Abstract

In the aftermath of the rediscovery of Einstein’s hole argument by Earman and Norton (1987), we hear that the ontological relational/substantival debate over the status of spacetime seems to have reached stable grounds. Despite Einstein’s early intention to cast GR’s spacetime as a relational entity à la Leibniz-Mach, most philosophers of science feel comfortable with the now standard sophisticated substantivalist (SS) account of spacetime. Furthermore, most philosophers share the impression that although relational accounts of certain highly restricted models of GR are viable, at a deep down level, they require substantival spacetime structures. SS claims that although manifold spacetime points do not enjoy the sort of robust existence provided by primitive identity, it is still natural to be realistic about the existence of spacetime as an independent entity in its own right. It is argued that since the bare manifold lacks the basic spacetime structures –such as geometry and inertia- one should count as an independent spacetime the couple manifold +metric (M, g). The metric tensor field of GR encodes inertial and metrical structure so, in a way, it plays the explanatory role that Newtonian absolute space played in classical dynamics. In a nutshell, according to the SS account of spacetime, one should view the metric field of GR as the modern version of a realistically constructed spacetime since it has the properties –or contains the structures- that Newtonian space had. I will try to dismantle the widespread impression that a relational account of full GR is implausible. To do so, I will start by highlighting that when turning back to the original Leibniz-Newton dispute one sees that substantivalism turns out prima facie triumphant since Newton was able to successfully formulate dynamics. However, to give relationalism a fair chance, one can also put forward the following hypothetical questions: What if Leibniz –or some leibnizian- had had a good relational theory? What role would geometry play in this type of theory? Would it be natural to view geometry and inertia as intrinsic properties of substantival space –if not spacetime? Would it still seem natural to interpret the metric field of GR along substantival lines regardless of the fact that it also encodes important material properties such as energy-momentum? After bringing these questions out into the light I will cast some important doubts on the substantival (SS) interpretation of the metric field. Perhaps the metric turns out to be viewed as a relational matter field. Finally, to strengthen the relational account of spacetime I expect to remove the possible remaining interpretative tension by briefly discussing the relevance of two important facts: i) Dynamical variables are usually linked to material objects in physical theories. The metric field of GR is a dynamical object so, I claim, it should be viewed as a matter field. ii) Barbour and Bertotti (BB2, 1982) have provided and alternative formulation of classical dynamics. They provide a “genuinely relational interpretation of dynamics” (Pooley & Brown 2001). Geometry and inertia become –contra SS- relational structures in BB2.