Peterfalvi (8.15.3)/(9.5): the §9 family inside the (8.15.3) family, base layer
DefinitionOddOrder_Peterfalvi_S11_NineElevenBridgeBasefeit-thompsonfinite-groupsgroup-theoryishida-odd-order
Formalizes the base layer of the subcoherence bridge showing that the §9 family sits inside the (8.15.3) family (Peterfalvi, Character Theory for the Odd Order Theorem, §8 (8.15.3), §9 (9.5), pp. 47-51), yielding the base coherence used by (9.11).
Main declarations: sOf_subset_inducedNonKernelFamily (The §9 family is contained in the (8.15.3) family); sOf_degreeSubfamily_coherent ((9.11) base coherence at §9 level); caseA_coherent_sOf_cprime_of_refuter (Peterfalvi (9.11), case (9.7.a), at §9 level); sOf_subset_inducedKernelFamily_bot (The §9 family sits inside the ⊥-kernel induced family).
The bundle also proves 2 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/Peterfalvi/S11_NineElevenBridgeBase.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_S07_PivotCoherence
import Definitions.Def_OddOrder_Peterfalvi_S10_SubcoherentTypeP
import Definitions.Def_OddOrder_Peterfalvi_S11_MaximalII_III_IV_ThetaCountAssembly
import Definitions.Def_OddOrder_Peterfalvi_S11_NineElevenCoherence
import Definitions.Def_OddOrder_Peterfalvi_S11_SingleFactorCentralizer
namespace OddOrder.Peterfalvi.S11
open OddOrder.GroupTheory
open OddOrder.RepresentationTheory
variable {G : Type*} [Group G]
/-- **`M' ⊴ M`, realised inside `↥M`**: `(derivedInG M).subgroupOf M` is the comap of a `map` along
the injective `M.subtype`, i.e. `commutator ↥M`.
The §11/§13 chain gets this instance transitively from its packaging's import closure; the §9 leaves
here need it explicitly to borrow the `S08.inducedKernelFamily_*` suite. -/
theorem derivedInG_subgroupOf_normal (M : Subgroup G) : ((derivedInG M).subgroupOf M).Normal := by
rw [derivedInG, Subgroup.subgroupOf,
Subgroup.comap_map_eq_self_of_injective M.subtype_injective]
infer_instance
/-- **`H` sits inside `M_σ`, realised in `HU`**: `hInHu data ≤ (M_σ.subgroupOf M).subgroupOf (HU)`.
Two applications of `Subgroup.subgroupOf` monotonicity to `M_F ≤ M_σ`
(`BG.Ch4.S15.maxNilpotentNormalHall_le_Msigma`, rewritten through `data.typeP.H_eq`). -/
theorem hInHu_le_Msigma_subgroupOf [Finite G] {M : Subgroup G}
(hG : OddOrder.BG.IsMinimalSimpleOdd G) (hM : M ∈ maximalSubgroups G)
(data : TypesIIIIIIVSetup M) :
hInHu data ≤ ((OddOrder.BG.Ch3.S10.Msigma M).subgroupOf M).subgroupOf (huSub data) := by
have hMF : data.H ≤ OddOrder.BG.Ch3.S10.Msigma M := by
rw [show data.H = data.typeP.H from rfl, data.typeP.H_eq]
exact OddOrder.BG.Ch4.S15.maxNilpotentNormalHall_le_Msigma hG hM
exact Subgroup.comap_mono (Subgroup.comap_mono hMF)
open scoped OddOrder.Peterfalvi.S12.FiniteInduce in
/-- **The §9 family is contained in the (8.15.3) family** (issue 1045): `𝒮(Y) ⊆ 𝒮_{(8.15.3)}`.
This is what lets Peterfalvi's own route to the (9.11) base coherence — (8.15.3) followed by (5.7)
— replace the repo's §10 μ-grid engine, and with it the type-III/IV restriction: the (8.15.3)
producer `S10.typePACore_subcoherent` carries no type hypothesis beyond `IsTypeP`.
The proof establishes membership over `huSub data` and then transports along
`huSub_eq_derivedInG_subgroupOf`; the two subgroups are only *propositionally* equal, so
`↥(huSub data)` and `↥((derivedInG M).subgroupOf M)` are distinct types and the characters do not
transfer definitionally. This is the same `▸`-transport idiom the repo already uses for
`S08.inducedKernelFamily` (`S15_SSetMemberRFamily`). -/
theorem sOf_subset_inducedNonKernelFamily [Finite G] {M : Subgroup G}
(hG : OddOrder.BG.IsMinimalSimpleOdd G) (hM : M ∈ maximalSubgroups G)
(data : TypesIIIIIIVSetup M) (Y : Subgroup G) :
sOf data Y ⊆ OddOrder.Peterfalvi.S10.inducedNonKernelFamily
((derivedInG M).subgroupOf M) ((OddOrder.BG.Ch3.S10.Msigma M).subgroupOf M) := by
have hKeq : huSub data = (derivedInG M).subgroupOf M :=
huSub_eq_derivedInG_subgroupOf data
have hle := hInHu_le_Msigma_subgroupOf hG hM data
-- first over `huSub data`, then transport
have hbase : sOf data Y ⊆ OddOrder.Peterfalvi.S10.inducedNonKernelFamily
(huSub data) ((OddOrder.BG.Ch3.S10.Msigma M).subgroupOf M) := by
rintro φ ⟨χ, hχ, rfl⟩
refine ⟨χ, ?_, induceHU_eq_induce data _⟩
-- `M_σ ⊄ Ker χ`: else `H ⊆ Ker χ` too, contradicting `χ ∈ 𝒳`
intro hsub
exact hχ.1 fun x hx => hsub (SetLike.mem_coe.mpr (hle (SetLike.mem_coe.mp hx)))
exact hKeq ▸ hbase
open scoped OddOrder.Peterfalvi.S12.FiniteInduce in
/-- **(9.11) base coherence at §9 level**: the degree-`d` irreducible cut of `𝒮(Y)` is coherent.
This is the step that takes the (9.11) base case off the §10 μ-grid engine. The repo obtained it
from `S12.Hypothesis.inducedFamily_degreeSubfamily_isCoherent`, which needs a `S13.Hypothesis` and
hence types III/IV; Peterfalvi instead routes it through **(8.15.3) then (5.7)**, neither of which
carries a type hypothesis. Here that route is assembled:
* the cut sits inside the (8.15.3) family — `sOf_subset_inducedNonKernelFamily`;
* it is conjugation-closed — `irrCut_conjClosed` (§9-level since issue 1045);
* it is finite — `sOf_finite` (§9-level since issue 1045);
* (5.3.b) + (5.7) — `S10.inducedNonKernelFamily_degreeSubfamily_coherent`.
⚠ `h2 : 2 ≤ ncard` stays exposed, as in the §8 companion and in
`S15.Hypothesis.sSetIrrDeg_coherent`: the (9.8.d) count gives `∃ ζ`, not two members.
`hKeq` pins Hypothesis (4.6)'s `K` to `M'`, and `hHle` asks only that its `H` *contain* `M_σ` —
the book's choice `H = M_s = M_σ` (`S10.typePACore_toHypothesis46_core`) is the equality case, and
the §10/§13 packaging, which instantiates (4.6.c) at the larger `H = K = M'`, also qualifies
(`inducedNonKernelFamily_mono`; issue 1045 着手順 3). They are hypotheses rather than definitional
assumptions so that no `Hypothesis46Core` has to be rebuilt here (rebuilding it is what makes the
two copies fail to be definitionally equal). -/
theorem sOf_degreeSubfamily_coherent [Finite G] {M : Subgroup G} {A : Set G}
(hodd : Odd (Nat.card ↥M))
(h46 : OddOrder.Peterfalvi.S06.Hypothesis46Core A M)
(dd : OddOrder.Peterfalvi.S10.DadeSupportHypothesisData M A)
(hG : OddOrder.BG.IsMinimalSimpleOdd G) (hM : M ∈ maximalSubgroups G)
(data : TypesIIIIIIVSetup M) (Y : Subgroup G) (d : ℕ)
(hKeq : h46.K = (derivedInG M).subgroupOf M)
(hHle : (OddOrder.BG.Ch3.S10.Msigma M).subgroupOf M ≤ h46.subH)
(hd0 : ((d : ℂ)) ≠ 0)
(h2 : 2 ≤ {φ : ClassFunction ↥M ℂ | φ ∈ sOf data Y ∧
IsIrreducibleCharacter φ ∧ ((φ : ↥M → ℂ) 1 = (d : ℂ))}.ncard)
(h1A : (1 : ↥M) ∉ OddOrder.Peterfalvi.S04.supportInSubgroup A M) :
Nonempty (OddOrder.Peterfalvi.S07.IsCoherent
(OddOrder.Peterfalvi.S10.inducedNonKernelFamily_subcoherent hodd h46 dd
(fun _ hx => OddOrder.Peterfalvi.S10.inducedNonKernelFamily_mono hHle
(hKeq ▸ sOf_subset_inducedNonKernelFamily hG hM data Y hx.1))
(fun _ hx => hx.2.1)
(fun _ hx => irrCut_conjClosed data Y d hx)).tau
{φ : ClassFunction ↥M ℂ | φ ∈ sOf data Y ∧
IsIrreducibleCharacter φ ∧ ((φ : ↥M → ℂ) 1 = (d : ℂ))}
(OddOrder.Peterfalvi.S04.supportInSubgroup A M)) :=
OddOrder.Peterfalvi.S10.inducedNonKernelFamily_degreeSubfamily_coherent hodd h46 dd _ _ _
((sOf_finite data Y).subset fun _ hx => hx.1) h2 (d : ℂ) (fun _ hx => hx.2.2) hd0 h1A
open scoped OddOrder.Peterfalvi.S12.FiniteInduce in
/-- **Peterfalvi (9.11), case (9.7.a), at §9 level**: reduction to the maximality refuter.
The §9-level form of `S13.caseA_coherent_sOf_H0Cprime_of_refuter`. Everything is stated over
Hypotheses (9.2)/(9.4)/(9.5) — `data`, `chief`, `chars` — with the Dade map `tau` and support `A0`
as explicit parameters, so **no type hypothesis appears anywhere**.
The §11 version reaches the same conclusion through `S13.Hypothesis`, whose `type_alt` pins types
III/IV. Tracing its proof shows that packaging is inessential: `hyp.C` is `cSub data chief`,
`hyp.H0Cprime` is `chief.H0 ⊔ cprimeSub data chief`, `hyp.base.tau`/`.A0` are parameters, the
finiteness bridge is `sOf_finite`, and the degree-`qa` base coherence — the one genuinely §10-bound
input, via `S12.Hypothesis.inducedFamily_degreeSubfamily_isCoherent` — becomes the parameter
`hbase`, which `sOf_degreeSubfamily_coherent` supplies along the book's (8.15.3) → (5.7) route.
Book argument (unchanged): the (9.8.d) count `caseA_character_count_exact` has positive lower
bound `(p−1)·[U:U′]`, so a degree-`qa` irreducible exists in `𝒮(H₀U′)`; it transports to
`𝒮(H₀C′)` along `H₀C′ ≤ H₀U′` (`C′ = [C,C] ≤ [U,U] = U′`), and the maximality skeleton
`S07.coherent_of_maximal_coherent_pair_refuted` closes it against `hrefute`. -/
theorem caseA_coherent_sOf_cprime_of_refuter [Finite G] {M : Subgroup G}
(hG : OddOrder.BG.IsMinimalSimpleOdd G)
{data : TypesIIIIIIVSetup M} {chief : ChiefFactorData data}
(chars : Section11CharacterData data chief)
(tau : OddOrder.Peterfalvi.S07.IntegralCharacterMap ↥M G) (A0 : Set ↥M)
(caseA : CliffordCaseAData chars)
(hbase : OddOrder.Peterfalvi.S07.IsCoherent tau
{φ : ClassFunction ↥M ℂ | φ ∈ sOf data (chief.H0 ⊔ cprimeSub data chief) ∧
IsIrreducibleCharacter φ ∧ ((φ : ↥M → ℂ) 1 = ((data.q * caseA.a : ℕ) : ℂ))} A0)
(hrefute : ∀ S₂ : Set (ClassFunction ↥M ℂ),
{φ : ClassFunction ↥M ℂ | φ ∈ sOf data (chief.H0 ⊔ cprimeSub data chief) ∧
IsIrreducibleCharacter φ ∧
((φ : ↥M → ℂ) 1 = ((data.q * caseA.a : ℕ) : ℂ))} ⊆ S₂ →
S₂ ⊆ sOf data (chief.H0 ⊔ cprimeSub data chief) →
OddOrder.Peterfalvi.S03.ClosedUnderConjugate S₂ →
Nonempty (OddOrder.Peterfalvi.S07.IsCoherent tau S₂ A0) →
(sOf data (chief.H0 ⊔ cprimeSub data chief) \ S₂).Nonempty →
(∀ χ ∈ sOf data (chief.H0 ⊔ cprimeSub data chief) \ S₂,
¬ Nonempty (OddOrder.Peterfalvi.S07.IsCoherent tau (S₂ ∪ {χ, χ.conj}) A0)) → False) :
Nonempty (OddOrder.Peterfalvi.S07.IsCoherent tau
(sOf data (chief.H0 ⊔ cprimeSub data chief)) A0) := by
classical
-- positivity of the (9.8.d) count lower bound `(p−1)·[U:U′]`
have hp1 : 0 < chief.p - 1 := Nat.sub_pos_of_lt chief.p_prime.one_lt
have hrel : 0 < (uprimeSub data).relIndex data.U :=
lt_of_lt_of_le (u_odd hG chars).pos (u_le_relIndex_uprimeSub_U chars)
have hNpos := lt_of_lt_of_le (mul_pos hp1 hrel) (caseA_character_count_exact hG caseA)
-- the (9.8.d) count set is nonempty: a degree-`qa` irreducible in `𝒮(H₀U′)`
have hne : {χ ∈ sOf data (chief.H0 ⊔ uprimeSub data) |
IsIrreducibleCharacter χ ∧ χ 1 = ((data.q * caseA.a : ℕ) : ℂ)}.Nonempty := by
apply Set.nonempty_of_ncard_ne_zero
intro h0
rw [h0, Nat.zero_mul] at hNpos
exact absurd hNpos (lt_irrefl 0)
-- `H₀C′ ≤ H₀U′` (`C′ = [C,C] ≤ [U,U] = U′` by derived-subgroup monotonicity)
have hle : chief.H0 ⊔ cprimeSub data chief ≤ chief.H0 ⊔ uprimeSub data := by
refine sup_le_sup_left ?_ chief.H0
change derivedInG (cSub data chief) ≤ derivedInG data.U
rw [derivedInG_eq_commutator (cSub data chief), derivedInG_eq_commutator data.U]
exact Subgroup.commutator_mono (cSub_le_U data chief) (cSub_le_U data chief)
-- assemble the reduction via the maximality skeleton
exact OddOrder.Peterfalvi.S07.coherent_of_maximal_coherent_pair_refuted
(S₁ := {φ : ClassFunction ↥M ℂ | φ ∈ sOf data (chief.H0 ⊔ cprimeSub data chief) ∧
IsIrreducibleCharacter φ ∧ ((φ : ↥M → ℂ) 1 = ((data.q * caseA.a : ℕ) : ℂ))})
(sOf_finite data _)
(sOf_closedUnderConjugate data _)
(fun _ hφ => hφ.1)
(fun _ hχ => irrCut_conjClosed data _ (data.q * caseA.a) hχ)
⟨hbase⟩
hrefute
open scoped OddOrder.Peterfalvi.S12.FiniteInduce in
/-- **The §9 family sits inside the `⊥`-kernel induced family**: `𝒮(Y) ⊆ S(⊥)`.
Composite of `sOf_subset_inducedNonKernelFamily` with
`S10.inducedNonKernelFamily_subset_inducedKernelFamily_bot`.
This is the §9-level replacement for the "world-bridge" step that the §11 chain performs with
`hyp.SOf_eq`/`hyp.sOf_subset_SOf` — the route through the `S13.Hypothesis` packaging. The §11
pivot lemmas (`caseB_sOf_memberRFamily`, `sOf_anchor_diff_support`, …) use that bridge only to
borrow the `S08.inducedKernelFamily_*` suite (support, no-real-characters, pairwise orthogonality,
`ZIrr` membership); with this lemma they can borrow the same suite without a `S13.Hypothesis`. -/
theorem sOf_subset_inducedKernelFamily_bot [Finite G] {M : Subgroup G}
(hG : OddOrder.BG.IsMinimalSimpleOdd G) (hM : M ∈ maximalSubgroups G)
(data : TypesIIIIIIVSetup M) (Y : Subgroup G) :
sOf data Y ⊆ OddOrder.Peterfalvi.S08.inducedKernelFamily
((derivedInG M).subgroupOf M) (⊥ : Subgroup ↥M) := fun _ hx =>
OddOrder.Peterfalvi.S10.inducedNonKernelFamily_subset_inducedKernelFamily_bot
(sOf_subset_inducedNonKernelFamily hG hM data Y hx)
open scoped OddOrder.Peterfalvi.S12.FiniteInduce in
/-- **A nonempty degree cut of `𝒮(Y)` has at least two members** — the `h2` input of
`sOf_degreeSubfamily_coherent`, from mere nonemptiness.
`G` has odd order, so no member of the `⊥`-kernel induced family is real
(`S08.inducedKernelFamily_hasNoRealCharacters`); the cut is conjugation-closed
(`irrCut_conjClosed`, degrees being natural numbers). So `φ` and `φ̄` are two *distinct* members.
⚠ This is what lets the §9 route match §13's: §13 obtains its base coherence from the §10 μ-grid
engine with only an existence witness, while the (8.15.3) + (5.7) route of issue 1045 asks for
`2 ≤ ncard`. The gap was apparent — for a conjugation-closed family in odd order the two are the
same condition. -/
theorem irrCut_two_le_ncard [Finite G] {M : Subgroup G}
(hG : OddOrder.BG.IsMinimalSimpleOdd G) (hM : M ∈ maximalSubgroups G)
(data : TypesIIIIIIVSetup M) (Y : Subgroup G) (d : ℕ)
(hne : {φ : ClassFunction ↥M ℂ | φ ∈ sOf data Y ∧
IsIrreducibleCharacter φ ∧ ((φ : ↥M → ℂ) 1 = (d : ℂ))}.Nonempty) :
2 ≤ {φ : ClassFunction ↥M ℂ | φ ∈ sOf data Y ∧
IsIrreducibleCharacter φ ∧ ((φ : ↥M → ℂ) 1 = (d : ℂ))}.ncard := by
classical
have := derivedInG_subgroupOf_normal M
obtain ⟨φ, hφ⟩ := hne
have hφc := irrCut_conjClosed data Y d hφ
have hne' : φ ≠ φ.conj := fun h =>
OddOrder.Peterfalvi.S08.inducedKernelFamily_hasNoRealCharacters
(hG.odd.of_dvd_nat (Subgroup.card_subgroup_dvd_card M)) (⊥ : Subgroup ↥M)
(sOf_subset_inducedKernelFamily_bot hG hM data Y hφ.1) h.symm
have hfin : {φ : ClassFunction ↥M ℂ | φ ∈ sOf data Y ∧
IsIrreducibleCharacter φ ∧ ((φ : ↥M → ℂ) 1 = (d : ℂ))}.Finite :=
(sOf_finite data Y).subset fun _ hx => hx.1
have hsub : ({φ, φ.conj} : Set (ClassFunction ↥M ℂ)) ⊆
{φ : ClassFunction ↥M ℂ | φ ∈ sOf data Y ∧
IsIrreducibleCharacter φ ∧ ((φ : ↥M → ℂ) 1 = (d : ℂ))} := by
rintro x (rfl | rfl)
· exact hφ
· exact hφc
have hpair : ({φ, φ.conj} : Set (ClassFunction ↥M ℂ)).ncard = 2 := by
rw [Set.ncard_pair hne']
exact hpair ▸ Set.ncard_le_ncard hsub hfin
open scoped OddOrder.Peterfalvi.S12.FiniteInduce in
/-- **Uniform-degree member differences are `A₀`-supported, at §9 level** (the `hsuppdiff` input of
the (9.11) caseB engine).
The §9 form of `S13.sOf_anchor_diff_support`. That version reaches the `⊥`-kernel induced family
through `hyp.SOf_eq`/`hyp.sOf_subset_SOf`, i.e. through `S13.Hypothesis`; here the same step is
`sOf_subset_inducedKernelFamily_bot`, and the (8.10) containment `(M')^# ⊆ A₀` — which the §13
version reads off `hyp.base.mderivSharp_subset_A0` — becomes the explicit parameter `hKsupp`.
Neither version needs a type hypothesis; the §13 binding was purely the packaging. -/
theorem sOf_anchor_diff_support [Finite G] {M : Subgroup G} {A0 : Set ↥M}
(hG : OddOrder.BG.IsMinimalSimpleOdd G) (hM : M ∈ maximalSubgroups G)
(data : TypesIIIIIIVSetup M) (Y : Subgroup G)
(hKsupp : ∀ x : ↥M, x ∈ (derivedInG M).subgroupOf M → x ≠ 1 → x ∈ A0)
(d : ℕ) (hunif : ∀ φ ∈ sOf data Y, ((φ : ClassFunction ↥M ℂ) : ↥M → ℂ) 1 = (d : ℂ))
{χ₁ x : ClassFunction ↥M ℂ}
(hχ₁mem : χ₁ ∈ sOf data Y) (hx : x ∈ sOf data Y) :
((x - χ₁ : ClassFunction ↥M ℂ)).support ⊆ A0 := by
have := derivedInG_subgroupOf_normal M
have h := OddOrder.Peterfalvi.S08.inducedKernelFamily_scaledDiff_support hKsupp
(sOf_subset_inducedKernelFamily_bot hG hM data Y hx)
(sOf_subset_inducedKernelFamily_bot hG hM data Y hχ₁mem) (d := 1)
(by rw [Nat.cast_one, one_mul, hunif x hx, hunif χ₁ hχ₁mem])
rwa [one_smul] at h
end OddOrder.Peterfalvi.S11
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/S11_NineElevenBridgeBase.lean