Countable compactness and finite powers of topological groups without convergent sequences (Q1763617): Difference between revisions

The main contribution of the article is to provide a partial answer to the following question of Comfort concerning countable compactness of powers of topological groups: Given any cardinal number \(\alpha\leq 2^{\mathfrak{c}}\), is there a topological group \(G\) such that \(G^\gamma\) is countably compact for all cardinals \(\gamma<\alpha\), but \(G^\alpha\) is not countably compact? The question was motivated by a long sequence of research results about the failure of countable compactness for topological spaces or topological groups by taking products or powers. Among the variations treating powers of topological groups seems to be the most challenging, and a positive answer to this question within ZFC would be an optimal result in this subject. In this article the author assumes \({\text{ MA}}_{\text{ countable}}\) and answers the question in the affirmative for all finite numbers. The construction of the examples allows the author to deduce the following stronger consequences. First of all the example groups have no non-trivial convergent sequences (in contrast to all the previously known examples witnessing failure of countable compactness by taking powers). The author also obtains the following interesting corollary. Under \(\text{ MA}_{\text{countable}}+2^{<{\mathfrak {c}}}=\mathfrak{c}\) for every positive integer \(n\geq 2\) there are \(2^{\mathfrak{c}}\) many non-homeomorphic group topologies on the additive group of \(2^{\mathfrak{c}}\) such that the \(n\)-th power of each group is countably compact but the \((n+1)\)-th power is not. Two other properties of topological groups are also considered in place of countable compactness and similar results are obtained. The first property is the existence of the so-called suitable sets in a topological group. Dikranjan has asked a question concerning this property similar to the above question of Comfort. The second property concerns the existence of winning strategies in certain two-player games played on topological spaces. There remains the question whether the same can be proved within ZFC or the negation is consistent.