There are non-perturbative string theory calculations out there, but they have no known relation to the real world, so can’t be used to predict anything.

I’m not sure what Douglas and Vafa work you are referring to. One guess would be toy d=2 string models and topological strings. Neither of these has any understood connection to 4d physics. Similarly with matrix models. There simply is no such thing as a non-perturbative formulation of string theory that allows one to calculate anything in the situation one cares about of four large dimensions.

Even worse, you not only don’t have a theory you can calculate with, you don’t even know the energy scale of the mythical theory you want to believe in. Is it the Planck scale, is it 1 Tev? It could be anything as long as it’s not too small since then it would be ruled out by experiment.

So, not only is there no prediction of an infinite tower of new states, there is no prediction of how many new states to expect from the theory, what their properties will be, or even at what energy scale they will occur.

true, the full tower of massive states would only be observable for vanishing coupling. At finite coupling string self-interactions lead to a collapse of highly excited states as computed in papers concerned with string/black hole correspondence, e.g.

Thibault Damour & Gabriele Veneziano, Self-gravitating fundamental strings and black-holes (1999)

This effect leads to the creation of ‘string balls‘:

Kingman Cheung, Black hole, string ball, and p-brane production at
hadronic supercolliders (2002)

(I recall that we had a very similar discussion before.)

Yes, the above paper mentiones large extra dimensions such that the string scale would come into reach of next-generation colliders. If the string scale is further away, this means that these experiments will be possible only at higher energies. But it is possible in principle.

Regarding the general criticism that only perturbative effects are known: Recent results by Douglas on D-branes and by Vafa and others are all concerned with non-perturbative effects. But I don’t understand these very technical paper well enough to say much more about them.

Regarding the fact that Gross and Witten say that a full non-perturbative formulation is still not known:

Sure. Robert Helling for instance (together with Nicolai and others) has shown that BFSS can only apply to a certain ’sector’ of M-theory. But there it is a fully non-perturbative formulation. Helling tells me that the advent of the AdS/CFT ‘bandwagon’ has caused many people to reduce research in BFSS and concentrate on Maldacena’s conjecture. Hence counting number of people and papers in these fields doesn’t say much about the field itself.

A few comments:

About non-perturbative string theory: I did quickly look at the matrix model references you gave and saw no real evidence there of any hope of extracting realistic, 4d physics out of them. Maybe you or someone else can make progress on this, but it looks to me like a very unpromising idea to pursue. The statement that there is no known consistent non-perturbative formulation of string theory that gives realistic physics (only four large dimensions, recognizable particle spectrum) is not controversial. Gross and Witten for years have made clear in all of their general talks about string theory that they feel they don’t have a good idea about what non-perturbative string theory is.

About Distler’s comments: He says that string theory “promises you” universal behavior of quantum gravitational effects at high energies. The problem is that this is still a promise, not reality, and will continue to be only a promise until either someone shows that perturbative string theory is a consistent approximation to something or finds a true non-perturbative theory.

About the string theory “prediction” of an infinite tower of states:

This is a “prediction” of perturbative string theory. Why should one believe that perturbation theory gets the properties of the higher mass states right but not the vacuum state? Isn’t this just because one can experimentally check whether a prediction of the vacuum state is right, but you can’t check “predictions” of higher energy states so you are free to claim them? Non-perturbatively you are going to have all sorts of things happening with these higher energy states. Won’t you start producing not just perturbative excitations of the string, but black holes, branes, and who knows what else? The high energy spectrum could be just about anything.

Ummmm, yes. Some people think it was rather too subtle, some not.

“I heard from Harvard yesterday that Lubos Motl has decided to take up professional wrestling now that his job there is being cancelled.”

I had thought about including news about Lubos, but he is pretty hard to parody.

I think it is a reasonable standpoint to say, as you implicitly do in your last comment, that quantum gravity as such is a problematic enterprise because it is not clear yet if and how the theory can be experimentally tested in practice.

If anyone feels that this is reason enough for him or her not to be interested in quantum gravity then that’s fine with me. Similar comments probably apply to research for instance in theoretical quantum computing as well as some aspects of (classical) cosmology.

At the moment we cannot check this, just as we cannot check if the core of Pluto isn’t made of green cheese.
>

Hence this prediction is not falsifyable within a reasonable amount of
time. Or is there any experiment that can produce this tower of
infinitely massive particles?

This is probably the most disturbing aspect of string theory. It may
make predictions in principle, like LQG predicts quantization of area
and volume, but no experiment will ever be able to disprove these
predictions, so they are not falsifyable. As Dan Friedan points out,
only large-distance physics is really science.

I find it very disturbing that most string theorists simply don’t seem
to understand that a theory is not testable unless there is some
experiment that can test it, and falsify it, within a reasonable
amount of time.

At the moment we cannot check this, just as we cannot check if the core of Pluto isn’t made of green cheese.

Please, Urs! Do you really claim that string theory makes hard
predictions, that can be conclusively tested, and possibly falsified,
within our lifetime or indeed within the present millenium? Would you
then please tell us exactly which experimental result that would
conclusively prove string theory wrong? No light Higgs at the
Tevatron? No sparticles or extradimensions at LHC? No proton decay at
SuperK? A positive cosmological constant?

Unless you can give us an example of a falsifyable string theory
prediction that can be tested within a reasonable amount of time, I
think you should stop insinuating that there is.

I agree with you that LQG cannot make such predictions neither, and
that it has other problems as well, but at least Ashtekar and Rovelli
don’t go around and claim that LQG is more predictive than QFT.