On finiteness of some verbal subgroups in profinite groups (Q1998971)

Let \(w=w(x_1, \ldots, x_k)\) be a group word, that is a nontrivial element of the free group on \(x_1, \ldots, x_k\), and \(G\) a group. Then, \(w\) can be viewed as a \(k\)-variable function defined on \(G\). The set of \(w\)-values in \(G\) is denoted \(G_w\) and the subgroup generated by \(G_w\) is denoted \(w(G)\). If \(G\) is a topological group then \(w(G)\) is taken to be the closed subgroup. The concept of conciseness of a word \(w\) in a class of groups \({\mathcal C}\), that is whether the finiteness of \(G_w\) implies the finiteness of \(w(G)\) for all \(G\) in \({\mathcal C}\), has a long history. More recently, a variation of conciseness for profinite groups has been considered. A word \(w\) is strongly concise in a class \({\mathcal C}\) of profinite groups if \(|G_w| < 2^{\aleph_0}\) implies that \(w(G)\) is finite for all \(G\) in \({\mathcal C}\). In this article, the authors consider a closely related question. They show that, for several families of words, if you suppose \(G\) is a profinite group with \(|G_w|< 2^{\aleph_0}\) and \(w(G)\) generated by finitely many \(w\)-values, then \(w(G)\) is finite. In the first theorem, the authors consider words of type \([y, _{\;n} v^q]\) and \([v^q, _{\;n} y]\) where \(v\) is the left normed commutator \([x_1, x_2, \ldots, x_k]\) and \([y, _{\;n} x] = [y,x, \ldots, x]\) with \(x\) repeated \(n\) times and \(k\), \(n\) and \(q\) all positive integers. The second theorem is more technical to state, but covers many families of words.