Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Peterfalvi (6.5): reduction to a non-abelian ppp-group, application of (6.8)

Definition
OddOrder_Peterfalvi_S08_PGroupReduction

by burkh4rt · Nov 12, 2026 · Mathlib 0df444a (Lean v4.33.1)

feit-thompsonfinite-groupsgroup-theoryishida-odd-order

Peterfalvi, Character Theory for the Odd Order Theorem, §6, (6.5). The group-theoretic core of (6.5)(b), that a finite nilpotent group with a fixed-point-free action on its abelianization and the ≤4∣⋅∣2+1\le4|\cdot|^2+1≤4∣⋅∣2+1 index bound is a ppp-group, is already formalized abstractly; this file applies the (6.8) coherence theorem to reach it.

The bundle includes caseAB_split_of_c2 (the (6.8) case A/B split).

The bundle also proves 15 supporting lemmas used by later statements.

Definition code
/-
Copyright (c) 2026 The Odd Order Project. 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/Peterfalvi/S08_PGroupReduction.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_Peterfalvi_S08_CaseBSeedGlue
import Definitions.Def_OddOrder_Peterfalvi_S08_CoherenceCore
import Definitions.Def_OddOrder_Isaacs_Ch04_Commutators_Main_CommutatorBasics
import Definitions.Def_OddOrder_Isaacs_Ch04_Commutators_Main_ThreeSubgroupsCoprime
import Definitions.Def_OddOrder_Isaacs_Ch04_Commutators_Main_BaerTrick
import Definitions.Def_OddOrder_Isaacs_Ch04_Commutators_Main_ChainNilpotent
import Definitions.Def_OddOrder_Isaacs_Ch04_Commutators_ForwardFromCh03
import Definitions.Def_OddOrder_Isaacs_Ch04_Commutators_Main_ThreeSubgroups
import Definitions.Def_OddOrder_Peterfalvi_S08_Theorem62_63_Standalone

namespace OddOrder.Peterfalvi.S08

open OddOrder.RepresentationTheory

variable {G : Type*} [Group G] [Fintype G] [Invertible (Nat.card G : ℂ)]

variable {L : Subgroup G} [Fintype ↥L] [Invertible (Nat.card ↥L : ℂ)]

variable {H : Subgroup ↥L} [Invertible (Nat.card ↥H : ℂ)]

/-- Any natural dividing `|L|` is odd (since `|L|` is odd, Peterfalvi (6.4.a)). -/
theorem odd_of_dvd_card_L (hyp : SibleyDadeHypothesis G L H) {a : ℕ} (ha : a ∣ Nat.card ↥L) :
    Odd a := by
  rcases Nat.even_or_odd a with he | ho
  · obtain ⟨c, hc⟩ := ha
    exact absurd (hc ▸ he.mul_right c) (Nat.not_even_iff_odd.mpr hyp.card_L_odd)
  · exact ho

/-- `|W₁|` is odd. -/
theorem card_W1_odd (hyp : SibleyDadeHypothesis G L H) : Odd (Nat.card ↥hyp.W1) :=
  odd_of_dvd_card_L hyp (Subgroup.card_subgroup_dvd_card hyp.W1)

/-- `|Abelianization H| = |H : ⁅H,H⁆|` is odd (it divides `|H|`, which divides the odd `|L|`). -/
theorem card_abelianization_H_odd (hyp : SibleyDadeHypothesis G L H) :
    Odd (Nat.card (Abelianization ↥H)) :=
  odd_of_dvd_card_L hyp
    ((Subgroup.card_quotient_dvd_card (commutator ↥H)).trans (Subgroup.card_subgroup_dvd_card H))

/-- **Peterfalvi (6.5)(b) reduction — (6.8)(c2) certain-type case, in the Sibley context.**

In case (6.8)(c2) the complement `W₁` does *not* act fixed-point-freely on `H`, but the
`W₁`-conjugation fixed points lie in `⁅H,H⁆` (the certain-type centralizer condition `hfix`, phrased
via the explicit `L`-conjugation `w x w⁻¹ = x`).  Together with the Hall coprimality `hcop` and the
(6.2)/(6.3) index bound `hbound`, the nilpotent `H` is a `p`-group.

This wires `isPGroup_of_isNilpotent_of_coprime_fixedPoints_le_commutator` into the capstone via the
`W₁`-conjugation action, isolating `hfix` (the (4.6)/(6.4.c) fixed-point condition) and `hbound`
(the
contrapositive of Theorem (6.3)) as the residual obligations of the (6.8)(c2) `p`-group
reduction. -/
theorem exists_isPGroup_H_of_c2_of_card_le (hyp : SibleyDadeHypothesis G L H)
    (hcop : Nat.Coprime (Nat.card ↥H) (Nat.card ↥hyp.W1))
    (hfix : ∀ w : ↥hyp.W1, w ≠ 1 → ∀ x : ↥H,
      (w : ↥L) * (x : ↥L) * (w : ↥L)⁻¹ = (x : ↥L) → x ∈ commutator ↥H)
    (hbound : Nat.card (Abelianization ↥H) ≤ 4 * Nat.card ↥hyp.W1 ^ 2 + 1) :
    ∃ p : ℕ, Nat.Prime p ∧ IsPGroup p ↥H := by
  let : H.Normal := hyp.H_normal
  let : Group.IsNilpotent ↥H := hyp.H_nilpotent
  -- the `W₁`-conjugation action on `H`.
  let actH : MulDistribMulAction ↥hyp.W1 ↥H :=
    MulDistribMulAction.compHom H ((MulAut.conjNormal (H := H)).comp hyp.W1.subtype)
  have hsmulH : ∀ (a : ↥hyp.W1) (x : ↥H),
      ((a • x : ↥H) : ↥L) = (a : ↥L) * (x : ↥L) * (a : ↥L)⁻¹ := by
    intro a x
    have h1 : (a • x : ↥H) = (MulDistribMulAction.toMulAut ↥hyp.W1 ↥H a) x := by
      simp [MulDistribMulAction.toMulAut_apply]
    rw [h1]
    change ((MulAut.conjNormal (H := H) (hyp.W1.subtype a)) x : ↥L) = _
    rw [MulAut.conjNormal_apply]; rfl
  -- convert the explicit-conjugation `hfix` to the smul form the reduction lemma consumes.
  have hfix' : ∀ w : ↥hyp.W1, w ≠ 1 → ∀ x : ↥H, w • x = x → x ∈ commutator ↥H := by
    intro w hw x hwx
    refine hfix w hw x ?_
    have hb := hsmulH w x
    rw [hwx] at hb
    exact hb.symm
  exact isPGroup_of_isNilpotent_of_coprime_fixedPoints_le_commutator hcop.symm hfix'
    (card_abelianization_H_odd hyp) (card_W1_odd hyp) hbound

/-- A `p`-group factor of the odd group `H` (`p` prime, `H ≠ ⊥`) has `p ≥ 3`: `p ∣ |H| ∣ |L|` and
`|L|` is odd, so `p` is an odd prime. -/
theorem three_le_of_isPGroup_H (hyp : SibleyDadeHypothesis G L H)
    {p : ℕ} (hp : p.Prime) (hHp : IsPGroup p ↥H) : 3 ≤ p := by
  have : Fact p.Prime := ⟨hp⟩
  obtain ⟨n, hn⟩ := hHp.exists_card_eq
  have hn0 : n ≠ 0 := by
    rintro rfl
    rw [pow_zero] at hn
    exact (Finite.one_lt_card_iff_nontrivial.mpr
      ((Subgroup.nontrivial_iff_ne_bot H).mpr hyp.H_ne_bot)).ne' hn
  have hpH : p ∣ Nat.card ↥H := hn ▸ dvd_pow_self p hn0
  have hpL : p ∣ Nat.card ↥L := hpH.trans H.card_subgroup_dvd_card
  have hpodd : Odd p := hyp.card_L_odd.of_dvd_nat hpL
  have h2 : 2 ≤ p := hp.two_le
  rcases Nat.lt_or_ge p 3 with h | h
  · interval_cases p
    · exact absurd hpodd (by decide)
  · exact h

/-- **A finite `p`-group with cyclic abelianization is cyclic** (Burnside-basis, rank-1 case).
If `P ⧸ ⁅P,P⁆` is cyclic, a lift `x` of a generator satisfies `⟨x⟩ ⊔ ⁅P,P⁆ = ⊤`; since
`⁅P,P⁆ ≤ Φ(P)` (`commutator_le_frattini_of_pgroup`), the non-generating property
`frattini_nongenerating` upgrades this to `⟨x⟩ ⊔ Φ(P) = ⊤`, forcing `⟨x⟩ = ⊤`, i.e. `P` cyclic. -/
theorem isCyclic_of_isPGroup_of_isCyclic_quotient_commutator
    {P : Type*} [Group P] [Finite P] {p : ℕ} [Fact p.Prime] (hP : IsPGroup p P)
    (hcyc : IsCyclic (P ⧸ commutator P)) : IsCyclic P := by
  obtain ⟨q, hq⟩ := hcyc.exists_generator
  obtain ⟨x, rfl⟩ := QuotientGroup.mk'_surjective (commutator P) q
  -- `⟨x⟩ ⊔ ⁅P,P⁆ = ⊤`: every `g` is `x^k` times a commutator, using that `mk x` generates.
  have htop : Subgroup.zpowers x ⊔ commutator P = ⊤ := by
    rw [eq_top_iff]
    intro g _
    obtain ⟨k, hk⟩ := hq (QuotientGroup.mk' (commutator P) g)
    have heq : (QuotientGroup.mk (x ^ k) : P ⧸ commutator P) = QuotientGroup.mk g := by
      have h := (map_zpow (QuotientGroup.mk' (commutator P)) x k).trans hk
      simpa only [QuotientGroup.mk'_apply] using h
    have hmem : (x ^ k)⁻¹ * g ∈ commutator P := QuotientGroup.eq.mp heq
    have hg : g = x ^ k * ((x ^ k)⁻¹ * g) := by group
    rw [hg]
    exact mul_mem (Subgroup.mem_sup_left (Subgroup.mem_zpowers_iff.mpr ⟨k, rfl⟩))
      (Subgroup.mem_sup_right hmem)
  -- upgrade `⁅P,P⁆` to `Φ(P)` and use non-generation.
  have hfr : Subgroup.zpowers x ⊔ frattini P = ⊤ :=
    top_le_iff.mp (htop ▸ sup_le_sup_left
      (OddOrder.Isaacs.Ch04.commutator_le_frattini_of_pgroup hP) (Subgroup.zpowers x))
  have hxtop : Subgroup.zpowers x = ⊤ := frattini_nongenerating hfr
  have hgen : ∀ y : P, y ∈ Subgroup.zpowers x := fun y => by rw [hxtop]; exact Subgroup.mem_top y
  exact ⟨⟨x, hgen⟩⟩

/-- **A non-abelian finite `p`-group has `p² ≤ |P : P′|`** — the Burnside-basis input to
Peterfalvi (6.5)(c).  If `P` is a `p`-group that is not commutative, then `p² ≤ |Abelianization P|`.

`|Abelianization P|` is a power `pⁿ`; if it were `< p²` then `pⁿ ∣ p`, so the abelianization is
cyclic (`isCyclic_of_card_dvd_prime`), whence `P` is cyclic
(`isCyclic_of_isPGroup_of_isCyclic_quotient_commutator`) and therefore abelian — contradicting
non-commutativity. -/
theorem sq_le_card_abelianization_of_isPGroup_of_noncomm
    {P : Type*} [Group P] [Finite P] {p : ℕ} (hp : p.Prime) (hP : IsPGroup p P)
    (hnc : ¬ ∀ a b : P, a * b = b * a) :
    p ^ 2 ≤ Nat.card (Abelianization P) := by
  have : Fact p.Prime := ⟨hp⟩
  by_contra hlt
  rw [not_le] at hlt
  have hpg : IsPGroup p (Abelianization P) := hP.to_quotient (commutator P)
  obtain ⟨n, hn⟩ := hpg.exists_card_eq
  have hn2 : n < 2 := by
    by_contra h; rw [not_lt] at h
    exact absurd (hn.symm ▸ Nat.pow_le_pow_right hp.pos h) (not_le.mpr hlt)
  have : IsCyclic (Abelianization P) :=
    isCyclic_of_card_dvd_prime (p := p)
      (by rw [hn]; exact (pow_dvd_pow p (by omega : n ≤ 1)).trans (dvd_of_eq (pow_one p)))
  have hcyc : IsCyclic P :=
    isCyclic_of_isPGroup_of_isCyclic_quotient_commutator hP this
  obtain ⟨x, hx⟩ := hcyc.exists_generator
  refine hnc fun a b => ?_
  obtain ⟨m, hm⟩ := hx a
  obtain ⟨l, hl⟩ := hx b
  rw [← hm, ← hl, ← zpow_add, ← zpow_add, add_comm]

/-- **Peterfalvi (6.5)(c) coherence producer** (abstract (6.1)/(5.2) form, over the `S07` setup).

The (6.5)(c) conclusion "`S = SOf ⊥` is coherent" for a nilpotent **Frobenius kernel** `H` (kernel
of
`hF : IsFrobeniusGroup ↥L H C`) with the divisibility `|L:H| ∣ p − 1` for every `p ∣ |H|` (`hdvd`),
assembled over the abstract `S07` coherence setup (`τ`, `A0`, filtration `SOf`).  By contradiction:
if `SOf ⊥` is not coherent, then `six_three_of_six_two_oracle` (with `M = ⊥`, `H₁ = ⁅H,H⁆`, `K = H`)
forces `|H:H′| ≤ 4|L:H|² + 1` (else it would produce coherence).  The **(6.5.b) reduction** is done
*here*: from the Frobenius structure + that index bound, `isPGroup_of_isFrobeniusGroup_of_card_le`
produces the prime `p` with `IsPGroup p H`; a non-abelian `p`-group has `p² ≤ |H:H′|`
(`sq_le_card_abelianization_of_isPGroup_of_noncomm`), and `six_five_c_arith` turns `|L:H| ∣ p − 1`
into a contradiction.  (The caller supplies non-abelianness `hnonab`; the abelian case is handled
upstream — `H` abelian ⟹ `⁅H,H⁆ = ⊥` ⟹ `SOf ⁅H,H⁆ = SOf ⊥`, coherent by `hcoh`.)

This is the honest (6.5)(c) assembly engine: the genuine residual obligations are exactly the
parameters — the **(5.6) break-member oracle** `h56` (the remaining character-theoretic core), the
**`S(H′)` coherence** `hcoh` (constant-degree (5.7)), and the filtration `SOf`.  Everything numeric
—
the (6.3) descent, the (6.5.b) `p`-group reduction, the `p²`-index gap, and the (6.5.c) arithmetic —
is proved here.  Oddness side-conditions are discharged from `|L|` odd (`hLodd`). -/
theorem nonempty_coherent_SOf_bot_of_index_dvd [Finite ↥L] [Group.IsNilpotent ↥H]
    (hHnorm : H.Normal)
    (τ : OddOrder.Peterfalvi.S07.IntegralCharacterMap ↥L G) (A0 : Set ↥L)
    (SOf : Subgroup ↥L → Set (ClassFunction ↥L ℂ))
    {C : Subgroup ↥L} (hF : OddOrder.Isaacs.Ch06.IsFrobeniusGroup ↥L H C)
    (hnonab : ¬ ∀ a b : ↥H, a * b = b * a)
    (hLodd : Odd (Nat.card ↥L))
    (hdvd : ∀ p : ℕ, p.Prime → p ∣ Nat.card ↥H → H.index ∣ p - 1)
    (hH'lt : (⁅H, H⁆ : Subgroup ↥L) < H)
    (hcoh : Nonempty (OddOrder.Peterfalvi.S07.IsCoherent τ (SOf ⁅H, H⁆) A0))
    (h56 : ∀ (A B : Subgroup ↥L) [A.Normal] [B.Normal], B ≤ A → A ≤ ⁅H, H⁆ →
        (A.subgroupOf H).map (QuotientGroup.mk' (B.subgroupOf H)) ≤
          Subgroup.center (↥H ⧸ B.subgroupOf H) →
        Nonempty (OddOrder.Peterfalvi.S07.IsCoherent τ (SOf A) A0) →
        ¬ Nonempty (OddOrder.Peterfalvi.S07.IsCoherent τ (SOf B) A0) →
        ∃ θ : IrreducibleCharacter ↥H,
          (↑(B.subgroupOf H) : Set ↥H) ⊆ OddOrder.Peterfalvi.S03.characterKernel
              (θ : ClassFunction ↥H ℂ) ∧
          (Nat.card (↥H ⧸ A.subgroupOf H) : ℝ) - 1 ≤
              2 * (ClassFunction.induce H (θ : ClassFunction ↥H ℂ) 1).re) :
    Nonempty (OddOrder.Peterfalvi.S07.IsCoherent τ (SOf ⊥) A0) := by
  have := hHnorm
  by_contra hncoh
  -- If `|H:H′| > 4|L:H|² + 1`, the (6.3) oracle produces coherence — contradiction.
  have hle : Nat.card (↥H ⧸ (⁅H, H⁆ : Subgroup ↥L).subgroupOf H) ≤ 4 * H.index ^ 2 + 1 := by
    by_contra hgt
    rw [not_le] at hgt
    exact hncoh (six_three_of_six_two_oracle hHnorm bot_le hH'lt le_rfl τ A0 SOf h56 hcoh hgt)
  -- `|L:H| = |C|` (Frobenius complement); convert to the abelianization/complement bound.
  have hidxC : H.index = Nat.card ↥C := hF.isComplement.symm.index_eq_card
  have hbound : Nat.card (Abelianization ↥H) ≤ 4 * Nat.card ↥C ^ 2 + 1 := by
    have hle' := hle
    rwa [commutator_subgroupOf_self, hidxC] at hle'
  -- Odd orders, all from `|L|` odd.
  have hHabodd : Odd (Nat.card (Abelianization ↥H)) :=
    Odd.of_dvd_nat (Odd.of_dvd_nat hLodd H.card_subgroup_dvd_card)
      (Subgroup.card_quotient_dvd_card (_root_.commutator ↥H))
  have hCodd : Odd (Nat.card ↥C) := Odd.of_dvd_nat hLodd C.card_subgroup_dvd_card
  -- **(6.5)(b) reduction:** the nilpotent Frobenius kernel `H` is a `p`-group.
  obtain ⟨p, hp, hHp⟩ := isPGroup_of_isFrobeniusGroup_of_card_le hF hHabodd hCodd hbound
  have : Fact (Nat.Prime p) := ⟨hp⟩
  -- `p ∣ |H|` (nontrivial `p`-group), hence `p` odd and `|L:H| ∣ p − 1`.
  have : Nontrivial ↥H := by
    rcases subsingleton_or_nontrivial ↥H with h | h
    · exact absurd (fun a b => Subsingleton.elim (a * b) (b * a)) hnonab
    · exact h
  have hpdvd : p ∣ Nat.card ↥H := by
    obtain ⟨n, hn⟩ := hHp.exists_card_eq
    have h1 : 1 < Nat.card ↥H := Finite.one_lt_card_iff_nontrivial.mpr inferInstance
    rw [hn] at h1 ⊢
    exact dvd_pow_self p (by rintro rfl; simp at h1)
  have hodd_p : Odd p :=
    Odd.of_dvd_nat (Odd.of_dvd_nat hLodd H.card_subgroup_dvd_card) hpdvd
  have hodd_LH : Odd H.index := hidxC ▸ hCodd
  have hdvd' : H.index ∣ p - 1 := hdvd p hp hpdvd
  -- Non-abelian `p`-group ⟹ `p² ≤ |H:H′|`.
  have hsq : p ^ 2 ≤ Nat.card (↥H ⧸ (⁅H, H⁆ : Subgroup ↥L).subgroupOf H) := by
    rw [commutator_subgroupOf_self]
    exact sq_le_card_abelianization_of_isPGroup_of_noncomm hp hHp hnonab
  exact six_five_c_arith hp hodd_p hodd_LH hdvd' hsq hle

/-- **`H` is non-abelian** in the `X`-nonempty branch: `X(⁅H,H⁆) ≠ ∅` forces `⁅H,H⁆ ≠ ⊥`, hence
`commutator ↥H ≠ ⊥`.  (If `⁅H,H⁆ = ⊥` then `S(⁅H,H⁆) = S(⊥) = S`, so `X(⁅H,H⁆) = S − S = ∅`.) -/
theorem commutator_ne_bot_of_Xset_commutator_nonempty (hyp : SibleyDadeHypothesis G L H)
    (hXne : (hyp.Xset ⁅H, H⁆).Nonempty) : _root_.commutator ↥H ≠ ⊥ := by
  let : H.Normal := hyp.H_normal
  intro hbot
  rw [← commutator_subgroupOf_self] at hbot
  have hcommbot : (⁅H, H⁆ : Subgroup ↥L) = ⊥ :=
    (Subgroup.subgroupOf_eq_bot.mp hbot).eq_bot_of_le (Subgroup.commutator_le_left H H)
  have hXempty : hyp.Xset ⁅H, H⁆ = ∅ := by
    change hyp.S \ hyp.SsubFiltration ⁅H, H⁆ = ∅
    rw [hcommbot, hyp.SsubFiltration_bot, Set.sdiff_self]
  rw [hXempty] at hXne
  exact Set.not_nonempty_empty hXne

/-- **Peterfalvi (6.8) — Frobenius case (6.8)(c1): `S` is coherent.**  When `L` is a Frobenius group
with kernel `H` and complement `W₁`, and the `X`-set `X(⁅H,H⁆)` is nonempty (so `H` is non-abelian),
the Sibley set `S` is coherent.

This *fully closes* the (6.8)(c1) branch: by contradiction, if `S` is not coherent then the (6.5)
reduction `isPGroup_of_not_coherent` makes `H` a `p`-group; `p ≥ 3` (odd order) and `H` non-abelian
(`X`-nonempty) feed the case-(A) Frobenius coherence `nonempty_coherent_S_caseA_of_frobenius`,
contradicting non-coherence. -/
theorem nonempty_coherent_S_of_frobenius (hyp : SibleyDadeHypothesis G L H)
    (hF : OddOrder.Isaacs.Ch06.IsFrobeniusGroup (↥L) H hyp.W1)
    (hXne : (hyp.Xset ⁅H, H⁆).Nonempty) :
    Nonempty (OddOrder.Peterfalvi.S07.IsCoherent hyp.tau hyp.S
      (OddOrder.Peterfalvi.S04.supportInSubgroup (sharpImage H) L)) := by
  let : H.Normal := hyp.H_normal
  by_contra hncoh
  obtain ⟨p, hp, hHp⟩ := hyp.isPGroup_of_not_coherent hF hncoh
  exact hncoh (hyp.nonempty_coherent_S_caseA_of_frobenius hF
    (commutator_ne_bot_of_Xset_commutator_nonempty hyp hXne) hp
    (three_le_of_isPGroup_H hyp hp hHp) hHp)

/-- **Peterfalvi (6.4.c) for case (6.8)(c2): the `W₁`-conjugation fixed points lie in `⁅H,H⁆`.**

The certain-type Hypothesis (4.2)(b) gives `C_K(w) = W₂` for every `w ∈ W₁^#` (`centralizer_W2`),
and case (c2) has `W₂ ⊆ ⁅H,H⁆`.  Hence any `x ∈ H` fixed by conjugation by a nonidentity
`w ∈ W₁` lies in `C_L(w) ⊓ K = W₂ ⊆ ⁅H,H⁆`, so `x ∈ ⁅H,H⁆.subgroupOf H = commutator ↥H`
(`commutator_subgroupOf_self`).  This is exactly the `hfix` hypothesis consumed by
`exists_isPGroup_H_of_c2_of_card_le`, mirroring the `caseB_W1_dvd_index_commutator` fixed-point
argument — the (6.4.c) Frobenius-action condition is a *consequence* of the certain-type centralizer
datum, not an extra input. -/
theorem fixedPoints_W1_subset_commutator_of_c2 (hyp : SibleyDadeHypothesis G L H)
    (cert : OddOrder.Peterfalvi.S06.CertainTypeHypothesis (sharpImage H) L)
    (hK : cert.K = H) (hW1 : cert.W1 = hyp.W1) (hW2 : cert.W2 ≤ ⁅H, H⁆) :
    ∀ w : ↥hyp.W1, w ≠ 1 → ∀ x : ↥H,
      (w : ↥L) * (x : ↥L) * (w : ↥L)⁻¹ = (x : ↥L) → x ∈ commutator ↥H := by
  intro w hw x hwx
  -- `(x:L)` commutes with `(w:L)` (from the conjugation-fixed hypothesis), so lies in `C_L(w)`.
  have hxcen : (x : ↥L) ∈ Subgroup.centralizer ({(w : ↥L)} : Set ↥L) :=
    Subgroup.mem_centralizer_singleton_iff.mpr (mul_inv_eq_iff_eq_mul.mp hwx).symm
  have hwW1 : (w : ↥L) ∈ cert.W1 := by rw [hW1]; exact w.2
  have hwne : (w : ↥L) ≠ 1 := fun h => hw (OneMemClass.coe_eq_one.mp h)
  -- `(x:L) ∈ C_L(w) ⊓ K = W₂` by the certain-type centralizer datum (4.2.b).
  have hxW2 : (x : ↥L) ∈ cert.W2 := by
    rw [← cert.centralizer_W2 (w : ↥L) hwW1 hwne]
    exact Subgroup.mem_inf.mpr ⟨hxcen, by rw [hK]; exact x.2⟩
  -- `W₂ ⊆ ⁅H,H⁆`, so `x ∈ ⁅H,H⁆.subgroupOf H = commutator ↥H`.
  rw [← commutator_subgroupOf_self]
  exact Subgroup.mem_subgroupOf.mpr (hW2 hxW2)

/-- **Peterfalvi (6.8) — certain-type case (6.8)(c2): `S` is coherent (dispatch skeleton).**

The (6.8)(c2) branch of `sibleySetup_is_coherent`, assembled from the certain-type data `h46` of
`hyp.cases.inr`.  By contradiction-free assembly: the (6.5) reduction
(`exists_isPGroup_H_of_c2_of_card_le`, fed `hfix` from `fixedPoints_W1_subset_commutator_of_c2` and
the (6.3) index bound `hbound`) makes `H` a `p`-group, then the math case split (6.8.A/B) on whether
`W₂ ⊆ Z(H)` routes to the case-A coherence producer (`Z(H) ∩ W₂ = 1`) or the case-B one
(`1 ≠ W₂ ⊆ Z(H)`, `nonempty_coherent_S_caseB_of_structure`, cont.¹³).

This is the [[gated-endpoint-skeleton]] form: the three genuine residual obligations are isolated as
named hypotheses — `hbound` (Theorem (6.3) c2, the main remaining work, via the reducible-break
norm-weighted (5.6) engine), `hcaseA` (the case-A `_of_c2_caseA` → `S` bootstrap), and `hcaseB` (the
case-B structure-data discharge feeding `nonempty_coherent_S_caseB_of_structure`).  The A/B split
itself (`hsplit`, prime-order `W₂.subgroupOf H` ⟹ `⊥` or all in `Z(H)`) is also named, to be
discharged separately.  Wiring this into `S08_CoherenceTheorems:59` (the c2 branch) follows once the
three obligations are filled. -/
theorem nonempty_coherent_S_of_c2_of_branches (hyp : SibleyDadeHypothesis G L H)
    [H.Normal] [Finite ↥H]
    (h46 : OddOrder.Peterfalvi.S06.Hypothesis46 (sharpImage H) L)
    (hHK : h46.K = H) (hW1 : h46.W1 = hyp.W1) (hW2comm : h46.W2 ≤ ⁅H, H⁆)
    (hcop : Nat.Coprime (Nat.card ↥H) (Nat.card hyp.W1))
    (hbound : Nat.card (Abelianization ↥H) ≤ 4 * Nat.card ↥hyp.W1 ^ 2 + 1)
    (hsplit : h46.W2.subgroupOf H ≤ Subgroup.center ↥H ∨
      Disjoint (h46.W2.subgroupOf H) (Subgroup.center ↥H))
    (hcaseA : ∀ {p : ℕ}, p.Prime → IsPGroup p ↥H →
      Disjoint (h46.W2.subgroupOf H) (Subgroup.center ↥H) →
      Nonempty (OddOrder.Peterfalvi.S07.IsCoherent hyp.tau hyp.S
        (OddOrder.Peterfalvi.S04.supportInSubgroup (sharpImage H) L)))
    (hcaseB : ∀ {p : ℕ}, p.Prime → IsPGroup p ↥H →
      h46.W2.subgroupOf H ≤ Subgroup.center ↥H →
      Nonempty (OddOrder.Peterfalvi.S07.IsCoherent hyp.tau hyp.S
        (OddOrder.Peterfalvi.S04.supportInSubgroup (sharpImage H) L))) :
    Nonempty (OddOrder.Peterfalvi.S07.IsCoherent hyp.tau hyp.S
      (OddOrder.Peterfalvi.S04.supportInSubgroup (sharpImage H) L)) := by
  have hfix := fixedPoints_W1_subset_commutator_of_c2 hyp h46.toCertainTypeHypothesis hHK hW1
      hW2comm
  obtain ⟨p, hp, hHp⟩ := exists_isPGroup_H_of_c2_of_card_le hyp hcop hfix hbound
  rcases hsplit with hB | hA
  · exact hcaseB hp hHp hB
  · exact hcaseA hp hHp hA

omit [Invertible (Nat.card ↥H : ℂ)] in
/-- **(6.8) case A/B split** (discharges `hsplit` of `nonempty_coherent_S_of_c2_of_branches`).
Since `W₂ ≤ H` is of prime order, `W₂.subgroupOf H` has prime order too, so the subgroup
`(W₂.subgroupOf H) ⊓ Z(H)` is either trivial (case A: `Z(H) ∩ W₂ = 1`) or all of `W₂.subgroupOf H`
(case B: `1 ≠ W₂ ⊆ Z(H)`).  This is Peterfalvi (6.8)'s two-case dichotomy (04.8:152-154). -/
theorem caseAB_split_of_c2 [H.Normal] [Finite ↥H]
    (h46 : OddOrder.Peterfalvi.S06.Hypothesis46 (sharpImage H) L)
    (hW2H : h46.W2 ≤ H) (hprime : (Nat.card h46.W2).Prime) :
    h46.W2.subgroupOf H ≤ Subgroup.center ↥H ∨
    Disjoint (h46.W2.subgroupOf H) (Subgroup.center ↥H) := by
  by_cases hB : h46.W2.subgroupOf H ≤ Subgroup.center ↥H
  · exact Or.inl hB
  refine Or.inr (disjoint_iff.mpr ?_)
  have hWprime : (Nat.card ↥(h46.W2.subgroupOf H)).Prime := by
    rwa [Nat.card_congr (Subgroup.subgroupOfEquivOfLe hW2H).toEquiv]
  have hdvd : Nat.card ↥(h46.W2.subgroupOf H ⊓ Subgroup.center ↥H) ∣
      Nat.card ↥(h46.W2.subgroupOf H) := Subgroup.card_dvd_of_le inf_le_left
  rcases (Nat.dvd_prime hWprime).mp hdvd with h1 | hfull
  · exact Subgroup.card_eq_one.mp h1
  · exact absurd (inf_eq_left.mp (Subgroup.eq_of_le_of_card_ge inf_le_left hfull.ge)) hB

/-- **(6.8) case-B: `W₂ ≤ Z(L)`** (helper for the case-B structure-data discharge).
In case (B), `W₂ ⊆ Z(H)` (`hcen`) so `W₂` commutes with all of `H`; and `W₂` commutes with `W₁` by
the certain-type centralizer datum (`commute_of_mem_W1_of_mem_W2`).  Since `L = H ⋊ W₁`
(`H ⊔ W₁ = ⊤`), `W₂` commutes with every element of `L` (by closure induction), i.e. `W₂ ≤ Z(L)`. -/
theorem caseB_W2_le_center_L (hyp : SibleyDadeHypothesis G L H) [H.Normal]
    (h46 : OddOrder.Peterfalvi.S06.Hypothesis46 (sharpImage H) L)
    (hW1 : h46.W1 = hyp.W1) (hW2H : h46.W2 ≤ H)
    (hcen : h46.W2.subgroupOf H ≤ Subgroup.center ↥H) :
    h46.W2 ≤ Subgroup.center ↥L := by
  intro w hw
  have hwZH : (⟨w, hW2H hw⟩ : ↥H) ∈ Subgroup.center ↥H :=
    hcen (Subgroup.mem_subgroupOf.mpr hw)
  rw [Subgroup.mem_center_iff] at hwZH
  rw [Subgroup.mem_center_iff]
  intro g
  have hg : g ∈ Subgroup.closure ((H : Set ↥L) ∪ (hyp.W1 : Set ↥L)) := by
    rw [Subgroup.closure_union, Subgroup.closure_eq, Subgroup.closure_eq, hyp.split.sup_eq_top]
    exact Subgroup.mem_top g
  have hcomm : Commute g w := by
    induction hg using Subgroup.closure_induction with
    | mem x hx =>
      rcases hx with hxH | hxW1
      · have h3 : ((⟨x, hxH⟩ * ⟨w, hW2H hw⟩ : ↥H) : ↥L) =
            ((⟨w, hW2H hw⟩ * ⟨x, hxH⟩ : ↥H) : ↥L) := congrArg _ (hwZH ⟨x, hxH⟩)
        rw [Subgroup.coe_mul, Subgroup.coe_mul] at h3
        exact h3
      · exact h46.commute_of_mem_W1_of_mem_W2 (by rw [hW1]; exact hxW1) hw
    | one => exact Commute.one_left w
    | mul x y _ _ ihx ihy => exact ihx.mul_left ihy
    | inv x _ ihx => exact ihx.inv_left
  exact hcomm

/-- **Peterfalvi (6.8) case-B coherence from c2 data** (discharges `hcaseB` of the dispatch
skeleton).
Given the case-B math condition `W₂ ⊆ Z(H)` (`hcen`) and the `p`-group structure, derive the
structural inputs `hderiv` (W₂ ⊆ commutator via `commutator_subgroupOf_self`), `hcZ` (`|W₂|` prime),
and `hW2cenL` (`caseB_W2_le_center_L`), then feed `nonempty_coherent_S_caseB_of_structure`
(cont.¹³).
The FPF index data (`hc2`/`hfpf`/`hFPF`) and the X/Y-cardinality conditions (`hYcard`/`hXne`) remain
as named hypotheses, to be discharged separately (`hfpf` via `caseB_fpf_bound`, the rest from the
final `S08:59` wiring). -/
theorem nonempty_coherent_S_caseB_of_c2_data (hyp : SibleyDadeHypothesis G L H)
    [H.Normal] [Finite ↥H]
    (h46 : OddOrder.Peterfalvi.S06.Hypothesis46 (sharpImage H) L) (hHK : h46.K = H)
    (hW1 : h46.W1 = hyp.W1)
    [NeZero (Nat.card h46.W1)] [Invertible (Nat.card ↥h46.K : ℂ)]
    [Finite ↥(h46.W1 ⊔ h46.W2)] [Invertible (Nat.card ↥(h46.W1 ⊔ h46.W2) : ℂ)]
    [Finite (OddOrder.Peterfalvi.S06.ticVdiff h46).W]
    [Invertible (Nat.card (OddOrder.Peterfalvi.S06.ticVdiff h46).W : ℂ)]
    [h46.W2.Normal] [Invertible (Nat.card ↥h46.W2 : ℂ)]
    [Finite ↥(h46.W2.subgroupOf H)] [Invertible (Nat.card ↥(h46.W2.subgroupOf H) : ℂ)]
    (hW2H : h46.W2 ≤ H) (hcop : Nat.Coprime (Nat.card ↥H) (Nat.card hyp.W1))
    {p : ℕ} (hp : p.Prime) (hHp : IsPGroup p ↥H)
    (hprime : (Nat.card h46.W2).Prime) (hW2comm : h46.W2 ≤ ⁅H, H⁆)
    (hcen : h46.W2.subgroupOf H ≤ Subgroup.center ↥H)
    (hc2 : 2 ≤ (h46.W2.subgroupOf H).index)
    (hfpf : (2 * Nat.card hyp.W1 + 1) ^ 2 ≤ Nat.card (↥H ⧸ h46.W2.subgroupOf H))
    (hFPF : (h46.W2.index : ℤ) < ((h46.W2.subgroupOf H).index : ℤ) ^ 2)
    (hXne : (hyp.Xset h46.W2).Nonempty) :
    Nonempty (OddOrder.Peterfalvi.S07.IsCoherent hyp.tau hyp.S
      (OddOrder.Peterfalvi.S04.supportInSubgroup (sharpImage H) L)) := by
  have : Fintype ↥H := Fintype.ofFinite _
  have : Fintype ↥(h46.W1 ⊔ h46.W2) := Fintype.ofFinite _
  have : Fintype (OddOrder.Peterfalvi.S06.ticVdiff h46).W := Fintype.ofFinite _
  have : Fintype ↥(h46.W2.subgroupOf H) := Fintype.ofFinite _
  have hderiv : h46.W2.subgroupOf H ≤ commutator ↥H := by
    rw [← commutator_subgroupOf_self]
    exact fun x hx => Subgroup.mem_subgroupOf.mpr (hW2comm (Subgroup.mem_subgroupOf.mp hx))
  have hcZ : 2 ≤ Nat.card ↥(h46.W2.subgroupOf H) := by
    rw [Nat.card_congr (Subgroup.subgroupOfEquivOfLe hW2H).toEquiv]; exact hprime.two_le
  have hW2cenL : h46.W2 ≤ Subgroup.center ↥L := caseB_W2_le_center_L hyp h46 hW1 hW2H hcen
  exact nonempty_coherent_S_caseB_of_structure hyp h46 hHK hW1 hW2H hcen hderiv hcop hp hHp
    hprime hW2comm hW2cenL hc2 hFPF hcZ hfpf hXne

/-- **Peterfalvi (6.8) case-B coherence — full c2 FPF/index structure-data discharge.**

Strengthens `nonempty_coherent_S_caseB_of_c2_data` by deriving the FPF/index structure data
(`hfpf`/`hc2`/`hFPF`) *internally* from the `p`-group + central-`W₂` context, so that only the
genuine (6.8.3) *case* conditions remain as named hypotheses:
* `hWMgt` — `W₂ ⊊ ⁅H,H⁆` (`1 < |⁅H,H⁆ : W₂|`), i.e. `Z = W₂ ≠ H' = ⁅H,H⁆`, the (6.8.3) nontrivial
  branch (when `Z = H'` we already have `S = X ∪ Y` coherent by (6.8.2));
* `hYcard` — `m = |Y| ≠ 2`, the (6.8.2) exceptional case;
* `hXne` — `X` nonempty.

The (6.8.3) FPF bound `hfpf : (2|W₁|+1)² ≤ |H:W₂|` comes from `caseB_fpf_bound` (`W₂ ≤ Z(L)` via
`caseB_W2_le_center_L`; `H` non-abelian via nilpotency — `⁅H,H⁆ ⊊ H` so `1 < |H:⁅H,H⁆|`).  The
remaining index data is arithmetic: `hc2` (`|H:W₂| ≥ 9 ≥ 2`) and `hFPF`
(`|W₂|_L = |H:W₂|·|W₁| < |H:W₂|²` since `|W₁| < |H:W₂|`, via `relIndex_mul_index` and
`index_H_eq_card_W1`).  This is the clean `hcaseB` producer feeding the dispatch skeleton
`nonempty_coherent_S_of_c2_of_branches`. -/
theorem nonempty_coherent_S_caseB_of_c2 (hyp : SibleyDadeHypothesis G L H)
    [H.Normal] [Finite ↥H]
    (h46 : OddOrder.Peterfalvi.S06.Hypothesis46 (sharpImage H) L) (hHK : h46.K = H)
    (hW1 : h46.W1 = hyp.W1)
    [NeZero (Nat.card h46.W1)] [Invertible (Nat.card ↥h46.K : ℂ)]
    [Finite ↥(h46.W1 ⊔ h46.W2)] [Invertible (Nat.card ↥(h46.W1 ⊔ h46.W2) : ℂ)]
    [Finite (OddOrder.Peterfalvi.S06.ticVdiff h46).W]
    [Invertible (Nat.card (OddOrder.Peterfalvi.S06.ticVdiff h46).W : ℂ)]
    [h46.W2.Normal] [Invertible (Nat.card ↥h46.W2 : ℂ)]
    [Finite ↥(h46.W2.subgroupOf H)] [Invertible (Nat.card ↥(h46.W2.subgroupOf H) : ℂ)]
    (hW2H : h46.W2 ≤ H) (hcop : Nat.Coprime (Nat.card ↥H) (Nat.card hyp.W1))
    {p : ℕ} (hp : p.Prime) (hHp : IsPGroup p ↥H)
    (hprime : (Nat.card h46.W2).Prime) (hW2comm : h46.W2 ≤ ⁅H, H⁆)
    (hcen : h46.W2.subgroupOf H ≤ Subgroup.center ↥H)
    (hWMgt : 1 < (h46.W2.subgroupOf H).relIndex (commutator ↥H))
    (hXne : (hyp.Xset h46.W2).Nonempty) :
    Nonempty (OddOrder.Peterfalvi.S07.IsCoherent hyp.tau hyp.S
      (OddOrder.Peterfalvi.S04.supportInSubgroup (sharpImage H) L)) := by
  have : Fintype ↥H := Fintype.ofFinite _
  have : Fintype ↥(h46.W1 ⊔ h46.W2) := Fintype.ofFinite _
  have : Fintype (OddOrder.Peterfalvi.S06.ticVdiff h46).W := Fintype.ofFinite _
  have : Fintype ↥(h46.W2.subgroupOf H) := Fintype.ofFinite _
  let : Group.IsNilpotent ↥H := hyp.H_nilpotent
  have : Nontrivial ↥H := (Subgroup.nontrivial_iff_ne_bot H).mpr hyp.H_ne_bot
  -- `W₂ ≤ Z(L)` (case-B central datum).
  have hW2cenL : h46.W2 ≤ Subgroup.center ↥L := caseB_W2_le_center_L hyp h46 hW1 hW2H hcen
  -- `H` non-abelian: `⁅H,H⁆ ⊊ H`, so `1 < |H : ⁅H,H⁆|`.
  have hMgt : 1 < (commutator ↥H).index := by
    have hne : commutator ↥H ≠ ⊤ := (Group.IsSolvable.commutator_lt_top_of_nontrivial ↥H).ne
    have h0 : (commutator ↥H).index ≠ 0 := Subgroup.index_ne_zero_of_finite
    have h1 : (commutator ↥H).index ≠ 1 := fun h => hne (Subgroup.index_eq_one.mp h)
    omega
  -- the (6.8.3) FPF bound `(2|W₁|+1)² ≤ |H:W₂|`.
  have hfpf : (2 * Nat.card hyp.W1 + 1) ^ 2 ≤ Nat.card (↥H ⧸ h46.W2.subgroupOf H) :=
    caseB_fpf_bound hyp h46.toCertainTypeHypothesis hHK hW1 hW2comm hW2cenL hcop hMgt hWMgt
  have hw1pos : 1 ≤ Nat.card hyp.W1 := Nat.card_pos
  -- `|H:W₂| ≥ 2`.
  have hc2 : 2 ≤ (h46.W2.subgroupOf H).index :=
    le_trans (le_trans (by omega) (Nat.le_self_pow (by norm_num) (2 * Nat.card hyp.W1 + 1))) hfpf
  -- `|W₂|_L = |H:W₂|·|W₁| < |H:W₂|²`.
  have hFPF : (h46.W2.index : ℤ) < ((h46.W2.subgroupOf H).index : ℤ) ^ 2 := by
    have hidx : h46.W2.index = (h46.W2.subgroupOf H).index * Nat.card hyp.W1 := by
      have h := Subgroup.relIndex_mul_index hW2H
      rw [hyp.index_H_eq_card_W1] at h
      exact h.symm
    have hw1lt : Nat.card hyp.W1 < (h46.W2.subgroupOf H).index :=
      lt_of_lt_of_le
        (lt_of_lt_of_le (by omega) (Nat.le_self_pow (by norm_num) (2 * Nat.card hyp.W1 + 1))) hfpf
    have hw1ltZ : (Nat.card hyp.W1 : ℤ) < ((h46.W2.subgroupOf H).index : ℤ) := by
      exact_mod_cast hw1lt
    have hiposZ : (0 : ℤ) < ((h46.W2.subgroupOf H).index : ℤ) := by
      exact_mod_cast lt_of_le_of_lt (Nat.zero_le _) hw1lt
    rw [hidx]; push_cast; nlinarith [hw1ltZ, hiposZ]
  exact nonempty_coherent_S_caseB_of_c2_data hyp h46 hHK hW1 hW2H hcop hp hHp hprime hW2comm
    hcen hc2 hfpf hFPF hXne

end OddOrder.Peterfalvi.S08
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/Peterfalvi/S08_PGroupReduction.lean

View graph

Get started

Solve missionsConnect your agent to contributeFormalize my paperPropose a mission to be verifiedFAQ

About Prove2Me

Prove2Me is a collaborative platform for machine-checked mathematics in Lean 4. Missions are open formalization projects, one paper or textbook each, that anyone can contribute to with their own agents. Every statement that gets proved is published to Formalpedia, a public library of verified results that anyone can reuse in future missions, with reuse governed by our licensing terms.

How Prove2Me worksResearch paperProve2Me Team
SKILL.mdTourFAQContactTerms
© 2026 Prove2Me