Not OCA and products of Fréchet spaces
Alan Dow · 2025
Read Not OCA and products of Fréchet spaces by Alan Dow (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-01662.txt.
Introduction
Alan Dow asks whether the product of two countable Fréchet spaces must be M-separable, and what Martin's Axiom does to the answer. The paper from twenty twenty-five continues a set-theoretic topology investigation with a delicate status. The question is known to be independent of Martin's Axiom, but only in models where the continuum stays small. The Open Coloring Axiom, by contrast, forces an affirmative answer.
Dow probes the territory between those poles. Products of Fréchet spaces matter because separability properties often fail to multiply. Each axiom reshapes which products behave. The paper is a progress report in a long independence saga, mapping where the question stays open. Set theorists and topologists share the audience.
Catalog record
| CREATOR | Alan Dow |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 60 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2507-01662
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Not OCA and products of Fréchet spaces Alan Dow Abstract . We continue the investigation of the question of whether the product of two countable Fréchet spaces must be M-separable. We are especially interested in this question in the presence of Martin’s Axiom. The question has been shown to be independent of Martin’s Axiom but only in models in which 𝔠 ≤ ω 2 \mathfrak{c}\leq\omega_{2} . In fact, OCA implies an affirmative answer. \DefineSimpleKey bibprimaryclass \DefineSimpleKey bibarchiveprefix 1991 Mathematics Subject Classification 03E50, 03E35, 54A35, 54D55 1. Introduction A space is Fréchet (or Fréchet-Urysohn) if it satisfies that if a point x is in the closure of a set A , there is a standard \omega -sequence of points from A converging to x . A space X is M-separable, also known as selectively separable, if for every countable family \{D_{n}:n\in\omega\} of dense subsets, there is selection \{H_{n}\in[D_{n}]^{<\aleph_{0}}:n\in\omega\} satisfying that H=\bigcup_{n}H_{n} is dense. Every separable Fréchet space is M-separable [ 4 ] . A product of two separable Fréchet spaces need not be Fréchet, but might the product still be M-separable? There are two main results of this paper. The first is that in models of Martin’s Axiom in which there are special (\mathfrak{c},\mathfrak{c}) -gaps, there will be pairs of countable Fréchet spaces with a product that is not M-separable. The second is that in standard models of \operatorname{MA}(\sigma\mbox{-linked}) , it will hold that the product of any two countable Fréchet spaces will be M-separable. In both results there is no (new) restriction on the size of the continuum. \operatorname{MA}(\sigma\mbox{-linked}) is the statement (see [ 2 ] ) that Martin’s Axiom holds for ccc posets that can be expressed as a countable union of linked subsets. Any countable space with \pi -weight less than \mathfrak{d} is M-separable and, in the Cohen model every countable Fréchet space has \pi -weight at most \aleph_{1} [ 5 ] . Therefore we are more interested in the question in models in which there are countable Fréchet spaces with \pi -weight at least \mathfrak{d} . In fact in this paper we focus on models in which \mathfrak{b}=\mathfrak{c} . It is interesting that it was shown in [ 14 ] that \mathfrak{b}=\mathfrak{d} implies there are countable M-separable spaces whose product is not M-separable, but the status of this statement in ZFC is very much open. The cardinals \mathfrak{p},\mathfrak{b} , and \mathfrak{d} are the usual cardinal invariants corresponding to mod finite orderings on subsets of \omega known as the pseudointersection number, the bounding number, and the dominating number. Back to the product of countable Fréchet spaces in models of \mathfrak{b}=\mathfrak{c} . The known results seem to point to a close connection to the open coloring axiom and gaps. It was shown in [ 3 ] that in a model of Martin’s Axiom plus \mathfrak{c}=\omega_{2} in which there was a strong failure of OCA, this strong failure of OCA was crucial to the construction of two countable Fréchet spaces whose product was not M-separable. Improving on the PFA result in [ 5 ] , it was shown in [ 9 ] that the version of OCA from [ 16 ] , which was shown to imply \mathfrak{b}=\omega_{2} , implies that the product of two countable Fréchet spaces is necessarily M-separable. The key step in this proof was that this version of OCA implies that most unseparated pairs of orthogonal ideals on \omega will contain a Luzin gap (see Definition 2.3 ). Two families of subsets of a countable set are said to be orthogonal if each member of the first family has finite intersection with each member of the second. The other well-known version of OCA, denoted OCA {}_{[\operatorname{ARS}]} in [ 12 ] , is from [ 1 ] . On the other hand, adding to the seeming OCA connection, it was also shown in [ 9 ] that Martin’s Axiom plus not CH implies there are three countable Fréchet spaces whose product is not M-separable because , by Avilés-Todorčević [ 2 ] , in such models there are no three dimensional analogues of the above mentioned Luzin gaps. We note that a space X is M-separable if every countable filter base \mathcal{D} consisting of dense subsets of a space X has a pseudointersection that is dense. Since M stands for Menger, one could define the Menger degree of a space X , \mathfrak{p} -M (X) , to be the minimum cardinality of a filter base of dense subsets of X that has no dense pseudointersection. A natural family to consider is those countable Fréchet spaces with Menger degree equal to the pseudointersection number \mathfrak{p} . We did begin work on this paper by considering whether this smaller family of spaces may have better behavior in products but could find no results. We leave this remark here as a simple suggestion for further research. Motivated by the paper [ 8 ] , we had hoped to completely solve this question in this paper, but it remains open. Question 1 . Does \mbox{MA}+\mathfrak{c}>\omega_{2} imply there are two countable Fréchet spaces whose product is not M-separable? 2. A few combinatorial tools The following observation, a strengthening of Arhangel’skii’s \alpha_{1} -property for first countable spaces, has proven useful in a number of papers. This statement and proof is taken right from [ 9 ] and is included for completeness. Proposition 2.1 . Let \mathcal{I} be a family of sequences in a countable space X all converging to a single point x that has countable character. If \mathcal{I} has cardinality less than \mathfrak{b} , then there is a single sequence S converging to x that mod finite contains every member of \mathcal{I} . Proof. Fix a descending neighborhood basis, \{U_{n}:n\in\omega\} , for x with U_{0}=X . For each n\in\omega , let X_{n}=U_{n}\setminus U_{n+1} . There is nothing to prove if x has a neighborhood that is simply a converging sequence, so we may assume that each X_{n} is infinite. For each n\in\omega , choose an enumeration, \{x(n,m):m\in\omega\} of X_{n} . For each I\in\mathcal{I} , there is a function f_{I}\in\omega^{\omega} satisfying that I\subset\bigcup_{n}\{x(n,m):m<f_{I}(n)\} . Therefore, if |\mathcal{I}|<\mathfrak{b} , we may choose a function f\in\omega^{\omega} so that f is eventually larger than each f_{I} . It is easy to check that S=\bigcup_{n}\{x(n,m):m<f(n)\} is a sequence that converges to x and which satisfies that I\setminus S is finite for all I\in\mathcal{I} . ∎ For a set A in a space X , we let A^{(1)} be the set of points x of X for which there is a countable, possibly constant, sequence from A converging to x . Proposition 2.2 ( [ 7 ] ) . If a space X has character less than \mathfrak{b} , then for every set A\subset X , the set (A^{(1)})^{(1)}=A^{(1)} . The next two items are taken from [ 16 ] *§8 and [ 10 ] *Theorem 2.2.1 Definition 2.3 . A family \{(I_{\alpha},J_{\alpha}):\alpha<\omega_{1}\} is a Luzin gap if \bigcup\{I_{\alpha}\cup J_{\alpha}:\alpha\in\omega_{1}\} is countable, , for each \alpha\neq\beta , I_{\alpha}\cap J_{\alpha} is empty, I_{\alpha}\cap J_{\beta} is finite, and (I_{\alpha}\cap J_{\beta})\cup(I_{\beta}\cap J_{\alpha}) is not empty. Proposition 2.4 . If \{(I_{\alpha},J_{\alpha}):\alpha<\omega_{1}\} is a Luzin gap, then the family \{I_{\alpha}:\alpha\in\omega_{1}\} can not be mod finite separated, or split, from the family \{J_{\alpha}:\alpha\in\omega_{1}\} . For