“…In fact, Shalev's results show that in the case in which A is a finite 2-group, the hypothesis that A, A g is finite for any g ∈ G can be omitted (if A is of order 2, then, by virtue of Shunkov's theorem [10], even the assumption that G is residually finite is not required). It seems, however, that the 2-group case is special and in general the hypothesis on A, A g , or something similar, is necessary.…”