Multiplicative subgroups in weakly locally finite division rings (Q2667219)

A division ring is said to be \textit{centrally finite} if it has finite dimension over its center; instead, if every its finite subset generates a division subring which is finite dimensional over the center then it is called a \textit{locally finite} division ring. It follows immediately from the definitions that every centrally finite division ring is locally finite. In 1986, in his PhD thesis [Locally supersoluble skew linear groups. London: Queen Mary College (PhD Thesis) (1986)], \textit{A. I. Stewart} introduced a wider class of division rings. These division rings are locally finite dimensional over a subfield that is not necessarily central, and such division rings are called \textit{Stewart's division rings} for short. Stewart also gave examples of division rings which are locally finite dimensional over a non-central subfield (see [Arch. Math. 49, 277--280 (1987; Zbl 0599.20081)] and [J. Lond. Math. Soc., II. Ser. 35, No. 1, 91--97 (1987; Zbl 0578.20048)]). These examples demonstrate that the class of locally finite division rings is properly contained in that of Stewart's division rings. Recently, the notion of a \textit{weakly locally finite division ring} was introduced in [\textit{T. T. Deo} et al., Acta Math. Vietnam. 44, No. 2, 553--569 (2019; Zbl 1456.16013)]. By definition, a division ring is weakly locally finite if every its finite subset generates over the prime subfield a centrally finite division ring. It was turned out that every locally finite division ring is weakly locally finite. In the same article, they also gave an example of a weakly locally finite division ring which is not locally finite. Weakly locally finite division rings are also called \textit{locally PI}. There is a famous example of a non-algebraic locally PI division ring which due to J. C. McConnell; however, the details were never published by him. So far the example have been published by R. A. Wadsworth in [\textit{C. J. McConnell} and \textit{A. R. Wadsworth}, ``Locally PI but not PI division rings of arbitrary GK-dimension'', Preprint, \url{arXiv:1803.09423v1}]. The upshot of above discussion is that, both classes of Stewart's division rings and of weakly locally finite division rings strictly contain that of locally finite division rings. An important observation pointed out in the paper under review is that every Stewart's division ring is weakly locally finite. Also, it is conjectured that the class of weakly locally finite division rings strictly contains that of Stewart's division rings. In the paper under review, the authors also study some algebraic structures of multiplicative subgroups in a weakly locally finite division ring. The obtained results can be summarised as follows. Let \(D\) be a weakly locally finite division ring with center \(F\), and \(D^*\) be the multiplicative group of \(D\). Then: \begin{itemize} \item[1.] Let \(G\) be a subgroup of \(D^*\). Then \(G\) does not contain a non-cyclic free subgroup iff \(G\) is abelain-by locally finite iff every finitely generated subgroup of \(G\) satisfies a group identity. \item[2.] Let \(H\leq G\) are non-central subgroup of \(D^*\). If \(H\) is almost subnormal in \(G\), then \(H\) contains a non-cyclic free subgroup iff so does \(H^G\). \item[3.] Every locally nilpotent subgroup of \(D^*\) is center-by-locally finite and also metabelian. \item[4.] Every locally supersolvable subgroup of \(D^*\) is solvable of class at most 3. \item[5.] Let \(G\) be a non-central almost subnormal subgroup of \(D^*\). Assume that \(M\) is a non-abelian maximal subgroup of \(G\). If \(M\) does not contain non-cyclic free subgroups, then \(D\) is a Stewart's division ring. And, if we assume further that \(M\) is algebraic over \(F\), then \(D\) is a locally finite division ring. \end{itemize}