« The Knowledge | Main | Book Review »
April 10, 2007
On not understanding physics
My understanding of modern physics isn't high. I saw an episode or Horizon on it about ten or twenty years ago. Oh, and I once knew enough about Schrödinger wave equations to get through A-level chemistry.
However, from time to time I do look at Not Even Wrong, Peter Woit's on-going record of his dim view of string theory. I'm not really sure why. Perhaps its because academic wrangling—although unedifying—can make a good spectator sport.
In this vein I did find the Bognadov affair (a.k.a. the reverse Sokal) quite fun. I think it's remarkable that, going solely on the account on John Baez's website there's no particular reason—from a lay perspective—for believing the claim that the Bognadoff's work was jargon-thick gibberish. Obviously sock puppetry doesn't suggest any great trustworthiness. But if a judgement here rests on Baez coming across as a more honest character then this doesn't go anywhere near the physics.
I'm not particularly trying to suggest that the Bognadoff's have been picked upon unfairly, or that Baez has revealed some weakness by arguing from (his own) authority. Instead, the lesson I'd take from this is that technical discussions often don't make much sense from the outside. The idea of trying to explain physics in layman's terms doesn't make much sense to me. To be blunt: if layman's terms were adequate then you wouldn't have much need for physicists. Not that this necessarily matters, unless…I'll I'll get back to that later.
I could go beyond saying that physics can only be understood in its own terms. Somewhere Bachelard says that it is the mathematics that physicists use—rather than the physicists themselves—that understands physics. Which is to say that physics has become a cognitive machine that no longer requires the consciousness of a physicist at its heart. Not that this would stop the machine working, and so might not be important, but… again I'll leave this hanging.
Ignoring everything I've just said, I want to make some sense of string theory. My understanding is that string theory attempts to combine quantum mechanics and the theory of relativity by assuming that the universe has—despite appearances—a dozen or so spatial dimensions. There no doubt a great deal more to it, but that will do for me.
This chimes with my dim recollections of this telly program from years ago that I mentioned: something about the equations describing electro-magnetism being the same as something else (weak nuclear force?) if the universe were assumed to be five dimensional.
The obvious question to ask of this is whether it is really true that the universe has extra dimensions (wrapped up small so no-one can see them), or whether this is just a trick to make simplify the mathematics. My prejudice would be towards saying that it if it doesn't make any difference then it doesn't really matter, but that I doubt I'd care even if it did.
String theory is criticised for not being testable. Part of this seems to be a claim that it isn't a theory at all: it is a label for a research programme that hopes to find a theory but hasn't done yet. (And that, as the decades roll by, it becomes less credible to defend it as being a “promising approach”.)
But apparently there is another problem for anyone trying to combine quantum mechanics with relativity: the predictions made by these theories only differ noticeably from those of classical mechanics under certain conditions, and the sets of conditions for the two theories don't overlap. There's no reason why in principle they couldn't; but—for the time being at least—they just don't. Any unified theory must give an account of how ‘quantum phenomena’ and ‘relativity phenomena’ interact; but this has to be done in the absence of any experimental data as to what this interaction might be like.
I think in any other subject this sort of thing might be taken as a hint to give up and go home. If it were me I'd say that there's not much point in spending a great dealing of time working out how you might predict the results of experiments that can't be performed: better to cross the bridge when you come to it; and if you don't come to it then you'd have saved a lot of effort by not bothering before-hand. But physics isn't any other subject. No, physics is special.
Stepping back a bit, one can ask: what is the problem that string theory is there to solve? Well, quantum mechanics and relativity—the two jewels in the crown of 20th century physics—are incompatible. OK, so why is this a problem?
Given a commitment to a realist interpretation of physics—that theories aren't just there to fit experimental data, they describe the underlying laws of nature—then it must be irksome to require two different theories: it's not as if the universe could obey different laws depending on how it was being considered. But then life is full of disappointments: there must be more to it than this.
What is at stake is to do with one of the original motivations for mathematical physics—an inheritance from the renaissance hermeticism from which it arose. It befits the dignity of the miracle that is Man that He—having been given granted an intellect almost divine—should be the one who comprehends the secrets of the cosmos; who reads from the book of nature in its own language; who hums along with the music of the spheres; who should come to know the mind of God. This is why physics must push onwards to an unambiguous, and supremely elegant, statement of the laws of nature: it is Man's sacred destiny to know them. And, at last, this is why it won't do to have a theory that absolutely no-one (or, more charitably, almost no-one) understands: this isn't book-keeping; this knowledge is beatific.
It's enough to give an aspirin a headache.
Posted by robin2 at April 10, 2007 07:34 AM