§8 capstone — the Proposition 8.9 two-source splice helpers #
GQ2.SectionEight.prop_8_9 asserts the boxed recursion system ClosedRecursion for both
sources B.bA (Γ_A) and B.bF (G_ℚ₂), sharing one witness (μ, G0, DT, phase). Its proof is
the final assembly of §8: prop_8_9_aux turns a per-source RecursionInputs bundle
(stageR136 + half139 + phase140, with (137)/(138) discharged internally) into
ClosedRecursion, and the two sources share the phase witness.
This file keeps the splice in a leaf of the import graph: prop_8_9_of reduces the conclusion to
the per-source inputs and shared witness, and the component lemmas construct those inputs from the
obstruction, half-torsor, and phase APIs. The final theorem is assembled downstream in
GQ2/Prop89Close.lean.
Inputs (per source s ∈ {A, F}, Γ_s ∈ {Γ_A, G_ℚ₂}) #
stageR136—RStageObstructionBuild.stageR136_ofRSepDatafrom a concreteRObstructionData- the source residues (
hsep_hom,hZcount) +hE2.
- the source residues (
half139—RecursionFrame.half139_of, discharged bycentralOver_equiv/liftsOver_equivlemma_8_6_local(G_ℚ₂) /lemma_8_6_gammaA(Γ_A) + theM-lift count (5.15/5.16).
phase140— thelemma_8_7/lemma_8_5/Prop 8.8/lemma_6_21/cor 5.17 chain.- the witness
(μ, G0, DT, phase)—phaseFamily/centralCoverOfCocycle. - side conditions
hfg/hscalar/hhead— the source's t.f.g.,#Hom(Γ,𝔽₂)=8, head surjectivity.
Prop 8.9, reduced to the per-source RecursionInputs + shared witness (the splice
backbone).
Given the shared phase witness (μ, G0, DT, phase), the two per-source side-condition triples, and
the two RecursionInputs bundles, the boxed system holds for both sources — each via
prop_8_9_aux. The component lemmas below construct the two RecursionInputs.
half139 reduced to the source's MLifts-level count (d3 bridge discharged) #
half139_via_radData strips the Prop. 8.9 assembly bridge plumbing (centralOver_equiv/liftsOver_equiv
over En.radData) off the half139 obligation, reducing it to the two pure MLifts source
facts for the transported lower map ρ' = rhoPrime …:
hlem86M— the source's Lemma 8.6 half-torsor identity2·#{central M-lifts} = #(M-lifts)(lemma_8_6_local✓ forG_ℚ₂;lemma_8_6_gammaA= the Γ_A half-torsor proof forΓ_A), andhMcountM— theM-lift count#(M-lifts) = |M_B|²(props 5.15/5.16).
So a caller feeds half139_of (hence RecursionInputs.half139) directly from the source
arithmetic, with no CentralOver/LiftsOver bookkeeping.
The zBC fibration over the lower exact-image map ρ: zBC = Σ_ρ #CentralOver(ρ). Both
the (139) and (140) counts rest on this (it is the first step inside half139_of); extracted here
so the (140) hfib datum zBC = μ·M gets it too — zBC = Σ_ρ #CentralOver = Σ_ρ μ·M_ρ = μ·M.
phase140 reduced to a clean "phase datum" (the lemma_8_5/8.7 analog of stageR136_of) #
hfib level 2 — the per-ρ μ-partition of the central M-lifts #
The per-ρ μ-partition (the Prop. 8.9 assembly, hfib level 2): in the zero-edge regime the central
M-lifts of a lower map ρ split into the fibres of the T-reduction map red_T, and each
(nonempty) fibre is a free Z¹_{Γ,ρ}(T)-torsor of size μ = #Z¹(T) (lemma_8_7_count,
Central constant by central_twist_iff). Hence the central-lift count factors as
(#achievable central T-reductions) · μ. Summed over the C-image ρ (via
zBC_eq_sum_centralOver, after transport through centralOver_equiv) and combined with the
μ-independence #Z¹(T) = μ, this is the (140) hfib datum zBC = μ·M fed to
phase140_ofPhaseData; here M = Σ_ρ #achievable central T-reductions is the constrained
count of lemma_8_5.
The per-ρ μ-partition, in bridge vocabulary (the Prop. 8.9 assembly): transporting
central_card_eq_reductions_mul_tcocycle through centralOver_equiv, the zBC-fibre
#CentralOver(ρ) (the summand of zBC_eq_sum_centralOver) factors as
(#achievable central T-reductions of ρ' = rhoPrime …) · #Z¹(T). This is the per-fibre form
of the (140) hfib datum: once #Z¹(T) = μ is shown ρ-independent, summing over ρ gives
zBC = μ · M with M = Σ_ρ #achievable central T-reductions (the lemma_8_5 count).
The (140) hfib datum, reduced to μ-independence (the Prop. 8.9 assembly). Summing the per-ρ
μ-partition (centralOver_card_eq_reductions_mul_tcocycle) over the C-image via
zBC_eq_sum_centralOver and factoring out the common μ (hypothesis hμ: the T-cocycle
count #Z¹(T) is ρ-independent — the source 5.15/5.16 fact) gives the (140) fibration
zBC = μ · M, with M = Σ_ρ #achievable central T-reductions. This is exactly the hfib
argument of phase140_ofPhaseData: the (140) fibration is now reduced to the single source
input hμ (and M is the lemma_8_5 constrained count fed to hgauss).
hgauss level 1 — aggregating the Gauss engine lemma_8_5 over the ρ-family #
The aggregated constrained-Gauss identity (the Prop. 8.9 assembly, hgauss level 1): summing the proved
Gauss engine lemma_8_5 over a finite index family I (the C-image ρ, each with its own
constraint (κ_i, ε_i)) and swapping the resulting double sum gives
2·|E^∨|·Σ_i N(κ_i,ε_i) = |I|·|W| + G(Q)·Σ_χ Σ_i (−1)^{χκ_i+ε_i+Q(a_χ)}.
Pure 𝔽₂-linear algebra — no frame data. This is the aggregation step of hgauss: with the
concrete correspondences Σ_i N(κ_i,ε_i) = M, |I| = e_Γ(C), |W| = |V|, |E^∨| = |D_T|,
G(Q) = G0, and the phase reindex Σ_i sign(χκ_i+ε_i+Q(a_χ)) = 2·nPhase(phase χ) − e_Γ(C)
(the Prop 8.8 / (135) content coupled to the witness), it becomes the hgauss hypothesis of
phase140_ofPhaseData.
The capstone (140) reducer — phase140 from the concrete correspondences #
Discharging the polar data a_χ from nonsingularity (the En.hns supply) #
The |V| = |M_B|/|T_B| match (the Prop. 8.9 assembly): the enrichment's descent M_B ↠ V with
ker = T_B gives |V| = |M_B|/|T_B| by the first isomorphism theorem — discharging the hWV
cardinality match of phase140_of_nonsingular directly from En (with W := En.Vmod).
Paper-tag ledger (auto-generated by paperforge; do not edit) #
- cor 5.17 = ⟦cor-adjointboundary⟧
- Lemma 8.6 = ⟦lem-radicaledge⟧
- Prop 8.8 = ⟦prop-phaseidentity⟧
- Prop 8.9 = ⟦thm-closedrecursion⟧ (= theorem 8.17 in current tex)