Marked length spectrum rigidity in groups with contracting elements (Q6986635)

This article develops a comprehensive theory of marked length spectrum (MLS) rigidity for isometric group actions on geodesic metric spaces endowed with contracting elements. Let \(G \curvearrowright X_i\) (for \(i = 1,2\)) be two such actions. The central result, Theorem~5.1, asserts that if the MLS functions \(\ell_{d_1}(g) = \ell_{d_2}(g)\) for all \(g \in G\), then the orbit maps \(Go_1 \to Go_2\) are rough isometries. A more quantitative refinement (Theorem~5.2) proves that if \(\lambda_1 \ell_{d_2}(g) \le \ell_{d_1}(g) \le \lambda_2 \ell_{d_2}(g)\) for all \(g \in G\), then the distance functions satisfy\N\[\N\lambda_1 d_2(g o_2, h o_2) - C \le d_1(g o_1, h o_1) \le \lambda_2 d_2(g o_2, h o_2) + C,\N\]\Nfor all \(g,h \in G\) and some uniform \(C \ge 0\).\N\NThe notion of contraction is formulated via coarse projections: a subset \(Y \subset X\) is \(C\)-contracting if for every geodesic segment \(\gamma\) with \(\gamma \cap N_C(Y) = \emptyset\), the projected diameter satisfies \(\operatorname{diam}(\pi_Y(\gamma)) \le C\). A group element \(g\) is called contracting if the orbit of \(\langle g \rangle\) defines a contracting quasi-geodesic.\N\NUsing the extension lemma (Lemma~3.1), the authors construct a finite set \(F \subset G\) of cardinality \(3\) such that such that for any \(g,h \in G\) there exists \(f \in F\) satisfying \[\left| d(o, gfh \cdot o) - d(o, g \cdot o) - d(o, h \cdot o) \right| \le \epsilon_0\] for a fixed constant \(\epsilon_0 > 0\). This property enables the construction of large families of so-called simultaneously contracting elements across both actions, which are then used to relate stable lengths and derive uniform distortion bounds.\N\NThe results are extended to relatively hyperbolic groups acting cusp-uniformly (Theorem~6.17), yielding rough isometries between the entire spaces. Further rigidity results are established from subgroups \(H \le G\), including normal subgroups with unbounded orbits (Theorem~7.2), confined subgroups under properness assumptions (Theorem~7.19) and geometrically dense subgroups satisfying an ``extension condition'' (Theorem~8.11).\N\NA final part of the paper (Section~9) addresses MLS rigidity from subsets \(E \subset G\) with large growth. The key tool is the Manhattan curve, defined via the convergence domain of the series\N\[\N\mathcal{P}(a,b) = \sum_{g \in G} \exp(-a\, d_2(o_2, g o_2) - b\, d_1(o_1, g o_1)),\N\]\Nwhose boundary is the graph of the function \(\theta(a) := \inf \{ b \in \mathbb{R} \mid \mathcal{P}(a,b) < \infty \}\).\N\NUnder convexity assumptions on this curve and assuming \(\delta_1(G) = \delta_2(G)\), Theorem~9.17 shows that \(G\) has MLS rigidity from any subset \(E \subset G\) satisfying \(\delta_c^{(1)}(E) = \delta_c^{(1)}(G)\), where \(\delta_c^{(1)}\) denotes the exponential growth rate of conjugacy classes with respect to \(\ell_{d_1}\).