Nina Emery is an assistant professor of philosophy at Brown University. She works on topics in metaphysics, epistemology, and philosophy of physics. She is especially interested in whether and how our best scientific theories should inform our understanding of entities like chance, time, and laws of nature. Her Ph.D. is from MIT.

At Brown Nina has helped found the Summer Immersion Program in Philosophy (SIPP@Brown), a two-week residential summer program for students from traditionally underrepresented groups in philosophy. If you know any talented undergraduate students who would benefit from such a program, please encourage them to apply to SIPP@Brown!

A Middle Ground for Metaphysics

Thanks so much to Meena for running this wonderful and inspiring series and for inviting me to contribute. I’m going to use this opportunity to say a bit about a question that underlies many of my research interests. The general version question is: what is the proper relationship between philosophy and science? A more specific version of the question, and the version in which I am especially interested, is: what is the proper relationship between metaphysics and physics?

There’s a way of thinking about this specific version of the question that gives rise to something of a dilemma for metaphysicians. For reasons that I will spell out below, I think this dilemma isn’t really all that problematic. But I also think that it underlies much of the continuing skepticism about the legitimacy of metaphysics as a discipline.

The supposed dilemma is this. On the one hand, suppose that our best metaphysical theories must accord with our best scientific theories in some robust sense. Then why do we need metaphysicians at all? Can’t we just ask the scientists what the consequences of their theories are, accept those consequences, and move on? At most metaphysicians are just translators, helping the rest of us understand the increasingly technical and complex language in which scientists, especially physicists, work. On the other hand, suppose our best metaphysical theories can come apart from our best scientific theories. Then what are the rules by which we investigate, construct, and judge metaphysical theories. Insofar as we think that metaphysics has a distinctive methodology, what exactly is is that methodology? And doesn’t the lack of progress and consensus in metaphysics suggest that whatever that methodology is, it is somehow defective?

My own view is that neither horn of this supposed dilemma is all that troubling. I think that the process of reading consequences off of our best scientific theories is philosophically interesting in and of itself. And I think that pessimism stemming from the lack of consensus and progress in metaphysics is overblown. But I’m not going to defend either of those claims here. Instead I want to point out that there is some substantive middle ground between the two positions described above. In addition to paying attention to the content and consequences of our best scientific theories, metaphysicians can also pay attention to the methodology of standard scientific practice. One way of being a scientifically-respectable metaphysician, then, is to take standard scientific practice as a good guide to successful inquiry into what the world is like, but to apply the principles embodied by that practice to questions that go beyond the standard scientific purview. This gives metaphysics as distinctive domain but also allows it to inherit some of the legitimacy of scientific practice.

All of that is fairly abstract. Here is a concrete example of this sort of middle-ground approach to metaphysics. The history of science appears to include many cases in which scientists—especially physicists—were convinced to admit the existence of entities that they previously would have thought were far too strange or too novel to countenance. Why? Because it turned out that entities of that sort were required in order to play a certain explanatory role. Think of the introduction of electromagnetic fields at the end of the 19^th-century. Or the introduction of particles that sometimes behaved like waves at the beginning of the 20^th. A contemporary example might be the widespread belief among cosmologists in the existence of dark energy—if there wasn’t such a thing then what would explain the accelerating rate of expansion of the universe?—despite their utter uncertainty about what dark energy is. In standard scientific practice, it seems, explanatory power trumps metaphysical weirdness.

But taking this methodology seriously would require many metaphysicians to reconsider, if not their favored positions, at least their understanding of the main challenges facing their favored positions. Take, for instance, the continuing popularity of Humean theories of laws and of chances. According to such theories, laws and chances can be reductively analyzed in terms of the Humean mosaic—the distribution of non-modal, non-causal, and non-dispositional properties throughout spacetime. At first glance, this sounds like a promising approach—primitive modal or causal or dispositional properties are strange sorts of things. Why not eliminate them from our metaphysics if we can?

The problem is that Humean theories of laws and chance have a notoriously difficult time showing how laws and chances, once they have been reductively analyzed in terms of the Humean mosaic, can play any robust explanatory role. How can the law of universal gravitation explain the fact that certain sorts of bodies exhibit certain sorts of behavior if the law of universal gravitation is in some sense nothing over and above the fact that those sorts of bodies exhibit that sort of behavior? How can the chance of a certain kind of radioactive decay explain the relative frequency with which that kind of decay occurs if the chance of that kind of decay is in some sense nothing over and above the relative frequency of that kind of decay?

Recently Humeans have made what I think may be real progress on this explanatory challenge. In particular, Barry Loewer has made the interesting suggestion that the challenge might be diffused if we distinguish between two kinds of explanation—scientific explanation on the one hand and metaphysical explanation on the other. (See Loewer, B. 2012. ‘Two Accounts of Laws and Time’, Philosophical Studies, 160 (1): 115-137.) It is very much an open question whether this sort of move succeeds—at the very least we need a much better understanding of metaphysical explanation before we lean too heavily on Loewer’s distinction. The relevant point for our purposes, however, is just this: insofar as we are taking seriously the fact that explanatory power trumps metaphysical weirdness, the Loewer strategy, if it works, is not just icing on the cake. It is crucial. Insofar as we take explanatory power to trump metaphysical weirdness, the explanatory challenge is the central challenge facing Humeans. If Loewer’s suggestion is the only helpful suggestion forthcoming, Humeans ought to drop whatever else they are working on in order to try to prove that it succeeds.

In fact, once we internalize the fact that in standard scientific practice explanatory power trumps metaphysical weirdness, we ought to be prepared for anything when it comes to the metaphysics of explanatorily important entities like laws and chances. Insofar as someone convinces us that Humean laws and chances cannot play a robust explanatory role, we ought not be Humeans. Insofar we become convinced that no reductive account of laws or chances—Humean or non-Humean—can meet the explanatory challenge, we ought to be primitivists about laws and chances.

Yes, on such an account laws and chances would be novel, and in virtue of being novel, would count as strange. But the very same thing could have been said about electromagnetic fields or quantum particles. The very same thing could be said about dark energy today.

Of course, even with this one example there’s plenty of work yet to be done. Surely it is a certain kind of explanatory power that is relevant in these cases. So what kind is it? And what about other relatively straightforward features of scientific practice? Scientists from Galileo to Einstein have thought it was important to eliminate what appears to be excess spacetime structure from our theories. Is there a way of understanding this move so that it has bearing on metaphysical theories more generally? And anyway, how are these methodological principles that we’re extrapolating from scientific practice to be justified?

But all of that just serves to make my point—there is interesting and well-defined philosophical work to be done here. My own view is that this sort of work is by no means the only interesting or well-defined work that there is to be done within the realm of metaphysics, but it is as good and uncontroversial a starting place as I can think of.