Finiteness of 3-manifolds associated with non-zero degree mappings (Q2448613)

The well known JSJ-decomposition and geometrisation asserts that any irreducible 3-manifold with incompressible, toral or empty boundary can be cut along tori into pieces that are either Seifert fibered or hyperbolizable. To formulate the JSJ-decomposition for irreducible 3-manifolds \(M\) with incompressible, higher genus boundary \(\partial M\) one needs the notion of ``boundary pattern'', which (in the notation of this paper) is a union of annuli in \(\partial M\). (So doubling \(M\) along the complement of the boundary pattern results in an irreducible 3-manifold with toral boundary.) It goes back to Johannson and Jaco-Shalen that an irreducible 3-manifold with boundary pattern has a JSJ-decomposition into pieces that are I-bundles, Seifert fibrations or ``simple'', where the latter ones in view of geometrisation admit a hyperbolic manifold with geodesic boundary and cusps. In particular, if \(S\subset M\) is an incompressible surface, then one can consider the manifold \(M_S\) obtained by cutting \(M\) along \(S\) and one can look at the union of Seifert fibrations and simple pieces in the JSJ-decomposition of \(M_S\). The authors call this the guts of \(M_S\). (One should be aware that there are different notions of ``guts'' around and some other authors exclude Seifert pieces from the guts. Also the notion of ``boundary pattern'' used in this paper is exactly the complement of the one used by Johannson.) The main result of the paper under review states that for a given closed, irreducible \(3\)-manifold \(M\) there are only finitely many \(3\)-manifolds with boundary pattern that may occur as \(guts(M_S)\) for a closed, incompressible (not necessarily connected) surface \(S\subset M\). This is then applied to prove finiteness results (for \(S\) running over all such surfaces) on the Thurston norms and absolute Gromov volumes of \(M_S\), where the latter is defined as half of the simplicial volume of the manifold obtained by doubling \(M_S\) along its boundary. These results are then applied to study the question how many closed \(3\)-manifolds \(N\) can be dominated by a given closed \(3\)-manifold \(M\). Here \(M\) is said to dominate \(N\) if there is a non-zero degree map from \(M\) to \(N\). The main result states that for a given \(M\) there is a finite set of hyperbolic or Seifert fibered manifolds such that the JSJ-pieces of all manifolds \(N\) dominated by \(M\) belong to that finite set provided that \(N\) does not support the geometries of \(S^3,\widetilde{PSL_2(\mathbb{R})},Nil\). Applications of this result include that a closed \(3\)-manifold may dominate only a finite number of integral homology spheres or that a compact \(3\)-manifold may be dominated by at most a finite number of knot complements in \(S^3\).