Slashdot Mirror
String Theory Put to the Test
secretsather writes to mention that scientists have come up with a definitive test that could prove or disprove string theory. The project is described as "Similar to the well known U.S. particle collider at Fermi Lab, the Large Hadron Collider, scheduled for November 2007, is expected to be the largest, and highest energy particle accelerator in existence; it will use liquid helium cooled superconducting magnets to produce electric fields that will propel particles to near light speeds in a 16.7 mile circular tunnel. They then introduce a new particle into the accelerator, which collides with the existing ones, scattering many other mysterious subatomic particles about."
Who do you get to be an expert to tell you something's not obvious? The least insightful person you can find? -J Roberts
I think it's funny how the article forgets to mention that the LH collider is located at the CERN (the European nuclear physics institute). As a matter of fact, it is not only in Switzerland, but extends to France as well. The article only mentions it is similar to the U.S. Fermilab accelerator, but then forgets to add that there are many kinds of accelerators world wide.
Funny, ain't it?
"The canonical forms of string theory include three mathematical assumptions--Lorentz invariance, analyticity and unitarity. Our test sets bounds on these assumptions." --Benjamin Grinstein
Don't quantum mechanics and GRT also include the above? Meaning if the experements don't confirm the above then more than just string theory is in trouble.
Of course analyticity probably has some very subtle meaning in string theory. Any one here in the know?
Everyone always seems eager to compare to epicycles any modern physics theory they don't care for. String theory, dark matter, what have you...
Physicists were led to string theory in a search for a consistent theory of quantum gravity, not in a search to make up the most complicated theory possible to fudge arbitrary data. For more on why string theory should be taken seriously as a solution to this problem, you can read a long analysis in a previous post of mine here. String theory itself cannot be modified to "fit" to a model; it is a unique theory with no adjustable parameters or interactions. However, you can construct various string models to fit observations, as you can presently using quantum field theory models like the Standard Model.
It is also not correct that string theory doesn't make testable predictions. This whole story is about testing predictions of certain string models. However, we can't presently test predictions of all string models at once, and thus rule out all of string theory. But then, the same is true of quantum field theory models as well; there are infinitely many such models that could be true but which we can't yet test.
There are background independent formulations of string theory, but none that give (4D, non-supersymmetric) GR in an obvious way. However, formal background independence is a matter of philosophical preference, not physical necessity. The basic idea may seem simple, but is overlaid by a lot of kludges such as supersymmetry to eliminate tachyons and fluxes to get a positive cosmological constant? I wouldn't call adding supersymmetry to eliminate tachyons a "kludge", anymore than, say, adding gauge invariance in QFT to eliminate non-renormalizability. As for the positive cosmological constant, I'm not up to date on what is necessary to get that to work out in string theory, but all quantum theories have had notable problems explaining any realistic value of the cosmological constant.
IAAPhysicist, an experimental high energy physicist to be more precise, and I don't like string theory much. I am not opposed to its study, I do not campaign to have funding removed from its proponents, in short, I do not hate it. I just don't care for it, and rather hope that it turns out to be wrong. OTOH, I don't really like the Standard Model (and extensions to it) much either. I think that something different from either is what is needed. Not being a theorist, I am not working on an alternative myself, but I have seen one or two things at various conferences, and thought (just gut reaction) that they looked very promising. One in particular that I found myself unaccountably fond of was a neat little statistical approach from a guy at tamu.
Anyway, my reason for disliking string theory is not at all that I find it "too elegant" or "too cute". You have most of the experimental hep people I know, including myself, pegged quite wrong there. In my opinion, and that of most of my colleagues that I have discussed it with (not a large percentage of all my colleagues), the problem with string theory is that it is not as cute or elegant as it thinks it is. It has precious few free parameters (contrast the standard model), and its first principles are strikingly simple. That ought to be elegance. However, the fact remains, as the GP said, that getting our observable (3,1) universe to appear, even just at low energies, from string theory is quite difficult. Why is this? Primarily because string theory does not tell us how the small extra dimensions are wrapped up around each other. The topology of space presents a huge theory space to search around in.
The standard model is criticised because it does not nail down the values of its free parameters (tautology), and if you don't have the right values of those parameters, then the theory does not describe our universe. However, we can perform experiments which measure various values which depend upon those parameters, and by so doing, obtain values for those parameters with ever increasing precision. Thus, we can find the values such that the standard model describes our universe. Furthermore, the standard model is not chaotic. If you are just a little bit off in the values of your parameters, then your theory describes a universe which is very like ours.
Now, take string theory. The topology of space winds up acting very much like free parameters. However, we can't do experiments to measure the "value" of the topology of space, so finding the right topology is, as I understand it, a huge trial and error process. Furthermore, as I understand it, even if you managed to define some notion of "closeness" to the correct topology, one topology which was "closer" to right than another one would not produce a universe which was necessarily any "closer" in its various properties to correct than the other one. In that sense, string theory is chaotic. So, for all its apparent elegance, it seems to me that string theory is a great deal uglier in the end that QFT and the standard model. This is why I and many others do not like string theory.
SIGSEGV caught, terminating
wait... not that kind of sig.