Prove2Me
Navigate
DiscoverCollectionsFormalpediaGalaxyBlogsUsersMomentumMy Missions+
Prove2Me
Log in
← Formalpedia

Narrow ppp-groups

Definition
OddOrder_GroupTheory_NarrowPGroup

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

feit-thompsonfinite-groupsgroup-theoryishida-odd-order

The shared narrow ppp-group predicate IsNarrow (Bender–Glauberman, Local Analysis for the Odd Order Theorem, §1 p. 2 and §5 p. 44), a cross-cutting concept used in BG §5 and cited in §6, §10, §16 and Peterfalvi §9.

Main definitions: IsNarrow (BG narrow ppp-group), IsMaximalElementaryAbelian (E∗(R)E^*(R)E∗(R), a maximal elementary abelian subgroup) and omega1UpperCentralTwo (W=Ω1(Z2(R))W=\Omega_1(Z_2(R))W=Ω1​(Z2​(R)), BG Lemma 4.5), with instances showing WWW and T=CR(W)T=C_R(W)T=CR​(W) are characteristic in RRR.

The bundle also proves 12 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/GroupTheory/NarrowPGroup.lean
-- Generated for Prove2Me by deleting declarations; the remaining commands are verbatim except:
-- `private` dropped where another file uses the declaration, and a solution's main theorem
-- renamed `_root_.solution` (platform convention).
import Definitions.Def_OddOrder_GroupTheory_PRank
import Definitions.Def_OddOrder_GroupTheory_CriticalSubgroup

namespace OddOrder.GroupTheory

/-- **BG narrow `p`-group** (§1 p.2 / §5 p.44). For a prime `p`, a `p`-group `R` is
*narrow* iff `r(R) ≤ 2` (`pRank R p ≤ 2`) or there is an order-`p` subgroup `R₀` whose
centralizer is generated by `R₀` and a cyclic subgroup `R₁`, with `R₀ ∩ R₁ = 1`.

The internal direct product `C_R(R₀) = R₀ × R₁` is encoded inside the centralizer by
`R₀ ⊓ R₁ = ⊥` and `Subgroup.centralizer R₀ = R₀ ⊔ R₁`. We deliberately do not use
`Subgroup.IsComplement' R₀ R₁`, since that would make `R₀` and `R₁` complements in the
ambient group `R`, forcing `R₀ ⊔ R₁ = ⊤` rather than only filling `C_R(R₀)`. The naïve
form `pRank R p ≤ 2` is used for `r(R) ≤ 2`; BG Theorem 5.3 / Corollary 5.4 supply the
equivalent characterisations when `r(R) ≥ 3`. -/
def IsNarrow (p : ℕ) (R : Type*) [Group R] : Prop :=
  pRank R p ≤ 2 ∨
    ∃ R₀ R₁ : Subgroup R, Nat.card R₀ = p ∧ IsCyclic R₁ ∧
      R₀ ⊓ R₁ = ⊥ ∧ Subgroup.centralizer (R₀ : Set R) = R₀ ⊔ R₁

/-- A `p`-group of rank at most `2` is narrow (the first disjunct of `IsNarrow`). -/
theorem isNarrow_of_pRank_le_two {p : ℕ} {R : Type*} [Group R] (h : pRank R p ≤ 2) :
    IsNarrow p R :=
  Or.inl h

/-- **`IsNarrow` is invariant under group isomorphism**: transport the rank bound or the
narrow witness `(R₀, R₁)` along `e : R ≃* R'`. Used to move narrowness between a Sylow
subgroup and its isomorphic images (BG Thm 5.6: `S ≅ S̄ ≤ G/O_{p'}(G)`,
`S̄ = O_p(Ḡ)`). -/
theorem IsNarrow.of_mulEquiv {p : ℕ} {R R' : Type*} [Group R] [Group R'] [Finite R]
    (e : R ≃* R') (h : IsNarrow p R) : IsNarrow p R' := by
  have : Finite R' := Finite.of_equiv R e.toEquiv
  rcases h with hrank | ⟨R₀, R₁, hcard, hcyc, hinf, hcent⟩
  · exact Or.inl (le_trans
      (pRank_le_of_injective (f := e.symm.toMonoidHom) e.symm.injective) hrank)
  · refine Or.inr ⟨R₀.map e.toMonoidHom, R₁.map e.toMonoidHom, ?_, ?_, ?_, ?_⟩
    · rw [← hcard]
      exact (Nat.card_congr
        (Subgroup.equivMapOfInjective R₀ e.toMonoidHom e.injective).toEquiv).symm
    · have := hcyc
      exact isCyclic_of_surjective _ (e.toMonoidHom.subgroupMap_surjective R₁)
    · rw [← Subgroup.map_inf R₀ R₁ e.toMonoidHom e.injective, hinf, Subgroup.map_bot]
    · have hmap_cent :
          Subgroup.centralizer ((R₀.map e.toMonoidHom : Subgroup R') : Set R') =
            (Subgroup.centralizer (R₀ : Set R)).map e.toMonoidHom := by
        ext x
        constructor
        · intro hx
          have hx' := Subgroup.mem_centralizer_iff.mp hx
          refine ⟨e.symm x, ?_, by simp⟩
          rw [SetLike.mem_coe, Subgroup.mem_centralizer_iff]
          intro r hr
          apply e.injective
          rw [map_mul, map_mul]
          simp only [MulEquiv.apply_symm_apply]
          exact hx' (e r) ⟨r, hr, rfl⟩
        · rintro ⟨y, hy, rfl⟩
          rw [SetLike.mem_coe, Subgroup.mem_centralizer_iff] at hy
          rw [Subgroup.mem_centralizer_iff]
          rintro _ ⟨r, hr, rfl⟩
          rw [← map_mul, ← map_mul, hy r hr]
      rw [hmap_cent, hcent, Subgroup.map_sup]

/-- **`E*(R)`: maximal elementary abelian subgroup** (BG §5 p.45). A subgroup `E ≤ R` is
`p`-elementary abelian and is contained in no strictly larger `p`-elementary abelian
subgroup of `R`.

The "no larger" condition is phrased as: any `p`-elementary abelian `F` with `E ≤ F`
satisfies `F = E`. BG's `E²(R) ∩ E*(R)` (elementary abelian of order `p²` not contained
in any `E_{p³}`) is expressed at use sites as
`Nat.card E = p ^ 2 ∧ IsMaximalElementaryAbelian p E` (no separate abbreviation). -/
def IsMaximalElementaryAbelian (p : ℕ) {R : Type*} [Group R] (E : Subgroup R) : Prop :=
  E.IsElementaryAbelian p ∧
    ∀ F : Subgroup R, F.IsElementaryAbelian p → E ≤ F → F = E

namespace IsMaximalElementaryAbelian

variable {p : ℕ} {R : Type*} [Group R] {E F : Subgroup R}

/-- A maximal elementary abelian subgroup is, in particular, elementary abelian. -/
theorem isElementaryAbelian (h : IsMaximalElementaryAbelian p E) : E.IsElementaryAbelian p :=
  h.1

/-- Maximality: any `p`-elementary abelian `F ⊇ E` collapses back to `E`. -/
theorem eq_of_le (h : IsMaximalElementaryAbelian p E) (hF : F.IsElementaryAbelian p)
    (hEF : E ≤ F) : F = E :=
  h.2 F hF hEF

/-- If an elementary abelian subgroup `F` centralizes a maximal elementary abelian
subgroup `E`, then adjoining `F` to `E` adds nothing. -/
theorem sup_eq_left_of_le_centralizer (h : IsMaximalElementaryAbelian p E)
    (hF : F.IsElementaryAbelian p) (hEF : E ≤ Subgroup.centralizer (F : Set R)) :
    E ⊔ F = E :=
  h.eq_of_le
    (Subgroup.IsElementaryAbelian.sup_of_le_centralizer h.isElementaryAbelian hF hEF)
    le_sup_left

/-- If an elementary abelian subgroup `F` centralizes a maximal elementary abelian
subgroup `E`, then `F ≤ E`. -/
theorem le_of_le_centralizer (h : IsMaximalElementaryAbelian p E)
    (hF : F.IsElementaryAbelian p) (hEF : E ≤ Subgroup.centralizer (F : Set R)) :
    F ≤ E := by
  have hsup : E ⊔ F = E := h.sup_eq_left_of_le_centralizer hF hEF
  intro x hx
  rw [← hsup]
  exact (le_sup_right : F ≤ E ⊔ F) hx

end IsMaximalElementaryAbelian

/-- **Existence of a maximal elementary abelian subgroup above a given one** (used by
BG Theorem 5.3 ⇐). In a finite group, every `p`-elementary abelian subgroup `E` is
contained in some `E* ∈ E*(R)`, i.e. a `p`-elementary abelian subgroup maximal under
inclusion.

This is a pure finite-poset maximality statement: take a `≤`-maximal element of the
finite subgroup lattice among the `p`-elementary abelian subgroups containing `E`
(`Finite.exists_le_maximal`); the maximality witness upgrades to
`IsMaximalElementaryAbelian` via antisymmetry. -/
theorem exists_maximalElementaryAbelian_ge
    {p : ℕ} {R : Type*} [Group R] [Finite R] {E : Subgroup R}
    (hE : E.IsElementaryAbelian p) :
    ∃ F : Subgroup R, E ≤ F ∧ IsMaximalElementaryAbelian p F := by
  obtain ⟨F, hEF, hFmax⟩ :=
    Finite.exists_le_maximal (p := fun F : Subgroup R => F.IsElementaryAbelian p) hE
  exact ⟨F, hEF, hFmax.1, fun G hG hFG => le_antisymm (hFmax.2 hG hFG) hFG⟩

section Omega1UpperCentralTwo

open scoped commutatorElement

/-- **BG Lemma 4.5** `W = Ω₁(Z₂(R))`. For a prime `p`, the subgroup of `R` generated
by the order-`p` elements of the second centre `Z₂(R) = upperCentralSeries R 2`,
realised as the image of `Ω₁(↥Z₂(R))` under `Z₂(R).subtype` (the `omega1Map`
template, since `Z₂(R)` is generally non-abelian). -/
def omega1UpperCentralTwo (R : Type*) [Group R] (p : ℕ) : Subgroup R :=
  (Omega ↥(Subgroup.upperCentralSeries R 2) p 1).map (Subgroup.upperCentralSeries R 2).subtype

/-- `W = Ω₁(Z₂(R)) ≤ Z₂(R)`. -/
theorem omega1UpperCentralTwo_le (R : Type*) [Group R] (p : ℕ) :
    omega1UpperCentralTwo R p ≤ Subgroup.upperCentralSeries R 2 :=
  Subgroup.map_subtype_le _

/-- `W = Ω₁(Z₂(R))` is characteristic in `R` (char-in-char: `Ω₁` is characteristic
in `Z₂(R)`, `Z₂(R)` is characteristic in `R`). -/
instance omega1UpperCentralTwo_characteristic {R : Type*} [Group R] {p : ℕ} :
    (omega1UpperCentralTwo R p).Characteristic :=
  characteristic_map_subtype_of_characteristic inferInstance Omega.characteristic

/-- `Z₂(R) = upperCentralSeries R 2` has class `≤ 2`: `⁅Z₂, Z₂⁆ ≤ Z(↥Z₂)`.

For `x, y ∈ Z₂(R) = upperCentralSeries R 2`, `mem_upperCentralSeries_succ_iff` (with
`n = 1`) gives `⁅x, y⁆ = x * y * x⁻¹ * y⁻¹ ∈ upperCentralSeries R 1 = Z(R)`, so
`⁅x, y⁆` commutes with all of `R`, in particular centralises `Z₂(R)`. Translating
along `Z₂(R).subtype` yields the claim. -/
theorem commutator_upperCentralSeries_two_le_center {R : Type*} [Group R] :
    _root_.commutator ↥(Subgroup.upperCentralSeries R 2) ≤
      Subgroup.center ↥(Subgroup.upperCentralSeries R 2) := by
  set Z := Subgroup.upperCentralSeries R 2 with hZ
  -- `⁅Z, Z⁆ ≤ Z(R)`: for `x, y ∈ Z = upperCentralSeries R 2`, `⁅x, y⁆ ∈ Z(R)`.
  have hZZ_le_center : ⁅Z, Z⁆ ≤ Subgroup.center R := by
    rw [Subgroup.commutator_le]
    intro x hx y _
    -- `x ∈ upperCentralSeries R (1 + 1)` ⇒ `⁅x, y⁆ ∈ upperCentralSeries R 1`.
    have hmem : ⁅x, y⁆ ∈ Subgroup.upperCentralSeries R 1 :=
      Subgroup.mem_upperCentralSeries_succ_iff.mp hx y
    rwa [Subgroup.upperCentralSeries_one] at hmem
  -- `Z(R)` centralises everything, so `⁅Z, Z⁆ ≤ centralizer (Z : Set R)`.
  have hZZ_cent : ⁅Z, Z⁆ ≤ Subgroup.centralizer (Z : Set R) := by
    intro x hx
    rw [Subgroup.mem_centralizer_iff]
    intro w _
    exact Subgroup.mem_center_iff.mp (hZZ_le_center hx) w
  -- Translate `⁅Z, Z⁆ ≤ centralizer Z` to `commutator ↥Z ≤ center ↥Z`.
  rw [← Z.map_subtype_le_map_subtype, Z.map_subtype_commutator]
  intro x hx
  have hxZ : x ∈ Z := Z.commutator_le_self hx
  rw [Subgroup.mem_map]
  refine ⟨⟨x, hxZ⟩, ?_, rfl⟩
  rw [Subgroup.mem_center_iff]
  intro k
  apply Subtype.ext
  rw [Subgroup.coe_mul, Subgroup.coe_mul]
  exact Subgroup.mem_centralizer_iff.mp (hZZ_cent hx) (k : R) k.2

/-- **BG Lemma 4.5(c)** (exponent-`p` half). For odd `p`, every element of
`W = Ω₁(Z₂(R))` has `p`-th power `1`. -/
theorem pow_eq_one_of_mem_omega1UpperCentralTwo
    {R : Type*} [Group R] {p : ℕ} (hp_odd : Odd p) {g : R}
    (hg : g ∈ omega1UpperCentralTwo R p) : g ^ p = 1 := by
  rw [omega1UpperCentralTwo, Subgroup.mem_map] at hg
  obtain ⟨z, hz, rfl⟩ := hg
  have hzp : z ^ p = 1 :=
    Omega.pow_eq_one_of_class_le_two hp_odd commutator_upperCentralSeries_two_le_center hz
  rw [Subgroup.coe_subtype, ← Subgroup.coe_pow, hzp, Subgroup.coe_one]

/-- **`T = C_R(W)` is characteristic** (BG §5). The centralizer of the
characteristic subgroup `W = Ω₁(Z₂(R))` is characteristic
(`Subgroup.characteristic_centralizer` + `omega1UpperCentralTwo_characteristic`). -/
instance centralizer_omega1UpperCentralTwo_characteristic
    {R : Type*} [Group R] {p : ℕ} :
    (Subgroup.centralizer (omega1UpperCentralTwo R p : Set R)).Characteristic :=
  Subgroup.characteristic_centralizer

/-- `Z(R) ⊆ T = C_R(W)` (BG §5): the centre centralises every subset, in
particular `W = Ω₁(Z₂(R))` (`Subgroup.center_le_centralizer`). -/
theorem center_le_centralizer_omega1UpperCentralTwo
    {R : Type*} [Group R] {p : ℕ} :
    Subgroup.center R ≤ Subgroup.centralizer (omega1UpperCentralTwo R p : Set R) :=
  Subgroup.center_le_centralizer _

/-- `R' ⊆ T = C_R(W)` (BG §5, mmd L1862), where `W = Ω₁(Z₂(R))`.

This only uses `W ≤ Z₂(R)`: for `w ∈ W`, the defining property of `Z₂(R)` gives
`[w,R] ≤ Z(R)`, hence also `[R,w] ≤ Z(R)`. The three-subgroups lemma then gives
`[[R,R],W]=1`, i.e. the commutator subgroup centralizes `W`. -/
theorem commutator_le_centralizer_omega1UpperCentralTwo
    {R : Type*} [Group R] {p : ℕ} :
    _root_.commutator R ≤ Subgroup.centralizer (omega1UpperCentralTwo R p : Set R) := by
  let W : Subgroup R := omega1UpperCentralTwo R p
  have hW_top_center : ⁅W, (⊤ : Subgroup R)⁆ ≤ Subgroup.center R := by
    rw [Subgroup.commutator_le]
    intro x hx y _
    have hxZ2 : x ∈ Subgroup.upperCentralSeries R 2 := omega1UpperCentralTwo_le R p hx
    have hmem : ⁅x, y⁆ ∈ Subgroup.upperCentralSeries R 1 :=
      Subgroup.mem_upperCentralSeries_succ_iff.mp hxZ2 y
    rwa [Subgroup.upperCentralSeries_one] at hmem
  have htop_W_center : ⁅(⊤ : Subgroup R), W⁆ ≤ Subgroup.center R := by
    rw [Subgroup.commutator_comm]
    exact hW_top_center
  have hW_top_top : ⁅⁅W, (⊤ : Subgroup R)⁆, (⊤ : Subgroup R)⁆ = ⊥ :=
    Subgroup.commutator_eq_bot_iff_le_centralizer.mpr
      (hW_top_center.trans (Subgroup.center_le_centralizer _))
  have htop_W_top : ⁅⁅(⊤ : Subgroup R), W⁆, (⊤ : Subgroup R)⁆ = ⊥ :=
    Subgroup.commutator_eq_bot_iff_le_centralizer.mpr
      (htop_W_center.trans (Subgroup.center_le_centralizer _))
  have hderived_W : ⁅⁅(⊤ : Subgroup R), (⊤ : Subgroup R)⁆, W⁆ = ⊥ :=
    Subgroup.commutator_commutator_eq_bot_of_rotate htop_W_top hW_top_top
  rw [_root_.commutator_def]
  exact Subgroup.commutator_eq_bot_iff_le_centralizer.mp hderived_W

end Omega1UpperCentralTwo

end OddOrder.GroupTheory
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/GroupTheory/NarrowPGroup.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