Codimension-1 subgroups and splittings of groups (Q1355590)

The author makes the following definitions: Let \(G\) be a finitely generated group and let \(H\) be a subgroup of \(G\). The number \(e(G,H)\) of ends of the pair \((G,H)\) is the number of ends of the Cayley graph of \(G\) modulo the action of \(H\). The subgroup \(H\) is said to be of codimension 1 if \(e(G,H)>1\). If \(H\) is a codimension-1 subgroup such that \(G\) splits as an amalgam over a subgroup commensurable with a subgroup of \(H\), then \(H\) is said to be associated to a splitting of \(G\). The author makes the following conjecture: Suppose \(H\) is a finitely generated, codimension-1 subgroup of a finitely generated group \(G\). Then either \(H\) is associated to a splitting of \(G\) or \(H\) has a codimension-1 subgroup. The main result of this paper is that the conjecture is true in the case where \(G\) is hyperbolic and \(H\) is quasi-convex. As a corollary, the author obtains that a codimension-1 quasi-convex subgroup satisfying property \(T\) is always associated to a splitting of the ambient hyperbolic group. The author proves also that if \(G\) is a hyperbolic group and \(H\) a quasi-convex subgroup, then either \(G\) has no codimension-1 subgroups or there exists a finite descending sequence of subgroups \(G=H_1>\dots>H_n\), so that \(H_i\) has codimension 1 in \(H_{i-1}\) and \(H_n\) is associated to a splitting of \(H_{n-1}\). In the meanwhile, the author settles a conjecture of Kropholler and Roller concerning quasi-convex subgroups of hyperbolic groups.