String Theory Fails Another Test, the “Supertest”

Wednesday’s CMS result finding no black holes in early LHC data has led to internet headlines such as String Theory Fails First Major Experimental Test (for what this really means, see here). At a talk today at CERN, yet another impressive new CMS result was announced, this one causing even more trouble for string theory (if you believe in purported LHC tests of string theory, that is…).

Back in 1997, Physics Today published an article by Gordon Kane with the title String Theory is Testable, Even Supertestable. It included as Figure 2 a detailed spectrum which was supposed to show the sort of thing that string theory predicts. Tevatron results have already caused trouble for many of these mass predictions. For example, gluinos are supposed to have a mass of 250 GeV, but the PDG lists a lower bound (under various assumptions) of 308 GeV. At CERN today, the CMS talk in the end-of-year LHC jamboree has a slide labeled “First SUSY Result at the LHC!”, showing dramatically larger exclusion ranges for possible squark and gluino masses. Over much of the relevant range, gluino masses are now excluded all the way up to 650 GeV. It looks like string theory has failed the “supertest”.

If you believe that string theory “predicts” low-energy supersymmetry, this is a serious failure. Completely independently of string theory, it’s a discouraging result for low-energy supersymmetry in general. The LHC has just dashed hopes that, at least for strongly-interacting particles, supersymmetry would show up just beyond the energy range accessible at the Tevatron.

This entry was posted in Experimental HEP News. Bookmark the permalink.

35 Responses to String Theory Fails Another Test, the “Supertest”

  1. Glenn says:

    So, what’s the ‘motive’ towards “Supersymmetry” anyway? I see why people wanted to unify forces (electric force vs. weak force vs. strong force vs. gravity) but I’m not sure why having bosons vs. fermions is such a bad thing?

    And where are we in the strong plus electroweak unification theories anyway? The last I heard, combining them lead to predictions of proton decay that haven’t panned out.

    I always thought String Theory (TM) was supposed to resolve all this stuff.

  2. Peter Woit says:

    Glenn,

    Supersymmetry is a long and complicated story, I wrote a whole book to address questions like yours. Not much has changed since the book

    To all: please try and restrict comments to the topic of the post. While a general purpose place for people to ask questions and discuss particle physics would be a desirable thing, this isn’t it.

  3. Eric says:

    Hi Glenn,

    The main phenomenological motivation for supersymmetry is that it solves the hierarchy problem, which means that it prevents the Higgs mass from receiving large quantum corrections that would push the electroweak scale up to the Planck scale. Also, it improves the running of the gauge couplings to that they seem to unify at high energy. Finally, it includes a natural dark matter candidate.

    Peter: Looking at the supersymmetry particle spectra from some well-motivated string models, it appears that the viable parameter space prior to any results from LHC always has the gluino in the 1-2 TeV range. So, I don’t think that the current constraints represent an actual test that string theory has failed.

  4. Peter Woit says:

    Eric,

    The favored gluino mass in string theory models seems to have moved up from Kane’s 250 GeV back in 1997….

    I don’t doubt that string theory models can be found with whatever gluino mass one wants. Today’s news is presumably just the start of a long story during which superpartner mass limits will keep moving up, with favored string theory models keeping pace.

  5. Glenn says:

    Peter,

    Thanks, I own the book, read it, and even followed along pretty well for most of it, I think.

    I need to put it on the list to read again, along with “The Trouble With Physics.”

  6. Giotis says:

    Well 1997 was a long time ago…

    Currently and as far as I know String theory from a top down approach does not favour low energy SUSY in the sense that it is not expected these kind of vacua to be a majority in the landscape in comparison to vacua with other (high) breaking scales. I could even say that the opposite is true although this is debatable.

    On the other hand low energy SUSY as it is well known is motivated mainly by theoretical considerations in order to solve problems like unification of couplings and the hierarchy. If it is excluded in LHC then one might naturally ask why we need SUSY after all if it doesn’t solve the problems of the SM we want to solve. So SUSY as the most prominent candidate theory for physics beyond the SM will be strongly questioned. This will have at least a physiological impact to Supersymmetric theories in general like String theory.

  7. Peter Woit says:

    Giotis,

    I see. If SUSY shows up at the LHC, string theorists will claim vindication. If it doesn’t show up at the LHC, they’ll just say it shows that the string theory landscape is right, that they knew very well SUSY wouldn’t be at LHC scales since most of the vacua are at high supersymmetry-breaking scales.

  8. Mark says:

    Peter, if anything, this exclusion plot should only make Gordon Kane very happy because it excludes a region with light sfermions.

  9. Giotis says:

    Peter I understand but here is the point:

    Let’s say that we want to find a solution to the hierarchy problem and we have two ways for solving it, fine tuning (like the fine tuning of CC in the Bousso Polchinski model) and Low energy SUSY. Now in String theory there are models which can produce both mechanisms. Fine tuning may seem quite unlikely at first but given the plethora of vacua that can realize it this may well be the prediction of String theory after all and not low energy SUSY. This would be the case if we could estimate that the vacua which we obtain from these models are much more frequent and thus statistically favoured over the low energy SUSY vacua.

    This is an over simplified picture of course and many other factors must be considered but it is the general idea. The fact that String theory is a Supersymmetric theory doesn’t imply that String theory favours SUSY at LHC. You must first check the models and the statistics of the corresponding vacua. Of course if SUSY is found at LHC this would be very important for String theory but only as a theoretical framework regardless of any ‘predictions’.

  10. anon says:

    Since landscape statistics is generally pulled out the nether regions of the body, the only way to judge any statement of the form ‘most vacua are X’ or ‘X is more likely than Y in the landscape’ is to see WHOSE nether regions it is pulled out from. A professor at Stanford — well, maybe I’ll give it 2 minutes of thought. A grad student — not so much.

    Given this is the only available criterion, I find it hard to judge statements made on a blog about what’s common on the landscape.. I mean, how respectable is your nether region?

  11. anon says:

    as a string theorist (on the more mathematical end) and as a human being (who knows the general structure of issues being debated here), it drives me up the wall that clearly intelligent and thoughtful people like ‘giotis’ go through all these ridiculous constructions in clear denial of what is staring them in the face.

    the world is so full of interesting ideas. science even more so, physics most so. Take a short walk and take the time to understand what a good non-high-energy physicist or scientist is thinking about — you’ll be amazed at (and reminded of) how interesting ideas are formed and work out.

    No twisting and turning and playing dodgeball for 20 years required.. You think up an interesting idea, try it, throw it out if it doesn’t work and go on to something else.

    It’ll really seem like a breath of fresh air, compared to the landscape cr*p people here at talking about.

  12. Mitchell Porter says:

    anon @ 9.24pm, do you consider string theory itself a dead end, or just landscape statistics?

  13. Verified Armonyous says:

    anon @ 9.24pm, we would be happy to test our theories in our tabletop colliders on Fridays, but even a string theorist surely knows it’s a bit more complicated than that. That’s unfortunate but it’s the only way when you’re asking the most fundamental questions.

    Glenn, Supersymmetry is the coolest thing you could expect to find at the TeV energy frontier. It is revolutionary (you heard about fermionic dimensions?), beautiful (people will wear the SUSY algebra in T-shirts :)), perturbatively well behaved (you can extend it to hugely large energy scales), solves neatly some SM problems, gives naturally a dark matter candidate, points to grand unification, passes switfly precision tests, etc.

  14. anon 924 says:

    landscape statistics and related attempts to connect strings to particle physics are bad mathematics and bad physics. i think there are so many interesting things to think about and do in string theory and hep-th in general. I know many around me who do this.

    On the other hand, it’s seems stupid that some people need to twist themselves into pretzels and wave their hands and waste their youth writing the most non-rigorous worthless papers, trying to justify this as physics testable at the LHC, just so funding levels remain higher than it would if this were classified as mathematics or mathematical physics.

    To all you younger landscape / string pheno people out there — do you really think your work will 1) be proven true? 2) be proven false? 3) neither but contributes a new physical IDEA? 4) none of the above?
    My opinion is that (4) holds for all of the landscape. It’s like biting into an air-filled pastry.

    Oh, please stop with the “can’t test grand theories so easily”. My complaint is that pheno end of string theory is quackery with no rigor, not misunderstood genius that has come way before its time.

  15. Mark says:

    Dear anon 924,
    String pheno is a huge field that has been around for a few decades before this landscape statistics bs came about. There are many far more important, in my opinion, and interesting problems in string pheno that people are working on. If you check out the talks at this year’s KITP program on string pheno you’ll see that landscape statistics is a rather marginal topic. On the other hand, understanding the mechanism of SUSY breaking is a very important problem and will remain one of the dominant topics in het in the years to come, especially if superpartners are discovered at the LHC. After years of research it is pretty clear that moduli stabilization and SUSY breaking are extremely closely related and one can already make some very specific statements about sparticle spectra based on very few known realistic scenarios.

  16. MathPhys says:

    “Fine tuning may seem quite unlikely at first but given the plethora of vacua that can realize it this may well be the prediction of String theory after all and not low energy SUSY. This would be the case if we could estimate that the vacua which we obtain from these models are much more frequent and thus statistically favoured over the low energy SUSY vacua.”

    The above statement goes against the grain of every intuition that we have gained since Pierre-Simon Laplace wrote his book on probability theory.

  17. anon 924 says:

    mark, i admit that string pheno is larger than landscape ramblings – but string pheno is often either rigorous and irrelevant to the LHC or very ad hoc and isn’t really ‘stringy’ anyway.

    For example, to follow your example, i think there is VERY VERY little string theory has to add to the story of SUSY breaking.. besides in gravity mediation perhaps. i don’t think any serious honest person is working on stringy models of SUSY breaking.. even people with stringy knowledge and skills like Seiberg.

    You can claim susy breaking is related to modulii stabilization in string theory but that’s like saying issue X is related to issue X + Y. True but I can ignore Y and work on X alone just fine.. modulii stabilization might have consequences for susy breaking but the latter can exist without the former.

  18. chris says:

    Mark,

    i respect your opinion, but i can’t help noticing the irony in a statement like

    “understanding the mechanism of SUSY breaking is a very important problem and will remain one of the dominant topics in het in the years to come, especially if superpartners are discovered at the LHC”

    posted in a thread on the first huge exclusion sweep of the LHC.

    i am asking this in honesty and not tongue in cheek (as i really would like to know): is your scientific judgement and gut feeling still pointing towards low energy SUSY?

    in the 70s and 80s SUSY certainly had its appeal. we knew much less then about the TeV scale and even electroweak symmetry breaking. but looking at what has happened experimentally in between makes SUSY look like a dinosaur to me. W and Zs were found and later the top, the Higgs limit was pushed to 115GeV, the cc was found and anisotropies in the CMB. our view of the world has considerably changed and expanded since the invention of susy but still no trace of remnants of this elusive symmetry. are you not bothered by this or, in other words, was it to be expected that susy is so hard to find?

  19. Mark says:

    Dear anon 924,
    The assumption that “modulii stabilization might have consequences for susy breaking but the latter can exist without the former.” is certainly not ruled out but appears to be very unlikely for very good reasons. Let us be conservative and assume that a given compactification is described by N=1 D=4 sugra. Then, decoupling moduli stabilization from SUSY breaking would require 1) Stabilizing all moduli by the superpotential only without relying on the Kahler potential. 2) Having an additional dial in the superpotential, essentially a constant, and using it to fine tune the total superpotential very close to zero so that the the resulting vacuum is very nearly SUSY Minkowski. 3) Breaking SUSY on top of that and decoupling the s-goldstino (aka a modulus) from the gravitino while maintaining a tiny CC (this is needed in gauge mediation to avoid a very light modulus) requires a rather contrived choice of a Kahler potential for the SUSY breaking field. While 1) is certainly possible, 2) is extremely non-generic because of the amount of fine tuning required to achieve SUSY Minkowski vacua. Furthermore, 1) + 2) would certainly kill any attempt to solve the strong CP problem via a QCD axion since all the axions would be rendered too heavy, as Joe Conlon has rigorously proved in the context of N=1 D=4. 4) Simply embedding a viable gauge mediation scenario into string theory is a rather daunting task in itself.
    Therefore, unless you are so heavily invested in gauge mediation that you are willing to give up on naturalness, it is pretty clear from the top down perspective that gravity mediation is much more natural and that the problem of moduli stabilization is intimately related to the question of SUSY breaking.

  20. Mark says:

    “i am asking this in honesty and not tongue in cheek (as i really would like to know): is your scientific judgement and gut feeling still pointing towards low energy SUSY?”
    Yes, there are at least four very compelling reasons to consider TeV scale superpartners. Stabilizing the gauge hierarchy, precision gauge coupling unification, radiative electroweak symmetry breaking, LSP as a dark matter candidate. The exclusion plot discussed in the posting considers a small portion in the parameter space of a very simplified, almost adhoc, set of GUT scale boundary conditions called msugra. Let us wait until the LHC collects at least 10/fb before we jump to premature conclusions.

    “was it to be expected that susy is so hard to find?” In retrospect, if the SSC had not been cancelled, SUSY could have already been found a decade ago.

  21. piscator says:

    anon924,

    If you have thought about the subject seriously, I find it hard to see how you can say string theory has little to add in susy breaking except perhaps in gravity mediation – so little except in the (arguably) best motivated type of susy breaking, where all the operators are coming from the Planck scale.

    In any case every serious person knows (and has known for years) that in any context with a UV string embedding an honest discussion of susy breaking requires addressing moduli stabilisation. You’re right it’s possible to work on susy breaking in global field theory without thinking about moduli stabilisation. It’s even possible to have a successful career doing so. There are lots of interesting results thereby obtained for susy breaking in worlds that don’t have gravity. None of this means this has any real physics value to susy breaking in this world. It’s like Feynman’s example of the flow of dry water, it may be interesting qft but its missing a crucial part.

  22. MathPhys says:

    Dear Mark,

    “In retrospect, if the SSC had not been cancelled, SUSY could have already been found a decade ago.”

    It’s statements like this that give string theory and theorists a bad name.

  23. Eric says:

    Dear MathPhys,

    Can you explain exactly how Mark’s factual statement possibly gives string theory and theorists a bad name? If TeV-scale SUSY does exist, it would most certainly have been found at the SSC had it been constructed. Since it was cancelled, theorists have more or less had to wait for the LHC, although there was always an outside chance that it could have been seen at LEP or the Tevatron.

  24. Thomas Larsson says:

    In retrospect, if the SSC had not been cancelled, low-energy SUSY could have already been ruled out a decade ago.

  25. MathPhys says:

    Dear Eric,

    To me, Mark’s statement implied that

    “There is supersymmetry, there is no doubt about it, but we were denied the chance to discover it”.

    It’s this messianic approach to science that gives the subject a bad name. A better statement would have included Larsson’s comment above

    “In retrospect, if the SSC had not been cancelled, low energy SUSY could have either been found or ruled out a decade ago.”

  26. Chris Oakley says:

    Not entirely. There is Occam’s razor. If I posit that there is a real Father Christmas who charges around the sky on a sleigh pulled by reindeer, climbing down chimneys to leave presents – but only for children who have been good, then the onus is on me to prove it – not on you to DISprove it. Similarly, all experimental evidence so far leads one to the conclusion that there is no supersymmetry – its only value is as a band aid to patch up ailing theories that probably should have been binned instead. So unless someone can come up with something more compelling the default position should be that it does not exist.

  27. MathPhys says:

    On the one hand, supersymmetry is by and large the reason why quantum physics (starting with Witten’s work on Morse theory in the early 80′s to mirror symmetry) has been spectacularly successful in pure mathematics over the past 30 years.

    On the other hand, after 35 years now, there is indeed not a shred of evidence of supersymmetry in the real world. It’s a very peculiar situation.

  28. chris says:

    Hi Mark,

    thanks for your honest answer. I see the standard arguments, but there is always this nagging feeling.

    Let’s take the unificaton of couplings. As far as I know, this is essentially still a 1-loop argument. But granted that it is true in the full, nonperturbative context, what does it tell us?
    I am thinking in comparison of the Weinberg angle. It is really close to the predicted SU(5) GUT value. So close in fact that Georgi and Glashow in their original paper were about as sure as one can be that this is the true symmetry group of nature. Alas, it has been firmly ruled out by now.

    So what does it tell me that 3 lines do intersect in one point instead of a larger region? unification looks not too bad if only very little is allowed to happen between the TeV and the GUT scale. I really find it very difficult to interpret anything furthter into it.

  29. Eric says:

    MathPhys,

    I think you are misinterpreting Mark’s statement. Saying that SUSY could have been discovered already if SSC had not been canceled is quite different than saying that it would have or should have, which is what you seem to think he said. I think it’s fair to say that whatever new physics exists at the TeV-scale would have been discovered by now if the SSC had been built, SUSY or otherwise.

  30. MathPhys says:

    Dear Eric,

    Yes, I agree with you. I took his statement more along the lines of “would have been discovered”. I also wish he would have added “or otherwise” to make his statement more balanced.

  31. Mark says:

    Dear Chris,
    “So what does it tell me that 3 lines do intersect in one point instead of a larger region?”
    As you know, the unification idea, whether it is SU(5) or SO(10), is well motivated by the fact that the matter multiplets of the SM nicely fit into into complete GUT multiplets. This is a higly non-trivial clue and should be considered very seriously. However, for this idea to work: 1) the three gauge coupling must unify at some scale; 2) the unification scale must be high enough to suppress the proton decay. As we know now, if the SM is all there is between the EW scale and the GUT scale, this does not work. The proton decays too fast while the couplings do not actually unify.

    Keep in mind that the MSSM was introduced to address the gauge hierarchy problem, first and foremost, and the other three properties I mentioned in my previous posting followed *automatically*.

    In fact, to get the couplings to unify at 1-loop one does not need the full MSSM! Including light Higgsinos only can do the job to satisfy requirement 1). However, in this case the unification scale will be around 10^14 GeV – the same as the original non-susy GUT. This is deadly because of the proton decay. Including the MSSM gauginos changes the relative running speed and raises the scale at which the couplings unify, which is nice because the proton lifetime becomes much longer, hence satisfying requirement 2). On the other hand, at 1-loop, the MSSM matter multiplets play no essential role (they only change the slopes and thus the value of the unified coupling) as they come in complete GUT multiplets. So, if you only want to use SUSY for the purpose of preserving the gauge coupling unification you can completely decouple all the squarks and sleptons! This idea is called split SUSY and completely gives up on addressing the hierarchy problem but some people are taking this possibility seriously. A long-lived gluino is a robust prediction of such a model and it is actually being searched for at the LHC. It is perfectly possible to add a few heavy complete vector-like GUT multiplets to the largangian and not spoil the perturbative unification. Now, when you start to include 2-loop effects life becomes more interesting. In this case the running becomes slightly more sensitive and depending on the details of, say the gaugino spectrum at the unification scale, the couplings can unify even more precisely at 2-loops.

  32. chris says:

    Dear Mark,

    thank you very much for your detailed response. As you explained so nicely it is indeed a very nice feature of the MSSM to address several problems of unification at once. although I do not see the gauge hierarchy problem as solved by the MSSM but rather shifted to another sector, it is indeed intriguing circumstantial evidence that this solution also results in a dark matter candidate and unification of couplings.

    but my problem really is with the circumstantial nature of this evidence and i would like to once again pick the unification of couplings as the example. since i first saw this plot, i always wondered about one thing: unification seems to happen at almost the planck scale (logarithmically). it is nice that susy at 1-loop brings the 3 couplings together, but to be completely compelling i would have somehow expected this intersection of 3 lines to – so to say – intersect with a 4th line, the naively computed fundamental scale (plank scale). you see, intersecting 3 lines is not such a big deal. as you so nicely pointed out yourself, you need just one new fermion with the appropriate couplings. that’s it. and the advantage of intersecting higher that 10^14 TeV is also a two-edged sword: yes, one can portray it as a success given the experimental knowledge we have, but in fact the success consists of nothing more than evading another experiment. from an unprejudiced, purely theoretical point of view unification at a lower scale would be preferred a priori because of its increased testability or , put differently, because it decreases the extent of the region one has to proclaim as being a dessert. that is, if this scale is not found to be the planck scale or given by another, independent experimental hint.

    to me it seems that susy does get some things right but not quite all of them. and i am sceptical as to whether the features it does get right are not just some rather universal ones.

  33. Mark says:

    Dear Chris,
    Once the three gauge couplings unify at ~10^16 GeV, they keep running as a single coupling, i.e. one line. This is because all the GUT scale extra states complete the formerly incomplete GUT multiplets. So the higgses and higgsinos instead of doublets become 5-plets and \bar 5-plets, the gauge bosons become 24-plets etc. So after the three couplings unify they no longer diverge and run as one. However, above the GUT scale the theory may no longer be effectively 4-dimensional so one needs to take into account the contributions from the KK modes, etc. This will change the shape of the running. The unification of the SU(5) gauge coupling and the gravitational coupling will eventually take place at a scale a few orders of magnitude above the GUT scale.

  34. Mark says:

    Dear Chris,

    “from an unprejudiced, purely theoretical point of view unification at a lower scale would be preferred a priori because of its increased testability or , put differently, because it decreases the extent of the region one has to proclaim as being a dessert.”

    I would qualify this point of view as motivated by experimental testability accessible with the current technology rather than a purely theoretical. I’m fairly certain that testing the proton decay lifetime in the context of SUSY GUTs will eventually be possible as new expreiments will come online.

    “although I do not see the gauge hierarchy problem as solved by the MSSM but rather shifted to another sector, it is indeed intriguing circumstantial evidence that this solution also results in a dark matter candidate and unification of couplings.”

    Indeed, SUSY does not “solve” the gauge hierarchy problem, rather it “stabilizes” the electroweak scale while it does not really explain its origin. In order to *really* explain the origin of the electroweak scale once needs to have a model of SUSY breaking where the only dimensionful input is the Planck scale. Back in the early 80s Ed Witten suggested an excellent idea where some strong gauge dynamics in the hidden sector generates a small scale of SUSY breaking in the visible sector via dimensional transmutation. This is referred to as dynamical SUSY breaking. I personally think that this is the best idea we have and I don’t agree with the landscapeologists that the EW scale is there because some fluxes are extremely fine-tuned.

    By the way, as I said in my previous posting, in addition to the precision unification and the LSP as a dark matter candidate, the MSSM also naturally explains why the electroweak symmetry is broken. Recall that in the SM the higgs potential has a negative mass squared term in order to generate a minimum where the higgs vev is non-zero. In the SM this tachyonic potential is put in by hand and has no dynamical explanation. In the MSSM, on the other hand, the negative mass squared term has a dynamical explanation. It is generated via radiative corrections, mainly due to the large value of the top Yukawa coupling. So, in the context of the MSSM, one can claim that at least one of the SM quarks must be very heavy in order for the EW symmetry to be broken. Again, this rather non trivial result was not the reason why the MSSM was introduced but comes automatically as a bonus.

  35. Pingback: LHC levert nog geen bewijsmateriaal voor snaartheorie