Laminations and \(2\)-filling rays on infinite type surfaces (Q6080095)

A surface is of infinite topological type if its fundamental group is not finitely generated. The mapping class groups of surfaces of infinite topological type are called \textit{big} mapping class groups and have recently been intensively studied, see [\textit{J. Aramayona} and \textit{N. G. Vlamis}, in: In the tradition of Thurston. Geometry and topology. Cham: Springer. 459--496 (2020; Zbl 1479.57037)]. A tool to study big mapping class groups has been the \textit{loop graph} \(L(S; p)\), where \(p\in S\) is an isolated puncture inside the surface of infinite type \(S\). To define this graph, we fix a complete hyperbolic metric on the surface. Vertices of the loop graph are isotopy classes of simple geodesics rays which are asymptotic in both directions to \(p\) (``loops''), and edges connect disjoint loops. This is analogous to the curve graph for finite type surfaces. The loop graph is a subgraph of a larger graph, called the \textit{ray-and-loop graph}, \(\mathcal{R}(S; p)\), which also comprises \textit{short} rays (proper simple geodesics rays asymptotic to \(p\) in exactly one direction) and \textit{long} rays (simple geodesics rays asymptotic to \(p\) which are neither loops nor short rays). In fact, one can show that \(L(S; p)\) is contained and quasi-isometric to a connected component of \(\mathcal{R}(S; p)\). All the other connected components are cliques of rays, and can be identified with the Gromov boundary of \(L(S; p)\). These rays are called high-filling and, in particular, they are filling: they intersect every short ray and loop. A natural question is if every filling ray has to be high-filling. This paper shows that this is \textit{not} the case, at least if \(S\) is the plane minus a Cantor set. A ray which is filling but not high-filling is called \(2\)-filling. This is equivalent to being a ray which intersects every loop, but is disjoint from a long ray which is in turn disjoint from a loop. The main result of this beautiful and richly illustrated paper is that there are \textit{uncountably} many distinct mapping class group orbits of \(2\)-filling rays on the plane minus a Cantor set. This result is achieved twice, by different methods. First, the authors use a direct combinatorial method. Then, they use train tracks, laminations and dynamics of pseudo-Anosov maps on half-translation surfaces. Finally, they show that these two, very different constructions, are actually two descriptions of the same one. Moreover, they show that \(2\)-filling rays lie in cliques of mutually disjoint rays, some of which are \(2\)-filling and some of which are not. They investigate these cliques, showing that one can obtain cliques of an arbitrarily large number of \(2\)-filling rays together with exactly one non-\(2\)-filling ray.