Where the Laws of Physics come from...

The Gline has an interesting link up to a controversy that has arisen from an Op-Ed piece written by pop science writer Paul Davies in the New York Times.

The problem is that Davies has said that :

"Clearly, then, both religion and science are founded on faith — namely, on belief in the existence of something outside the universe, like an unexplained God or an unexplained set of physical laws, maybe even a huge ensemble of unseen universes, too. For that reason, both monotheistic religion and orthodox science fail to provide a complete account of physical existence."

This has prompted some irritated responses, including this quite typical one from Sean at Cosmic Variance.

The thing is that we've been here before.  Philosophers have been asking these kinds of questions since Aristotle, and I should know because, once upon a time, I actually did some research as a philosopher into the question of where the laws of physics comes from.

The philosopher Elizabeth Anscombe once pointed out that one of the great difficulties in moral philosophy is overcoming the inherited idea that all moral laws come from some kind of divine law-giver who issues commandments from on high.  The same is true of natural law as the very name implies that something is forcing nature to behave in a certain way.

Are There Such Things as Laws of Nature Anyway?

David Hume, the father of the problem of induction, pointed out that when we look at events unfolding around us, we never actually see one event causing another.  We only ever see two events.  For example, if you hit a billiard ball and it goes towards another billiard ball and makes it go into a pocket, you do not see one billiard ball causing the second one to move, you simply see one billiard ball moving up the table and then a second billiard ball going into a pocket.  You don’t see causation, you just see conjunction... stuff happening one thing after the next.

If you expand this idea you’ll realise that you never actually see the laws of physics.  What you see are patterns of things happening such as water always boiling when you heat it to a certain temperature.  In fact, we could, at this moment, be living in a world without any laws of physics at all.  Such worlds can, catchily for any SF writers that might be out there, be referred to as Hume Worlds... worlds where stuff just happens at random.  Such worlds wouldn’t need a theory of natural law.

If we lived in such a world then all we would mean when we spoke of a “law of physics” would be a pattern... the fact that all observed instances of something had behaved in a certain way.  E.g. every raven we’ve seen has been black and every piece of water we’ve seen has boiled at 100 degrees celsius at sea level.  The problem with this theory is that there would be no way of distinguishing between the stuff we want to be physical laws and stuff that has always happened to be true.  For example, imagine a world in which all dogs born at sea had been labradors.  If all a law of physics is is a universal truth then we’d want to say that it was a natural laws that all dogs born at sea are labradors.  However, part of the fun of a natural law is that it allows us to make predictions about the past.  So we’re not just saying that all water, up until now, has boiled at 100 degrees C, but all water we boil tomorrow will behave that way too.  Do we really want to say that all dogs in the future that are born at sea will be labradors?  what If a scotty has puppies at sea... will it give birth to labrador pups?  of course not.  We want something more to our theories of natural law than just universal truth.

What about logical truth?  what about the idea that a law of nature is any universal truth that cannot logically be false?  well, the problem with this is that it’s opening the door to mathematics.  It’s possible to come up with closed deductive systems in which 2+2 is not = 4.  So we then have to ask which logical system are we talking about?  well... the one that governs the world.  But this is a circular theory; a law of nature is a universal truth that is a law of nature by virtue of the laws governing nature.  It tells us nothing.  Furthermore, such laws can’t be logical as there is nothing absurd about the idea of different laws of physics, they might well yield completely different-looking universes but they’re not contradictory in anyway... they’re not illogical.

(if you want to read more about such theories then check out A.J. Ayer’s paper “What is a Law of Nature?” in Revue Internationale de Philosophie 10, 1956 and for the logical system stuff, take a look at F. P. Ramsey’s Foundations)

Metaphysical theories of Physics.

The key concept to understand in such discussions is Saul Kripke’s idea of restricted necessity.  By this, I mean statements that are necessarily true within a certain context... namely our universe.  This fits the bill quite nicely as we want our laws of nature to be necessarily true in so far as we can use them to make predictions, but we don’t want them to be so necessarily true that it would be logically absurd for them to be false.  The easiest way to think of necessity is in terms of possible worlds (as popularised by David Lewis) and say that something is a law if it is universally true in all worlds that share our laws of nature.  The problem with this assessment is that it circular... it tells us that our world is linked to a number of possible worlds by virtue of sharing physical laws but it does not tell us how those worlds interact... it is just describing the situation... not explaining.

Metaphysicians have responded to this criticism by attempting to come up with a theory of natural law that is mechanistic rather than descriptive and reliant upon weird relations between metaphysical entities.  The most famous adherrent of this approach is a guy called Armstrong who argued that laws of nature are relations between universals, which are a bit like Platonic forms, they are the metaphysical essences of different entities.  So, under this view, there’s just something about water molecules that make them behave a certain way when they are heated.  they have a special necessitation relationship between them which means that if you take some water and you introduce it to 100 degree heat, you will necessarily get boiling.

The problem with these metaphysical theories is (aside from their internal problems) is that they posit the existence of entities and relations for which there is not a shred of evidence.  There’s more evidence for memes than there is for the existence of universals.  This is where the faith comes into it.

(if you check out Bas van Fraassen’s Laws and Symmetry you’ll find all of these theories nicely discussed and dealt with)

If we consider what it is we mean when we say that something is bound by the laws of physics to happen, we have a pretty clear idea that we’re not just talking about accidental universal truth.  We want there to be something that makes enetities governed by physical law behave in a certain manner.

This is why Davies claims that scientists are operating under faith.  A theory of natural law, though less grandiose than a religion (it doesn’t requite mind, intention, beardy guys in clouds or any of the other detritus that comes with christianity) operates on the same level as a religion and therefore cannot be tested, and cannot be falsified (I bet you were wondering when Popper was going to come into this... Ayer’s good on the same subject).

Where Davies is mistaken is that he is confusing the belief that there are laws of physics, with belief in the existence of metaphysical forces making objects act in a certain way.  These are two very different philosophical positions.

Modern quantum physics can allow us to make predictions accurate to 10 digital places, suggesting that there’s more to the laws of physics than dumb patterns.  So the evidence exists that there’s something going on here.  but we have no grounds for choosing between different theories of natural law.

The situation is similar to being in one’s bedroom and suddenly hearing someone cough on the landing outside one’s bedroom.  One has clear evidence that someone is out there coughing, but the only way for us to believe that it’s a friend, a flatmate or a burglar is to make a leap of faith and pick a conclusion with no evidence to support it.

Similarly, in the Schroedinger’s Cat thought experiment, we know that the box contains a cat that is either alive or dead, but we have no means of reliably guessing whether the cat is alive or the cat is dead.

In the first situation it is rational to believe that there is someone on the landing and in the second situation it is rational to believe that there’s a live or dead cat in the box.  Faith does not enter into it until we are asked to guess as to what is out there on the landing or inside the box.

In order for science to function and for us to be able to make predictions, we do not need to have faith in a particular theory of natural law.  We simply need to consider the evidence that suggests that we’re not living in a Hume World, and that evidence grows each time we make a correct prediction.

Comments

[Serdar]

Thanks for the shout-out.

My main beef was that even after a second close reading I couldn't figure out what his precise point was, or what ought to be done about it. Then I realized it was essentially the "science is just another form of religion" canard and felt like I'd been sent on the intellectual equivalent of a Chinese Fire Drill.

[Jonathan McCalmont]

It's more that he's saying that there are parts of the scientific infrastructure that can't be proved according to the rules that govern the content of the superstructure. It's a bit like a epistemological Goedel's proof.

Therefore, the only way you can completely believe in the entire edifice of science is through some degree of faith.

The problem is that you don't NEED to believe in those bits that lie outside of science... science allows you to say "something's going on there but we don't yet have an answer to that question".

I once read an article from a physics journal from the 1800's and it spoke about the stars as these entities we could never know anything about except by speculation. Now though, we know loads of things about the stars. What has changed is that we've pushed back the frontiers of science.

The day might very well come when we can detect the metaphysical stuff that dictates the laws of nature. Until then scientists can happily be agnostic about how it is that physics works. That's the correct scientific procedure and no need for faith.

[A.R.Yngve]

I wish more people understood Goedel's Theorem, a.k.a. the Incompleteness Theorem.

Goedel's Theorem simply says "This theorem cannot be proven within the limits of its own logical system." This does not mean you must add faith to science. Faith doesn't really remove the issue.

It means, with unbreakable certainty, that a complete scientific explanation of the universe no matter how much our knowledge of it expands is logically and practically impossible. Science works, mind you... but it can never become all-knowing.

Also, this does not imply our knowledge will soon run into some sort of cosmic wall. I'm fairly convinced our knowledge of the universe will continue to expand without limit. Which ought to make scientists happy: they'll never run out of work. ;-)

BTW: Have you noticed how, with depressing regularity, educated people who ought to know better say things like "Soon we will know everything" or "A Theory of Everything is at hand"? What the hell are they thinking? That is trying to turn science into religion....

[Jonathan McCalmont]

I was actually being metaphorical. Goedel's proof only applies to closed logical systems, not open empirical models like scientific knowledge, so in truth I suspect that Goedel's proof means fuck all as far as science is concerned. It was just a handy way of thinking about the problem.

[Ryan]

Very Deep. You know, Physicist Victor Stenger thinks that the laws of physics are derived from physical symmetries. I've written something on this under "Answering the Big questions"

http://www.godriddance.com

-Ryan