I’ve been thinking a lot lately about the nature of natural laws (along with Greg Lusk), and I’ve been tripping over the metaphysics. I was trained as an undergraduate in both history and physics, so my default stance is some sort of empiricism. The main empiricist understanding of laws of nature is that they are observed regularities in the world. Kepler’s laws are laws because of the reliable behaviour of the planets. This gives us a nice, spare metaphysics (à la Hume): we don’t have to invent some new class of things to explain laws, they are just there is the world as collections of data. This quickly runs into a problem, of course, because we don’t want to say that all regularities are laws. It may be true that all the buildings on my street are made of brick, but that doesn’t make it a law of nature that that’s true. It is an accidental regularity; a regularity that just happens to be true. If someone put up a wooden-framed house, no law would be broken.
So the task of the regularity-theorist is to try and identify what has to be added to the notion of regularity that will allow us to separate laws of nature from accidental generalizations. This has led some regularity-theorists to argue that one thing that separates laws from accidental generalizations is their use by scientists (see e.g. Dretske 1977). That is, what counts as a law is contingent.
This contingency is not a happy position for those philosophers who want laws to be somehow objective, mind-independent, and eternal. Scientific realists are not fans of contingency. And so realist philosophers have tried to offer an explanation that is (in their minds) objective, eternal, etc. and these explanations revolve around supposing the existence of something more than just observed regularities.
One such realist account is offered by Chris Swoyer (1982). He proposes that laws are expressions of relations between properties. (There is a subtlety here about laws as mind-independent things vs. law-like statements that are expressed by humans. For this discussion I’ll ignore the issue.) Swoyer thinks that the properties things have—colour, mass, energy—are the key players in laws, and that properties exist in the world independently of any object that might exemplify them.
To Swoyer it is worth accepting the existence of properties because they help explain things about laws that the regularity account cannot. For example, regularities cannot be confirmed by their instances. This mean that when I look at planets and see that they move according to Kepler’s laws, I have found one more object that fits in the class of things that exhibit this regularity. But this is just making the data in my regularity bigger, it’s not helping me say something like All planets behave this way. Or, if I saw a new planet, it would behave this way. Our observation gets folded into the regularity, and does not explain it. But of course we want to say that when we observe something that fits a law that we have confirmed it. Swoyer argues that we can explain Kepler’s laws in terms of it holding between objects with a certain property (mass and a range of angular momenta). It is the fact that his properties explain that Swoyer thinks we should believe in them.
Indeed, the only reason I can see for supposing that there are such things as properties at all is that a philosophical theory of them has explanatory value. (original emphasis, 204)
Without going any further into Swoyer’s account, we can see a problem. Properties may help philosopher’s explain laws, but how can we argue that the best thing for philosophers is what dictates out metaphysics?
Swoyer is just pushing the anthropocentrism around. For regularity theorists laws are contingent, meaning that hey are relative to the minds of scientists trying to prove or disprove them. The people at the centre of the regularity theorist’s anthropocentrism are scientists. For Swoyer, laws are not longer dependent on scientists’ minds, and he thinks he’s done away with contingency. But he’s just moved the centre to himself. Now laws (and quite a bit more of our ontology) are contingent on what philosophers think is best for them.
This is hardly an improvement.
If I may, let me try to run a defense of Swoyer. You’ll have to see if you think there is any merit in it.
I tend to think there is a difference in the type of anthropocentrism in the two cases. After Kant, (responsible) metaphysical arguments have to relate to subjectivity in some way or other. This, I take it, is the kind of anthropocentrism that Swoyer is committed to – we approach metaphysics by trying to figure out a coherent picture of our world (your “what is useful to philosophers”), even if that picture goes beyond the available empirical evidence.
The kind of anthropocentrism that Swoyer rejects is of a different kind. He rejects the idea that our metaphysics should include reference to what particular people think at a particular time. If “law” is a substantive metaphysical category (you may want to reject that premise as an empiricist) then we don’t want what falls under that category to depend on fashions in certain groups of people.
The difference between the two is important. The first, Kantian type of anthropocentrism makes our metaphysics dependent on the conditions for the possibility of coherent subjective experience. The second makes metaphysics depend on the beliefs of individuals. You may want to argue that these two kinds shade in to each other, and that’s fine, so long as you recognize the qualitative difference.
Hi Cory,
Thanks for the comments!
What I’d probably say is that I don’t think Swoyer is in that Kantian-mode. Kant (of whom I know very little, so please correct me if I’m wrong) was looking for the necessary conditions for metaphysics, right? From what you’ve said, I don’ think Swoyer is using that kind of responsible subjectivity. He isn’t trying to figure out the most coherent picture of the world (with some mind-independent notion of coherence). He’s explicitly, repeatably arguing from pragmatism.
Another way of looking at it (again apologies if I’ve got Kant wrong) is that your “responsible” metaphysics is looking for the only answer that solves some problem. That is, just like theoretical physicists, they seek unique answers under strict constraints. Swoyer is arguing from utility, not necessity.
Let me know what you think,
Aa
I probably am being unreasonably charitable to Swoyer by taking his work to be in a Kantian mode. Kant tried to find conditions that were strictly necessary preconditions for any cognition. If there is any case to be made for Swoyer, it is that he is looking for important preconditions for some kinds of cognition – like counterfactual reasoning. I took his argument not to be a pragmatic one (we need universal properties because they’re handy for doing physics) but a transcendental one (we need universal properties because they are a precondition of certain types of reasoning). But I think you’re right – to count as a transcendental argument, he’d have to show that they are strictly necessary for reason, not just handy.