Many string theorists seem to have decided to react to the criticisms of string theory that have recently been getting a lot of attention by going to the press with claims to have experimental tests of string theory that can be performed in the very near future.

The latest of these claims has nothing at all to do with the aspect of string theory that has come under criticism, its failure as a unified theory of gravity and particle physics, but instead involves the conjectural use of string theory as an approximation method in QCD. The main problem with this idea so far is that it involves not QCD but a related theory (N=4 supersymmetric Yang-Mills), and it is very unclear exactly what the relation is between the calculation and the real world. For some earlier comments about this, see here. John Baez also has a summary, and the Backreaction blog of Sabine Hossenfelder and Stefan Scherer has a very extensive explanation here.

Last week the AIP Physics News site carried a story entitled String Theory Explains RHIC Jet Suppression, which dealt with recent work by Hong Liu, Krishna Rajagopal and Urs Wiedemann concerning the jet quenching parameter, which describes how charmed quarks move through a quark-gluon plasma. In the AIP story, Rajagopal claims that their calculation “agrees closely with the experimentally observed value”, and that other related calculations “make a specific testable prediction using string theory.” This story was picked up by Scientific American, which has a story by JR Minkel. According to Scientific American, “trying to fit the QCD-like theory to reality makes the results only semi-precise, Rajagopal says,” alluding to the problem of doing the calculation in the wrong theory. Maldacena is quoted about this as follows:

*It’s like saying you are trying to study water, but instead you are studying alcohol… We certainly know it’s not the correct theory, but maybe it behaves in the same way.
*

Theorist Ulrich Heinz is even more skeptical:

Even if any of the numbers worked out by accident, I don’t think it would validate the approach… If they predict the color of an apple, and somebody looks at a pear and finds it has the same color, would you say that the prediction was correct?

Even if any of the numbers worked out by accident, I don’t think it would validate the approach… If they predict the color of an apple, and somebody looks at a pear and finds it has the same color, would you say that the prediction was correct?

Other recent claims by Shiu et al. and Distler et al. to be able to make predictions using the string theory approach to unifying physics are covered in the latest issue of Plus magazine, in a story entitled Stringent Tests. According to this story

*It seems that string theory, so far the strongest contender for a physical “theory of everything”, may soon be put to the test for the first time. Two separate teams of physicists have just published work describing how to compare the theory’s predictions with reality….*

*Neither of the two new tests will be capable of verifying string theory once and for all. If the results concur with its predictions, then this is just some further evidence for its correctness, not absolute proof. But the tests’ ability to falsify string theory, or at least certain aspects of it, means that a philosophical barrier has been overcome.*

On a somewhat related note, I’ll soon be traveling to Italy, giving talks in Rome and Pisa on the topic of “Is String Theory Testable?”.

The only way to sort out this mess is to REALLY figure out QCD. Why don’t we spend our time doing that instead of playing cowboys?

Fancy new string-theory-inspired methods to calculate amplitudes in supersymmetric theories do have relevance to the real world. For one thing, purely gluonic QCD tree amplitudes actually coincide with the analogous ones in susy theories. At loop level, one can add appropriately chosen susy amplitudes in a clever way and end up with a qcd amplitude.

(i’m not for a minute defending the ridiculous pseudo-predictions you’ve reported on recently, Peter)

There’s no denying that string theory has inspired a lot of important ideas, both in math and in physics. The field would be a lot healthier if people would just admit that trying to get a TOE out of string theory doesn’t work, and if they really want to keep doing string theory, instead focus on trying to better understand those parts of the theory that might lead to something interesting. Trying to better understand QCD via AdS/CFT is a perfectly reasonable thing to be doing, although over-enthusiastic claims about where that project is don’t help matters. And it would be nice if there were more encouragement for people trying to understand QCD or other non-perturbative QFTs using non-string theory methods.

“The field would be a lot healthier if people would just admit that trying to get a TOE out of string theory doesn’t work.”

It depends on your definition of “TOE”.

If by “everything” you mean something like “the Standard Model, along with whatever (if anything) supplements it at the next 10 orders of magnitude (or so) in energy”, then I think that the vast majority of praciticing string theorists (it’s hard to think of any exceptions) would happily admit this.

However, many string theorists have a broader definition of “everything”, roughly something along the lines of “the set of all possible consistent quantum theories of gravity”.

A String Theorist,

Interesting to hear that almost all string theorists have given up on getting particle physics out of string theory. Maybe you’re right, most of the ones I talk to are certainly pretty pessimistic about it, although not necessarily willing to give up on the idea.

Besides the usual arguments about whether string theory has actually produced a “consistent quantum theory of gravity”, I’m not sure I see the point of a research program devoted to generating “the set of all possible consistent quantum theories of gravity” based on the idea that they’re all extremely complicated, ugly and untestable. This will certainly keep one busy for a long time, but why do it?

Kea,

There is a very tiny group of people who still try to figure out QCD. Unfortunately, it’s a tough problem. The best we can do now is try to solve related models which we hope are relevant, much as the string theorists do. Whose models are better is in the eye of the beholder, and will

continue to be so until QCD is finally understood.

String theory is testable, even

supertestable

If a theory predicts something, that only validates the theory if there is evidence the theory is the only one possible. Ptolemy’s epicycles-based mathematical model of the universe allowed predictions of where the planets would be at any time. As a spin-off, it also led to new mathematical methods. It’s practitioners were extremely enthusiastic about it being the final theory of everything, so they tried to block all nonsense of alternative ideas without investigating them (one turned out to be more predictive!).

Falsifiability isn’t the key. Newton’s theory failed to predict the precession of the perhelion of Mercury, but that didn’t lead to dumping the theory. Today the Pioneer anomaly doesn’t falsify general relativity. An error in prediction is just an anomaly. This would be true for string theory if it made a predictive error. The real problem for string theory is that it isn’t based on an facts.

Time and again the scientific method in physics has been to

collect data

summarise the data by empirical maths

check the maths by predictive extrapolations

come up with a theory for the maths that predicts more

check it experimentally/observationally

The basis of M-theory is religious belief in spin-2 gravitons, planck scale unification, and the 6/7 extra-dimensions which are required to make string explain these speculations. It’s not factual physics.

Peter: You mentioned your trip to Italy which reminded me of a famous Italian story about strings: Pinocchio! In the Disney version, Pinocchio sings: “I’ve got no strings to hold me down…”

The best way to test string theory is to construct a realistic MSSM model derived from string theory that can be tested by LHC. In this regard, please see our upcoming paper (Chen, Li, Mayes, Nanopoulos).

This post conflates a number of questions that are

really quite distinct:

1) Can string theory predictions for physics beyond the

standard model be tested? (Do such predictions exist?)

2) Can string dualities (like AdS/CFT) be tested?

3) Are AdS/CFT results useful for QCD phenomenology?

I would argue that the answers are:

1) Maybe. A definitive test in the near future would

certainly seem to require a lucky break.

2) Absolutely.

3) Yes. I agree with you that “semi-precise” in is an

unfortunate choice of words. At present, this is clearly

a qualitative rather than a quantitative exercise.

“I’m not sure I see the point of a research program devoted to generating “the set of all possible consistent quantum theories of gravity” based on the idea that they’re all extremely complicated, ugly and untestable. This will certainly keep one busy for a long time, but why do it?”

They are not *all* extremely complicated, ugly and untestable. Some of them, such as for example topological string theories or maximally supersymmetric ones, can be quite simple and beautiful. For the same reasons, these are also the most well-studied (“under the lamppost”).

As far as testability, there is absolutely no reason to expect any observational consequences of quantum gravity below 10^19 GeV. We might get luck, of course (large extra dimensions, braneworld, cosmological signatures, etc…)

Thomas,

I think some of the people promoting this work to the press would actually like to see some of these issues conflated in the public mind…

You’re also conflating different things, by using the term “test” in quite different contexts.

I’ll agree with you about 2), duality conjectures like AdS/CFT coming from string theory can be tested. Whenever you manage to compute things that correspond on either side of the duality, whether they match is a test of the conjecture. But this is something completely different from an experimental test of a theory. You’re testing the conjecture that two theories are related, not testing whether a theory has anything to do with the real world. If you go around telling people “string theory is testable”, meaning that you have tests concerning whether a certain string theory is related to QCD, you are being misleading.

About 1), that’s what I’m writing a talk about. I’d claim that the conventional understanding of the phrase “test a theory” means to use distinctive features of the theory to make falsifiable predictions about what you will see if you do a specific experiment, acknowledging the theory is wrong if you don’t see what you predicted. By this definition, string theory is not testable. Attempts to use different definitions of “testable” again sometimes seem to be motivated by a desire to obscure this point.

As for 3), sure, this is an uncontrolled approximation, maybe it’s qualitatively useful. That’s fine and an interesting thing to work on, but people should avoid misleading the public about what is going on.

Oops, I clicked submit again before I was finished:

We might get lucky, but this fact has been clear for decades — it’s not as if there has been some massive conspiracy on the part of string theorists to cover up elementary dimensional analyisis.

“This will certainly keep one busy for a long time, but why do it?”

Different people will give you different answers, but fundamentally the reason people study string theory is that it interests them. Some people study the distribution of prime numbers, others study pre-Columbian musical rituals. Most people find both of those topics, as well as string theory, to be uninteresting wastes of time. But they don’t start blogs, or write books, about their prejudices.

The March 5 observation by AST misses the point. Of course, many people study many things for their own sake. The difference is that they do not make extraordinary claims about what they are doing. No one studying prime numbers has asserted that they are on the verge of explaining huge swathes of the natural world, at least, not since Pythagoras. If string theorists said “Hey, this is just esoteric math stuff that is fun” there would be no real issue. In fact, it appears that many make profound claims, and other string theorists implicitly endorse those claims by not opposing them.

However, many string theorists have a broader definition of “everything”, roughly something along the lines of “the set of all possible consistent quantum theories of gravity”.Some days ago, I stumbled across the Wikipedia entry about Paul Ehrenfest. I learned a number of things, e.g. that Ehrenfest’s suicide probably had something to do with his youngest son having Down’s syndrome. What really caught my eye, however, was the title of his 1912 inaugural lecture: “About the crises of the light-ether hypothesis”. My point is that in 1912, i.e.

* 25 years after the Michelson-Morley experiment,

* 12 years after Planck’s formula for blackbody radiation,

* 7 years after Einstein’s explanation of the photo-electric effect and discovery of special relativity,

mainstream physicists realized that there was a crisis in ether theory.

How could an obviously wrong idea keep its hold for such a long time? Clearly because most physicists considered it as “the set of all possible consistent theories of light”.

It seems to me that string theory is following the trail of ether theory, exactly one century behind.

Hi Peter,

thanks for the link.

Trying to better understand QCD via AdS/CFT is a perfectly reasonable thing to be doing, although over-enthusiastic claims about where that project is don’t help matters.I think the sentence you quoted: ‘make a specific testable prediction using string theory’ has been very carefully formulated. Krishna Rajagopal gave a colloquium here at PI some weeks ago about his calculations, and he has definitely not made any over-enthusiastic claims. For me it looks like this: there is a model that has turned out to agree with some observables. Now people are using it to make actual predictions to see how far its range of application extends. Whether or not one can justify

whythis is (or isn’t) a useful approximation (in certain regions of the phase diagram) is a different question. For sure one would like to understand that better.In case somebody is interested, there’s video, audio and slides at the PI websites

Probing the Properties of Quark-Gluon Plasma

(I hope the link works, they keep moving the item IDs which drives me nuts. In case the link is broken, search the seminar series by speaker.)

What concerns me most about the AdS/QCD hype is that suddenly so many people jump on the topic. For one, even if it turns out to be the greatest model for heavy ion collisions ever, there will be only limited demand on people working on it. But what’s worse is that it means all these people have previously worked on something so fascinating that they are willing to drop it from one day to the next.

Best,

B.

Bee,

Thanks for the explanation and link to Rajagopal’s talk.

A String Theorist,

The problem is that the non-ugly string theory backgrounds provably don’t relate to the real world. If you want to amuse yourself by studying string theory backgrounds that you know can’t ever be related to the real world, purely for your intellectual entertainment, that’s fine, but you should be honest about it, and then compete for resources with lots of other people who have their own arcane intellectual interests. Somehow I’ve never seen this description of string theory research on a grant proposal or research statement from someone trying to get a job.

Some of these things (especially topological strings) do involve some wonderful mathematical structures, and they’re under active research by many people in the mathematics community. If this is what you want to do, great. I do suggest though that you might want to try and get a job in a math department. You’d find there some very smart people who know a lot about what is already known about the relevant mathematical structures, and an active, healthy research field.

Your talk in Pisa is somewhat funny:

https://indico.pi.infn.it/categoryDisplay.py?categId=24

March:

29 Raphael Bousso (Berkeley), Predictions in the Landscape

20 Paolo Gambino TBA

19 – 21 Supersymmetry, Supergravity, Superstrings

15 Peter Woit Is String Theory Testable?

First you, and then, when the audience has learned something, comes Bousso!

It is important, that you traven to europe. Since some departments think of installing LARGE string departments.

It would be important that you say, string theory should be installed, but only with small groups, for investigation of QCD or purely mathematical physics and under all circumstances, not as the only game in town for fundamental particle physics.

Otherwise it is the danger, that Europe makes the same mistakes as theorists did in the US.

“Somehow I’ve never seen this description of string theory research on a grant proposal or research statement from someone trying to get a job.”

How often do you think it happens that people who study pre-columbian music rituals or prime number distributions write grant applications reading “my reseach is completely esoteric and without practical applications, focusing on some details of a field which most people consider to be an uninteresting waste of time”?

grant,

Obviously my comment was a bit sarcastic. In grant or job applications, people try and make the case for the widest possible significance of their work. But the honest ones don’t write things that they privately acknowledge to be untrue. Claiming that what you are doing is going to lead to predictions about beyond the standard model physics when you know very well this is not possible is not honest.

Look, I know people who study prime number distributions, and have seen their grant and job applications. These documents make a case for studying prime number distributions for their intrinsic interest and give legitimate arguments about the significance of this kind of research for the rest of mathematics.

Peter, this is hilarious! Are you seriously saying that you will give talks on a subject on which you’ve never ever worked in your entire life? Writing a blog about it is fine because one can blog just about anything but giving a talk at a university??? And you complain about the lack of professional attitudes on the part of certain string theorists?

puzzled,

If you’re unhappy with the concept of my giving these talks, take it up with the physicists who invited me to do so, I can assure you that the talks were not my idea. Quite a few physicists read this blog, and many of them seem to find what I have to say well-informed and worth listening to. Also, I hate to break this to you, but it’s not the first time I’ve given a public talk about this at a university.

As a relevant example, I don’t see why only people who have written papers on the landscape should be giving talks about it and the problems it raises. It might be a good idea to hear from people who have experience as Ph.Ds and postdocs in particle theory, have spent a lot of time learning about the issue and debating it with others, coming to the conclusion that “working” on the landscape by writing papers about it is not a good idea.

And, you know, it’s really tiresome how many string theorists seem to be unwilling to put their name publicly to their critical comments here. That is something I find highly unprofessional.

Eric,

Sorry your comment initially got caught in my spam filter for some unknown reason.

If the LHC sees super-partners, how can you tell they came from a string theory.?In particular, what values of the MSSM parameter space are predicted by string theory, or does string theory say nothing about this?

A propos your suggestion that string theorists should look for jobs in math departments:

Of course as you know several string theorists have gone to math departments. I think this movement will pick up steam in the next few years. In fact, this year is the first year that (barely) more than half of the “string theory” jobs listed on the rumor page are actually in math departments.

It is very nice to see the overall moderate tone of the post and remarks. I was impressed by Peter’s sentence “There’s no denying that string theory has inspired a lot of important ideas, both in math and in physics”. Recognizing this while playing a role of “devil’s advocate” regarding string theory is perfectly fine.

Attempts to make predictions via string theory should be welcomed by everybody. As we often try to make predictions on humans based on mice, predicting the behavior of water based on alcohol and of apples based on pears is reasonable.

A String Theorist Said:

When String Theory was initially promoted as “the only serious candidate to TOE”, string theorists defined “everything” like

everything(not just extension of SM or a quantum gravity theory) in papers, lectures, and books.String Theory was promoted in that way to general public also.

It is fine if you want change now the meaning of “everything” in basis to the recent String Theory

fiascoas a sensible TOE. It is fine if now the superb String Theory is now to be promoted like just some kind of QCD recipe.It is not fine you want change history.

The prime number theorist might well include this in the grant application:

My theory, with a little more work, will solve important problems in Physics. See:

“Applications of Statistical Mechanics in Prime Number Theory”

Marek Wolf

http://secamlocal.ex.ac.uk/people/staff/mrwatkin/zeta/wolfgas.htm

Wolf begins by briefly referring to the general phenomenon of the ‘number theory – physics crossover’. Two references to the application of prime number theory to quantum chaology in particular are given [1,2,3]. Two mathematical properties which are normally associated with physical systems have been detected in the distribution of prime numbers by Wolf himself, namely multifractality [4] and 1/f noise [5]. A paper which numerically calculates Lyapunov exponents (familiar to chaos theorists) for the distribution of primes is mentioned [6]. Finally an article is cited [7] wherein the authors apply the Wiener-Khintchine formula, which relates spectral densities and the autocorrelation function, to the problem of the distribution of pairs of primes {p,p+2}.

Pingback: Aaron’s and Lubos’s anti-Loop Quantum Gravity propaganda « Quantum field theory

Juan R:

Your quote about the “theory of everything” is precisely in accord with my original post.

String theory IS a theory of everything, including many things (such as topological strings, an example I gave above) which do not actually “exist” in our universe.

But even though it is the deepest possible theory of physics, it’s not going to tell you what the next winning lottery number is (any more than it will tell you what the mass of the electron is).

No, it’s not a theory because it doesn’t explain anything about real phenomena. It links gravitons nobody has ever seen to Planck scale unification nobody can see, using extra-dimensions that have to be explained away by a Calabi-Yau manifold which gives the string a massive landscape of 10^500 solutions of particle physics.

Explaining (by further speculation) a few speculations about gravitons and unification isn’t a theory of everything. It’s not even a ‘theory’ about speculations, it’s just vague hype that isn’t tied down to any known facts. The claim it’s the “deepest possible theory of physics” suggests you have a disproof of LQG and every suggested alternative. Where are these disproofs?