As a mature age Ph.D. student myself I had already resigned myself to leaving the academy, it being unwelcoming to the unexpected. I had hped that was just understandable age prejudice and had was not endemic.

Perhaps I’m a hopeless romantic but Lee’s indictment of modern science is scary and I have yet to see a single rebuttal to the meat of the main thesis presented in TTWP.

Dear Lee, This description shows that the convention is not to state or emphasize what was not proved but rather what was proved. Professional ethics is largely based on conventions. According to your own description, even Feynman was unable to change this convention. (In general physics journals, not string theory journals.) This only supports my opinion that your critique regarding string theory’s methodology has little to do with ethics and maybe it has little to do with string theory as well.

(As we saw again and again the questions of what was proved and what was not proved and what is the interpretation are complicated and disputed; be that in the complicated issue of finiteness of string theory, or the less complicated question if the “free lunch theorem” has anything to do with evolution. It looks that the editors and not Feynman were correct in their approach. It is hard enough for authors to tell clearly what was shown in their paper and what is the interpretation. It is OK to leave the “what was not shown” part for others (or for blogs.))

Finally I would hope that our taking the time to sort out and clarify the actual logic behind this complicated situation so as to make clear what remains to do would be taken as friendly and supportive of string theory.

This is exactly the point, there is a missing assumption in your argument. I do agree there is a decent, heuristic (not rigorous) argument that ‘a quantum gravity theory on AdS should always be dual to a conformal field theory’. In my paper on the subject with Arnsdorf we called this argument conformal induction. But there is no non-perturbative definition of string theory with asymptotically AdS boundary conditions, and indeed, except for certain extremal limits, nor is there any perturbative definition. (I agree there is a definition of supergravity and agree that the argument of conformal induction establishes a relationship between it and a CFT which is a limit of N=4 SyM.) So you cannot use this argument to reason from a theory that so far has not been shown to exist.

One can reason in several ways:

1) one can state that one is making an additional assumption

A) there is a full non-perturbative definition of string theory with asymptotically AdS boundary conditions.

If you assume A you can use this assumption to learn things that will be true about this theory, if it exists. But because you made this assumption you cannot use the same argument to demonstrate A, because you are then assuming what needs to be proving. This is the fallacy in your argument.

or

2) You can evaluate the results on AdS/CFT as to how strong evidence they give for A). In that case you also cannot use A to strengthen the case for A.

or

3) You can try to actually formulate the theory that A) asserts exists. This means give a genuine non-perturbative formulation of string theory. This is what I think we should try harder to do.

or

4) You can try to state a weaker conjecture, which is that string theory is DEFINED by it being dual to N=4 SYM. This is different from the conjecture that there are two theories which are equivalent. To go this route however, one needs 1) a full non-perturbative definition of N=4 SYM, which doesn’t exist, because its hard to make lattice regularizations of SUSY theories. 2) a precise description of the observables of the bulk quantum gravity theory in the precise Hilbert space of N=4 SYM.

Let me emphasize that I don’t rule out the possibility that 3) or 4) are doable and that there really is a precise non-perturbative string theory in this context. A real demonstration of 3) or 4) would be tremendous progress.

All I am asking for is that we be straightforward about what has and has not been shown and we do not reason in fallacious ways that assume the existence of the theory we are aiming to construct. I would claim that, among other things, to do so slows down progress because it makes people complacent about what still is not known, so they don’t try so hard to do the big steps that real progress requires.

Finally I would hope that our taking the time to sort out and clarify the actual logic behind this complicated situation so as to make clear what remains to do would be taken as friendly and supportive of string theory. If people were not so defensive they would see this as helpful. This was the spirit that my work on this with Arnsdorf was done.

Thanks,

Lee