ℤ₂-powering on pro-2 groups #
This file defines x ^ u for a 2-adic exponent u ∈ ℤ₂ and x in any pro-2 group, with the
laws consumed by the proofs in §3.
Contents #
(i) The projection ℤ̂ ↠ ℤ₂ and the seam isomorphism.
zhatProjTwo : ℤ̂ → Multiplicative ℤ₂— defined asγ ↦ (ofAdd 1) ^ᶻ γ, i.e. as aẑ-power (the profinite-exponentiation API), so continuity/hom-ness/ofInt-anchors come from the existing API.zhatProjTwo_surjective,ker_zhatProjTwo : ker = proPKernel 2 ℤ̂.ztwoEquivPadic : maxProPQuotient 2 ℤ̂ ≃ₜ* Multiplicative ℤ₂, pinned bymaxProPMk (ofInt 1) ↦ ofAdd 1. SinceZtwo := maxProPQuotient 2 Zhat(GQ2/BoundaryFrame.lean, definitionally), this is theιofSectionThree.prop_3_10_local_markedand the identification Prop. 3.2'snuTwo-surjectivity can compose with.
(ii) ℤ₂-powering on pro-2 groups.
zpowZtwo (hP : IsProP 2 P) (x : P) (u : ℤ_[2]) : P, bundled aszpowZtwoHom hP x : Multiplicative ℤ₂ →ₜ* P; built fromzpowHatHomthrough (i) via the universal propertymaxProPHomEquiv.- Anchors:
zpowZtwo_intCast(= x ^ non integer exponents), exponent additivity (it is a hom), the composition lawzpowZtwo_zpowZtwo : (x^u)^v = x^{uv}, naturalitymap_zpowZtwo, and the uniqueness principlezpowZtwoHom_unique(a continuous homℤ₂ → Pis determined by its value at1— the tool for identifying constructed maps with powerings). - Bijectivity:
zpowZtwo_bijective(x ↦ x^ubijective foru ∈ ℤ₂ˣ, inversex ↦ x^{u⁻¹}) andpow_bijective_of_odd(odd integer powers; the Lemmas 3.6–3.8 proof's "cube roots", inversex ↦ x^{m⁻¹}).
(iii) The η-facts (shared the Lemmas 3.4–3.5 proof lemma_3_5_injective / the Lemmas 3.6–3.8 proof prop_3_8_classification
prerequisite).
isProP_two_unitsPadicInt : IsProP 2 ℤ₂ˣ— soη ^ umakes sense forη ∈ ℤ₂ˣ,u ∈ ℤ₂.zpowZtwo_injective_of_norm—η-injectivity: if‖η − 1‖ = 2⁻²(i.e.v₂(η−1) = 2; the paper'sη = (−3)⁻¹, cf.norm_inv_neg_three_sub_one) thenu ↦ η ^ uis injective. This is the consumable form of the paper's "ηtopologically generates1 + 4ℤ₂" (Lemma 3.5's injectivity row and Prop. 3.8'sȲ-coordinate forcing use exactly injectivity).
(iv) Pro-2 Frattini/Burnside criterion.
The complementary statement "surjective on G/Φ(G) implies surjective" is proved in
GQ2/FrattiniCriterion.lean in index-p detection form.
Conventions #
Multiplicative throughout: exponents live in Multiplicative ℤ_[2] (group law = addition of
exponents), matching Zhat's convention; zpowZtwo takes the additive u : ℤ_[2] and
wraps it. 2-specific facts use p = 2 instances (Fact (Nat.Prime 2) is found automatically).
The density/extension workhorse #
ℤ ⊆ ℤ₂ (indeed ℕ ⊆ ℤ₂) is dense, so a continuous monoid hom out of Multiplicative ℤ₂
into a Hausdorff monoid is determined by its value at ofAdd 1. This single lemma powers all
the algebraic laws of zpowZtwo below.
Two continuous monoid homs Multiplicative ℤ₂ →* M (M Hausdorff) agreeing at ofAdd 1
are equal. (Via DenseRange.addChar_eq_of_eval_one_eq and
AddChar ℤ₂ M ≃ (Multiplicative ℤ₂ →* M).)
ℤ₂-powering of an element of 2-power order #
The elementary building block: if g ^ 2^k = 1 then u ↦ g ^ (u mod 2^k) is a continuous
monoid hom Multiplicative ℤ₂ →* Q. Used for the kernel computation of zhatProjTwo (where
Q is a finite quotient of ℤ̂), and reusable for any finite 2-group target.
u ↦ g ^ (u mod 2^k) for an element g with g ^ 2^k = 1.
Equations
- GQ2.powZModTwoHom g k hg = { toFun := fun (u : Multiplicative ℤ_[2]) => g ^ ((PadicInt.toZModPow k) (Multiplicative.toAdd u)).val, map_one' := ⋯, map_mul' := ⋯ }
Instances For
The fibres of toZModPow k : ℤ₂ → ℤ/2^k are open (cosets of the ball span {2^k}).
Every toZModPow-preimage is open (union of open fibres): toZModPow k is locally
constant.
powZModTwoHom is continuous: it factors through the locally constant toZModPow k.
The canonical projection ℤ̂ ↠ ℤ₂ (part (i)) #
zhatProjTwo := γ ↦ (ofAdd 1) ^ᶻ γ — the ẑ-power of 1 ∈ ℤ₂ (the profinite-exponentiation API machinery), which on
integer exponents is just n ↦ n : ℤ → ℤ₂.
The dense embedding ℤ → ℤ̂, bundled (multiplicative convention, as everywhere in
GQ2/Zhat.lean).
Equations
- GQ2.Zhat.ofIntHom = { toFun := fun (n : Multiplicative ℤ) => GQ2.Zhat.ofInt (Multiplicative.toAdd n), map_one' := GQ2.Zhat.ofInt_zero, map_mul' := GQ2.Zhat.ofIntHom._proof_1 }
Instances For
ofInt turns integer powers of ofInt 1 into ofInt of the integer.
The canonical projection ℤ̂ → ℤ₂ (multiplicatively: onto Multiplicative ℤ₂), as the
ẑ-power morphism of ofAdd 1: γ ↦ (ofAdd 1) ^ᶻ γ. On ℤ ⊆ ℤ̂ it is the inclusion
ℤ ⊆ ℤ₂ (zhatProjTwo_ofInt); it realizes ℤ₂ as the maximal pro-2 quotient of ℤ̂
(ker_zhatProjTwo, ztwoEquivPadic).
Equations
- GQ2.zhatProjTwo = ProfiniteGrp.Hom.hom (GQ2.zpowHatHom (Multiplicative.ofAdd 1))
Instances For
zhatProjTwo is surjective (closed range ⊇ the dense ℕ ⊆ ℤ₂).
The kernel of zhatProjTwo is contained in the pro-2 kernel: any element of ℤ̂ killed by
the projection to ℤ₂ dies in every finite 2-group quotient. Proof idea: for an open
normal U with 2-group quotient, the quotient map q : ℤ̂ → ℤ̂/U and the composite of
zhatProjTwo with "ℤ₂-powering of q(ofInt 1)" (powZModTwoHom) are continuous maps
agreeing on the dense ℤ ⊆ ℤ̂, hence equal (funext_ofInt); the second visibly kills
ker zhatProjTwo.
ker zhatProjTwo = proPKernel 2 ℤ̂: the projection to ℤ₂ realizes exactly the
maximal pro-2 quotient of ℤ̂ (the missing direction is proPKernel_le_ker for the pro-2
target ℤ₂, GQ2/PropOneOne.lean).
The seam isomorphism maxPro2(ℤ̂) ≅ ℤ₂ (part (i), packaged) #
Ztwo := maxProPQuotient 2 Zhat in GQ2/BoundaryFrame.lean (definitionally), so
ztwoEquivPadic is the ι : Ztwo ≃ₜ* Multiplicative ℤ₂ of
SectionThree.prop_3_10_local_marked, pinned by ztwoOne ↦ ofAdd 1
(ztwoEquivPadic_ofInt_one).
The descent of zhatProjTwo to the maximal pro-2 quotient.
Equations
- GQ2.ztwoDescend = GQ2.quotientLift (GQ2.proPKernel 2 ↑GQ2.Zhat.toProfinite.toTop) GQ2.zhatProjTwo GQ2.ztwoDescend._proof_2
Instances For
maxPro2(ℤ̂) ≅ ℤ₂ — the peripheral-action interface "nice-to-have", and the ι-seam of
prop_3_10_local_marked/Prop. 3.2's nuTwo (as Ztwo = maxProPQuotient 2 Zhat definitionally).
Equations
Instances For
The generator pin: ztwoOne = maxProPMk (ofInt 1) ↦ ofAdd 1 (the normalization
prop_3_10_local_marked requires of ι).
ℤ₂-powering on pro-2 groups (part (ii)) #
For x in a pro-2 profinite group P, the ẑ-power morphism zpowHatHom x : ℤ̂ ⟶ P (the profinite-exponentiation API)
kills proPKernel 2 ℤ̂ (universal property, since P is pro-2), so through
ker zhatProjTwo = proPKernel 2 ℤ̂ (part (i)) it factors through ℤ₂: this is zpowZtwoHom.
Everything else follows from the extension-by-density workhorse multPadicIntHom_ext.
Evaluation of the maxProPHomEquiv-descended hom on classes: the descent of F at
maxProPMk γ is F γ. (The .symm of the universal property is quotientLift.)
The ℤ₂-power morphism of x in a pro-2 group P: the unique continuous extension of
n ↦ xⁿ to exponents in ℤ₂, as a continuous hom Multiplicative ℤ₂ →ₜ* P. Built by
factoring zpowHatHom x : ℤ̂ ⟶ P through zhatProjTwo via the universal property of the
maximal pro-2 quotient and the part-(i) isomorphism.
Equations
- GQ2.zpowZtwoHom hP x = ((GQ2.maxProPHomEquiv hP).symm (ProfiniteGrp.Hom.hom (GQ2.zpowHatHom x))).comp ↑GQ2.ztwoEquivPadic.symm
Instances For
x ^ u for a 2-adic exponent u ∈ ℤ₂ (x in a pro-2 group; additive exponent
convention: zpowZtwo hP x (u + v) = zpowZtwo hP x u * zpowZtwo hP x v).
Equations
- GQ2.zpowZtwo hP x u = (GQ2.zpowZtwoHom hP x) (Multiplicative.ofAdd u)
Instances For
The generator pin: x ^ (1 : ℤ₂) = x.
ℤ₂-powers extend ordinary integer powers.
Exponent additivity (the hom law, in zpowZtwo clothing).
Uniqueness/identification principle: any continuous hom φ : ℤ₂ → P is the
ℤ₂-powering of its value at 1. (The tool for recognizing constructed maps as powerings —
e.g. the Lemmas 3.6–3.8 proof's Θ_b legs.)
Powering the identity: 1 ^ u = 1.
Unit-power bijectivity: for u ∈ ℤ₂ˣ, x ↦ x ^ u is a bijection of P, with inverse
x ↦ x ^ u⁻¹.
Odd integers are units of ℤ₂.
Naturality: continuous homs of pro-2 groups commute with ℤ₂-powers.
Bridge to ẑ-powers: on a pro-2 group, x ^ᶻ γ computes the ℤ₂-power at the
projection of γ — the profinite exponents of the profinite-exponentiation API (e.g. ω₂) factor through ℤ₂ here.
ℤ₂ˣ is pro-2, and the η-injectivity (part (iii)) #
This is a shared prerequisite for the Lemmas 3.4–3.8 proofs. Level-tracking is done
algebraically (divisibility by
2^k, with unit witnesses for exact levels) — p = 2 is what makes ℤ₂ˣ pro-2: every unit is
≡ 1 (mod 2), and squaring gains exactly one level. Real norms appear only once, bridging to
the open-ball description of a neighbourhood in ℤ₂ˣ.
1 + 2x is a unit of ℤ₂ (its residue mod 2 is 1).
Every 2-adic unit is ≡ 1 (mod 2) (p = 2: the residue field has one nonzero class).
Level gain under 2-power powers: 2^{k+1} ∣ u^{2^k} − 1 for every u ∈ ℤ₂ˣ
(squaring 1 + 2^m a gains exactly one level: (1+2^m a)² = 1 + 2^{m+1}(a + 2^{m-1}a²)).
Exact-level version (unit-witnessed): if η − 1 = 4a with a ∈ ℤ₂ˣ (i.e.
v₂(η−1) = 2 exactly), then η^{2^k} − 1 = 2^{k+2}·(unit) — the level grows by exactly one
per squaring, so no 2-power power of η is 1.
The divisibility-to-metric bridge: 2^n ∣ x − 1 means x is within 2^{-n} of 1.
ℤ₂ˣ is a pro-2 group: every finite continuous quotient of the 2-adic units is a
2-group. (Uniform annihilation: any open U ∋ 1 contains all units within 2^{-(K+1)} of 1
— in both the value and the inverse coordinate of the Units topology — and
u^{2^K} ≡ 1 (mod 2^{K+1}) for every unit u.)
The profinite-group instances on ℤ₂ˣ #
Mathlib provides IsTopologicalGroup αˣ ([ContinuousMul α]), CompactSpace αˣ
([T1Space α] [ContinuousMul α] [CompactSpace α]), and T2Space αˣ; total disconnectedness
transfers along the embedProduct embedding. With these, zpowZtwo applies to ℤ₂ˣ.
The η-injectivity #
η-injectivity: if η − 1 = 4a with a ∈ ℤ₂ˣ (equivalently v₂(η − 1) = 2 — the
paper's "η topologically generates 1 + 4ℤ₂", in the form Lemma 3.5's injectivity row and
Prop. 3.8's Ȳ-coordinate forcing consume), then u ↦ η ^ u is injective on ℤ₂-exponents.
Proof: a nonzero exponent factors as c = w·2^m (w ∈ ℤ₂ˣ, m = v₂(c)); η^c = (η^{2^m})^w,
so η^c = 1 forces η^{2^m} = 1 (unit powers are bijective), contradicting the exact level
η^{2^m} − 1 = 2^{m+2}·(unit) ≠ 0.
The paper's η = (−3)⁻¹ has exact level 2: η − 1 = 4·η (witness a = η itself),
since η − 1 = η(1 − (−3))·… = 4η.
The form consumed by the Lemmas 3.4–3.8 proofs: u ↦ ((−3)⁻¹) ^ u is injective
(η = y⁻¹ for the class
y = −3 of Lemma 3.5 / Prop. 3.8).
Paper-tag ledger (auto-generated by paperforge; do not edit) #
- Lemma 3.5 = ⟦lem-markedinitialform⟧
- Prop 3.8 = ⟦prop-orientationlift⟧ (= proposition 3.9 in current tex)