Lower bound on growth of non-elementary subgroups in relatively hyperbolic groups (Q2170265)

Let \(S\) be a finite symmetric generating set of the group \(H\) and \(d_{s}\) the corresponding word metric. The (logarithmic) growth rate of \(H\) with respect to \(S\) is defined to be \(\omega (H, S) = \lim_{n \rightarrow \infty }\frac{\log \sharp (S^{\leq n})}{n}\), where \(S^{\leq n}\) denotes the sphere with center 1 and ray \(n\) for all \(n\in \mathbb{N}\cup\{ 0 \}\). The spectrum of growth rates is defined as \[ \Omega (H) = \{ \omega (H, S)\mid \sharp S<\infty\,\,\, \langle S\rangle = H \}. \] A group \(H\) has uniform exponential growth if \(\inf \Omega (H)>0\). \textit{J. S. Wilson} [Invent. Math. 155, No. 2, 287--303 (2004; Zbl 1065.20054)] constructed the first example of a group \(H\) with non-uniform exponential growth. Therefore, if a group has uniform exponential growth, it is interesting to ask whether \(\Omega (H)\) obtains the minimum. In [Ann. Sci. Éc. Norm. Supér. (4) 35, No. 4, 477--488 (2002; Zbl 1018.20036)] it showed by \textit{A. Sambusetti} that the answer is negative for the free product of two non-Hopfian groups. Notice that free products are special cases of relatively hyperbolic groups. On the other hand for the case of hyperbolic groups the answer if affirmative, a result obtained in [\textit{K. Fujiwara} and \textit{Z. Sela},``The rates of growth in a hyperbolic group'', Preprint, url{arXiv:2002.10278}]. Instead studying \(\Omega (H)\) it is interesting to study \[ \Theta (H) = \{ \frac{\omega (H, S)}{\log \sharp S}\mid \sharp S<\infty \,\,\, \langle S\rangle = H \}. \] Of course \(\Theta (H)\subset [0, 1]\) and we have that \(1\in \Theta (H)\) if and only if \(H\) is a free group [\textit{R. Grigorchuk} and \textit{P. de la Harpe}, J. Dyn. Control Syst. 3, No. 1, 51--89 (1997; Zbl 0949.20033)]. Analogous to the question of uniform exponential growth, we can ask for which groups \(\Theta (H)\) admits a positive infimum. In fact, \(\inf \Theta (H)> 0\) has been obtained for hyperbolic groups in [\textit{G. N. Arzhantseva} and \textit{I. G. Lysenok}, J. Lond. Math. Soc., II. Ser. 73, No. 1, 109--125 (2006; Zbl 1093.20024)]. In this paper it is obtained a generalization of the previous result in to the class of relatively hyperbolic groups. There are many equivalent definitions of a relatively hyperbolic group. Here the authors quote one appropriate to formulate their main results (definition 2.10 in the paper). Let \(G\) be a group acting properly on a proper hyperbolic space \(X\). The limit set \(\Lambda H\) of a subgroup \(H\) of \(G\) is the set of accumulation points in \(\partial X\) of any \(H\)-orbit in \(X\). A subgroup \(H\) is called elementary if its limit set contains at most two points. So a subgroup is non-elementary if its limit set contains at least three points. If \(\Lambda H\) consists of only one point \(p\), then \(H\) is called parabolic and the point is said to be a parabolic point. Theorem A. Assume that \(G\) is a non-elementary relatively hyperbolic group. Then there exists a constant \(\kappa = \kappa (G)\in (0, 1]\), such that for any non-elementary subgroup \(H\) with a finite symmetric generating set \(S\), we have \(\omega (H, S)\geq \kappa\cdot \log \sharp S\). Let \(\mathcal{P}\) denotes the set of maximal parabolic subgroups in \(G\). The group \(G\) has locally uniform exponential growth if every finitely generated subgroup of exponential growth has uniform exponential growth. Theorem B. Any non-elementary finitely generated subgroup \(H\) of a non-elementary relatively hyperbolic group \(G\) has uniform exponential growth. Moreover, the pair \((G, \mathcal{P})\) has locally uniform exponential growth if and only if every maximal parabolic subgroup \(P\in \mathcal{P}\) has locally uniform exponential growth. In [\textit{X. Xie}, Proc. Am. Math. Soc. 135, No. 3, 695--704 (2007; Zbl 1121.20033)] it is proved that relatively hyperbolic groups have uniform exponential growth. Theorem B is a strengthening of this result. As it is noted by the authors Theorem B has been proved independently and almost simultaneously in [\textit{R. Kropholler}, ``Extensions of hyperbolic groups have locally uniform exponential growth'', Preprint, \url{arXiv:2012.14880}]. Also they refer that follow closely the strategy of [\textit{G. N. Arzhantseva} and \textit{I. G. Lysenok}, J. Lond. Math. Soc., II. Ser. 73, No. 1, 109--125 (2006; Zbl 1093.20024)]. But to overcome the difficulties during transition from hyperbolic groups to relatively hyperbolic groups they adapted techniques from [\textit{X. Xie}, Proc. Am. Math. Soc. 135, No. 3, 695--704 (2007; Zbl 1121.20033)] and strengthening a result in [\textit{M. Koubi}, Ann. Inst. Fourier 48, No. 5, 1441--1453 (1998; Zbl 0914.20033)].