Central quotients of biautomatic groups (Q1359432)

The main theorem of this paper is that if \(G\) is a biautomatic group and \(C\) is a central subgroup of \(G\), then \(G/C\) is biautomatic. This implies that direct factors of biautomatic groups are biautomatic, answering a question posed by S. Gersten and H. Short. It also gives some information about an analogue of Thurston's geometrization theorem for 3-manifolds which conceivably holds for biautomatic groups: is it true that every biautomatic group either is word hyperbolic or has a \(\mathbb{Z}^2\) subgroup? The main theorem implies that if a biautomatic group contains a subgroup with an infinite index central \(\mathbb{Z}\) subgroup, then \(G\) contains a \(\mathbb{Z}^2\) subgroup. So the analogue can now be stated with a stronger hypothesis: if every infinite cyclic subgroup in a biautomatic group has finite index in its centralizer, is the group word hyperbolic? Still another application shows that if there exists a biautomatic group with an infinitely generated abelian subgroup, then either there exists one with an infinite rank abelian subgroup, or there exists one with an infinite torsion abelian subgroup. The author notes that some recent preprints of W. Neumann and L. Reeves contain results that give a sharpening of his main theorem.