A nonhereditary Borel-cover \(\gamma\)-set (Q1769693)

A (nontrivial open) {cover} \(\mathcal{U}\) of a set \(X\) is a family of open subsets of \(X\) such that \(X=\bigcup\mathcal{U}\) but \(X\not\in\mathcal{U}\). A cover \(\mathcal{U}\) of \(X\) is an {\(\omega\)-cover} of \(X\) if every finite subset of \(X\) is contained in an element of \(\mathcal{U}\). \(\mathcal{U}\) is a {\(\gamma\)-cover} of \(X\) if \(\mathcal{U}\) is infinite and each element of \(X\) is contained in all but finitely many members of \(\mathcal{U}\). These notions were introduced by \textit{J. Gerlits} and \textit{Zs. Nagy} [``Some properties of \(C(X)\). I'', Topology Appl. 14, 151--161 (1982; Zbl 0503.54020)] and are motivated by the study of closure properties in function spaces. The author considers the analogous types of covers where ``open'' is replaced by ``countable Borel''. More specifically, he considers the following property. A set \(X\) is a {Borel-cover \(\gamma\)-set} if every countable \(\omega\)-cover of \(X\) by Borel sets contains a \(\gamma\)-cover. Answering a question raised by \textit{T. Bartoszyński} and \textit{B. Tsaban} [``Hereditary topological diagonalizations and the Menger-Hurewicz conjectures'', Proc. Am. Math. Soc. (to appear)], he proves that, assuming the Continuum Hypothesis, the property of being a Borel-cover \(\gamma\)-set is not hereditary for subsets. This is an anomaly in the sense that many similar properties were proved to be hereditary in Bartoszyński and Tsaban's cited paper. In fact, the author proves the following assertions: (1) Assuming the Continuum Hypothesis, there exists a Borel-cover \(\gamma\)-set of cardinality continuum. (A modification of a construction of Todorčevic.) (2) If there exists a Borel-cover \(\gamma\)-set of cardinality continuum, then there is one with a subset \(Y\) which is not a Borel-cover \(\gamma\)-set. (Using a nice trick involving independent sets.) (3) The set \(Y\) in (2) actually has an {open} \(\omega\)-cover with no \(\tau\)-subcover: \(\mathcal{U}\) is a {\(\tau\)-cover} of \(X\) if every element of \(X\) is in infinitely many elements of \(\mathcal{U}\), and for every \(x,y\in X\) at least one of the sets \(\{U\in\mathcal{U}: x\in U, y\notin U\}\) or \(\{U\in\mathcal{U}: x\notin U, y\in U\}\) is finite. (So every \(\gamma\)-cover is a \(\tau\)-cover.) (4) The last assertion solves two more questions raised by Bartoszyński and Tsaban's cited paper. The main result also implies a negative solution to a question raised by \textit{L. Bukovský, I. Recław} and \textit{M. Repický} [``Spaces not distinguishing convergences of real-valued functions'', Topology Appl. 112, 13--40 (2001; Zbl 1067.54028)], as was pointed out by \textit{B. Tsaban} [``A \(\gamma\)- and \(\sigma\)-set need not be hereditary'', SPM Bulletin 3, 5 (2003)]. Several interesting problems are posed in the paper.