The Laws of Physics Lie

Aka over at Physicality of Worlds has made an interesting post about, among other things, the laws of physics.

Aka says :

Our intuitive grasp of physics is (as intuition tends to be) based on generalizations of experiences of how things usually work. This is usually a good thing.

Too bad that experience we build with our senses in a particular environment is not very helpful when it comes to understanding the underlying principles. This is why people often find physics difficult -- it is counterintuitive. I'm not talking about quantum mechanics here, you can probably recall how people scratch their heads in high school physics when trying to grasp Newton's laws. (You mean that if I throw a ball it will continue in a straight line until an external force acts on it? That's not true! Everyone knows that it will lose speed and eventually fall to the ground. OK, so without gravity it would not fall, but it must come to a stop, since it left my hand and nothing is pushing it. What friction? No, I'm not buying this!)

The problem is not that the laws of physics are counter-intuitive, it's that they are, strictly speaking, false.  For example, if you take Newton's law of gravitation F = G(mm'/r^2) you'll find that the law simply doesn't apply to anything real because there will never be a situation in which the only force acting upon two masses is gravity.  The law only holds true ceteris paribus, which means that if you did have a situation in which the only force acting upon two masses was gravity then the force would behave in the way described by the law.  So the law of gravitation is true all things being equal, despite the fact that things never are equal.  Indeed, if you then want to do anything useful with the law you have to plug in friction, acceleration and so on.

This point was first made by a philosopher of science called Nancy Cartwright nearly 25 years ago.  Cartwright went on to argue that in order for the laws of physics to be true, you had to have not generalised laws but particular laws that apply in specific situations.  So when attempting to describe a thrown ball's motion through the air, rather than wheeling out these different laws governing acceleration, gravity and friction, you should really have a law that describes how the ball will behave without needing to break the ball's trajectory down into the effects of a series of different laws.

The reason why we tend to think about physics in such a counter-intuitive manner is because ultimately, the laws of physics are a load of mathematical models which can fit together to describe real world events but in and of themselves, frequently don't describe anything real.  One danger of this type of thinking is that you can wind up thinking that there are no laws of physics... there are just empirically adequate mathematical models.  In other words, it's possible to think that there are no laws of physics, but we know how to make the sums work, which is pretty much the line taken by another philosopher of science called Bas van Fraassen who argued that we don't actually need the concept of a law of physics to explain things.

Where van Fraassen is correct is that it's not clear how much the concept of a law of physics actually explains stuff.  You can take a law of physics to just be an observation but the problem is that what weight does that observation carry when it comes to predicting stuff in the future?  Ao while it might be a case that all dogs born at sea have been Labradors, does it follow that if two scotty dogs got it on during a cruise then their puppies would be Labradors?  of course not... you need some metaphysical weight behind the concept of the law... you want to say that not only has this object always behaved this way, it will continue to behave this way in the future.

Which is where you get into the metaphysics of natural law and that way leads to madness as beyond knowing that we need some kind of metaphysical buttress, there's actually no way of working out how a law of nature might function because metaphysics is, by its nature, impervious to any kind of empirical testing.  Not that that's stopped philosophers trying their best to come up with theories...

Comments

[David Harmon]

Funny, those supposedly inapplicable "laws" do a pretty good job of letting us designing new gadgets! (Did you know that GPS requires relativistic corrections for the satellite's orbits?).

The only metaphysics needed for science is the basic idea that there are, in fact, discoverable laws governing the universe. Beyond that, the first test of science is reproducibility -- but the second is technology. Both have been spectacularly successful in our time.

[Jonathan McCalmont]

As someone once put it "Show me a relativist in a plane at 35,000 feet and I'll show you a hypocrite".

I'm not arguing that there are no laws of physics. I'm just wondering how it is that they work. What is it about the universe that makes water boil at 100 degrees and not 140 degrees? What forces the atoms and molecules to behave that was every time?

There was a book that argued that there was natural selection. Out of all the possible universes that could be created it's only those with laws a bit like ours that allow stars and therefore black holes to form. A black hole can, in theory at least, pinch itself off and create another universe. But what is it that makes that little pocket of reality operate one way rather than another?

That's a metaphysical question. It's literally after physics.

It's also unanswerable because it's impossible to choose between two metaphysical theories.