In the years before the LHC start-up, one heavily promoted claim that “yes, string theory can too make predictions, and here’s what it predicts the LHC will see” was based on a class of models known as “F-theory”. Detailed superpartner mass spectra were produced and shown around the world at conferences and departmental colloquia. For an example, take a look at figures 3 and 29 of this paper.
Early LHC results showed that nothing like these mass spectra corresponds to reality. For the latest such results, see Resonaances, which describes new SUSY limits from ATLAS, now wildly out of agreement with the F-theory “predictions”.
At this week’s F-theory Workshop, there seems to have been little acknowledgement of this failure. I didn’t notice either any reference to the fate of the “predictions”, or even an attempt to come up with new, updated ones. The closest I could find was this comment by Michael Dine in a discussion of the state of F-theory phenomenology”:
A lot of us I think are resigned to the idea that maybe there’s supersymmetry and it’s going to look tuned, or maybe there’s not low energy supersymmetry. I think a challenge I’ve always said for string theory is to try and think about theories without supersymmetry and that has proven to be hard. But you know, that’s certainly a direction which maybe we’re being confronted with.
So, the long-standing ideology that supersymmetry stabilizes the weak scale, and seeing its effects will finally give evidence for string theory unification looks like it is crumbling. With this hope gone, string theory unification becomes a completely unpredictive subject, with no hope of connection to experiment. One has an infinite array of mathematically highly complex models one can spend time studying, but it’s hard to characterize doing so as any recognizable form of physical science.
This situation hasn’t slowed down string phenomenologists, who will follow up the F-theory workshop with a summer school for graduate students to train them in the failed techniques of the subject. I have a hard time understanding why any sensible graduate student would want to attend such a thing, or why any responsible advisor would encourage them to do so.
In case anyone’s curious that quote from Michael Dine begins shortly after minute 52:30 of this video discussion:
http://media.scgp.stonybrook.edu/video/video.php?f=20120322_1_qtp.mp4
If you drag the button to minute 53 you miss the start of it.
This was the Thursday 22 March 10:15-11:15 am session with Dine, Moore, Seiberg,
titled “A Critical Look at Phenomenological Issues in F-theory”
that you can find the link to in the overall program here:
http://scgp.stonybrook.edu/archives/1493
sigh, such a sad situation. how do these guys get funding! Someone should really attend
some of these talks and ask these hard questions.
OTOH I know a colleague who has been looking for jobs for last seven years without luck and only because the work he does is not trendy.
Great post IMHO. I am surprised at the level of dismay – after all, susy proponents could just claim we need higher energy (but maybe it is absolutely essential to see things at LHC scale? I don’t know – I’m kind of ignorant). I have to be honest: I really like SUSY except for the experimental non-validation.
But, to be frank, you have been puzzled for 20 years as to behaviour of responsible advisors. I mean that in a neutral, not negative way.
“One has an infinite array of mathematically highly complex models one can spend time studying, but it’s hard to characterize doing so as any recognizable form of physical science.”
“We seem to be at a critical point in the history of science, in which we must alter our conception of goals and of what makes a physical theory acceptable.”
Stephen Hawking, THE GRAND DESIGN.
You may reach a point where your physical theories cannot be verified directly via experiment due to their very nature and the nature of their general predictions. What do you do then? You abandon the theory? Of course not, the theory could be right. Then you need new criteria to decide when a theory is acceptable as scientific. These criteria are set by the people who are doing the actual scientific work i.e. the physics community. You can’t live in the 19th or 20th century forever following blindly a strict dogma of what science is and what isn’t. This just stalls progress.
Giotis,
Science without experimental verification isn’t science. Period. It may very well be that some theories can’t be tested with current (or future) technology, but then don’t call those scientific theories. Call it math, call it natural philosophy, or whatever. But don’t call it science.
Giotis: You can’t live in the 19th or 20th century forever following blindly a strict dogma of what science is and what isn’t. This just stalls progress.
Progress? Progress in what, exactly?
Of course not, the theory could be right.
Could be right? In what sense? Have you utterly forgotten that the whole point of testability is to have a way to determine from empirical evidence that a theory is wrong, and that no amount of confirmation, if it could be had, cold ever definitively establish that a theory is “right”? If “right” means anything, it means “has been subjected to many empirical tests, and has passed them”. That’s all you can hope for, and string theory hasn’t even found its way to the examination room. However, it sounds like F-theory has, and it flunked. I guess that would be something, if the grade wasn’t being disputed or ignored. 🙂
abbyyorker,
You can just claim that SUSY breaking is at a higher mass scale, but then you give up the main argument used for why you want SUSY. It starts to become just one more unmotivated, complicated extension of the SM which doesn’t explain anything. Why spend any time on such things?
Giotis,
You’re joking, right?
There is an important theoretical question to be answered, which is: how is it possible to reconcile quantum field theory and general relativity? Some string theorists claim that string theory answers this, but judging from the number of people working on how information gets out of black holes, for the most part, they seem to have either given up on the question or declared victory and moved on.
And maybe, before we start working on the question of how to mathematically reconcile quantum field theory with general relativity, we should really try to answer the question of how to mathematically reconcile quantum field theory with itself.
Re: Giotis’ comment, a senior mathematician/string theorist once said almost exactly the same thing to me, following it up with “The definition of `science’ was invented to satisfy Marxism. It has nothing to do with modern physics.”
I responded with “or you could say that modern physics has nothing to do with science” and _that_ was met with genuine shock. And a protracted silence.
I just want to make a clear distinction between constructive criticism which is a healthy endeavour and a Taliban-like preaching in the name of some scientific orthodoxy which demands the theoretical research to stop if it doesn’t follow some strict definitions and dogmas. These rules were made by the physics community and could be bent or re-adjusted by the same community if necessary in order to face the new difficult challenges of Planck scale physics.
To play devil’s advocate and also to ask a question . . .
Pretty much everyone here is being critical of F-theory, which is respected and worked on by many of the brightest in the scientific community. It has made a resurgence since 2008, at which point many good people who hadn’t worked on it started working on it because of apparent promise. This all happened after Peter’s book.
Who of you critics can actually state cleanly what F-theory is, what the status of the field is, and why people believe it is promising compared to (say) the heterotic or type II superstring.
I’m interested to hear your responses.
P.,
Personally I don’t see any point in arguments over the relative merits of heterotic/type II/F-theory models since they’ve all now failed more or less equally miserably, and all show no hope of leading anywhere.
I claim no expertise in F-theory, beyond that coming from sitting through several promotional talks (the topic is a favorite at math/physics conferences), and spending some time carefully looking at the supposed “LHC predictions” of the theory before writing about them here. At one of such talks, a prominent string theorist I know who was sitting next to me after the talk explained to me in detail how it showed that I was just wrong: string theory does make predictions about LHC physics, and the LHC results would change my mind. I was polite.
The resurgence of F-theory since 2008 was based to a large degree on the strong and explicit claims being made for LHC predictions based on F-theory models. I’m still waiting to hear the people who made those claims acknowledge what has now happened.
Peter,
It’s a bit irresponsible of a prominent string theorist to claim that F-theory definitely makes LHC predictions, in my opinion. On the other hand, there are definite reasons to believe that it gives rise to more promising models for particle physics (GUTs in particular) than the weakly coupled type II string, so it’s simply wrong to say that they’ve all failed “equally miserably.” Regarding the failure of the authors to address what their papers said in light of experimental data, I can’t speak to why they haven’t. I did take a look at Heckman’s talk, though, and it’s interesting to notice that F-theory did give rise to the correct Theta_13 angle that was recently measured in the PMNS matrix.
I took a look at some of the other videos and recent abstracts of speakers, and I think it’s important to point out that most people in the field are currently working on very hard problems having to do with the global aspects of F-theory compactifications. Many of the properties of the “local” GUTs studied over the last few years have things to say about GUT physics, but of course GUT physics isn’t everything and there are other effects in the vacua which must be taken into account. This is probably why the workshop at Simons didn’t seem as phenomenological as similar programs in the past.
P.
“It’s a bit irresponsible of a prominent string theorist to claim that F-theory definitely makes LHC predictions…. F-theory did give rise to the correct Theta_13 angle that was recently measured…”
Right. Glad to hear you’re opposed to irresponsible claims about F-theory.
As for why “the workshop at Simons didn’t seem as phenomenological as similar programs in the past”, I think it would be a reasonable conjecture that maybe the unmentioned LHC results that torpedoed and sank all the F-theory “predictions” of recent years might have something to do with this. Then again, it was a Math/Physics program, not a pure physics one.
P, can you point me to the F-theory paper which made the correct prediction about theta_13? I am curious. Also you said that F-theory predicts GUTs. So what does it predict for proton lifetime and which modes should one look for a signature?
Offtopic, but the Krauss thread is closed. If you have access to the New York Times, philosopher David Albert takes on Krauss.
http://www.nytimes.com/2012/03/25/books/review/a-universe-from-nothing-by-lawrence-m-krauss.html
@Peter:
Right, I am opposed to those claims about F-theory that are, in fact, irresponsible. Given the current state of the field, I would put “definite LHC predictions” on that list.
The problem is that the overall situation is much more nuanced, particularly as the field is still developing (again, F-theory for particle physics only since 2008), and to make blanket statements is not appropriate and doesn’t do justice to the dozens of very bright people working on the subject.
As I’m familiar with some of the literature, it’s worth correcting your conjecture about the mathematical (not obviously phenomenological) nature of the Simons workshop.
Most of the phenomenological scenarios of 2008 and on were so-called “local” models that focus on the physics near a stack of GUT 7-branes, where (say) the SU(5) gauge theory lives. This approach has its merits, as much of the GUT physics is described locally (GUT matter, Yukawas, and so on), but there are important “global” issues of the physics and of the compactification which are missed by the local picture. This isn’t surprising: the local picture isn’t a full blown F-theory compactification, but instead is a subset of the mathematical data which controls the GUT physics.
The reason for the more mathematical and “global” Simons workshop seems to have been to report on progress in understanding these global issues, including the description of chirality inducing G-fluxes. It’s only by understanding these global issues that one might hope to address moduli stabilization and (ideally) vacuum selection. It’s not that people have abandoned hope in the role of F-theory for phenomenology. It’s that they’re studying hard mathematical aspects which give rise to a deeper understanding.
@Shantanu:
I’m not sure, it’s likely a paper by Heckman and Vafa, though, and has neutrino or PMNS somewhere in the abstract. If you can’t find it I’ll let me know and I’ll google around a bit.
@ Shantanu, see http://arxiv.org/abs/0904.1419 for \theta_{13} prediction
Anonyrat,
Thanks. I read that over my coffee this morning and just wrote a short posting about it.
P,
it is by no means irresponsible to produce concrete LHC predictions. on the contrary, that is the essence of the theory being predictive.
what is irresponsible is the failure to acknowledge that these predictions and, consequently, the underlying model is now falsified.
Chris,
Sorry, I wasn’t clear enough. You missed the point.
The local F-theory models that produced LHC “predictions” are “local models”, not globally defined F-theory compactifications. This approach has it’s advantages, though, as many aspects of GUT physics can be understood, but global effects can change the physics and are completely missed by the F-theory models you’re referring to. People are making progress in understanding the global aspects, and they matter.
I completely agree that making concrete LHC predictions from a very well understood theory is of crucial importance and is absolutely the responsible thing to do! The point is that F-theory is still getting there (yet again, concrete particle aspects only since 2008) . . . and the authors of the mentioned predictions tried to understand as much of the low energy physics as they could from the local picture. But we know the local picture isn’t the whole story.
P
As a mathematician (not in math physics) I wonder how it feels to invest enormous amounts of time working on theoretical physics models which may end up being complete failures.
Math people have the advantage of using pure esthetics (and correctness) as the main criteria of interest- so anything you do which is interesting to you (if you have reasonably good taste) will not only be of interest (eventually!) to the community but will remain so in perpetuity.
Maybe math physics on the speculative fringe should give up on the supposed connections to the real world and ALL of the hype and just try to do good work instead?
Question is, whether it still makes sense to do it in a “physics way”- without precise definitions and proofs….standards are way way different. But if the hype were removed it would help clear the air of junk. Maybe that’s part of what Peter is about: trying to ridicule the hype out of HEP.
Bugsy,
That’s about right (although I think I reserve ridicule for only the ridiculous cases…)
F-theory is a good example. Lots of algebraic geometers have gotten involved in it, because questions about algebraic geometry arise in the subject. The hype about bogus progress can encourage them to spend their time and effort on questions of little real mathematical interest, whereas this time and effort would be better spent pursuing issues of mathematical interest that come up. This is independent of the question of rigor and proof. That’s another question, one where most mathematicians are aware of the dangers of getting too far away from what is precisely understood.
Peter,
Again, you’re not being specific enough. What precisely do you mean about the “the hype about bogus progress”? If you mean specific, completely unique LHC predictions, then I agree.
But from the point of view of the relationship between gauge theories in F-theory and mathematics, the relationship is deep, which is why the algebraic geometers work on it. For example, the heterotic F-theory duality gives an equivalence of moduli spaces of vector bundles on Calabi-Yau manifolds and elliptic K3 fibrations (roughly). The study of such moduli spaces actually governs gauge theories, including both the gauge symmetry and matter content in four dimensions. If string theory has anything to say about the real world, these things are clearly of great importance.
Mathematicians aren’t studying F-theory because of grandiose claims about local models. It is interesting to them because well-defined geometric moduli spaces correspond to concrete things in low energy gauge theories.
P
P,
I don’t disagree. I just think all mathematicians who get involved in this area need to realize that some of the talks they’re exposed to on this subject need to be consumed with a massive grain of salt. Especially ones that claim F-theory “predictions” about observable physics.
I really don’t see the point in the especially caustic rhetoric denouncing the mainstream BSM theories. You really should start complaining if/after the LHC significantly disagrees with the SM in an important way and this can be explained by theories simpler than SUSY.
SA,
If and when the LHC data significantly disagrees with the SM, I and everyone else will be too excited to complain about anything.
Right now though, I don’t see what’s wrong with pointing out which “mainstream BSM” models promoted as making LHC predictions have been shown to be wrong by experiment. When this happens, maybe they should stop being “mainstream” (if they ever were..), don’t you think? Or, if the predictions were bogus, maybe the question of why that was so should be examined. I understand that people whose models have recently died may still be in one of the earlier stages of grief and not yet at acceptance and ready to move on, but they should be encouraged to do so.
Peter,
Agreed, and this is where it is the responsibility of physicists to be honest with them: string vacua do make low energy predictions (and even sometimes stringy ones depending on M_string), but the big “predictivity” problem involves not individual vacua but instead the landscape and vacuum selection. Hopefully there is some mechanism. Of course, this is an extremely difficult but important problem and is even a problem in quantum field theory – effectively the same as asking “why the standard model?” from an enormous class of possible quantum field theories.
The role of mathematicians in this process includes helping to understand string vacua and the low energy predictions of certain vacua or classes of vacua. To say that a certain vacuum or class of vacua look something like the world (e.g. F-theory vacua can at certain regions in their moduli space) is not the same as addressing the landscape problem.
Do you agree?
P
P,
Sorry, but I think the problem is much worse than you describe it. I just don’t believe that constructing ever more complicated “string vacua”, designed to evade obvious conflict with experiment and using poorly controlled approximations to claim “predictions” about physics is a worthwhile activity. It’s been going on for nearly 30 years, and the results are clear. I’m quite serious that I think training graduate students to do this is irresponsible. The hope that somehow a “vacuum selection principle” will be found that will change the situation is based on nothing but wishful thinking.
The people doing this had a shot at showing they could get somewhere. They just needed to say something, anything, about what the LHC would see. That’s gone now, time to fold tent and give up.
Peter,
I think this is where we fundamentally disagree, and that’s fine.
For the last 30 years, every 3-5 years (or even shorter) there is a very serious development in the understanding of string theory that improves the our understanding of string vacua and the associated low energy physics. So it’s not just that people have been beating their heads against the wall for thirty years. There has been progress in understanding (I can cite papers and ideas, if you wish), and to me the irresponsible thing is to ignore this progress and completely stop the subject.
That being said, I agree that training hordes of graduate students in these techniques probably isn’t the way to go. The real progress in the subject over the last 15 years has been made by a relatively small number of people (order 100-200 worldwide?) and if the problem is to be solved, I don’t think it’s just by adding bodies. There are other interesting things to study as well, particularly with the data coming from LHC.
Cheers,
P
P,
My impression from following the field is that, yes, every 3-5 years there’s a new idea, but that these new ideas all go in the wrong direction: they show you some new sort of “string vacuum” to look at. The whole field started out in 84-5 with some very tight constraints on what a string vacuum could be, and since then the possibilities just keep multiplying. This is not a good thing…
In any case, there’s no danger that people will be forced to stop doing this, based on what I or anyone else thinks. It will likely be a major industry for quite a while, until an actual good idea about BSM physics and unification appears, at which point it will get dropped like a hot potato…
Peter,
Insofar as the only problems with the LHC are theoretical in nature, it makes sense to develop theories that mitigate these issues and explore their consequences. The most obvious extensions of the SM have been considered already: fourth generation, technicolor, more trivial extensions of the gauge group, more dimensions, or SUSY.
In the case of SUSY, all of the predictions of course depend on the mass spectrum, and the current limits on superpartner masses simply do not disqualify even the simplest SUSY models from solving the theoretical concerns that motivated them.
Er, sorry by LHC I meant SM.
SA,
“the current limits on superpartner masses simply do not disqualify even the simplest SUSY models from solving the theoretical concerns that motivated them.”
I just don’t think that is true. The main motivation given for SUSY pre-LHC was that it would stabilize the weak scale. Assuming this, you can go back and listen to talks from people like Arkani-Hamed explaining that SUSY should already have been seen at the Tevatron, so it was a sure thing that it would be there in the early LHC data. Now that it’s not, people are trying desperately to find ways around their earlier predictions. I don’t think it’s going to work. The argument that “well, we didn’t really mean it when we said SUSY had to show up at the LHC if it was going to stabilize the electroweak scale”, or “so what if SUSY doesn’t stabilize the electroweak scale” is going to be tried by SUSY proponents, but I don’t think it’s going to fly.
Peter,
This is not really a matter of opinion. We can be specific here. If the light CP-even Higgs has a mass of about 125 GeV, then even in the case of minimal supergravity, one can stabilize the electroweak scale while being totally consistent with other experiments. The task is even simpler if we allow the gaugino fields to have different masses at the high scale.
Indeed, we have no evidence to believe that SUSY is correct. However, the only accurate statement to make at this point is that if SUSY is correct, the scalars must be very heavy, in most models. It would of course be useless to keep studying SUSY if additionally we need the spinors to be very massive also. This is not the case yet, and certainly hasn’t been so due to Tevatron data.
I am not sure exactly what statement you are referring to by Nima, but it’s possible that he is referring to a very specific model or he is considering other qualities such as naturalness.
I would like to add that the main issue that I see with the minimal forms of local SUSY is that there is a tension between the heavy scalars implied by both the apparent Higgs mass and indirect limits from B-physics, and the need for larger corrections to g_\mu.
SA,
See for instance his talk at Strings 2005 (link and discussion here:
http://www.math.columbia.edu/~woit/wordpress/?p=3492
It seems to me that the definition of “minimal SUSY” you are using is a new one, which actually means “obscure corner of the large SUSY parameter space that is the hardest for the experiments to rule out, so they haven’t got there yet”.
Peter,
No. Look at the mSUGRA model. Suppose that the Higgs is 125 GeV. Then, you are in the region to which you are referring. Coincidentally, the other constraints are satisfied.
SA,
I see. But I guess according to Arkani-Hamed, these would just be models too fine-tuned for SUSY to explain what is stabilizing the electroweak scale, so the answer is that SUSY doesn’t do that.
It’s unclear to me then what SUSY actually does for you other than make your model much much more complicated than the SM. I suppose if the LHC sees no hint of SUSY, these models will still live, but there won’t be much motivation for them.
Yes, exactly!