The "generic" family of classical sequential growth dynamics for causal sets [1] provides cosmological models of causal sets which are a testing ground for ideas about the, as yet unknown, quantum theory. In particular we can investigate how general covariance manifests itself and address the problem of identifying and interpreting covariant "observables" in quantum gravity. The problem becomes, in this setting, that of identifying measurable covariant collections of causal sets, to each of which corresponds the question: "Does the causal set that occurs belong to this collection?" It has for answer the probability measure of the collection. Answerable covariant questions, then, correspond to measurable collections of causal sets which are independent of the labelings of the causal sets. However, what the transition probabilities of the classical sequential growth dynamics provide directly is a measure on the space of labeled causal sets and the physical interpretation of the covariant measurable collections is consequently obscured. We show that there is a physically meaningful characterisation of the class of measurable covariant sets as unions and differences of "stem sets".

No abstract

A change of spatial topology in a causal, compact spacetime cannot occur when the metric is globally Lorentzian. One can however construct a causal metric from a Riemannian metric and a Morse function on the background cobordism manifold, which is Lorentzian almost everywhere except that it is degenerate at each critical point of the function. We investigate causal structure in the neighbourhood of such a degeneracy, when the auxiliary Riemannian metric is taken to be Cartesian flat in appropriate coordinates. For these geometries, we verify Borde and Sorkin's conjecture that causal discontinuity occurs if and only if the Morse index is 1 or n − 1.

No abstract

We consider certain interesting processes in quantum gravity which involve a change of spatial topology. We use Morse theory and the machinery of handlebodies to characterise topology changes as suggested by Sorkin. Our results support the view that that the pair production of Kaluza-Klein monopoles and the nucleation of various higher dimensional objects are allowed transitions with non-zero amplitude. a dowker@ic.ac.uk