The Campbell--Magaard Theorem is inadequate and inappropriate as a protective theorem for relativistic field equations

Abstract

Given a particular prescription for the Einstein field equations (EFE's), it is important to have general protective theorems that lend support to it. The prescription of data on a timelike hypersurface for the (n + 1)-d EFE's arises in `noncompact Kaluza--Klein theory', and in certain kinds of braneworlds and low-energy string theory. The Campbell--Magaard (CM) theorem, which asserts local existence (and, with extra conditions, uniqueness) of analytic embeddings of completely general n-d manifolds into vacuum (n + 1)-d manifolds, has often recently been invoked as a protective theorem for such prescriptions. But in this paper I argue that there are problems with loosening the CM thoerem of restrictive meanings in its statement, which is worthwhile thing to do in pursuit of the proposed applications. While I remedy some problems by identifying the required topology, delineating what `local' can be taken to mean, and offering a new, more robust and covariant proof, other problems remain unsurmountable. The theorem lends only inadequate support, both because it offers no guarantee of continuous dependence on the data and because it disregards causality. Furthermore, the theorem is only for the analytic functions which renders it inappropriate for the study of the relativistic equations of modern physics. Unfortunately, there are no known general theorems that offer adequate protection to the proposed applications' prescription. I conclude by making some suggestions for more modest progress.