John Stachel

Towards the end of the career of many a distinguished scientist, or shortly after his or her death, an edition of the scientist's articles is published under the title: ‘The Collected Papers of…’. While not wishing to slight either the ceremonial importance or real utility of such collections, they must be clearly distinguished from the sort of editions on which theCollected Papers of Albert Einsteinis modelled. The former are primarily intended to make the published papers of a great scientist easily accessible to other scholars and students working in the same field as an aid to their research. The reader is provided with little or no help in understanding or evaluating the historical role these writings played in the development of this field, the circumstances leading to their creation, or how they fit into the life of their creator.

David Malament's (1977) well-known result, which is often taken to show the uniqueness of the Poincare-Einstein convention for defining simultaneity, involves an unwarranted physical assumption: that any simultaneity relation must remain invariant under temporal reflections. Once that assumption is removed, his other criteria for defining simultaneity are also satisfied by membership in the same backward (forward) null cone of the family of such cones with vertices on an inertial path. What is then unique about the Poincare-Einstein convention is that it is independent of the choice of inertial path in a given inertial frame, confirming a remark in Einstein 1905. Similarly, what is unique about the backward (forward) null cone definition is that it is independent of the state of motion of an observer at a point on the inertial path.

In the paper I wish to begin to explore the consequences for metaphysics of thinking that a good physical theory should be background-independent. More generally I want to ask whether the conception of time not as a background but as an active component of the physical universe has any significant consequences for metaphysics. I think that a natural conception of space and time is to regard them as a (possibly infinite) container or stage for the events that make up the history of the universe. They are not part of the contents of the container nor are they actors or props in the action on the stage. They are an inert but necessary background. This conception plays a part in metaphysical argument. And reasons to doubt that conception may undermine some of those arguments. In this paper I would like to examine this conception and the possible consequences of so doing for certain metaphysical questions.

After introductory surveys of Poincaré’s role in the Dreyfus case and of his “Fourth Geometry,” I turn to the main question. The problem confronting both Poincaré and Einstein was how to reconcile the phenomena of electrodynamics, notably the optical principle of relativity, with the principles of Newtonian mechanics. I show that, on such questions as the existence and role of the ether and the relation between kinematics and dynamics, Poincaré and Einstein held diametrically opposed views. Poincaré did everything to preserve the old viewpoint, while Einstein abandoned the absolute time and proposed a new mechanics.