Non-simple universal knots

A link or knot in S 3 is universal if it serves as common branching set for all closed, oriented 3-manifolds. A knot is simple if its exterior space is simple, i.e. any incompressible torus or annulus is parallel to the boundary. No iterated torus knot or link is universal, but we know of many knots and links that are universal. The natural problem is to describe the class of universal knots, and this was asked by one of the authors in his address to the `Symposium of Kleinian groups, 3-manifolds and Hyperbolic Geometry' held in Durham, U. K., during July 1984. In the problem session of the same symposium W. Thurston asked if a non-simple knot can be universal and more concretely, if a cable knot can be universal. The question had the interest of testing whether the universality property has anything to do with the hyperbolic structure of some knots. That this is not the case is shown in this paper, where we give infinitely many examples of double, composite and cable knots that are universal.