOrder.GroupTheory.IsNarrow q ↥(P : Subgroup ↥M) :=
OddOrder.BG.Ch3.S10.isNarrow_sylow_of_not_mem_beta hG hM hqπ hqβ P
have eQP : ↥Q ≃* ↥(P : Subgroup ↥M) :=
(MulEquiv.subgroupCongr hQP).trans
(Subgroup.equivMapOfInjective _ M.subtype M.subtype_injective).symm
have hQnarrow : OddOrder.GroupTheory.IsNarrow q ↥Q :=
OddOrder.GroupTheory.IsNarrow.of_mulEquiv eQP.symm hPnarrow
exact D_centralizes_Q_of_narrow hG hq_odd hQpg hQnarrow hQne hKne hFrobFPF hKnormD hcop hKsolv
hDKN hqD
/-- **Theorem 15.2 step 5 — `q ∈ β(M)`, gated-endpoint skeleton** (mmd L4202): "if `q ∉ β(M)`,
then Theorem 5.5(a) shows `(DK)' = D` centralizes `Q`, a contradiction". The contradiction is
clean: `D` centralizing `Q` means `Q ≤ C_G(D)`, i.e. `C_Q(D) = Q`, against the established
`C_Q(D) = Q₀ ⊊ Q` (`M_σ` non-nilpotent). Reduces `q ∈ β(M)` to the single Theorem-5.5 input
`hDcent` (`q ∉ β(M) → D ⊆ C_G(Q)`) and the proper-centralizer fact `hQ0` (`¬ Q ⊆ C_G(D)`). -/
theorem mem_beta_of_inputs {M Q D : Subgroup G} {q : ℕ}
(hQ0 : ¬ Q ≤ Subgroup.centralizer (D : Set G))
(hDcent : q ∉ OddOrder.BG.Ch3.S10.beta M → D ≤ Subgroup.centralizer (Q : Set G)) :
q ∈ OddOrder.BG.Ch3.S10.beta M := by
by_contra hq
have hDQ := hDcent hq
rw [← Subgroup.commutator_eq_bot_iff_le_centralizer] at hDQ
exact hQ0 (Subgroup.commutator_eq_bot_iff_le_centralizer.mp
(by rwa [Subgroup.commutator_comm] at hDQ))
/-- **Theorem 15.2(f) — `M_F` non-cyclic, gated-endpoint skeleton** (mmd L4202): `M_F` is
non-cyclic because it contains the non-cyclic section `Q̄ = Q/Q₀` (the elementary abelian chief
factor of order `q^p`, `p ≥ 2`). If `M_F` were cyclic, then so would be its subgroup `Q`
(`Subgroup.isCyclic_of_le`) and the quotient `Q/Q₀` (`isCyclic_of_surjective`), against `hQbar`.
Reduces `¬ IsCyclic M_F` to `Q ⊆ M_F` (Theorem 15.2(c)) and `¬ IsCyclic (Q/Q₀)` (from `|Q̄| = q^p`,
`p ≥ 2`). -/
theorem not_isCyclic_MF_of_inputs {M Q Q0 : Subgroup G} [(Q0.subgroupOf Q).Normal]
(hQMF : Q ≤ MF M) (hQbar : ¬ IsCyclic (↥Q ⧸ Q0.subgroupOf Q)) :
¬ IsCyclic ↥(MF M) := by
intro hcyc
have := hcyc
have : IsCyclic ↥Q := Subgroup.isCyclic_of_le hQMF
exact hQbar (isCyclic_of_surjective (QuotientGroup.mk' (Q0.subgroupOf Q))
(QuotientGroup.mk'_surjective _))
/-- A finite elementary-abelian `q`-group of order exceeding `q` is not cyclic (`§14`-independent,
reusable; generalises `not_isCyclic_of_card_prime_sq` to any order `> q`). A cyclic group has
`Monoid.exponent = Nat.card`, while elementary-abelianness forces the exponent to divide `q`, so
`Nat.card ∣ q`. -/
theorem not_isCyclic_of_lt_card {q : ℕ} (hq : q.Prime) {Mod : Type*} [Group Mod] [Finite Mod]
(h : OddOrder.GroupTheory.IsElementaryAbelian q Mod) (hlt : q < Nat.card Mod) :
¬ IsCyclic Mod := by
intro hcyc
have hExp_eq : Monoid.exponent Mod = Nat.card Mod := IsCyclic.exponent_eq_card
have hExp_dvd : Monoid.exponent Mod ∣ q := by
rw [Monoid.exponent_dvd_iff_forall_pow_eq_one]; exact h.pow_eq_one
rw [hExp_eq] at hExp_dvd
exact (Nat.not_dvd_of_pos_of_lt hq.pos hlt) hExp_dvd
/-- **Theorem 15.2(f) conjunct `¬ IsCyclic M_F`, gated-endpoint skeleton** (mmd L4202): assembles
`not_isCyclic_MF_of_inputs` with the engine output `[Q : Q₀] = q^n` (`n = |K| ≥ 2`, since `|K|` is
the prime `p`). The chief factor `Q̄ = Q/Q₀` (elementary abelian of order `q^n > q`) is non-cyclic
(`not_isCyclic_of_lt_card`), and `Q ⊆ M_F` (`hQMF`, Theorem 15.2(c)) lifts this to `M_F`. -/
theorem not_isCyclic_MF_of_chiefFactor_inputs [Finite G] {M Q Q0 : Subgroup G}
[(Q0.subgroupOf Q).Normal] {q n : ℕ} (hq : q.Prime) (hn : 2 ≤ n)
(hEA : OddOrder.GroupTheory.IsElementaryAbelian q (↥Q ⧸ Q0.subgroupOf Q))
(hindex : (Q0.subgroupOf Q).index = q ^ n) (hQMF : Q ≤ MF M) :
¬ IsCyclic ↥(MF M) := by
refine not_isCyclic_MF_of_inputs hQMF (not_isCyclic_of_lt_card hq hEA ?_)
rw [← Subgroup.index_eq_card, hindex]
calc q = q ^ 1 := (pow_one q).symm
_ < q ^ n := pow_lt_pow_right₀ hq.one_lt (by omega)
/-- **Theorem 15.2(g) `F(M) ⊆ M_σ`** (mmd L4198), from the same `σ`-gap as the `(g)` equality:
`F(M) = Q ⊔ (C_G(Q) ⊓ M)` (`fittingInAmbient_eq_sup_centralizer_inf_of_le_Msigma`), with
`Q = O_q(M) ⊆ M_σ` (`opiCoreInG_singleton_le_Msigma_of_mem_sigma`, `q ∈ σ(M)`) and
`C_M(Q) ⊆ M_σ` (the forward input `hCle`). So `F(M) ⊆ M_σ`. -/
theorem fittingInAmbient_le_Msigma_of_le_Msigma [Finite G]
(hG : OddOrder.BG.IsMinimalSimpleOdd G) {M Q : Subgroup G} {q : ℕ} [Fact q.Prime]
(hM : M ∈ maximalSubgroups G) (hMnormQ : M ≤ Subgroup.normalizer (Q : Set G))
(hQ : Q = opiCoreInG ({q} : Set ℕ) M) (hqσ : q ∈ OddOrder.BG.Ch3.S10.sigma M)
[(Q.subgroupOf (OddOrder.BG.Ch3.S10.Msigma M)).Normal]
[Group.IsNilpotent (↥(OddOrder.BG.Ch3.S10.Msigma M) ⧸
Q.subgroupOf (OddOrder.BG.Ch3.S10.Msigma M))]
(hCle : Subgroup.centralizer (Q : Set G) ⊓ M ≤ OddOrder.BG.Ch3.S10.Msigma M) :
fittingInAmbient M ≤ OddOrder.BG.Ch3.S10.Msigma M := by
rw [fittingInAmbient_eq_sup_centralizer_inf_of_le_Msigma hG hM hMnormQ hQ hCle]
refine sup_le ?_ hCle
rw [hQ]
exact OddOrder.BG.Ch3.S10.opiCoreInG_singleton_le_Msigma_of_mem_sigma hqσ
open scoped commutatorElement in
/-- **Theorem 15.2 conjunct 3 — `M'' ⊆ F(M)`, gated-endpoint skeleton** (mmd L4198-4201): the
chain `M'' = M_σ' ⊆ Q D' ⊆ C_{M_σ}(Q̄) = F(M)`.
After identifying `M'' = M_σ'` (conjunct 2, `h2 : M_σ = M'`), the proof reduces `M_σ' ⊆ F(M)` to
three structural ingredients:
* `hsigmaprime : M_σ' ⊆ Q ⊔ ⁅D, D⁆` — the derived subgroup of the semidirect `M_σ = Q ⋊ D`
(`Q ◁ M_σ`, complement `D`) lands in `Q · D'` (the `M_σ = QD` structure consequence);
* `hQab` / `hDcomm` — both `Q` and `D' = ⁅D, D⁆` *centralize the chief factor* `Q̄ = Q/Q₀`:
`Q̄` is (elementary) abelian (`⁅Q, Q⁆ ⊆ Q₀`, `hQab`) and `D' ⊆ C_D(Q̄)` is the engine output
`chiefFactor_card_and_commutator_of_inputs` (`∀ g ∈ ⁅D, D⁆, ∀ x ∈ Q, ⁅g, x⁆ ∈ Q₀`, mmd 15.2(g));
* `hsecFit : C_{M_σ}(Q̄) ⊆ F(M)` — the *section-centralizer* containment, the genuinely BG-specific
forward input. This is the mmd's "Proposition 1.5(d) yields `F(M) = Q C_M(Q) = C_{M_σ}(Q̄)`"
(`D` nilpotent while `M_σ` is not), which bundles the `σ`-uniqueness gap `C_M(Q) ⊆ M_σ` (cf.
`fittingInAmbient_eq_sup_centralizer_inf_of_le_Msigma`) with the Prop 1.5(d) section identity.
Note `D'` only centralizes the *section* `Q̄`, not `Q` itself, so the full-centralizer helper
`centralizer_inf_le_fittingInAmbient_of_le_Msigma` is too weak here; the section form is needed.
Both `Q` and `D'` therefore lie in `C_{M_σ}(Q̄)` (they centralize `Q̄` and sit inside `M_σ`), whence
in `F(M)` by `hsecFit`, so `M_σ' ⊆ Q ⊔ D' ⊆ F(M)`. Once the step-4 core supplies `hsigmaprime`
(QD structure) and `hsecFit` (Prop 1.5(d) + `σ`-gap), conjunct 3 becomes unconditional. -/
theorem derivedDerived_le_fittingInAmbient_of_inputs [Finite G] {M Q Q0 D : Subgroup G}
(h2 : OddOrder.BG.Ch3.S10.Msigma M = derivedInG M)
(hsigmaprime : derivedInG (OddOrder.BG.Ch3.S10.Msigma M) ≤ Q ⊔ ⁅D, D⁆)
(hQsig : Q ≤ OddOrder.BG.Ch3.S10.Msigma M)
(hDsig : D ≤ OddOrder.BG.Ch3.S10.Msigma M)
(hQab : ∀ x ∈ Q, ∀ y ∈ Q, ⁅x, y⁆ ∈ Q0)
(hDcomm : ∀ g ∈ ⁅D, D⁆, ∀ x ∈ Q, ⁅g, x⁆ ∈ Q0)
(hsecFit : ∀ x ∈ OddOrder.BG.Ch3.S10.Msigma M,
(∀ y ∈ Q, ⁅x, y⁆ ∈ Q0) → x ∈ fittingInAmbient M) :
derivedInG (derivedInG M) ≤ fittingInAmbient M := by
-- `M'' = M_σ'` (conjunct 2): rewrite the inner `derivedInG M` to `M_σ`.
rw [← h2]
-- `M_σ' ⊆ Q ⊔ D'`; show each of `Q`, `D'` lands in `F(M)` via the section-Fitting input.
refine hsigmaprime.trans (sup_le ?_ ?_)
· -- `Q ⊆ F(M)`: each `x ∈ Q` lies in `M_σ` and centralizes `Q̄` (`Q̄` abelian).
intro x hx
exact hsecFit x (hQsig hx) (fun y hy => hQab x hx y hy)
· -- `D' ⊆ F(M)`: each `g ∈ ⁅D, D⁆` lies in `M_σ` and centralizes `Q̄` (engine output).
have hDDsig : ⁅D, D⁆ ≤ OddOrder.BG.Ch3.S10.Msigma M := by
rw [Subgroup.commutator_le]
intro a ha b hb
rw [commutatorElement_def]
exact mul_mem (mul_mem (mul_mem (hDsig ha) (hDsig hb)) (inv_mem (hDsig ha)))
(inv_mem (hDsig hb))
intro g hg
exact hsecFit g (hDDsig hg) (fun x hx => hDcomm g hg x hx)
/-- **BG Corollary 15.5, "Lemma 1"**: `O_{σ(M)}(F(M)) = F(M_σ)` (`§14`-independent).
`≤`: `O_σ(F(M)) ≤ O_σ(M) = M_σ` (`opiCoreInG_fittingInG_le_opiCoreInG`); it is nilpotent (subgroup
of `F(M)`) and normal in `M` (characteristic in `F(M) ◁ M`), hence normal in `M_σ`, so a nilpotent
normal subgroup of `M_σ` lands in `F(M_σ)`. `≥`: `F(M_σ)` is characteristic in `M_σ ◁ M` hence
normal in `M`, nilpotent, so `F(M_σ) ≤ F(M)` (`fittingInG_le_fittingInG_of_le_normalizer`); it is a
`σ`-group (`≤ M_σ`) and normal in `F(M)`, so `F(M_σ) ≤ O_σ(F(M))`. -/
theorem opiCoreInG_sigma_fittingInAmbient_eq_fittingInAmbient_Msigma [Finite G]
{M : Subgroup G} :
opiCoreInG (OddOrder.BG.Ch3.S10.sigma M) (fittingInAmbient M) =
fittingInAmbient (OddOrder.BG.Ch3.S10.Msigma M) := by
set σ := OddOrder.BG.Ch3.S10.sigma M with hσ
-- `M` normalizes both `M_σ` and `O_σ(F(M))`.
have hM_norm_Mσ : M ≤ Subgroup.normalizer ((OddOrder.BG.Ch3.S10.Msigma M : Subgroup G) : Set G) :=
OddOrder.GroupTheory.le_normalizer_opiCoreInG σ M
have hMσ_le_M : OddOrder.BG.Ch3.S10.Msigma M ≤ M := OddOrder.BG.Ch3.S10.Msigma_le M
refine le_antisymm ?_ ?_
· -- `O_σ(F(M)) ≤ F(M_σ)`.
set N : Subgroup G := opiCoreInG σ (fittingInAmbient M) with hN
-- `N ≤ M_σ`.
have hN_Mσ : N ≤ OddOrder.BG.Ch3.S10.Msigma M := by
have := OddOrder.BG.Ch2.S08.opiCoreInG_fittingInG_le_opiCoreInG σ M
rwa [show opiCoreInG σ M = OddOrder.BG.Ch3.S10.Msigma M from rfl] at this
-- `N ◁ M` (characteristic in `F(M)`), hence `N ◁ M_σ`.
have hM_norm_N : M ≤ Subgroup.normalizer (N : Set G) :=
(Subgroup.normal_subgroupOf_iff_le_normalizer
((OddOrder.GroupTheory.opiCoreInG_le σ (fittingInAmbient M)).trans
(OddOrder.BG.Ch2.S08.fittingInG_le M))).mp
(OddOrder.BG.Ch2.S08.opiCoreInG_fittingInG_subgroupOf_normal σ M)
have hNnorm_Mσ : (N.subgroupOf (OddOrder.BG.Ch3.S10.Msigma M)).Normal :=
(Subgroup.normal_subgroupOf_iff_le_normalizer hN_Mσ).mpr (hMσ_le_M.trans hM_norm_N)
-- `N` is nilpotent (subgroup of the nilpotent `F(M)`).
have : Group.IsNilpotent ↥N := by
have : Group.IsNilpotent ↥(fittingInAmbient M) :=
OddOrder.BG.Ch2.S08.fittingInG_isNilpotent M
exact Group.nilpotent_of_mulEquiv
(Subgroup.subgroupOfEquivOfLe (OddOrder.GroupTheory.opiCoreInG_le σ (fittingInAmbient M)))
-- Nilpotent normal subgroup of `M_σ` lands in `F(M_σ)`.
have : Group.IsNilpotent ↥(N.subgroupOf (OddOrder.BG.Ch3.S10.Msigma M)) :=
Group.nilpotent_of_mulEquiv (Subgroup.subgroupOfEquivOfLe hN_Mσ).symm
have := hNnorm_Mσ
calc N = (N.subgroupOf (OddOrder.BG.Ch3.S10.Msigma M)).map
(OddOrder.BG.Ch3.S10.Msigma M).subtype :=
(Subgroup.map_subgroupOf_eq_of_le hN_Mσ).symm
_ ≤ (OddOrder.Isaacs.Ch01.fitting ↥(OddOrder.BG.Ch3.S10.Msigma M)).map
(OddOrder.BG.Ch3.S10.Msigma M).subtype :=
Subgroup.map_mono OddOrder.Isaacs.Ch01.nilpotent_normal_le_fitting
_ = fittingInAmbient (OddOrder.BG.Ch3.S10.Msigma M) := rfl
· -- `F(M_σ) ≤ O_σ(F(M))`.
-- `F(M_σ) ≤ F(M)` (`F(M_σ)` characteristic in `M_σ ◁ M`, nilpotent).
have hFMσ_le_FM : fittingInAmbient (OddOrder.BG.Ch3.S10.Msigma M) ≤ fittingInAmbient M :=
OddOrder.BG.Ch2.S08.fittingInG_le_fittingInG_of_le_normalizer hMσ_le_M hM_norm_Mσ
-- `F(M_σ) ≤ M_σ`, a `σ`-group.
have hFMσ_le_Mσ : fittingInAmbient (OddOrder.BG.Ch3.S10.Msigma M) ≤
OddOrder.BG.Ch3.S10.Msigma M := OddOrder.BG.Ch2.S08.fittingInG_le _
have hFMσ_pi : Subgroup.IsPiSubgroup σ (fittingInAmbient (OddOrder.BG.Ch3.S10.Msigma M)) :=
fun r hr => OddOrder.BG.Ch3.S10.Msigma_isPiGroup M r
(Nat.primeFactors_mono (Subgroup.card_dvd_of_le hFMσ_le_Mσ) Nat.card_pos.ne' hr)
-- `F(M_σ) ◁ F(M)` (since `M` normalizes `F(M_σ)` and `F(M) ≤ M`).
have hM_norm_FMσ : M ≤ Subgroup.normalizer
((fittingInAmbient (OddOrder.BG.Ch3.S10.Msigma M) : Subgroup G) : Set G) := fun x hx =>
OddOrder.BG.Ch2.S08.mem_normalizer_fittingInG_of_mem_normalizer (hM_norm_Mσ hx)
have hFMσ_norm_FM : ((fittingInAmbient (OddOrder.BG.Ch3.S10.Msigma M)).subgroupOf
(fittingInAmbient M)).Normal :=
(Subgroup.normal_subgroupOf_iff_le_normalizer hFMσ_le_FM).mpr
((OddOrder.BG.Ch2.S08.fittingInG_le M).trans hM_norm_FMσ)
exact OddOrder.GroupTheory.le_opiCoreInG_of_normal_of_isPiSubgroup hFMσ_le_FM
hFMσ_norm_FM hFMσ_pi
/-- Lattice absorption used in BG Corollary 15.5 (Case I): if `C = A ⊔ X` with `A ≤ B`, then
`C ⊔ B = X ⊔ B`. Pure lattice fact (kept generic to avoid `whnf` on the underlying `Subgroup`
`set`-locals in the main proof). -/
theorem sup_eq_sup_of_eq_sup_of_le {α : Type*} [Lattice α] {C A X B : α}
(hC : C = A ⊔ X) (hA : A ≤ B) : C ⊔ B = X ⊔ B := by
subst hC
rw [sup_right_comm, sup_eq_right.mpr hA, sup_comm]
/-- **Nilpotent normal subgroup lands in the ambient Fitting subgroup** (`§14`-independent,
reusable): if `N ≤ M`, `N.subgroupOf M ⊴ M`, and `N` is nilpotent, then `N ≤ F(M)`
(`fittingInAmbient M`). The relative `N.subgroupOf M` is a nilpotent normal subgroup of `↥M`,
so it lies in `fitting ↥M` (`nilpotent_normal_le_fitting`); mapping back gives the claim. -/
theorem le_fittingInAmbient_of_subgroupOf_normal_of_isNilpotent [Finite G] {M N : Subgroup G}
(hNM : N ≤ M) (hNnorm : (N.subgroupOf M).Normal) [Group.IsNilpotent ↥N] :
N ≤ fittingInAmbient M := by
have := hNnorm
have : Group.IsNilpotent ↥(N.subgroupOf M) :=
Group.nilpotent_of_mulEquiv (Subgroup.subgroupOfEquivOfLe hNM).symm
calc N = (N.subgroupOf M).map M.subtype := (Subgroup.map_subgroupOf_eq_of_le hNM).symm
_ ≤ (OddOrder.Isaacs.Ch01.fitting ↥M).map M.subtype :=
Subgroup.map_mono OddOrder.Isaacs.Ch01.nilpotent_normal_le_fitting
_ = fittingInAmbient M := rfl
/-- **BG Theorem 15.2** (mmd L4112): if `M_F` is strictly smaller than `M_sigma`,
then `M` is type `P1` and has the normal `q`-subgroup / minimal chief factor
structure described in the text. -/
theorem mf_ne_msigma_typeP1_structure [Finite G]
(hG : OddOrder.BG.IsMinimalSimpleOdd G) {M K Kstar : Subgroup G}
(hM : M ∈ maximalSubgroups G) (hne : MF M ≠ OddOrder.BG.Ch3.S10.Msigma M)
(hKM : K ≤ M)
(hK : Ch03.IsHallSubgroup (S14.kappa M) (K.subgroupOf M))
(hKstar : Kstar = OddOrder.BG.Ch3.S10.Msigma M ⊓ Subgroup.centralizer (K : Set G)) :
S14.IsTypeP1 M ∧
∃ Q Q0 D : Subgroup G, ∃ p q : ℕ,
p.Prime ∧ q.Prime ∧ Nat.card ↥K = p ∧ Nat.card ↥Kstar = q ∧
q ∈ S14.piSet (MF M) ∧ q ∈ OddOrder.BG.Ch3.S10.beta M ∧
Kstar ≤ MF M ∧
Q ≤ MF M ∧ M ≤ Subgroup.normalizer (Q : Set G) ∧
Subgroup.IsComplement' (Q.subgroupOf (OddOrder.BG.Ch3.S10.Msigma M))
(D.subgroupOf (OddOrder.BG.Ch3.S10.Msigma M)) ∧
Group.IsNilpotent ↥D ∧
Q0 = Q ⊓ Subgroup.centralizer (D : Set G) ∧
M ≤ Subgroup.normalizer (Q0 : Set G) ∧
-- mmd 15.2(f): the chief factor `Q̄ = Q/Q0` is elementary abelian of order `q^p`
-- (faithfulness fix, Lane G 2026-06-16: the previous scaffold wrote
-- `Nat.card ↥(Q.subgroupOf (Q ⊔ Q0))`, which is `|Q|` since `Q0 = Q ⊓ C(D) ⊆ Q` forces
-- `Q ⊔ Q0 = Q`; the intended `|Q̄| = |Q : Q0|` is `(Q0.subgroupOf Q).index`).
(Q0.subgroupOf Q).index = q ^ p ∧
OddOrder.BG.Ch3.S10.Msigma M = derivedInG M ∧
derivedInG (derivedInG M) ≤ fittingInAmbient M ∧
-- mmd 15.2(g) "F(M) ⊂ M_σ": the Fitting subgroup is contained in the σ-core.
fittingInAmbient M ≤ OddOrder.BG.Ch3.S10.Msigma M ∧
-- mmd 15.2(g) "F(M) = Q C_M(Q)": the Fitting subgroup is the product of the normal
-- `q`-subgroup `Q` and its `M`-centralizer (`Q` self-centralizing up to `C_M(Q)`).
fittingInAmbient M = Q ⊔ (Subgroup.centralizer (Q : Set G) ⊓ M) ∧
-- mmd 15.2(f): `M_F ⊇ Q̄`, an elementary abelian section of order `q^p` (rank `p ≥ 3`),
-- so `M_F` is non-cyclic. Breaks the 15.5↔15.6 circularity (Corollary 15.6's proof needs
-- this without citing Corollary 15.5).
¬ IsCyclic ↥(MF M) := by
classical
-- **Setup.** `p = |K|`, `q = |K*|`; `M` is type `P₁`, `q` prime, `M_σ = M'`.
set q : ℕ := Nat.card ↥Kstar with hqdef
have hP1 : S14.IsTypeP1 M := isTypeP1_of_mf_ne_msigma hG hM hne
have hP : S14.IsTypeP M := hP1.1
have hq_prime : q.Prime := kstar_card_prime_of_inputs hG hM hP1 hKM hK hKstar
have : Fact q.Prime := ⟨hq_prime⟩
have hMσderived : OddOrder.BG.Ch3.S10.Msigma M = derivedInG M :=
typeP1_msigma_eq_derivedInG hG hM hP1 hKM hK hKstar
obtain ⟨hcomplDer, _, _⟩ := S14.typeP_duality hG hM hP hKM hK hKstar
have hcomplMσ : Subgroup.IsComplement' ((OddOrder.BG.Ch3.S10.Msigma M).subgroupOf M)
(K.subgroupOf M) := by rw [hMσderived]; exact hcomplDer
have hMσM : OddOrder.BG.Ch3.S10.Msigma M ≤ M := OddOrder.BG.Ch3.S10.Msigma_le M
have : Group.IsSolvable ↥M := hG.isSolvable_of_mem_maximalSubgroups hM
have hMσnotnil : ¬ Group.IsNilpotent ↥(OddOrder.BG.Ch3.S10.Msigma M) := fun hnil =>
hne ((maxNilpotentNormalHall_eq_Msigma_iff_isNilpotent hG hM).mpr hnil)
have hKstarMσ : Kstar ≤ OddOrder.BG.Ch3.S10.Msigma M := by rw [hKstar]; exact inf_le_left
have hqMσ : q ∣ Nat.card ↥(OddOrder.BG.Ch3.S10.Msigma M) := by
rw [hqdef]; exact Subgroup.card_dvd_of_le hKstarMσ
have hqσ : q ∈ OddOrder.BG.Ch3.S10.sigma M :=
OddOrder.BG.Ch3.S10.Msigma_isPiGroup M q
(Nat.mem_primeFactors.mpr ⟨hq_prime, hqMσ, Nat.card_pos.ne'⟩)
set Q : Subgroup G := opiCoreInG ({q} : Set ℕ) M with hQdef
have hQMσ : Q ≤ OddOrder.BG.Ch3.S10.Msigma M := by
rw [hQdef]; exact OddOrder.BG.Ch3.S10.opiCoreInG_singleton_le_Msigma_of_mem_sigma hqσ
have hMnormQ : M ≤ Subgroup.normalizer (Q : Set G) := by
rw [hQdef]; exact OddOrder.GroupTheory.le_normalizer_opiCoreInG _ M
have hQM : Q ≤ M := hQMσ.trans hMσM
have hcop_sub : Nat.Coprime (Nat.card ↥((OddOrder.BG.Ch3.S10.Msigma M).subgroupOf M))
(Nat.card ↥(K.subgroupOf M)) := by
have h := coprime_card_derived_kappaHall_of_isComplement' hK hcomplDer
rwa [hMσderived]
have hcond2 := actsPrimeManner_subgroupOf_of_typeP hG hM hP hKM hK hKstar
have hqG : (Nat.card ↥(OddOrder.BG.Ch3.S10.Msigma M
⊓ Subgroup.centralizer (K : Set G))).Prime := by rw [← hKstar]; exact hq_prime
have hKstarQ : Kstar ≤ Q := by
have h := kstar_le_opiCore_of_inputs hG hM hKM hcomplMσ hcop_sub.symm hcond2 hne hqG
rw [← hKstar, ← hqdef, ← hQdef] at h; exact h
have hcopKMσ : Nat.Coprime (Nat.card ↥K) (Nat.card ↥(OddOrder.BG.Ch3.S10.Msigma M)) := by
have e1 : Nat.card ↥(K.subgroupOf M) = Nat.card ↥K :=
Nat.card_congr (Subgroup.subgroupOfEquivOfLe hKM).toEquiv
have e2 : Nat.card ↥((OddOrder.BG.Ch3.S10.Msigma M).subgroupOf M)
= Nat.card ↥(OddOrder.BG.Ch3.S10.Msigma M) :=
Nat.card_congr (Subgroup.subgroupOfEquivOfLe hMσM).toEquiv
rw [e1, e2] at hcop_sub; exact hcop_sub.symm
have hKMσdisj : Disjoint K (OddOrder.BG.Ch3.S10.Msigma M) := by
rw [disjoint_iff]
exact Subgroup.card_eq_one.mp (Nat.eq_one_of_dvd_coprimes hcopKMσ
(Subgroup.card_dvd_of_le inf_le_left) (Subgroup.card_dvd_of_le inf_le_right))
have hKne : K ≠ ⊥ := by
intro hK0
apply hMσnotnil
have hKstareq : Kstar = OddOrder.BG.Ch3.S10.Msigma M := by
rw [hKstar, hK0]
have hc : Subgroup.centralizer ((⊥ : Subgroup G) : Set G) = ⊤ := by
rw [Subgroup.coe_bot]; ext g; simp [Subgroup.mem_centralizer_iff]
rw [hc, inf_top_eq]
have hcardMσ : Nat.card ↥(OddOrder.BG.Ch3.S10.Msigma M) = q ^ 1 := by
rw [pow_one, hqdef, hKstareq]
exact (IsPGroup.of_card hcardMσ).isNilpotent
have hQpg : IsPGroup q ↥Q := by rw [hQdef]; exact isPGroup_opiCoreInG_singleton M
have hQneMσ : Q ≠ OddOrder.BG.Ch3.S10.Msigma M := by
intro hQeq; exact hMσnotnil (hQeq ▸ hQpg.isNilpotent)
-- **The `K`-invariant complement `D` of `Q` in `M_σ`.**
obtain ⟨D, hDMσ, hKnormD, hQDdisj, hcomplD, hDnil, hDne, hDq'⟩ :=
exists_kInvariant_qComplement hG hM hP hKM hK hKstar hQdef hQMσ hMnormQ hKstarQ hQneMσ hKne
hKMσdisj hcopKMσ
-- **The chief factor `Q₀ = C_Q(D) ⊴ M` and the Theorem 3.10 engine outputs.**
obtain ⟨hMNQ0, hKstarNotQ0, hQ0ltQ, hmin⟩ :=
chiefFactor_Q0_normal_minimal_of_inputs hG hM hP1 hKM hK hKstar hQdef hQMσ hMnormQ hKstarQ
hQneMσ hKne hKMσdisj hcopKMσ hMσnotnil hDq' hDMσ hKnormD hQDdisj hcomplD hDnil hDne
set Q0 : Subgroup G := Q ⊓ Subgroup.centralizer (D : Set G) with hQ0def
have hQ0Q : Q0 ≤ Q := hQ0ltQ.le
have hQ0nQ : (Q0.subgroupOf Q).Normal :=
(Subgroup.normal_subgroupOf_iff_le_normalizer hQ0Q).mpr (hQM.trans hMNQ0)
obtain ⟨hKprime, hindex, hDcomm, hEA⟩ :=
chiefFactor_engine_of_inputs hG hM hP1 hKM hK hKstar hQdef hQMσ hMnormQ hKstarQ hKne hKMσdisj
hcopKMσ hDq' hDMσ hKnormD hQDdisj hDne hQ0def hMNQ0 hKstarNotQ0 hQ0ltQ hmin
-- **Fitting subgroup (Theorem 15.2(g)).**
have hQnMσ : (Q.subgroupOf (OddOrder.BG.Ch3.S10.Msigma M)).Normal :=
(Subgroup.normal_subgroupOf_iff_le_normalizer hQMσ).mpr (hMσM.trans hMnormQ)
have hNilMσQ : Group.IsNilpotent (↥(OddOrder.BG.Ch3.S10.Msigma M) ⧸
Q.subgroupOf (OddOrder.BG.Ch3.S10.Msigma M)) :=
msigma_quotient_isNilpotent_of_inputs hG hM hP hKM hK hKstar hQMσ hMnormQ hKstarQ hQneMσ hKne
hKMσdisj hcopKMσ
have hprime := actsPrimeManner_of_typeP hG hM hP hKM hK hKstar
have hQgtq : q < Nat.card ↥Q := by
have h1 : (Q0.subgroupOf Q).index ≤ Nat.card ↥Q :=
Nat.le_of_dvd Nat.card_pos (Subgroup.index_dvd_card _)
have h2 : q < q ^ Nat.card ↥K := by
calc q = q ^ 1 := (pow_one q).symm
_ < q ^ Nat.card ↥K := Nat.pow_lt_pow_right hq_prime.one_lt hKprime.two_le
rw [hindex] at h1; exact lt_of_lt_of_le h2 h1
have hKstarneQ : Kstar ≠ Q := by
intro h
have hqeq : q = Nat.card ↥Q := by rw [hqdef, h]
rw [hqeq] at hQgtq; exact lt_irrefl _ hQgtq
have hCle : Subgroup.centralizer (Q : Set G) ⊓ M ≤ OddOrder.BG.Ch3.S10.Msigma M :=
centralizer_le_Msigma_of_primeManner hG hM hP1 hKM hK hprime hQMσ hMnormQ hKstarQ hKstarneQ
have cC17 : fittingInAmbient M ≤ OddOrder.BG.Ch3.S10.Msigma M :=
fittingInAmbient_le_Msigma_of_le_Msigma hG hM hMnormQ hQdef hqσ hCle
have cC18 : fittingInAmbient M = Q ⊔ (Subgroup.centralizer (Q : Set G) ⊓ M) :=
fittingInAmbient_eq_sup_centralizer_inf_of_le_Msigma hG hM hMnormQ hQdef hCle
have hcopDQ : Nat.Coprime (Nat.card ↥D) (Nat.card ↥Q) := by
obtain ⟨n, hn⟩ := hQpg.exists_card_eq
rw [hn]
rcases Nat.eq_zero_or_pos n with h0 | h0
· subst h0; simp
· rw [Nat.coprime_pow_right_iff h0]
exact ((Fact.out : q.Prime).coprime_iff_not_dvd.mpr (fun hd =>
hDq' (Nat.mem_primeFactors.mpr ⟨Fact.out, hd, Nat.card_pos.ne'⟩))).symm
have hQab : ∀ x ∈ Q, ∀ y ∈ Q, ⁅x, y⁆ ∈ Q0 := by
intro x hxQ y hyQ
have hcomm := hEA.comm (QuotientGroup.mk (⟨x, hxQ⟩ : ↥Q))
(QuotientGroup.mk (⟨y, hyQ⟩ : ↥Q))
have h1 : QuotientGroup.mk (⁅(⟨x, hxQ⟩ : ↥Q), (⟨y, hyQ⟩ : ↥Q)⁆) =
(1 : ↥Q ⧸ Q0.subgroupOf Q) := by
rw [← QuotientGroup.mk'_apply, map_commutatorElement, commutatorElement_eq_one_iff_mul_comm]
exact hcomm
rw [QuotientGroup.eq_one_iff, Subgroup.mem_subgroupOf] at h1
have h2 : ((⁅(⟨x, hxQ⟩ : ↥Q), (⟨y, hyQ⟩ : ↥Q)⁆ : ↥Q) : G) = ⁅x, y⁆ := by
push_cast [commutatorElement_def]; rfl
rwa [h2] at h1
-- `M_σ' ⊆ Q ⊔ ⁅D, D⁆`: the derived subgroup of the semidirect `M_σ = Q ⋊ D` (`Q` normal,
-- `D` complement). `derivedInG_le_sup_of_normal` (S13) is exactly this normal-target argument:
-- modulo the normal `Q` the quotient `M_σ/Q` is the image of `D`, so its derived subgroup is the
-- image of `D' = ⁅D, D⁆`, and pulling back gives `M_σ' ⊆ Q ⊔ ⁅D, D⁆`. (The `M_σ = QD`
-- decomposition `Q ⊔ D = M_σ` is read off the complement `hcomplD.sup_eq_top`.)
have hsigmaprime : derivedInG (OddOrder.BG.Ch3.S10.Msigma M) ≤ Q ⊔ ⁅D, D⁆ := by
have hsup : Q ⊔ D = OddOrder.BG.Ch3.S10.Msigma M := by
have h := congrArg (Subgroup.map (OddOrder.BG.Ch3.S10.Msigma M).subtype) hcomplD.sup_eq_top
rwa [Subgroup.map_sup, Subgroup.subgroupOf_map_subtype, Subgroup.subgroupOf_map_subtype,
inf_eq_left.mpr hQMσ, inf_eq_left.mpr hDMσ, ← MonoidHom.range_eq_map,
Subgroup.range_subtype] at h
have hle := OddOrder.BG.Ch3.S13.derivedInG_le_sup_of_normal hQMσ hDMσ hsup hQnMσ
rwa [show derivedInG D = ⁅D, D⁆ from Subgroup.map_subtype_commutator D] at hle
have hsecFit : ∀ x ∈ OddOrder.BG.Ch3.S10.Msigma M, (∀ y ∈ Q, ⁅x, y⁆ ∈ Q0) →
x ∈ fittingInAmbient M :=
centralizer_msigma_quotient_le_fittingInAmbient hG hM hQMσ hDMσ hQ0def hQ0Q hcomplD hMnormQ
hMNQ0 hQpg hDnil hcopDQ hQab
have cC16 : derivedInG (derivedInG M) ≤ fittingInAmbient M :=
derivedDerived_le_fittingInAmbient_of_inputs hMσderived hsigmaprime hQMσ hDMσ hQab hDcomm
hsecFit
-- **`q ∈ β(M)` (conjunct 6).**
have hQne : Q ≠ ⊥ := by
intro h0
have hKstar0 : Kstar = ⊥ := le_bot_iff.mp (h0 ▸ hKstarQ)
have : q = 1 := by rw [hqdef, hKstar0, Subgroup.card_bot]
exact hq_prime.ne_one this
have hQ0notC : ¬ Q ≤ Subgroup.centralizer (D : Set G) := by
intro hle
exact (ne_of_lt hQ0ltQ) (le_antisymm hQ0Q (le_inf le_rfl hle))
have hqπ : q ∈ (Nat.card ↥M).primeFactors :=
Nat.mem_primeFactors.mpr ⟨hq_prime, hqMσ.trans (Subgroup.card_dvd_of_le hMσM), Nat.card_pos.ne'⟩
have hq_odd : Odd q := hG.odd.of_dvd_nat ((Nat.mem_primeFactors.mp hqπ).2.1.trans
(Subgroup.card_subgroup_dvd_card M))
have hcopKD : Nat.Coprime (Nat.card ↥K) (Nat.card ↥D) :=
hcopKMσ.coprime_dvd_right (Subgroup.card_dvd_of_le hDMσ)
have hDKN : D ⊔ K ≤ Subgroup.normalizer (Q : Set G) :=
sup_le (hDMσ.trans (hMσM.trans hMnormQ)) (hKM.trans hMnormQ)
have hqD : ¬ q ∣ Nat.card ↥D := fun hd =>
hDq' (Nat.mem_primeFactors.mpr ⟨Fact.out, hd, Nat.card_pos.ne'⟩)
have hKsolv : Group.IsSolvable ↥K :=
Group.isSolvable_of_isSolvable_injective (Subgroup.inclusion_injective hKM)
have hqK : ¬ q ∣ Nat.card ↥K := fun hd =>
hq_prime.one_lt.ne' (Nat.eq_one_of_dvd_coprimes hcopKMσ hd hqMσ)
-- Sylow witness `P : Sylow q ↥M` with `Q = P.map M.subtype`.
have hidx_M : ¬ q ∣ (Q.subgroupOf M).index := by
have hMcard : Nat.card ↥M = Nat.card ↥(OddOrder.BG.Ch3.S10.Msigma M) * Nat.card ↥K := by
have h := hcomplMσ.card_mul_card
rw [Nat.card_congr (Subgroup.subgroupOfEquivOfLe hMσM).toEquiv,
Nat.card_congr (Subgroup.subgroupOfEquivOfLe hKM).toEquiv] at h
exact h.symm
have hMσcard : Nat.card ↥(OddOrder.BG.Ch3.S10.Msigma M) = Nat.card ↥Q * Nat.card ↥D := by
have h := hcomplD.card_mul_card
rw [Nat.card_congr (Subgroup.subgroupOfEquivOfLe hQMσ).toEquiv,
Nat.card_congr (Subgroup.subgroupOfEquivOfLe hDMσ).toEquiv] at h
exact h.symm
have hidxeq : (Q.subgroupOf M).index = Nat.card ↥D * Nat.card ↥K := by
have hmul := Subgroup.card_mul_index (Q.subgroupOf M)
rw [Nat.card_congr (Subgroup.subgroupOfEquivOfLe hQM).toEquiv, hMcard, hMσcard] at hmul
refine Nat.eq_of_mul_eq_mul_left Nat.card_pos (?_ : Nat.card ↥Q * _ = Nat.card ↥Q * _)
rw [hmul]; ring
rw [hidxeq]
exact fun hdvd => ((Nat.Prime.dvd_mul hq_prime).mp hdvd).elim hqD hqK
obtain ⟨P, hQP⟩ := exists_sylow_eq_opiCore hQdef hQM hMnormQ hQpg hidx_M
-- the `G`-level fixed-point-free condition, lifted from the `↥(D ⊔ K)`-Frobenius group.
have hfrob := isFrobeniusGroup_DK_of_primeManner (M := M) hprime hDMσ hKstarQ hQDdisj.symm hKnormD
(hKMσdisj.symm.mono_left hDMσ) hDne hKne
have hFrobFPF : ∀ a ∈ K, a ≠ 1 → ∀ n ∈ D, n ≠ 1 → a * n * a⁻¹ ≠ n := by
intro a haK ha1 n hnD hn1 heq
have haDK : a ∈ D ⊔ K := (le_sup_right : K ≤ D ⊔ K) haK
have hnDK : n ∈ D ⊔ K := (le_sup_left : D ≤ D ⊔ K) hnD
refine hfrob.conj_frobenius ⟨a, haDK⟩ (Subgroup.mem_subgroupOf.mpr haK)
(fun h => ha1 (congrArg Subtype.val h)) ⟨n, hnDK⟩ (Subgroup.mem_subgroupOf.mpr hnD)
(fun h => hn1 (congrArg Subtype.val h)) (Subtype.ext ?_)
change (a : G) * (n : G) * (a : G)⁻¹ = (n : G)
exact heq
have cC6 : q ∈ OddOrder.BG.Ch3.S10.beta M :=
mem_beta_of_inputs hQ0notC (D_centralizes_Q_of_not_mem_beta hG hM hq_odd hQpg hQne hqπ P hQP
hKne hFrobFPF hKnormD hcopKD hKsolv hDKN hqD)
-- **`Q, K* ⊆ M_F` and `q ∈ π(M_F)` (conjuncts 5,7,8).**
have hQhall : OddOrder.Isaacs.Ch03.IsHallSubgroup (Nat.card ↥Q).primeFactors
(Q.subgroupOf M) := by
have hpf : ((Nat.card ↥Q).primeFactors : Set ℕ) = ({q} : Set ℕ) := by
obtain ⟨n, hn⟩ := hQpg.exists_card_eq
have hn0 : n ≠ 0 := by
rintro rfl; rw [pow_zero] at hn; rw [hn] at hQgtq
have := hq_prime.two_le; omega
rw [hn, Nat.primeFactors_prime_pow hn0 hq_prime, Finset.coe_singleton]
rw [hpf]
exact isHallSubgroup_subgroupOf_of_pgroup_of_not_dvd_index hQM hQpg hidx_M
have hQnilM : Group.IsNilpotent ↥(Q.subgroupOf M) :=
(hQpg.comap_subtype).isNilpotent
have cC8 : Q ≤ MF M := le_maxNilpotentNormalHall hQM
((Subgroup.normal_subgroupOf_iff_le_normalizer hQM).mpr hMnormQ) hQnilM hQhall
have cC7 : Kstar ≤ MF M := hKstarQ.trans cC8
have cC5 : q ∈ S14.piSet (MF M) := by
rw [S14.piSet, Set.mem_ofPred_eq]
refine Nat.mem_primeFactors.mpr ⟨hq_prime, ?_, Nat.card_pos.ne'⟩
rw [hqdef]; exact Subgroup.card_dvd_of_le cC7
-- **`¬ IsCyclic M_F` (conjunct 19).**
have cC19 : ¬ IsCyclic ↥(MF M) :=
not_isCyclic_MF_of_chiefFactor_inputs hq_prime hKprime.two_le hEA hindex cC8
-- **Assemble.**
exact ⟨hP1, Q, Q0, D, Nat.card ↥K, q, hKprime, hq_prime, rfl, hqdef.symm, cC5, cC6, cC7, cC8,
hMnormQ, hcomplD, hDnil, hQ0def, hMNQ0, hindex, hMσderived, cC16, cC17, cC18, cC19⟩
/-- **BG Corollary 15.5** (mmd L4225): the decomposition `F(M) = F(M_σ) × Y` with
`Y = O_{σ(M)'}(F(M))` a cyclic `τ₂(M)`-subgroup, together with `F(M) = C_M(M_F)·M_F`,
`M'' ⊆ F(M)`, `M_F ⊆ M'`, and `K ≠ 1 → F(M) ⊆ M'`. Direct products are encoded by the
commuting/trivial-intersection package.
Faithfulness fix (Lane G): the previous scaffold parametrized an arbitrary `H ≤ M_F` (mmd
fixes `H = M_F`) and used `M_F(M_σ)` where the textbook has the Fitting subgroup `F(M_σ)`
(`fittingInAmbient (Msigma M)`); the dropped conjuncts (a)/(b)/(d) are restored. The `M'/M_F`
nilpotent clause of (c) is still deferred (quotient API).
`M_F` cyclic ⟹ `F(M)` cyclic exposure (Lane G 2026-06-15): the final conjunct records the
derived consequence that Corollary 15.6's proof cites ("if `M_F` is cyclic, then `F(M)` is
cyclic by Corollary 15.5"). It follows from (a)/(b): when `M_σ` is nilpotent, `F(M_σ) = M_σ =
M_F` (`fittingInAmbient_eq_self_of_isNilpotent`), so `F(M) = M_F × Y` is a product of coprime
cyclic factors (`isCyclic_prod_iff`); otherwise `M_F` is non-cyclic (Theorem 15.2) and the
implication is vacuous. This supplies the `hFcyc` hypothesis of `typeP_kstar_in_mf_of_inputs`. -/
theorem fitting_decomposition [Finite G] (hG : OddOrder.BG.IsMinimalSimpleOdd G)
{M : Subgroup G} (hM : M ∈ maximalSubgroups G) :
∃ Y : Subgroup G,
-- (a) `Y = O_{σ(M)'}(F(M))` is a cyclic `τ₂(M)`-subgroup of `F(M)`.
IsCyclic ↥Y ∧ (↑(Nat.card ↥Y).primeFactors ⊆ tau2 M) ∧ Y ≤ fittingInAmbient M ∧
-- (b) `M'' ⊆ F(M) = C_M(M_F)·M_F = F(M_σ) × Y`.
derivedInG (derivedInG M) ≤ fittingInAmbient M ∧
fittingInAmbient M = (Subgroup.centralizer (MF M : Set G) ⊓ M) ⊔ MF M ∧
fittingInAmbient M = fittingInAmbient (OddOrder.BG.Ch3.S10.Msigma M) ⊔ Y ∧
fittingInAmbient (OddOrder.BG.Ch3.S10.Msigma M) ⊓ Y = ⊥ ∧
⁅fittingInAmbient (OddOrder.BG.Ch3.S10.Msigma M), Y⁆ = ⊥ ∧
-- (c) `M_F ⊆ M'`. ⚠ The book's second half of (c), "`M'/M_F` is nilpotent", is **proved**
-- as `S16.derivedInG_quotient_maxNilpotentNormalHall_isNilpotent`; it cannot be bundled
-- here because that proof lives downstream (`S16_MainResults/TypeBridges.lean` imports
-- §15), so putting it in this statement would be an import cycle. (The old comment said
-- it was "deferred — quotient API"; stale, corrected in the issue 0177 §15 audit.)
MF M ≤ derivedInG M ∧
-- (d) if `K ≠ 1` (i.e. `M` is not of type `F`), then `F(M) ⊆ M'`.
(¬ S14.IsTypeF M → fittingInAmbient M ≤ derivedInG M) ∧
-- The derived consequence Corollary 15.6's proof cites ("`F(M)` is cyclic by Cor 15.5"):
-- via the `F(M) = F(M_σ) × Y` decomposition (both factors cyclic, coprime orders when
-- `M_σ` is nilpotent so `F(M_σ) = M_σ = M_F`; otherwise `M_F` is non-cyclic, vacuous).
(IsCyclic ↥(MF M) → IsCyclic ↥(fittingInAmbient M)) := by
classical
have : Group.IsSolvable ↥M := hG.isSolvable_of_mem_maximalSubgroups hM
set σ := OddOrder.BG.Ch3.S10.sigma M with hσ
set Mσ := OddOrder.BG.Ch3.S10.Msigma M with hMσ
set F := fittingInAmbient M with hF
set FMσ := fittingInAmbient Mσ with hFMσ
-- `Y := O_{σ'}(F(M))`, the `σ'`-Hall part of the Fitting subgroup.
set Y : Subgroup G := opiCoreInG σᶜ F with hY
have hFnil : Group.IsNilpotent ↥F := OddOrder.BG.Ch2.S08.fittingInG_isNilpotent M
-- ## Case-independent facts.
-- Lemma 1: `O_σ(F(M)) = F(M_σ)`.
have hL1 : opiCoreInG σ F = FMσ :=
opiCoreInG_sigma_fittingInAmbient_eq_fittingInAmbient_Msigma
-- Nilpotent Hall splitting `O_σ(F) ⊔ Y = F`.
have hsplit : opiCoreInG σ F ⊔ Y = F := opiCoreInG_sup_compl_eq_of_isNilpotent σ
-- Conjunct 3: `Y ≤ F`.
have h3 : Y ≤ F := OddOrder.GroupTheory.opiCoreInG_le σᶜ F
-- Conjunct 6: `F = F(M_σ) ⊔ Y`.
have h6 : F = FMσ ⊔ Y := by rw [← hL1, hsplit]
-- Conjunct 7: `F(M_σ) ⊓ Y = ⊥`.
have h7 : FMσ ⊓ Y = ⊥ := by
rw [← hL1, hY]
exact OddOrder.GroupTheory.inf_eq_bot_of_isPiSubgroup_of_isPiSubgroup_compl
(OddOrder.GroupTheory.isPiSubgroup_opiCoreInG σ F)
(OddOrder.GroupTheory.isPiSubgroup_opiCoreInG σᶜ F)
-- Conjunct 8: `⁅F(M_σ), Y⁆ = ⊥`.
have h8 : ⁅FMσ, Y⁆ = ⊥ := by
rw [← hL1, hY]; exact OddOrder.BG.Ch2.S08.opiCoreInG_commutator_compl_eq_bot σ F
-- Conjunct 9: `M_F ≤ M'`.
have h9 : MF M ≤ derivedInG M := maxNilpotentNormalHall_le_derived hG hM
-- `Y` is a `σ'`-group, `M_σ` is a `σ`-group, so they are coprime.
have hYpi : Subgroup.IsPiSubgroup σᶜ Y := OddOrder.GroupTheory.isPiSubgroup_opiCoreInG σᶜ F
-- `M ≤ N_G(M_σ)` and `M ≤ N_G(Y)` (the latter since `Y` is characteristic in `F(M) ◁ M`).
have hM_norm_Mσ : M ≤ Subgroup.normalizer (Mσ : Set G) :=
OddOrder.GroupTheory.le_normalizer_opiCoreInG σ M
have hMσ_le_M : Mσ ≤ M := OddOrder.BG.Ch3.S10.Msigma_le M
have hM_norm_Y : M ≤ Subgroup.normalizer (Y : Set G) :=
(Subgroup.normal_subgroupOf_iff_le_normalizer
(h3.trans (OddOrder.BG.Ch2.S08.fittingInG_le M))).mp
(by rw [hY]; exact OddOrder.BG.Ch2.S08.opiCoreInG_fittingInG_subgroupOf_normal σᶜ M)
-- ## Lemma 15.1 inputs (a `κ`-Hall `K` and a `(κ∪σ)ᶜ`-Hall `U`), via Hall's theorem in `↥M`.
obtain ⟨K', hK'⟩ := Ch03.hall_E_exists (G := ↥M) (S14.kappa M)
set K : Subgroup G := K'.map M.subtype with hK
have hKof : K.subgroupOf M = K' :=
Subgroup.comap_map_eq_self_of_injective M.subtype_injective K'
have hKHall : Ch03.IsHallSubgroup (S14.kappa M) (K.subgroupOf M) := by rw [hKof]; exact hK'
obtain ⟨U', hU'⟩ := Ch03.hall_E_exists (G := ↥M) ((S14.kappa M ∪ σ)ᶜ)
set U : Subgroup G := U'.map M.subtype with hU
have hUof : U.subgroupOf M = U' :=
Subgroup.comap_map_eq_self_of_injective M.subtype_injective U'
have hUHall : Ch03.IsHallSubgroup ((S14.kappa M ∪ σ)ᶜ) (U.subgroupOf M) := by
rw [hUof]; exact hU'
-- Lemma 15.1 conclusion (with `Kstar := M_σ ⊓ C_M(K)`).
obtain ⟨_, _, _, hMddσ, hKguard, _, _, _⟩ :=
typeP_auxiliary_structure hG hM (hK ▸ Subgroup.map_subtype_le K')
(hU ▸ Subgroup.map_subtype_le U') hKHall rfl hUHall
-- Conjunct 4 / 10 helper: `M'' ≤ M_σ` (Lemma 15.1, unconditional).
have hMdd_Mσ : derivedInG (derivedInG M) ≤ Mσ := hMddσ
-- ## Case split on whether `M_σ` is nilpotent (`M_F = M_σ`).
by_cases hcase : MF M = Mσ
· -- ### Case I: `M_σ` nilpotent, `M_F = M_σ`, `F(M_σ) = M_σ`.
have hMσnil : Group.IsNilpotent ↥Mσ :=
(maxNilpotentNormalHall_eq_Msigma_iff_isNilpotent hG hM).mp hcase
have hFMσ_eq : FMσ = Mσ := fittingInAmbient_eq_self_of_isNilpotent
-- `M_σ ≤ F(M)` (nilpotent normal subgroup of `M`).
have hMσ_le_F : Mσ ≤ F := by rw [← hFMσ_eq]; rw [h6]; exact le_sup_left
-- `[M_σ, Y] = ⊥`, so `Y` centralizes `M_σ`; together with `Y ≤ M`, `Y ≤ C_G(M_σ) ⊓ M`.
have hMσY : ⁅Mσ, Y⁆ = ⊥ := by rw [← hFMσ_eq]; exact h8
have hY_cent : Y ≤ Subgroup.centralizer (Mσ : Set G) := by
rw [Subgroup.commutator_comm] at hMσY
exact Subgroup.commutator_eq_bot_iff_le_centralizer.mp hMσY
have hY_le_M : Y ≤ M := h3.trans (OddOrder.BG.Ch2.S08.fittingInG_le M)
-- Corollary 15.3(a) at `H := M_σ`: `C_M(M_σ) = (C_G(M_σ) ⊓ M_σ) ⊔ X`, `X` cyclic `τ₂`.
-- Sorry-free via `mf_centralizer_msigma_decomp` (Prop 14.2(b1)(e) + Schur–Zassenhaus +
-- Lemma 15.1(c)); this de-axiomatises the A(8) `FittingIsTI` chain (issue 8016).
obtain ⟨X, hXcyc, hXτ₂, hCeq⟩ := mf_centralizer_msigma_decomp hG hM
-- `C := C_G(M_σ) ⊓ M`, `A := C_G(M_σ) ⊓ M_σ`.
set C : Subgroup G := Subgroup.centralizer (Mσ : Set G) ⊓ M with hCdef
set A : Subgroup G := Subgroup.centralizer (Mσ : Set G) ⊓ Mσ with hAdef
have hY_C : Y ≤ C := le_inf hY_cent hY_le_M
have hA_C : A ≤ C := inf_le_inf_left _ hMσ_le_M
have hX_C : X ≤ C := le_sup_right.trans hCeq.ge
-- `A ⊴ C` (so we can form the cyclic quotient `C/A`).
have hC_norm_A : C ≤ Subgroup.normalizer (A : Set G) := by
have h1 : C ≤ Subgroup.normalizer (Subgroup.centralizer (Mσ : Set G)) :=
inf_le_left.trans Subgroup.le_normalizer
have h2 : C ≤ Subgroup.normalizer (Mσ : Set G) := inf_le_right.trans hM_norm_Mσ
exact (le_inf h1 h2).trans Subgroup.inf_normalizer_le_normalizer_inf
have hA_normal : (A.subgroupOf C).Normal :=
(Subgroup.normal_subgroupOf_iff_le_normalizer_inf).mpr
(by rw [inf_eq_left.mpr hA_C]; exact hC_norm_A)
-- `A ≤ M_σ` is a `σ`-group; `Y` is a `σ'`-group; hence `|A|` and `|Y|` are coprime.
have hA_pi : ∀ r ∈ (Nat.card ↥A).primeFactors, r ∈ σ := fun r hr =>
OddOrder.BG.Ch3.S10.Msigma_isPiGroup M r
(Nat.primeFactors_mono (Subgroup.card_dvd_of_le (inf_le_right : A ≤ Mσ))
Nat.card_pos.ne' hr)
have hY_pi' : ∀ r ∈ (Nat.card ↥Y).primeFactors, r ∉ σ := fun r hr =>
(Set.mem_compl_iff _ _).mp (hYpi r hr)
-- `Y ⊓ A = ⊥` (coprime orders).
have hY_inf_A : Y ⊓ A = ⊥ :=
(Subgroup.disjoint_of_coprime_natCard
((Ch03.Nat.coprime_of_isPiGroup_of_isPiGroup_compl Nat.card_pos.ne' Nat.card_pos.ne'
hA_pi hY_pi').symm)).eq_bot
-- Embed `Y` into the cyclic quotient `C/A` (`C/A` is a quotient image of the cyclic `X`).
have haxtop : A.subgroupOf C ⊔ X.subgroupOf C = ⊤ := by
rw [← Subgroup.subgroupOf_sup hA_C hX_C, show A ⊔ X = C from hCeq.symm,
Subgroup.subgroupOf_self]
have hYc_inf_a : Y.subgroupOf C ⊓ A.subgroupOf C = ⊥ := by
rw [Subgroup.subgroupOf, Subgroup.subgroupOf, ← Subgroup.comap_inf, hY_inf_A,
MonoidHom.comap_bot]
exact C.ker_subtype
have hinj : Function.Injective
((QuotientGroup.mk' (A.subgroupOf C)).comp (Y.subgroupOf C).subtype) := by
rw [← MonoidHom.ker_eq_bot_iff, eq_bot_iff]
intro y hy
rw [MonoidHom.mem_ker, MonoidHom.comp_apply, QuotientGroup.mk'_apply,
QuotientGroup.eq_one_iff] at hy
have hmem : (Y.subgroupOf C).subtype y ∈ Y.subgroupOf C ⊓ A.subgroupOf C := ⟨y.2, hy⟩
rw [hYc_inf_a, Subgroup.mem_bot] at hmem
rw [Subgroup.mem_bot]; exact Subtype.ext hmem
have hxcyc : IsCyclic ↥(X.subgroupOf C) := by
have : IsCyclic ↥X := hXcyc
exact isCyclic_of_surjective _ (Subgroup.subgroupOfEquivOfLe hX_C).symm.surjective
have hquot_cyc : IsCyclic (↥C ⧸ A.subgroupOf C) := by
have hsurj : Function.Surjective
((QuotientGroup.mk' (A.subgroupOf C)).comp (X.subgroupOf C).subtype) := by
rw [← MonoidHom.range_eq_top, MonoidHom.range_comp, Subgroup.range_subtype]
have h1 : (A.subgroupOf C ⊔ X.subgroupOf C).map (QuotientGroup.mk' (A.subgroupOf C)) =
⊤ := by
rw [haxtop, Subgroup.map_top_of_surjective _ (QuotientGroup.mk'_surjective _)]
have hkerbot : (A.subgroupOf C).map (QuotientGroup.mk' (A.subgroupOf C)) = ⊥ := by
rw [Subgroup.map_eq_bot_iff, QuotientGroup.ker_mk']
rw [Subgroup.map_sup, hkerbot, bot_sup_eq] at h1
rw [h1]
exact isCyclic_of_surjective _ hsurj
-- Conjunct 1: `Y` is cyclic (`Y ≅ Y.subgroupOf C ↪ C/A` cyclic).
have hYcyc' : IsCyclic ↥(Y.subgroupOf C) :=
isCyclic_of_surjective _ (MonoidHom.ofInjective hinj).symm.surjective
have hYcyc : IsCyclic ↥Y :=
isCyclic_of_surjective _ (Subgroup.subgroupOfEquivOfLe hY_C).surjective
-- Conjunct 2: `π(Y) ⊆ τ₂` (`q ∣ |Y| ∣ [C:A] ∣ |X|`, `π(X) ⊆ τ₂`).
have hYτ₂ : (↑(Nat.card ↥Y).primeFactors : Set ℕ) ⊆ tau2 M := by
intro q hq
have hqp : q.Prime := Nat.prime_of_mem_primeFactors hq
have hcardYc : Nat.card ↥(Y.subgroupOf C) = Nat.card ↥Y :=
Nat.card_congr (Subgroup.subgroupOfEquivOfLe hY_C).toEquiv
have hdvd1 : Nat.card ↥Y ∣ (A.subgroupOf C).index := by
rw [← hcardYc, Subgroup.index_eq_card]
exact Subgroup.card_dvd_of_injective _ hinj
have hdvd2 : (A.subgroupOf C).index ∣ Nat.card ↥X := by
have hidx : (A.subgroupOf C).index = (A.subgroupOf C).relIndex (X.subgroupOf C) := by
rw [← Subgroup.relIndex_top_right, ← haxtop, Subgroup.relIndex_sup_left]
have hcardx : Nat.card ↥(X.subgroupOf C) = Nat.card ↥X :=
Nat.card_congr (Subgroup.subgroupOfEquivOfLe hX_C).toEquiv
rw [hidx, ← hcardx]
exact Subgroup.relIndex_dvd_card (A.subgroupOf C) (X.subgroupOf C)
have hqX : q ∈ (Nat.card ↥X).primeFactors :=
Nat.mem_primeFactors.mpr ⟨hqp,
((Nat.dvd_of_mem_primeFactors hq).trans hdvd1).trans hdvd2, Nat.card_pos.ne'⟩
exact hXτ₂ hqX
-- Conjunct 4: `M'' ⊆ F(M)` (`M'' ≤ M_σ ≤ F(M)`, `M_σ` nilpotent).
have h4 : derivedInG (derivedInG M) ≤ F := hMdd_Mσ.trans hMσ_le_F
-- Conjunct 5: `F(M) = (C_G(M_F) ⊓ M) ⊔ M_F = (C_G(M_σ) ⊓ M) ⊔ M_σ`.
have h5 : F = (Subgroup.centralizer (MF M : Set G) ⊓ M) ⊔ MF M := by
rw [hcase]
-- `M ≤ N_G(C_G(M_σ))`: normalizing `M_σ` normalizes its centralizer.
have hM_norm_CMσ : M ≤ Subgroup.normalizer (Subgroup.centralizer (Mσ : Set G)) :=
hM_norm_Mσ.trans (normalizer_le_normalizer_centralizer Mσ)
-- `C := C_G(M_σ) ⊓ M ⊴ M`.
have hC_norm : M ≤ Subgroup.normalizer (C : Set G) :=
(le_inf hM_norm_CMσ Subgroup.le_normalizer).trans Subgroup.inf_normalizer_le_normalizer_inf
have hC_normal : (C.subgroupOf M).Normal :=
(Subgroup.normal_subgroupOf_iff_le_normalizer inf_le_right).mpr hC_norm
refine le_antisymm ?_ ?_
· -- `F ⊆ (C_G(M_σ) ⊓ M) ⊔ M_σ`: `F = M_σ ⊔ Y`, `M_σ ≤ M_σ`, `Y ≤ C`.
rw [h6, hFMσ_eq]
exact sup_le (le_sup_right) (hY_C.trans le_sup_left)
· -- `(C_G(M_σ) ⊓ M) ⊔ M_σ ⊆ F`: `M_σ ≤ F`, and `C ⊔ M_σ` is nilpotent normal (`= X ⊔ M_σ`).
refine sup_le ?_ hMσ_le_F
-- `C ⊔ M_σ ⊴ M` and is nilpotent, hence `⊆ F(M)`.
have hCMσ_le_M : C ⊔ Mσ ≤ M := sup_le inf_le_right hMσ_le_M
have hCMσ_norm : ((C ⊔ Mσ).subgroupOf M).Normal := by
rw [Subgroup.normal_subgroupOf_iff_le_normalizer hCMσ_le_M]
exact le_trans (le_inf hC_norm hM_norm_Mσ)
(Subgroup.normalizer_inf_normalizer_le_normalizer_sup C Mσ)
-- `C ⊔ M_σ = X ⊔ M_σ` (since `C = A ⊔ X` and `A ≤ M_σ`).
have hA_le_Mσ : A ≤ Mσ := inf_le_right
have hCMσ_eq : C ⊔ Mσ = X ⊔ Mσ := sup_eq_sup_of_eq_sup_of_le hCeq hA_le_Mσ
-- `X ⊔ M_σ` nilpotent: `X` cyclic, `M_σ` nilpotent, `[X, M_σ] = ⊥` (`X ≤ C_G(M_σ)`).
have hXcent : ⁅X, Mσ⁆ = ⊥ := by
have hXle : X ≤ Subgroup.centralizer (Mσ : Set G) := hX_C.trans inf_le_left
exact Subgroup.commutator_eq_bot_iff_le_centralizer.mpr hXle
have hXcyc' : IsCyclic ↥X := hXcyc
let : CommGroup ↥X := IsCyclic.commGroup
have : Group.IsNilpotent ↥X := CommGroup.isNilpotent
have hCMσ_nil : Group.IsNilpotent ↥(C ⊔ Mσ) := by
rw [hCMσ_eq]; exact isNilpotent_sup_of_commutator_eq_bot hXcent
have := hCMσ_norm
have hCMσ_le_F : C ⊔ Mσ ≤ F :=
le_fittingInAmbient_of_subgroupOf_normal_of_isNilpotent hCMσ_le_M hCMσ_norm
exact le_sup_left.trans hCMσ_le_F
-- Conjunct 10: `¬ TypeF → F(M) ⊆ M'`. `M = K M'`, `M/M' ≅ K` (`κ`-group), `π(Y) ⊆ τ₂`,
-- and `κ ∩ τ₂ = ∅`, so `Y ≤ M'`; with `M_σ ≤ M'` this gives `F ⊆ M'`.
have h10 : ¬ S14.IsTypeF M → F ≤ derivedInG M := by
intro hnotF
have hP : S14.IsTypeP M := by
rw [S14.isTypeF_iff_not_isTypeP] at hnotF; exact not_not.mp hnotF
-- `K ≠ ⊥`: some `κ`-prime divides `|M|`, but a trivial `κ`-Hall would push it to the index.
have hKne : K ≠ ⊥ := by
obtain ⟨p, hpκ⟩ := hP
obtain ⟨hpprime, -, P, hPmem, hPM, -⟩ := id hpκ
have : Fact p.Prime := ⟨hpprime⟩
-- `p ∣ |M|` (a rank-one elementary abelian `p`-subgroup `P ≤ M`).
have hpcardP : Nat.card ↥P = p := by
have := (OddOrder.GroupTheory.mem_elemAbelianOfRank.mp hPmem).2
rwa [pow_one] at this
have hpM : p ∈ (Nat.card ↥M).primeFactors :=
Nat.mem_primeFactors.mpr ⟨hpprime,
hpcardP ▸ Subgroup.card_dvd_of_le hPM, Nat.card_pos.ne'⟩
intro hKbot
-- `K = ⊥` ⟹ `(K.subgroupOf M).index = |↥M|`, so `p` divides the index of the `κ`-Hall.
have hidx : (K.subgroupOf M).index = Nat.card ↥M := by
rw [hKbot, Subgroup.bot_subgroupOf, Subgroup.index_bot]
have hpidx : p ∈ (K.subgroupOf M).index.primeFactors := by rw [hidx]; exact hpM
exact hKHall.2 p hpidx hpκ
obtain ⟨hMderiv, _, hcompl, _⟩ := hKguard hKne
-- `Y ≤ M'`: image of the normal `τ₂`-subgroup `Y` in the abelian `M/M'` (order `|K|`, a
-- `κ`-number) is trivial because `τ₂ ∩ κ = ∅`.
have hY_le_deriv : Y ≤ derivedInG M := by
have hYM : Y ≤ M := h3.trans (OddOrder.BG.Ch2.S08.fittingInG_le M)
set Hsub : Subgroup ↥M := Y.subgroupOf M with hHsub
set D : Subgroup ↥M := (derivedInG M).subgroupOf M with hDdef
have hDcomm : D = commutator ↥M :=
Subgroup.comap_map_eq_self_of_injective M.subtype_injective (commutator ↥M)
have hDnorm : D.Normal := by rw [hDcomm]; infer_instance
-- `[M : M'] = |K|` (complement), a `κ`-number.
have hDindex : D.index = Nat.card ↥(K.subgroupOf M) := hcompl.symm.index_eq_card
-- `Coprime |Y| [M:M']` (`π(Y) ⊆ τ₂`, `π(K) ⊆ κ`, `τ₂ ∩ κ = ∅`).
have hKpi : ∀ r ∈ (Nat.card ↥(K.subgroupOf M)).primeFactors, r ∈ S14.kappa M :=
fun r hr => hKHall.1 r hr
have hcop : Nat.Coprime (Nat.card ↥Hsub) D.index := by
rw [hDindex]
refine Ch03.Nat.coprime_of_isPiGroup_of_isPiGroup_compl Nat.card_pos.ne'
Nat.card_pos.ne' (π := tau2 M) (fun r hr => ?_) (fun r hr => ?_)
· -- `π(Y) ⊆ τ₂`.
have : r ∈ (Nat.card ↥Y).primeFactors := by
rwa [hHsub, Nat.card_congr (Subgroup.subgroupOfEquivOfLe hYM).toEquiv] at hr
exact hYτ₂ this
· -- `κ ∩ τ₂ = ∅`: a `κ`-prime has rank one, a `τ₂`-prime has rank two.
intro hrτ₂
have hrκ : r ∈ S14.kappa M := hKpi r hr
have hr1 : pRank ↥M r = 1 := by
rcases S14.kappa_subset_tau1_union_tau3 hrκ with h | h
· exact ((mem_tau1_iff M r).mp h).2.2
· exact ((mem_tau3_iff M r).mp h).2.2
have hr2 : pRank ↥M r = 2 := ((mem_tau2_iff M r).mp hrτ₂).2
rw [hr1] at hr2; exact absurd hr2 (by norm_num)
-- `Y.subgroupOf M ≤ commutator ↥M`: image in the abelianization is trivial.
have hHnorm : Hsub.Normal :=
(Subgroup.normal_subgroupOf_iff_le_normalizer hYM).mpr hM_norm_Y
have hcard_img_dvd_Y : Nat.card ↥(Hsub.map (QuotientGroup.mk' D)) ∣ Nat.card ↥Hsub :=
Subgroup.card_map_dvd Hsub (QuotientGroup.mk' D)
have hcard_img_dvd_idx : Nat.card ↥(Hsub.map (QuotientGroup.mk' D)) ∣ D.index := by
rw [Subgroup.index_eq_card]
exact Subgroup.card_subgroup_dvd_card _
have hcard_img_one : Nat.card ↥(Hsub.map (QuotientGroup.mk' D)) = 1 :=
Nat.eq_one_of_dvd_coprimes hcop hcard_img_dvd_Y hcard_img_dvd_idx
have himg_bot : Hsub.map (QuotientGroup.mk' D) = ⊥ :=
Subgroup.card_eq_one.mp hcard_img_one
have hHsub_le_D : Hsub ≤ D := by
rw [Subgroup.map_eq_bot_iff, QuotientGroup.ker_mk'] at himg_bot
exact himg_bot
-- Transport back to `G`: `Y ≤ M'`.
calc Y = Hsub.map M.subtype := (Subgroup.map_subgroupOf_eq_of_le hYM).symm
_ ≤ D.map M.subtype := Subgroup.map_mono hHsub_le_D
_ = derivedInG M := Subgroup.map_subgroupOf_eq_of_le (Subgroup.map_subtype_le _)
rw [h6, hFMσ_eq]
exact sup_le ((OddOrder.BG.Ch3.S10.Msigma_le_derived hG hM)) hY_le_deriv
-- Conjunct 11: `M_F` cyclic → `F(M)` cyclic (`F = M_F × Y`, both cyclic coprime).
have h11 : IsCyclic ↥(MF M) → IsCyclic ↥F := by
intro hMFcyc
have := hMFcyc
have := hYcyc
-- `F = M_F ⊔ Y` with `M_F ⊓ Y = ⊥`, `[M_F, Y] = ⊥`, coprime orders.
have hMFY_inf : MF M ⊓ Y = ⊥ := by rw [hcase, ← hFMσ_eq]; exact h7
have hMFY_comm : ⁅(MF M : Subgroup G), Y⁆ = ⊥ := by rw [hcase, ← hFMσ_eq]; exact h8
have hFeq : F = MF M ⊔ Y := by rw [h6, hFMσ_eq, hcase]
have hcop : Nat.Coprime (Nat.card ↥(MF M)) (Nat.card ↥Y) := by
refine Ch03.Nat.coprime_of_isPiGroup_of_isPiGroup_compl Nat.card_pos.ne' Nat.card_pos.ne'
(π := σ) (fun r hr => ?_) hY_pi'
rw [hcase] at hr
exact OddOrder.BG.Ch3.S10.Msigma_isPiGroup M r hr
-- Cyclic product (orderOf approach, mirroring `S06.isCyclic_sup`).
obtain ⟨a, ha⟩ := IsCyclic.exists_generator (α := ↥(MF M))
obtain ⟨b, hb⟩ := IsCyclic.exists_generator (α := ↥Y)
have hMF_le : MF M ≤ MF M ⊔ Y := le_sup_left
have hY_le : Y ≤ MF M ⊔ Y := le_sup_right
have hoa : orderOf (Subgroup.inclusion hMF_le a) = Nat.card ↥(MF M) := by
rw [orderOf_injective _ (Subgroup.inclusion_injective _) a,
orderOf_eq_card_of_forall_mem_zpowers ha]
have hob : orderOf (Subgroup.inclusion hY_le b) = Nat.card ↥Y := by
rw [orderOf_injective _ (Subgroup.inclusion_injective _) b,
orderOf_eq_card_of_forall_mem_zpowers hb]
have hMFnorm_Y : MF M ≤ Subgroup.normalizer (Y : Set G) :=
(Subgroup.commutator_eq_bot_iff_le_centralizer.mp hMFY_comm).trans
(OddOrder.Isaacs.Ch07.centralizer_le_normalizer Y)
have hcardsup : Nat.card ↥(MF M ⊔ Y) = Nat.card ↥(MF M) * Nat.card ↥Y := by
have hprod := Subgroup.card_HK_mul_card_inf_eq_card_mul_card (MF M) Y
rw [hMFY_inf, Subgroup.card_bot, mul_one] at hprod
rwa [show ((MF M : Set G) * (Y : Set G)) = ((MF M ⊔ Y : Subgroup G) : Set G) from
(Subgroup.coe_mul_of_left_le_normalizer_right (MF M) Y hMFnorm_Y).symm] at hprod
have hcomm : Commute (Subgroup.inclusion hMF_le a) (Subgroup.inclusion hY_le b) := by
have hab : ((a : G)) * (b : G) = (b : G) * (a : G) :=
(Subgroup.mem_centralizer_iff.mp
(Subgroup.commutator_eq_bot_iff_le_centralizer.mp hMFY_comm a.2) (b : G) b.2).symm
exact Subtype.ext (by
simp only [Subgroup.coe_mul, Subgroup.coe_inclusion]; exact hab)
rw [hFeq]
refine isCyclic_of_orderOf_eq_card
(Subgroup.inclusion hMF_le a * Subgroup.inclusion hY_le b) ?_
rw [hcomm.orderOf_mul_eq_mul_orderOf_of_coprime (by rw [hoa, hob]; exact hcop),
hoa, hob, hcardsup]
exact ⟨Y, hYcyc, hYτ₂, h3, h4, h5, h6, h7, h8, h9, h10, h11⟩
· -- ### Case II: `M_σ` not nilpotent, `M_F ≠ M_σ`, so `M` is type `P1` and `F(M) ⊆ M_σ`.
obtain ⟨_hP1, Q, _Q0, _D, _p, _q, _, _, _, _, _, _, _, hQsubMF, _, _, _, _, _, _, hMσderiv,
_, hFsubMσ, hFQ, hMFnc⟩ :=
mf_ne_msigma_typeP1_structure hG hM hcase (Subgroup.map_subtype_le K') hKHall rfl
-- In Case II: `F(M) ⊆ M_σ`, so `Y = O_{σ'}(F(M)) = ⊥` and `F(M) = F(M_σ)`.
-- `Y = ⊥`: `F(M) ⊆ M_σ` is a `σ`-group, so its `σ'`-Hall core is trivial.
have hYbot : Y = ⊥ := by
rw [hY]
refine OddOrder.GroupTheory.opiCoreInG_compl_eq_bot_of_isPiSubgroup ?_
intro r hr
exact OddOrder.BG.Ch3.S10.Msigma_isPiGroup M r
(Nat.primeFactors_mono (Subgroup.card_dvd_of_le hFsubMσ) Nat.card_pos.ne' hr)
-- `F(M) = F(M_σ)`: `F(M) ◁ M` nilpotent `⊆ M_σ ⟹ ◁ M_σ ⟹ ⊆ F(M_σ)`; `F(M_σ) ⊆ F(M)` (Lemma 1).
have hFMσ_eq : F = FMσ := by
refine le_antisymm ?_ ?_
· -- `F ⊆ M_σ`, `F ◁ M_σ` (since `F ◁ M`), `F` nilpotent ⟹ `F ⊆ F(M_σ)`.
have hF_norm_Mσ : (F.subgroupOf Mσ).Normal :=
(Subgroup.normal_subgroupOf_iff_le_normalizer hFsubMσ).mpr
(hMσ_le_M.trans ((Subgroup.normal_subgroupOf_iff_le_normalizer
(OddOrder.BG.Ch2.S08.fittingInG_le M)).mp
(OddOrder.BG.Ch2.S08.fittingInG_subgroupOf_normal M)))
exact le_fittingInAmbient_of_subgroupOf_normal_of_isNilpotent hFsubMσ hF_norm_Mσ
· -- `F(M_σ) ⊆ O_σ(F(M)) ⊆ F(M)`.
rw [← hL1]; exact OddOrder.GroupTheory.opiCoreInG_le σ F
-- Now assemble. With `Y = ⊥`, conjuncts (a) and the `× Y` split collapse.
have hYcyc : IsCyclic ↥Y := by rw [hYbot]; infer_instance
have hYτ₂ : (↑(Nat.card ↥Y).primeFactors : Set ℕ) ⊆ tau2 M := by
rw [hYbot, Subgroup.card_bot]; simp
-- Conjunct 4: `M'' ⊆ F(M)`.
have h4 : derivedInG (derivedInG M) ≤ F := by
have hMdd_F : derivedInG (derivedInG M) ≤ fittingInAmbient M := ‹_›
exact hMdd_F
-- Conjunct 5 (Case II, `M_F ≠ M_σ`): `F(M) = (C_G(M_F) ⊓ M) ⊔ M_F` (mmd 15.2(g)
-- "(b) `F(M) = C_M(H)H`" with `H = M_F`).
have h5 : F = (Subgroup.centralizer (MF M : Set G) ⊓ M) ⊔ MF M := by
have hMF_le_F : MF M ≤ F := maxNilpotentNormalHall_le_fittingInG M
have hQ_le_MF : Q ≤ MF M := hQsubMF
-- `C_M(M_F) ⊆ C_M(Q) ⊆ Q ⊔ C_M(Q) = F(M)` (Theorem 15.2(g) equality `hFQ`).
have hCMF_le_F : Subgroup.centralizer (MF M : Set G) ⊓ M ≤ F := by
have hsub : Subgroup.centralizer (MF M : Set G) ⊓ M ≤
Subgroup.centralizer (Q : Set G) ⊓ M := by
refine inf_le_inf_right _ ?_
intro x hx
rw [Subgroup.mem_centralizer_iff] at hx ⊢
exact fun g hg => hx g (hQ_le_MF hg)
rw [hF, hFQ]; exact hsub.trans le_sup_right
refine le_antisymm ?_ (sup_le hCMF_le_F hMF_le_F)
-- `⊆` (mmd 15.2(g)): the `σ'`-free, type-`P1` structural step `F(M) ⊆ C_M(M_F)·M_F`.
-- Strategy (general, §14-independent): `M_F` is the full Hall `π(M_F)`-part of `F(M)`, so the
-- nilpotent `F(M)` splits as `F(M) = M_F × O_{π(M_F)'}(F(M))`, and the second factor
-- centralizes `M_F` (distinct Hall components of a nilpotent group commute).
set π : Set ℕ := ↑(Nat.card ↥(MF M)).primeFactors with hπ
-- `F ≤ M` and `M ≤ N_G(F)` (`F` is normal in `M`).
have hF_le_M : F ≤ M := OddOrder.BG.Ch2.S08.fittingInG_le M
have hM_norm_F : M ≤ Subgroup.normalizer (F : Set G) :=
(Subgroup.normal_subgroupOf_iff_le_normalizer (OddOrder.BG.Ch2.S08.fittingInG_le M)).mp
(OddOrder.BG.Ch2.S08.fittingInG_subgroupOf_normal M)
-- `M_F ≤ O_π(F)`: `M_F ≤ F`, `(M_F).subgroupOf F ⊴ F` (as `F ≤ M ≤ N_G(M_F)`), `M_F` a
-- `π`-group.
have hMF_norm_F : (((MF M).subgroupOf F)).Normal :=
(Subgroup.normal_subgroupOf_iff_le_normalizer hMF_le_F).mpr
(hF_le_M.trans (maxNilpotentNormalHall_le_normalizer M))
have hMF_pi : Subgroup.IsPiSubgroup π (MF M) := fun p hp => by
rw [hπ]; exact Finset.mem_coe.mpr hp
have hMF_le_Oπ : MF M ≤ OddOrder.GroupTheory.opiCoreInG π F :=
OddOrder.GroupTheory.le_opiCoreInG_of_normal_of_isPiSubgroup hMF_le_F hMF_norm_F hMF_pi
-- `O_π(F) ≤ M_F`: `O_π(F)` is a normal `π`-subgroup of `↥M`, and `M_F` is `π`-Hall in `↥M`.
have hOπ_le_F : OddOrder.GroupTheory.opiCoreInG π F ≤ F :=
OddOrder.GroupTheory.opiCoreInG_le π F
have hOπ_le_M : OddOrder.GroupTheory.opiCoreInG π F ≤ M := hOπ_le_F.trans hF_le_M
have hObar_norm : ((OddOrder.GroupTheory.opiCoreInG π F).subgroupOf M).Normal :=
(Subgroup.normal_subgroupOf_iff_le_normalizer hOπ_le_M).mpr
(OddOrder.GroupTheory.le_normalizer_opiCoreInG_of_le_normalizer π hM_norm_F)
have hObar_pi :
Ch03.Subgroup.IsPiGroup π ((OddOrder.GroupTheory.opiCoreInG π F).subgroupOf M) := by
intro p hp
have hcardO : Nat.card ↥((OddOrder.GroupTheory.opiCoreInG π F).subgroupOf M) =
Nat.card ↥(OddOrder.GroupTheory.opiCoreInG π F) :=
Nat.card_congr (Subgroup.subgroupOfEquivOfLe hOπ_le_M).toEquiv
exact OddOrder.GroupTheory.isPiSubgroup_opiCoreInG π F p (by rwa [hcardO] at hp)
have hObar_le_Hbar :
(OddOrder.GroupTheory.opiCoreInG π F).subgroupOf M ≤ (MF M).subgroupOf M :=
Ch03.Subgroup.IsPiGroup.normal_le_hall hObar_pi (maxNilpotentNormalHall_isHall M)
have hMF_le_M : MF M ≤ M := maxNilpotentNormalHall_le M
have hOπ_le_MF : OddOrder.GroupTheory.opiCoreInG π F ≤ MF M := by
have := Subgroup.map_mono (f := M.subtype) hObar_le_Hbar
rwa [Subgroup.map_subgroupOf_eq_of_le hOπ_le_M,
Subgroup.map_subgroupOf_eq_of_le hMF_le_M] at this
have hOπ_eq_MF : OddOrder.GroupTheory.opiCoreInG π F = MF M :=
le_antisymm hOπ_le_MF hMF_le_Oπ
-- `F = O_π(F) ⊔ O_{π'}(F) = M_F ⊔ O_{π'}(F)`.
have hsplit : OddOrder.GroupTheory.opiCoreInG π F ⊔
OddOrder.GroupTheory.opiCoreInG πᶜ F = F :=
opiCoreInG_sup_compl_eq_of_isNilpotent π
-- `O_{π'}(F)` centralizes `M_F = O_π(F)`, and lies in `M`, so `≤ C_G(M_F) ⊓ M`.
have hcomm : ⁅OddOrder.GroupTheory.opiCoreInG π F,
OddOrder.GroupTheory.opiCoreInG πᶜ F⁆ = ⊥ :=
OddOrder.BG.Ch2.S08.opiCoreInG_commutator_compl_eq_bot π F
have hOπ'_cent : OddOrder.GroupTheory.opiCoreInG πᶜ F ≤
Subgroup.centralizer (MF M : Set G) := by
have hcomm' : ⁅OddOrder.GroupTheory.opiCoreInG πᶜ F,
OddOrder.GroupTheory.opiCoreInG π F⁆ = ⊥ := by
rw [Subgroup.commutator_comm]; exact hcomm
rw [← hOπ_eq_MF]
exact Subgroup.commutator_eq_bot_iff_le_centralizer.mp hcomm'
have hOπ'_le_M : OddOrder.GroupTheory.opiCoreInG πᶜ F ≤ M :=
(OddOrder.GroupTheory.opiCoreInG_le πᶜ F).trans hF_le_M
have hOπ'_le : OddOrder.GroupTheory.opiCoreInG πᶜ F ≤
Subgroup.centralizer (MF M : Set G) ⊓ M := le_inf hOπ'_cent hOπ'_le_M
-- Assemble: `F = M_F ⊔ O_{π'}(F) ≤ (C_G(M_F) ⊓ M) ⊔ M_F`.
calc F = OddOrder.GroupTheory.opiCoreInG π F ⊔
OddOrder.GroupTheory.opiCoreInG πᶜ F := hsplit.symm
_ = MF M ⊔ OddOrder.GroupTheory.opiCoreInG πᶜ F := by rw [hOπ_eq_MF]
_ ≤ MF M ⊔ (Subgroup.centralizer (MF M : Set G) ⊓ M) := sup_le_sup_left hOπ'_le _
_ = (Subgroup.centralizer (MF M : Set G) ⊓ M) ⊔ MF M := sup_comm _ _
-- Conjunct 10: `¬ TypeF → F(M) ⊆ M'`. `F(M) ⊆ M_σ = M'` (Theorem 15.2).
have h10 : ¬ S14.IsTypeF M → F ≤ derivedInG M := fun _ => hFsubMσ.trans hMσderiv.le
-- Conjunct 11: `M_F` cyclic → `F(M)` cyclic. Vacuous: `M_F` is non-cyclic (Theorem 15.2).
have h11 : IsCyclic ↥(MF M) → IsCyclic ↥F := fun h => absurd h hMFnc
exact ⟨Y, hYcyc, hYτ₂, h3, h4, h5, h6, h7, h8, h9, h10, h11⟩
/-- **§14-independent assembly of BG Corollary 15.6** from its §14/§15 inputs taken as
hypotheses. This packages the *logic* of Corollary 15.6 (mmd L4232) with no citation of the §14
scaffold at all — it was written while §14 was still a `sorry`-stated skeleton. §14 has since
landed, and `typeP_kstar_in_mf` discharges each hypothesis by a single citation and applies this
skeleton. Hypothesis provenance (mmd L4434 dependency table):
* `hKne` (`K* ≠ 1`) ← Proposition 14.2(c) (`typeP_structure`, conjunct `Kstar ≠ ⊥`);
* `hcyc` (`K K*` cyclic) ← Theorem 14.7(d) (`typeP_duality`, conjunct `IsCyclic (K ⊔ Kstar)`);
* `hKsubMF` (`K* ⊆ M_F`) ← Theorem 15.2(b)(c) (case-split on `M_F = M_σ`);
* `hcompl`/`hcop` (`M = K M'`, `K ∩ M' = 1`, coprime) ← Theorem 14.7(h) / Lemma 15.1's `K ≠ 1`
clause;
* `hFcyc` (`M_F` cyclic ⟹ `F(M)` cyclic) ← Corollary 15.5 (the consequence its proof of 15.6
cites).
The two nontrivial steps are unconditional: `K* ⊆ M''`
(`Msigma_inf_centralizer_le_derivedDerived_of_isComplement'`) and the `M_F`-not-cyclic
contradiction (`fittingInAmbient_cyclic_imp_derivedDerived_eq_bot`, giving `M'' = 1`, against
`K* ⊆ M''` and `K* ≠ 1`). -/
theorem typeP_kstar_in_mf_of_inputs [Finite G] (hG : OddOrder.BG.IsMinimalSimpleOdd G)
{M K Kstar : Subgroup G} (hM : M ∈ maximalSubgroups G)
(hKstar : Kstar = OddOrder.BG.Ch3.S10.Msigma M ⊓ Subgroup.centralizer (K : Set G))
(hKne : Kstar ≠ ⊥) (hcyc : IsCyclic ↥(K ⊔ Kstar)) (hKsubMF : Kstar ≤ MF M)
(hcompl : Subgroup.IsComplement' ((derivedInG M).subgroupOf M) (K.subgroupOf M))
(hcop : Nat.Coprime (Nat.card ↥((derivedInG M).subgroupOf M))
(Nat.card ↥(K.subgroupOf M)))
(hFcyc : IsCyclic ↥(MF M) → IsCyclic ↥(fittingInAmbient M)) :
Kstar ≠ ⊥ ∧ IsCyclic ↥Kstar ∧ Kstar ≤ MF M ∧
Kstar ≤ derivedInG (derivedInG M) ∧ ¬ IsCyclic ↥(MF M) := by
have : Group.IsSolvable ↥M := hG.isSolvable_of_mem_maximalSubgroups hM
have := hcyc
-- `K* ⊆ M''`: the §14-independent conjunct-4 engine.
have hKsubdd : Kstar ≤ derivedInG (derivedInG M) := by
rw [hKstar]
exact Msigma_inf_centralizer_le_derivedDerived_of_isComplement' hG hM hcompl hcop
refine ⟨hKne, Subgroup.isCyclic_of_le (le_sup_right : Kstar ≤ K ⊔ Kstar), hKsubMF,
hKsubdd, ?_⟩
-- `M_F` not cyclic: else `F(M)` cyclic ⟹ `M'' = 1`, but `K* ⊆ M''` and `K* ≠ 1`.
intro hcycMF
have hMdd : derivedInG (derivedInG M) = ⊥ :=
fittingInAmbient_cyclic_imp_derivedDerived_eq_bot (hFcyc hcycMF)
exact hKne (le_bot_iff.mp (hMdd ▸ hKsubdd))
/-- **BG Corollary 15.6** (mmd L4174): for a type-P maximal subgroup, `Kstar` is
nontrivial cyclic and lies in `M_F`, while `M_F` itself is not cyclic. -/
theorem typeP_kstar_in_mf [Finite G] (hG : OddOrder.BG.IsMinimalSimpleOdd G)
{M K Kstar : Subgroup G} (hM : M ∈ maximalSubgroups G) (hP : S14.IsTypeP M)
(hKM : K ≤ M)
(hK : Ch03.IsHallSubgroup (S14.kappa M) (K.subgroupOf M))
(hKstar : Kstar = OddOrder.BG.Ch3.S10.Msigma M ⊓ Subgroup.centralizer (K : Set G)) :
Kstar ≠ ⊥ ∧ IsCyclic ↥Kstar ∧ Kstar ≤ MF M ∧
Kstar ≤ derivedInG (derivedInG M) ∧ ¬ IsCyclic ↥(MF M) := by
classical
have : Group.IsSolvable ↥M := hG.isSolvable_of_mem_maximalSubgroups hM
-- A Hall `(κ ∪ σ)ᶜ`-subgroup `U` of `M` exists by solvability (Hall's theorem); this is the
-- `U`-factor of the type-`P` decomposition `M = K U M_σ` needed to invoke Proposition 14.2.
obtain ⟨U', hU'⟩ := Ch03.hall_E_exists (G := ↥M)
((S14.kappa M ∪ OddOrder.BG.Ch3.S10.sigma M)ᶜ)
have hUeq : (U'.map M.subtype).subgroupOf M = U' :=
Subgroup.comap_map_eq_self_of_injective M.subtype_injective U'
have hU : Ch03.IsHallSubgroup ((S14.kappa M ∪ OddOrder.BG.Ch3.S10.sigma M)ᶜ)
((U'.map M.subtype).subgroupOf M) := by rw [hUeq]; exact hU'
-- `K* ≠ 1`: Proposition 14.2(c) (`typeP_structure`).
have hKne : Kstar ≠ ⊥ := (typeP_structure hG hM hP hKM hK hKstar hU).2.1
-- `K K*` cyclic and the `M = K M'` complement / coprime data: Theorem 14.7(d),(h).
obtain ⟨hcompl, hcop, _Mstar, ⟨_, _, _, _, hcyc, _, _, _⟩, _⟩ :=
typeP_duality hG hM hP hKM hK hKstar
-- `K* ⊆ M_F`: Theorem 15.2 when `M_F ≠ M_σ`, else `K* ⊆ M_σ = M_F` directly.
have hKsubMF : Kstar ≤ MF M := by
by_cases hMF : MF M = OddOrder.BG.Ch3.S10.Msigma M
· rw [hKstar, hMF]; exact inf_le_left
· obtain ⟨_, _, _, _, _, _, _, _, _, _, _, _, hk, _⟩ :=
mf_ne_msigma_typeP1_structure hG hM hMF hKM hK hKstar
exact hk
-- `M_F` cyclic ⟹ `F(M)` cyclic: Corollary 15.5 (`fitting_decomposition`, last conjunct).
obtain ⟨_, _, _, _, _, _, _, _, _, _, _, hFcyc⟩ := fitting_decomposition hG hM
exact typeP_kstar_in_mf_of_inputs hG hM hKstar hKne hcyc hKsubMF hcompl hcop hFcyc
end OddOrder.BG.Ch4.S15
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/Ch4_FamilyOfMaximal/S15_MF/Corollary155.lean