Set-theoretic aspects of periodic FC-groups. -- Extraspecial \(p\)-groups and Kurepa trees (Q1320181)

Centre-by-finite groups were characterized by \textit{B. H. Neumann} [Math. Z. 63, 76-96 (1955; Zbl 0064.25201)] as those groups in which each (abelian) subgroup has finitely many conjugates. This is a theorem about FC-groups and one can ask whether the following are equivalent for an FC- group \(G\) and an infinite cardinal \(\alpha\): (1) \(|G/Z(G)|<\alpha\), (2) \(|G:N_ G(U)|<\alpha\), for all subgroups \(U\leq G\), (3) \(|G: N_ G(A)|<\alpha\), for all abelian subgroups \(A\) of \(G\). In this interesting article the cases \(\alpha = \omega_ 1\) and \(\omega_ 2\) are considered and rather different results are obtained for the two cases. For \(\omega_ 1\) we have Theorem A: Assuming CH, there is an FC-group \(G\) with \(|G/Z(G)|=\omega_ 1\) but \(|G:N_ G(A)|<\omega_ 1\), for all abelian \(A\leq G\). But Theorem B: It is consistent that there is no such example. For finite-by-abelian groups it is proved that (1), (2), (3) are equivalent for \(\alpha = \omega_ 1\) and this suggests a possible approach for proving the equivalence of (1) and (2). For \(\omega_ 2\) we have Theorem D: It is consistent that there is an extraspecial \(p\)-group \(G\) with \(|G|=\omega_ 2\) and \(|G:|<\omega_ 2\) for all maximal abelian subgroups \(M \leq G\) (and hence \(|G:N_ G(A)|<\omega_ 2\) for all abelian subgroups \(A \leq G\)). This example, which requires the existence of a Kurepa tree, can be thought of as dual to Ehrenfeucht-Faber extraspecial groups in which all abelian subgroups have cardinality less than \(|G|\). Theorem E: Assuming the consistency of ZFC + `there is an inaccessible cardinal' it is consistent that (1), (2), (3) are equivalent for \(\alpha = \omega_ 2\).