A two-dimensional version of the Goldschmidt-Sims conjecture. (Q1414663)

The Goldschmidt-Sims conjecture asserts that given two prime numbers \(p,q\) and a bi-regular, \(p,q\)-tree \(T\), there is only a finite number of conjugacy classes in \(\Aut(T)\) of (edge transitive) lattices \(\Gamma\subset\Aut(T)\). The paper under review proves a certain ``two-dimensional'' analogue of this conjecture. Namely, let \(T_1,T_2\) be two regular trees and \(\Delta=T_1\times T_2\). The theory of uniform lattices in \(\Aut(\Delta)\) analogous to that of irreducible lattices in higher rank semisimple Lie groups was developed by \textit{M. Burger} and \textit{S. Mozes} [Publ. Math., Inst. Hautes Études Sci. 92, 113-150 (2000; Zbl 1007.22012)] and \textit{M. Burger}, \textit{S. Mozes} and \textit{R. Zimmer} [in preparation]. The author of the present paper shows [Theorem 1.4] that if \(\Delta=T_1\times T_2\) is a product of two regular trees of prime valences, then there is a finite number of conjugacy classes in \(\Aut(\Delta)\) of irreducible lattices \(\Gamma\subset\Aut(\Delta)\) such that \(X=\Gamma\setminus\Delta\) is a square. In fact, the paper contains a more general theorem [Theorem 1.6], where \(X=\Gamma\setminus\Delta\) is allowed to be a square complex which is locally a product and locally connected in both directions. This more general theorem follows from Theorem 1.4 and the following dichotomy [Theorem 1.7]: Given \(\Gamma\subset\Aut(\Delta)\) as in Theorem 1.4 acting without inversions on both trees and \(X=\Gamma\setminus\Delta\), assume the universal cover \(\widetilde X\) of \(X\) is a product of two trees, then either 1) \(X\) is a square or 2) \(\Gamma\) is torsion-free, \(\widetilde X=\Delta\) and \(\Gamma\cong \pi_1(X,\cdot)\). The proof of this dichotomy utilizes a construction of a square complex of groups based on \textit{A. Haefliger}'s theory [Group theory from a geometrical viewpoint, World Scientific, Singapore. 504-540 (1991; Zbl 0858.57013)] with morphisms in the sense of \textit{H. Bass'} covering theory for graphs of groups [J. Pure Appl. Algebra 89, No. 1-2, 3-47 (1993; Zbl 0805.57001)]. The paper is well written.