On Loeb and sequential spaces in \textbf{ZF} (Q776389): Difference between revisions

In what follows, all results are \(\mathbf{ZF}\) theorems, so the intended environment is choiceless set theory. \(\mathbf{AC}\) denotes the Axiom of Choice and \(\mathbf{ZFC} = \mathbf{ZF} + \mathbf{AC}\). \(\mathbf{CAC}\) denotes the Countable Axiom of Choice, which states that every infinite countable family of non-empty sets has a choice function. \(\mathbf{CAC}_\omega\) denotes the restriction of \(\mathbf{CAC}\) to countable families of non-empty countable sets and \(\mathbf{CAC}_\omega(\mathbb{R})\) denotes the restriction of \(\mathbf{CAC}_\omega\) to countable families of non-empty countable subsets of \(\mathbb{R}\). A metric space \(\mathbf{X} = (X,d)\) is said to be \textit{Cantor complete} if every decreasing sequence of non-empty closed sets whose diameters converge to zero has non-empty intersection. A topological space \(\mathbf{X} = (X,\tau)\) is said to be \textit{Cantor completely metrizable} if there exists a Cantor complete metric \(d\) on \(X\) which induces the topology \(\tau\) on \(X\). A topological space is said to be a \textit{Loeb} space if the family of all non-empty closed sets has a choice function. A subset \(A\) of a topological space \(X\) is said to be \textit{sequentially closed} if it is closed under limits of converging sequences (in \(X\)) of its points. A topological space is said to be \textit{sequential} if every sequentially closed subset of \(X\) is closed. In the paper under review, the authors investigate Loeb and sequential spaces in the absence of the Axiom of Choice. Typical results of the paper are as follows: \((i)\) Every Cantor completely metrizable second countable space is Loeb. \((ii)\) If \(\mathbf{X}\) is a Cantor completely metrizable second countable space, then \(\mathbf{X}^\omega\) is a Cantor completely metrizable second countable Loeb space. \((iii)\) \(\mathbf{AC}\) if, and only if, every sequential Cantor completely metrizable space is Loeb. \((iv)\) The statement ``Every sequential subspace of \(\mathbb{R}\) is Loeb''\, implies \(\mathbf{CAC}_\omega(\mathbb{R})\). \((v)\) \(\mathbb{R}\) is sequential if, and only if, every second countable compact Hausdorff space is sequential. The paper finishes by posing some open questions. For instance: Is there a model of \(\mathbf{ZF}\) in which there are metrizable sequential spaces \(\mathbf{X}\) and \(\mathbf{Y}\) such that \(\mathbf{X} \times \mathbf{Y}\) is not sequential? Must \(\mathbb{R}^\omega\) be sequential in every model in which \(\mathbb{R}\) is sequential? Must \(\mathbb{N}^\omega\) be sequential in every model of \(\mathbf{ZF}\) in which \(\mathbb{R}\) is sequential?