Analytically heavy spaces: analytic Cantor and analytic Baire theorems (Q624389): Difference between revisions

Motivated by some observations on the central role of analyticity in apparently unconnected areas of analysis, topology and mathematical logic, this work aims to prove some variants of the classical Baire theorem (which says that in a Polish space the intersection of countably many open dense sets is dense) in the context of analytic and \(\mathcal{K}\)-analytic sets. This supports the heuristic point of view that completeness assumptions can often be dropped and replaced with a suitable form of analyticity. In the paper under review, the author extends a result of van Mill and obtains a generalization of a theorem due to Oxtoby. This immediately allows to prove a \(\mathcal{K}\)-analytic version of the Baire theorem (Theorem 5) in which Polish spaces are replaced with regular Hausdorff spaces and the role of open dense sets is played by dense \(\mathcal{K}\)-analytic and ``heavy'' (i.e.\ everywhere large in the sense of some \(\sigma\)-ideal) sets. Then various related separation theorems are derived, together with a fine-topology version of the Baire theorem (Theorem 6) which gives a unified approach embracing classical as well as ``generalized'' Gandy-Harrington topologies (including e.g. the Ellentuck topology). The arguments involved mainly rely on the notion of localization (with respect to some \(\sigma\)-ideal) and on some restatements of the Cantor Theorem in \(\mathcal{K}\)-analytic form, which allow to show the weak \(\alpha\)-favourability (with respect to a multiple-target form of the Choquet Banach-Mazur game) of dense \(\mathcal{K}\)-analytic heavy subsets of a regular space.