On the residual finiteness growths of particular hyperbolic manifold groups (Q2520773): Difference between revisions

It has been known for a long time that linear groups, in particular the fundamental groups of (complete) hyperbolic manifolds, are residually finite. Other similar properties such as LERFness and its refinements have been the focus of more recent attention and are now known to hold for many (by no means all) fundamental groups of hyperbolic manifolds. In another direction a natural question is to attempt a quantification of the phenomenon of residual finiteness, and this is what this paper is concerned with. To this end the following type of functions were introduced in [J. Algebra 323, No. 3, 729--737 (2010; Zbl 1222.20020)] by \textit{K. Bou-Rabee}: let \(G\) be residually finite and for \(g \in G\) a nontrivial element let \(D_G(g) = \min_H |G/H|\) where the minimum is taken over all normal subgroups \(H \leq G\) such that \(g \not\in H\). Then for any function \(f\) from \(G\) to \(\mathbb R_+\) one can define the associated residual finiteness growth: it is the function sending \(n \geq 0\) to the maximal value of \(D_G(g)\) for \(g\) with \(f(g) \leq n\). A natural choice for \(f\) when \(G\) is finitely generated is a word length function; for fundamental groups of compact hyperbolic manifolds one can take the length of closed geodesics to define another function (in this case both possibilities give essentially the same answer). In this paper the author proves that for a certain class of hyperbolic manifolds in dimensions 3 and 4 the residual finiteness growth is at most linear. Geometrically, this means that if \(M\) is a compact hyperbolic manifold to which the result applies there is \(C > 0\) such that if \(\gamma \subset M\) is a closed geodesic of length \(\ell\) there exists a finite cover \(M' \to M\) of degree at most \(C\ell\) such that \(\gamma\) does not lift to \(M'\). There is also an upper bound for \(C\) in terms of the geometry of \(M\). The class of manifolds for which the result is proven are those which appear as totally geodesic submanifolds in finite covers of right-angled polytopes (viewed as orbifolds). This includes of course cocompact reflection groups in hyperbolic spaces of dimension 3 and 4 and groups commensurable to such, but also many others such as Bianchi groups. The proof of the main results proceeds by explicitely constructing, for a given closed geodesic, a finite cover of the orbifold to which it does not lift, by gluing copies of the fundamental polyhedron. The precise technique used is the same as in the author's previous work on surfaces [\textit{P. Patel}, Proc. Am. Math. Soc. 142, No. 8, 2891--2906 (2014; Zbl 1311.57003)]. Of course the computations on hyperbolic geometry are more complicated the higher the dimension gets.