Alphabeta Math
LemmaStatement: Literature-sourcedProof: AI-adaptedprecheck passjudge pass (gpt-5.6-terra)audited 2026-09-14
How statement and proof provenance work

The first chip identifies the source of the statement or construction; the second identifies the source of its local proof or verification.

  • Literature-sourced: the exact statement appears in a cited source; only wording and notation differ.
  • AI-adapted: a semantically identical restatement of literature-sourced material, modulo indexing, notation, and boundary cases adopted by the library.
  • AI-generated: a genuinely novel statement formulated by AI, with no source for the claim itself.

These labels describe origin, not correctness: citations and verification chips remain separate evidence.

Nondentability produces a vector measure without density

Statement

Assume the Axiom of Choice. If a Banach space X contains a nondentable nonempty bounded closed convex set, then on the Lebesgue interval there is an absolutely continuous bounded-variation X-valued vector measure whose range lies in a closed separable subspace of X and which has no Bochner density.

Facts & Assumptions

[A1]

The Axiom of Choice supplies arbitrary and recursive choices (The Axiom of Choice).

[L1]

Dentability is the existence of slices of arbitrarily small norm diameter (Dentable bounded set and slice).

[L2]

Under AC, dominated Hahn--Banach establishes HB, and under HB a point outside a nonempty closed convex set in a real or complex normed space can be uniformly strictly separated from it by a bounded functional (Hahn-Banach dominated extension theorem for real vector spaces, Relative geometric Hahn–Banach with the exact open, closed, and compact hypotheses).

[L3]

RNP requires a Bochner density for every absolutely continuous bounded-variation vector measure over a finite scalar measure (Radon--Nikodym property).

[L4]

A Bochner-integrable function has integrable simple approximants (Bochner-integrable function) and its integral obeys the norm inequality (Bochner integral norm inequality).

[L5]

Continuity from below and measurable set-difference calculus hold for measures (Continuity from below for measures, Measure of a set difference when the smaller set has finite measure), and the Lebesgue sigma-algebra is the completion of the Borel one (L(Rn) is exactly the completion of the restriction of λn to the Borel sets).

Proof

technique · contradiction

Given: A nondentable nonempty bounded closed convex set C⊆X and AC.

1.1givenA1L1L2construct

Convert nondentability into a uniformly separated convex bush. Choose η>0 such that no slice of C has diameter below η, and put r=η/4. For x∈C, if x∉conv⁡‾(C∖B(x,r)), [L2] gives a slice lying inside B(x,r) and hence of diameter at most 2r<η, a contradiction. Thus every x∈C belongs to that closed convex hull. Enlarge to D=C+B(0,r/2). Given z=x+y∈D, approximate x by a finite convex combination ∑iαixi=x+e of points outside B(x,r) with error e satisfying ∥e∥+∥y∥<r/2, and put zi=xi+y−e. Then z=∑iαizi, every zi∈D, and ∥zi−z∥≥r−∥e∥>r/2. With δ=r/2, [A1] recursively chooses such finite successor families from an initial z0∈C. The resulting node set is countable, bounded, and every child is at least δ from its parent.

2.1A1step 1.1construct

Realize the bush as a separated interval martingale. Starting with M0≡z0, partition every atom at level n−1 into finitely many half-open subintervals in the successor proportions (αi), put the corresponding child value on each, and overlay the dyadic grid of mesh 2−n. Let Pn be the resulting refining finite interval partition. Parent averages equal parent values, so (Mn) is a martingale on these finite algebras; it is uniformly bounded and ∥Mn−Mn−1∥≥δ away from the finitely many endpoints. The union algebra R=⋃nσ(Pn) contains every dyadic interval algebra.

3.1A1L5step 2.1

Define and extend the dominated vector measure. For A∈σ(Pn) set ν0(A)=∫AMn dλ. The martingale identity makes this independent of n, and if K=sup⁡n∥Mn∥∞, then ∥ν0(A)−ν0(B)∥≤Kλ(A△B). The class of Borel sets approximable in symmetric-difference measure by R is a sigma-algebra: complements preserve the distance, and countable unions reduce by [L5] to one large finite union. It contains the dyadic algebra and hence all Borel sets; the completion clause in [L5] adds Lebesgue sets. Choose such approximants. Completeness of X gives a unique extension ν with ∥ν(E)∥≤Kλ(E); the same estimate proves norm countable additivity. Summing it over finite partitions gives ∣ν∣(E)≤Kλ(E), so ν≪λ and has bounded variation. Every algebra value is a finite linear combination of bush nodes, hence every extended value lies in the closed separable span Y of the countable node set.

4.1assume-contraL3L4step 3.1

Assume a density and identify all its finite-partition averages. Suppose f∈L1([0,1];X) satisfies ν(E)=∫Ef for all Lebesgue E. For every atom A of Pn, the construction gives ∫Af=ν(A)=∫AMn=λ(A)Mn∣A. Thus the atomwise averaging operator Qn applied to f equals Mn.

5.1L4step 2.1step 3.1step 4.1

Prove that the atomwise averages of a Bochner density converge in L1. Choose an integrable simple s with ∫∥f−s∥<ε by [L4]. By the approximation proved in step 3.1 and finiteness of its level family, approximate the level sets of s by sets in R, obtaining an R-simple t with ∫∥f−t∥<2ε. For all sufficiently large n, Qnt=t. The norm inequality on each atom shows that Qn is an L1 contraction, so ∥Qnf−f∥1≤∥Qn(f−t)∥1+∥t−f∥1<4ε. Hence Mn=Qnf→f in L1.

6.1L3step 2.1step 3.1contradiction: step 5.1∎

Contradict the fixed separation and conclude. [discharge-contradiction: step 5.1, L3, step 2.1, step 4.1, step 5.1] Step 5.1 would imply ∥Mn−Mn−1∥1→0, whereas step 2.1 gives ∥Mn−Mn−1∥1≥δ for every n. Thus no density exists. The measure in step 3.1 is the witness required by [L3], with separable range. The exact uses of [A1] are strict separation through [L2], recursive bush and partition choices, and the countable approximation choices in extending ν0. The empty interval endpoints form null sets; the one-child case cannot occur because every child is δ-separated.

Depends on

Used by

Dependency tree · two levels

42 results within two dependency steps of this one, each drawn at its shortest distance from it. An arrow runs from a result to what uses it, so the chart reads left to right and ends at this result, which carries a heavier outline. Every node is a link to that result. Click elsewhere on the chart to enlarge it.

Sources