In 1922 in The Principle of Relativity, Whitehead presented an alternative theory of gravitation in response to Einstein’s general relativity. To the latter, he objected on philosophical grounds—specifically, that Einstein’s notion of a variable spacetime geometry contingent on the presence of matter (a) confounds theories of measurement, and, more generally, (b) is unacceptable within the bounds of a reasonable epistemology. Whitehead offered instead a theory based within a comprehensive philosophy of nature. The formulal Whitehead adopted for the gravitational field has (...) been described as involving both the flat metric nu, of Minkowski spacetime and a dynamic metric gu, dependent on the presence of source masses. The ontological relationship between the two must be fleshed out in the context of Whitehead’s philosophy of nature. The relationship is of some importance, not only in casting Whitehead’s theory within its proper metaphysical context vis-d—vis Einstein, but also in judging how the theory has faired empirically with respect to general relativity (GR hereafter). It makes the same predictions as GR with respect to the perihelion advance, the deflection of light rays and the gravitational red-shift; indeed, Eddington (1924) has shown that it is equivalent to the Schwarzschild solution of Einstein’s held equations for the one-body problem. However, it also appears to predict an anisotropy in the locally measured gravitational constant y that is in conflict.. (shrink)

It is generally acknowledged that the requirement that the laws of a spacetime theory be covariant under a general coordinate transformation is a restriction on the form but not the content of the theory. The prevalent view in the physics community holds that the substantive version of general covariance – exhibited, for example, by Einstein’s general theory of relativity – consists in the requirement that diffeomorphism invariance is a gauge symmetry of the theory. This conception of general covariance is explained (...) and confronted by two challenges. One challenge claims, in effect, that substantive general covariance is not deserving of the name since, just as it is possible to rewrite any spacetime so that it satisfies formal general covariance, so it is also possible to rewrite the theory so that it satisfies the proffered version of substantive general covariance. The other challenge claims that the proffered version of substantive general covariance is not strong enough to guarantee the intended meaning of general covariance. Both challenges are discussed in terms of concrete examples. It is argued that both challenges fail but, at the same time, that they help to clarify what is at stake on the seemingly never ending dispute over the nature and status of general covariance. (shrink)