New topological methods to solve equations over groups (Q507054)

The present paper is an important contribution to the theory of equations over groups. It is impossible to list here the variety of methods and ideas which pervade the proofs of the main results. Just a brief comment may illustrate the reasons of its scientific importance: Conjecture: Let \(G\) be a group and \(w \in G * F_n\) be an equation in \(n\) variables with coefficients in \(G\), where \(G*F_n\) denotes the free product of \(G\) with the free group on \(n\) generators. If \(w\) is non-singular, then it is solvable over \(G\). In addition, if \(G\) is finite, then a solution can be found in a finite extension. In the case \(n=1\) the conjecture above is known as the Kervaire-Laudenbach Conjecture and was proved by \textit{M. Gerstenhaber} and \textit{O. S. Rothaus} for locally residually finite groups. They showed in [Proc. Natl. Acad. Sci. USA 48, 1531--1533 (1962; Zbl 0112.02504)] that if \(G\) is finite, then every non-singular equation in one variable can be solved over \(G\) (and in fact in some finite extension of \(G\)). The results of the paper under review generalize the methods of Gerstenhaber-Rothaus, improving the above conjecture to wider classes of groups. The significant intuition has been to involve the notion of hyperlinear group, opening a new perspective on the use of cohomological methods in the theory of equations over groups and in the theory of extensions.