Monday, March 14, 2011

Back to Cartwright

I seem to keep coming back to reading "How the Laws of Physics Lie" and "The Dappled World".
I think the reason is that I don't usually get very far. I get bogged down, skip ahead, and feel annoyed. I get a little out of it, but its dense, and the meaning of an essay is often not what I thought it was going to be. I didn't realize that the examples such as the BCS theory of superconductivity, and some laser physics (Also the measurement problem and the relations between classical and quantum systems) were gone into in such detail. The exposition is actually pretty clear, but you don't often know so much detail is on its way.

Anyway, I've finally been getting to the end and actually understanding more of it. In the end, it is more inspirational for a practicing physicist than I thought. The essay on the Quantum measurement problem in HLPL ends with stating that one should take the formalism of quantum statistical mechanics seriously which treats mixtures and superpositions on the same footing. This leaves physicists with the job of figuring out which cases have unitary evolution and which don't. And I think this is in line with the way one analyzes a new phenomena. Take a look at it. Use what we already know. Make some guesses. Test them. Refine them. An approach to learning about new things that involves actually studying those things.

And the discussion about capacities, vs. laws, is not so radical in the end. Perhaps from a philosophy perspective it is. But thinking of electric charge and gravitational mass as imbuing capacities on electrons rather than laying out laws of motion is just the way one thinks about these things. We learn about certain aspects of things, and then we may try to isolate it such that just that one property comes into play. This is what fundamental experiments are about. And we do learn some real about these things. The question may arise as to whether the same effect occurs in different environments, or exactly how the forces or whatever other properties combine when acting in concert. And this way of looking at physical knowledge puts it in the same box as other knowledge we have. There still may be questions of reduction or relationships between different kinds of knowledge, but the starting point seems good.

I must say I'm still not sure I get what she means by nomological machine. She makes a "strong claim" in "The Dappled World" that behind every regularity in the world is a nomological machine. Let's say we are in Switzerland where trains are highly reliable. The matching of the trains to the schedule is a nomological machine? I suppose it is. And the fact that my car (usually) works is a nomological machine. And atoms are nomological machines. And proteins are nomological machines. And a market is a nomological machine. And a beam of light is a nomological machine.

Anyway, just to remind that getting into certain details can be demotivating if one gets stuck in them, but if a topic actually relates to the world and the way things really work, then understanding it will lead to more tools and more clarity in the end. Which is to say that previously, reading this made me shut down certain ways of thinking, but now I find it mainly adds.

Apparently Philip Anderson may have had some trouble following "Dappled World" as well.

I just read Anderson's review (see here). Wow, that's pretty intense. He refers a lot to the feeling he gets about it. "One gets the feeling" he states a lot, without actually quoting many passages, or covering arguments. I must admit, that I sometimes had these same "feelings". But reading more closely, I often found that Cartwright was saying something more precise and more interesting than I initially thought.

Here are other reviews of The Dappled World.

Finally, once one digests some of the seemingly right arguments, one would like to see engagement of other authors. Here's some essays doing this. In particular, I'd like to understand this one by Carl Hoefer in defense of "fundamentalism". He concludes with
To engineers and experimentalists, I commend Cartwright’s philosophy of science wholeheartedly. But I hope to have made space for theoreticians and philosophers of physics to keep their faith in a world with fundamental physical laws.
As a quick summary of Hoefer's article, he says that one can keep the fundamentalist approach but deny that some kinds of reductionism may be possible. What I wonder then is about the terminology of laws. Cartwright already says that we learn about capacities. I'd assume that the hydrogen atom in the dewar and the hydrogen atom in the hallway and the hydrogren atom in a distant galaxy all have the same capacities. (It does seem a bit odd to say that a hydrogen atom has the capacity to form the states specified by the Shroedinger equation with the central force potential term in its Hamiltonian, but maybe this is just what one must say.) I guess the question to address is how to relate such facts to programs in which we take electrodynamics, quantum mechanics, fluid mechanics, thermal physics, statistical mechanics, etc. and try to say that in some sense these theories "govern" matter.

Another way to get at the question is to take this fact about hydrogen seriously and see how much it says. Using the capacity language, we say that we know the capacities of hydrogen atoms, and also that we find hydrogen atoms all of the place. And less us take it a step further, and say we know the capacities of molecules, and that we also find these things all over the place. Is physics the domain that is responsible for such knowledge? Its certainly a good part of it, but I think its reasonable to say that the methods and ideas of chemistry are also involved in this knowledge. So I'd need to understand a little better what is meant by a law in the expression "fundamental law" to see if it captures the knowledge we have about the atomic and molecular basis of matter.