TIME FOR EMPIRICIST METAPHYSICS
The question of whether time is absolute or relative concerns the relationship between time and change. The general question is whether time is an aspect of material change, or somehow independent of matter. In a world such as Descartes’s, where all change is accounted for in terms of local motion, the question is whether time is an aspect of the relative motions of bodies, or somehow distinct from these. Sharpening the question a little, this becomes: Are there any actual relative motions that can serve as clocks for the material goings-on in the world, such that our project (of providing an account of all the rich variety of the world as we experience it) is tractable ? Newton’s answer is, of course, no; according to Newton, time is absolute, not relative.
My point here is not Newton’s answer, but the method by which this question is to be addressed. The project of the Principia demands the conceptual distinction between absolute and relative time, because (if all bodies are perhaps interacting by means of forces, as yet unknown) there is no material system to which we can turn a priori as a clock. It is then an empirical matter whether there is in fact any material system that acts as a perfect clock relative to the absolute time parameter of the project. Thus, in Newton’s hands, the question of whether time is relative or absolute has been turned into an empirically tractable question, and one whose answer depends on the details of empirical enquiry; the appropriate arguments for deciding the question are no longer those based on broadly a priori considerations.
A brief comparison with Aristotle may perhaps be helpful. For Aristotle, time is an aspect of material change, and the motion of the outermost heavenly sphere is the change by which all other changes are measured. In his view, it is not possible to discover empirically that the motion of the heavens (or, equivalently kinematically, the rotation of the Earth) is not uniform and regular. Newton’s method leaves the question open. It begins from an assumption of at least approximate uniformity in this rotation, and then harnesses the details in the phenomena to construct a time parameter and to examine the speed of the Earth’s rotation with respect to this parameter. As it turns out, it is very hard to tell that the rotation of the Earth (i.e., sidereal time) is not perfectly uniform, that is, that it is not a perfect clock relative to Newtons time parameter. It took until the twentieth century to show beyond a doubt that there are tiny irregularities in the rotation of the Earth. But again, it is the mode of argument rather than the conclusion that I want to emphasize. We are not arguing from general principles, or from experience in general, in order to establish our conclusion; rather, we are having to pay attention to the minute details of empirical enquiry in order to address the question of whether there are actual material changes of which time is an aspect, so that these material changes constitute a perfect clock, or whether there are no perfect clocks so that time itself is distinct from material change. It is in this sense that we are doing empiricist metaphysics.
Gassendi’s discussion of place, space, and time (in The Syntagma; see Gassendi 1972, 383-98) provides another helpful contrast. Newton’s treatment bears strong resemblances to Gassendi’s discussion, conceptually and in its phrasing (see McGuire 1966). However, two striking differences are pertinent to our purposes. First, Gassendi treats place and space first, and proposes that we can learn about time by drawing on analogies with space. Second, the arguments Gassendi gives are of the general kind that, I am emphasizing, are not to be found in Newton. He argues in general conceptual terms for why time must be prior to and distinct from the motions of material things. It is a methodology of argumentation that Aristotle would have recognized. Gorham (2012) situates Newton’s philosophy of time within this tradition of metaphysical arguments for and against the possibility of empty space, in which conclusions about time are drawn by analogy with conclusions about space. Gorham points out that Newton does not argue for absolute time using the kind of metaphysical arguments traditionally offered in support of absolute space. I think that the absence of traditional metaphysical arguments for absolute time in Newton’s philosophy of time is because the kind of argument he is offering is different. Newton is offering a methodology in which these questions become empirically tractable, so that a different type of argumentation is required. And I think this represents a big change in philosophy of time. In the wake of Newton’s Principia, those who accept Descartes’s project (as characterized above) have no choice but to develop their philosophy of time, and their account of the relationship between time and material change, in a way that takes into account the types of argument developed by Newton rather than only those that proceed by general conceptual argumentation.
Before moving on, one final comment is in order about the conclusion that Newton draws, that is, that time is absolute. Newton’s type of argument does not prove that there are no relative motions that constitute a perfect clock, of course. Rather, his argument relies on a conceptual distinction between absolute and relative time, and then provides an empirical argument for the claim that in fact time turns out to be distinct from relative motions of material bodies. And therefore, as such, it is open to empirical refutation, by the finding of a material system that constitutes a perfect clock. But a good empiricist metaphysics proceeds on the basis of empirical evidence and remains open to revision on the basis of future empirical findings, so this is no objection to the claim that, so far as we have been able to ascertain, time is not relative and is distinct from material change.
-  Moreover, the finding of such a clock would generate further questions about the relationship between thatclock and material sub-systems of the universe such that the ticking of the perfect clock could be constitutiveof time for these sub-systems.