Wiki
Wiki

Research notes on every problem, and a library of the papers behind them. Built from the open erdos repository.

Updated

../


Source. E. Glazer, Erdős Problem 501 after adding ω2\omega_2 random reals, draft rev10, the opening paragraph of Section 4 and Lemma 4.1 (Borel reading), physical pp. 4--5, in the eight-page PDF held by its library source card, Glazer (2026). The source gives a six-line proof; the version here expands it and names the standard facts it rests on.

Standing. This is an author-recorded reconstruction. It is not an independent review and changes no status and assigns no tier. The facts labeled (R1)--(R5) below are imported, not proved here.

Conventions and imported facts

For a set Θ\Theta of coordinates, let μΘ\mu_\Theta be the completion of the product of the fair-coin measures on 2Θ2^\Theta. The measure algebra B(Θ)\mathbb B(\Theta) is the Boolean algebra of μΘ\mu_\Theta-measurable subsets of 2Θ2^\Theta modulo μΘ\mu_\Theta-null sets; a condition is a nonzero element, and forcing with B(Θ)\mathbb B(\Theta) over a ground model MM is forcing with this complete Boolean algebra, with Boolean values ∥φ∥∈B(Θ)\|\varphi\|\in\mathbb B(\Theta). The source writes Bω2\mathbb B_{\omega_2} for B(ω2×ω)\mathbb B(\omega_2\times\omega). For S⊆ΘS\subseteq\Theta and u∈2Θu\in2^\Theta, u↾S∈2Su\restriction S\in2^S is the restriction.

The following standard facts are used on this page and the later forcing pages. For random forcing and measure algebras the source lists K. Kunen, Random and Cohen reals, Handbook of Set-Theoretic Topology (1984), 887--911; the Boolean-valued forcing facts are in T. Jech, Set Theory, third millennium edition, Chapters 14--15, and the descriptive set theory in A. S. Kechris, Classical Descriptive Set Theory (1995).

  • (R1) Countable supports. B(Θ)\mathbb B(\Theta) is a complete Boolean algebra with the countable chain condition, and every μΘ\mu_\Theta-measurable set is μΘ\mu_\Theta-almost equal to a set of the form W×2Θ∖SW\times2^{\Theta\setminus S} with S⊆ΘS\subseteq\Theta countable and W⊆2SW\subseteq2^S Borel. A countable SS supports an element of B(Θ)\mathbb B(\Theta) if the element has such a representative, and supports a name if it supports every Boolean value occurring in the name (the source's wording); a support may always be enlarged.
  • (R2) The generic point. G˙\dot G denotes the canonical name for the point uG∈2Θu_G\in2^\Theta with uG(θ)=1u_G(\theta)=1 if and only if [{u:u(θ)=1}]∈G[\{u:u(\theta)=1\}]\in G. For countable S⊆ΘS\subseteq\Theta and Borel W⊆2SW\subseteq2^S coded in MM, ∥G˙↾S∈W∥=[W×2Θ∖S]\|\dot G\restriction S\in W\|=[W\times2^{\Theta\setminus S}], where WW is reinterpreted in the extension from its code. In particular a Borel W⊆2SW\subseteq2^S is μS\mu_S-null if and only if ⊩G˙↾S∉W\Vdash\dot G\restriction S\notin W, and the condition [W×2Θ∖S][W\times2^{\Theta\setminus S}], when nonzero, forces G˙↾S∈W\dot G\restriction S\in W.
  • (R3) Forcing theorem and maximum principle. If a condition qq forces ∃x φ(x)\exists x\,\varphi(x), there is a name x˙\dot x with q⊩φ(x˙)q\Vdash\varphi(\dot x); names may be mixed along a partition of unity.
  • (R4) Absoluteness. Standard Borel spaces, Borel sets and Borel maps coded in MM are reinterpreted in M[G]M[G] from the same codes, a coded preimage, complement or countable union being reinterpreted as the preimage, complement or union of the reinterpretations; Borel statements about points of MM are absolute between MM and M[G]M[G]; and a Π11\Pi^1_1 statement about the coded objects that holds in MM holds in M[G]M[G] (Mostowski's absoluteness theorem, T. Jech, Set Theory, third millennium edition, Chapter 25), in particular that a coded Borel map is injective, carries a coded set into a coded set, or is inverse to another coded map. A name for an element of a standard Borel space XX is a name z˙\dot z with ⊩z˙∈X\Vdash\dot z\in X for the reinterpreted XX.
  • (R5) Borel isomorphism. Every standard Borel space is Borel isomorphic to a Borel subset of 2ω2^\omega.

Statement

The following is provable in ZFC. Let XX be a standard Borel space and z˙\dot z a B(Θ)\mathbb B(\Theta)-name for an element of XX. There are a countable S⊆ΘS\subseteq\Theta and a Borel map F ⁣:2S→XF\colon2^S\to X such that

⊩B(Θ)z˙=F(G˙↾S).\Vdash_{\mathbb B(\Theta)}\dot z=F(\dot G\restriction S).

We say that such an SS reads z˙\dot z through FF. If SS reads z˙\dot z through FF and S⊆S′S\subseteq S' is countable, then S′S' reads z˙\dot z through u↦F(u↾S)u\mapsto F(u\restriction S), because (G˙↾S′)↾S=G˙↾S(\dot G\restriction S')\restriction S=\dot G\restriction S.

Proof

The case X=2ωX=2^\omega. For each n<ωn<\omega the Boolean value bn=∥z˙(n)=1∥b_n=\|\dot z(n)=1\| lies in B(Θ)\mathbb B(\Theta). By (R1) choose a countable Sn⊆ΘS_n\subseteq\Theta and a Borel Wn⊆2SnW_n\subseteq2^{S_n} with bn=[Wn×2Θ∖Sn]b_n=[W_n\times2^{\Theta\setminus S_n}]. Put S=⋃nSnS=\bigcup_nS_n, countable, and define F ⁣:2S→2ωF\colon2^S\to2^\omega by

F(u)(n)=1  ⟺  u↾Sn∈Wn.F(u)(n)=1\iff u\restriction S_n\in W_n.

Each coordinate of FF is the indicator of a Borel set, so FF is Borel. By (R2), ∥G˙↾Sn∈Wn∥=bn=∥z˙(n)=1∥\|\dot G\restriction S_n\in W_n\|=b_n=\|\dot z(n)=1\| for every nn, so ⊩z˙(n)=F(G˙↾S)(n)\Vdash\dot z(n)=F(\dot G\restriction S)(n) for every nn, and hence ⊩z˙=F(G˙↾S)\Vdash\dot z=F(\dot G\restriction S).

General XX. By (R5) fix a Borel isomorphism ι\iota of XX onto a Borel set X′⊆2ωX'\subseteq2^\omega. Then ι(z˙)\iota(\dot z) is a name for an element of 2ω2^\omega; by the first case obtain countable SS and Borel F0 ⁣:2S→2ωF_0\colon2^S\to2^\omega with ⊩ι(z˙)=F0(G˙↾S)\Vdash\iota(\dot z)=F_0(\dot G\restriction S). The set N=F0−1(2ω∖X′)N=F_0^{-1}(2^\omega\setminus X') is Borel, and by (R2) and (R4) ∥G˙↾S∈N∥=∥ι(z˙)∉X′∥=0\|\dot G\restriction S\in N\|=\|\iota(\dot z)\notin X'\|=0, so NN is μS\mu_S-null. Fix x0∈Xx_0\in X and define F(u)=ι−1(F0(u))F(u)=\iota^{-1}(F_0(u)) for u∉Nu\notin N and F(u)=x0F(u)=x_0 for u∈Nu\in N. Then FF is Borel, and since ⊩G˙↾S∉N\Vdash\dot G\restriction S\notin N, ⊩z˙=F(G˙↾S)\Vdash\dot z=F(\dot G\restriction S).

Boundary. The lemma is applied in Proposition 4.4 to names for elements of OZ\mathcal O^{\mathbb Z} and in Lemma 4.5 to a name for a Borel code.