Witten Colloquium

I was down in Princeton today and went to hear Witten’s physics department colloquium on the topic of “Supersymmetry: Pro or Con”. He spent most of the hour going over the 25 year-old hierarchy argument for supersymmetry (that supersymmetry provides a reason for the Higgs to be much lighter than the Planck scale, since it is paired with a fermion whose mass can be protected by an approximate chiral symmetry).

He gave the following arguments for believing in GUTs:

1. Can naturally get small neutrino masses via the see-saw mechanism.
2. Coupling constant unification to 1%
3. Tentative evidence from CMB that fluctuations come from GUT scale.

Actually none of these seem to me very convincing (and to claim 1% coupling constant unification I think he has to use 1-loop results, at 2-loops it is more like 5-10% off, but this may depend on exactly what you are comparing to what).

His points in favor of supersymmetry were:

1. Solves hierarchy problem.
2. Coupling constant unification again.
3. Prediction of top mass from supersymmetric SO(10) GUT.
4. Supersymmetry is consistent with all accelerator data.
5. Lowest mass superpartner a good candidate for dark matter.
6. Part of string theory.

Again none of these are really convincing. If you don’t believe in GUTs, the GUT scale is irrelevant, and since we don’t understand quantum gravity, the significance of the Planck scale is also unclear. I’m no expert on supersymmetric GUT “predictions”, but they seem to depend on lots of choices for the details of the GUT, how its symmetry breaks, and how fermions get masses from the symmetry breaking. Saying that supersymmetry is consistent with all accelerator data is kind of strange since the standard model without supersymmetry is consistent with all accelerator data and there is no evidence for supersymmetry. You can guess what I think of his last argument.

His points against supersymmetry were:

1. The Higgs mass bound is already embarassingly high, need some fine-tuning to get a Higgs that massive in a supersymmetric theory.
2. Supersymmetry spoils many of the experimental successes of the standard model since it generically has experimentally disallowed amounts of violation of CP, baryon and lepton number conservation, flavor-changing neutral currents.
3. No good picture of how to break supersymmetry.

Well, for me the con has it over the pro, but Witten still seems to hold out hope that supersymmetry will be found at the LHC. At the end of his talk, he discussed what he called the “worst case scenario”; that LHC sees a Higgs particle, but nothing else: no supersymmetry, no technicolor, no Little Higgs, no extra dimensions. He said that if this happens people will look for anthropic explanations of the hierarchy problem, whereas if the LHC found something that explained the hierarchy problem, they might be encouraged to look again for non-anthropic answers to the cosmological constant problem (which he claimed was analogous to the hierarchy problem). He did say “I hope it is wrong” about the anthropic explanation of the cosmological constant.

On the anthropic front, Michael Dine is claiming that maybe the statistical analysis of the landscape will “predict” that supersymmetry breaking is at a low energy scale. The arguments he gives sound to me like a complete joke, and from what I remember Michael Douglas was recently claiming that the same kind of analysis indicated that supersymmetry was broken at a high energy scale. One other funny thing about Dine: he doesn’t say that the landscape makes predictions, but that it is “the first predictive framework we have encountered”. This is a guy who for nearly twenty years has been giving talks on “superstring phenomenology” and claiming that any day now string theory would make predictions. I wonder why in all of those previous talks he neglected to mention that not only were there no predictions from string theory, there wasn’t even a “predictive framework”.

This entry was posted in Uncategorized. Bookmark the permalink.

27 Responses to Witten Colloquium

  1. Chris W. says:

    A new paper by Sean Carroll and Jennifer Chen deals with some of the same issues, and cites both DKS and Albrecht and Sorbo:

    Spontaneous Inflation and the Origin of the Arrow of Time (hep-th/0410270)

  2. Chris W. says:

    The authors of hep-th/0410213 acknowledge helpful input from Andreas Albrecht, who posted the following interesting preprint in May:

    Can the universe afford inflation?

    Cosmic inflation is envisioned as the “most likely” start for the observed universe. To give substance to this claim, a framework is needed in which inflation can compete with other scenarios and the relative likelihood of all scenarios can be quantified. The most concrete scheme to date for performing such a comparison [due to Dyson, Kleban, and Susskind (DKS)] shows inflation to be strongly disfavored. We analyze the source of this failure for inflation and present an alternative calculation, based on more traditional semiclassical methods, that results in inflation being exponentially favored. We argue that reconciling the two contrasting approaches presents interesting fundamental challenges, and is likely to have a major impact on ideas about the early universe.

  3. Arun says:


    It has been shown that a SM electroweak fit including the anomalous magnetic moment of the muon a and
    the branching ratio Br(b ! Xs
    ) yields a probability of about 5%. The total 2 is improved in the MSSM,
    mainly because of a, but the probability does increase only marginally due to the larger number of free
    parameters in the MSSM. However, in both cases the 3  discrepancy in sin2 W from Ab
    FB and ALR is
    contributing to the low probability. Since at present no arguments to doubt any of the measurements can
    be found, we tested the Particle Data Groups procedure to rescale the errors of these two measurements by
    the corresponding pulls. This yields considerably improved 2 values, both in the SM and MSSM, without
    significantly changing the fitted parameters.

  4. Chris W. says:

    Serenus: I should have remembered your comment and checked the URLs in it before I posted. You beat me to the punch on hep-th/0410213.

  5. Thomas Larsson says:

    It is strange that nobody seems to know about the 2-loop calculation. Maybe it is just a rumor. Maybe Witten didn’t mention it because it isn’t true, or because it is true but he doesn’t know about it.

    The relative error depends on how you parametrize things, but the comparison between the 1- and 2-loop results does not. If x = 1 + e and y = x^7, then y = 1 + 7e, so the error in y is 7 times bigger than the error in x. However, that the 2-loop error is ten times as big as the 1-loop error (if that is true) does not depend on parametrization, since (7e’)/(7e) = e’/e.

  6. Thomas Larsson says:


    That the two-loop calculation was off by 10% is something that Aaron Bergman claimed in this forum – at least I, and presumably everybody else, interpreted him in that way. Maybe you should check with him.

  7. Chris W. says:

    See this:
    Birth of the Universe from the Landscape of String Theory
    (4 pages, posted 10/20/2004):

    “We show that a unique, most probable and stable solution for the wavefunction of the universe, with a very small cosmological constant …, can be predicted from the supersymmetric minisuperspace with N vacua, of the landscape of string theory without referring to the anthropic principle. ….”

    ..from the abstract, with correction of typos. (The authors are at UNC-Chapel Hill; Laura Mersini-Houghton is an assistant professor there.)

    String theory as such has little to do with their argument, other than providing a motivation for consideration of the landscape. The authors consider a SUSY minisuperspace, which leads to a lattice-like solution, and motivates them to employ a condensed matter analogy. It’s only 4 pages long, and well worth reading.

    Another recent analysis based on a condensed matter analogy has received some attention:

    Cosmological constant and vacuum energy
    (G.E. Volovik)  gr-qc/0405012

  8. Fred says:

    I’d point out that neutrino masses are a complete mystery if you don’t invoke a GUT theory.

    Thats the one area the standard model notably fails experimental bounds, and why people should be interested in additions.

    As far as SUSY goes, the coupling constant unification is a decent argument, but probably not the most convincing to me, particularly b/c its still a perturbative mess. However in GUT phenomenology there are various naturalness arguments in the Higgs sector, that favor (in simplicity) some sort of minimal SUSY model. Again the nastiness comes in the representation you choose, and I think its still pretty much an open question in physics phenomenology.

  9. Lubo Motl says:

    That’s very funny. Good joke. As far as the available data go, the previous posting could have been written by me, just to make fun out of myself – because everyone who is serious and observant enough knows that the arguments are usually on my side. 😉

  10. Anonymous says:

    “I think you’re right. Among thousands of exchanges I’ve had in my life, all of them had some facts and rational arguments.”

    Well, that’s a good start! Now, the next thing for you to do is to have some of those facts and rational arguments on *your* side.

  11. Lubo Motl says:

    Dear anonymous,

    I think you’re right. Among thousands of exchanges I’ve had in my life, all of them had some facts and rational arguments.

    Once again: some more quantitative data on SUSY and gauge coupling unification are on motl.blogspot.com

    All the best

  12. Anonymous says:

    In yet another piece of inadvertent humor from Lubos M, we have him writing about global warming on his blog:
    “The actual reply from SHS has surprised me, shocked me, terrified me. It was a rather long e-mail, but it contained nothing else than personal insults against the four “sceptics” listed above”
    Yes, it’s a terrible thing when serious scientists are reduced to personal insults. Lucky thing that LM would never stoop so low, isn’t it?

  13. I am wondering why so few consider the possibility that the dead alley in the particle theory might derive from wrong assumptions, which were made already three decades ago and became soon sacred by the illusory belief on the linear progress of big science.

    What created strong visceral reactions in me at that time was that quarks and leptons of single generation were forced into single multiplet although masses are widely different. The same was done even for different fermion generations having completely different mass scales. This
    is certainly the least imaginative approach to unification that one can personally imagine. Therefore I believed that the observation that proton did not decay had been taken as an important message from the elegance loving Nature.

    No one however cared although the message is easy to decipher: if quark and lepton numbers are separately conserved, they could do it as chiralities of 4+N dimensional spinors. Color would not be spin like quantum number but more like a partial wave in compactified dimensions. Completely different strategy to unification
    would open up.

    Anyone interested on details of this option can do this at
    http://www.physics.helsinki.fi/~matpitka/ . See also
    http://www.physics.helsinki.fi/~matpitka/newtgd.html for the newest developments in the understanding of particle spectrum and mass calculations.

    Matti Pitkanen

  14. sol says:


    Only I can be considered in that vain:).

    Truth is, I didn’t know that Lubos was to refer to Nima. I have been following this line of thinking for a bit now(google sci pysics strings early days:). Witten’s reference, keeps me looking along these lines and the latest work.

    Short Range Test’s of Inverse Square Law

    If your going to rebuttal, please do so on the reasons why not, so I can stay informed, and so, that others can too.

    The poetic inflections, that such comments leave from personalities(Lubos or Peter’s) are in no way my repsonsibility, or, are they your’s.:)

    The Standard Model and Beyond:)

  15. D R Lunsford says:

    It’s the sol and Lubos core reference dump!

  16. sol says:

    Little attention has been given here Peter. As a method to testability, it had to be considered. They approached this properly and of course await for confirmation. It was not considered devoid, as a philosphical approach?

    When symmetry breaks down, by Edward Witten

    Other ideas about electroweak-symmetry
    breaking go even further afield. One line
    of thought links this problem to extra
    dimensions of space-time, subnuclear in
    size, but observable at accelerators
    . This
    approach is probably a long shot, but the
    pay-off would be huge discovering extra
    dimensions could give us the chance for
    direct experimental tests of the quantum
    nature of gravity and black holes


  17. Lubos Motl says:

    Some more quantitative comments about the gauge coupling unification


  18. sol says:

    Sometimes a nice map helps out for reference.

    If all these events were taking place, lets say from a true vacuum perspective, would this imply that the false vacuum existed in the first place?

  19. Lubos Motl says:

    I recommend the paper


    and its references [1-4]. It makes it pretty clear the the unification works great if you compare the assumed values at the GUT scale, and the known value at the Z scale.

    On the other hand, the alpha3 coupling at low energies is 12-20 percent different, which shows some neglected running at *low* energies, below the Z scale, and it is usually assigned to a light gluino.

    It’s sort of embarassing if it’s true and Gordon Kane lives with his serious misconception. 😉

  20. Lubos Motl says:

    I think that I found the/a paper that contains a similar unusual statement.

    It is the last (or 2nd last) paper of John Hagelin before he joined Maharishi Mahesh-Yogi. 😉


    I leave it to the reader to decide whether this 10-page paper with 4 citations is trustworthy. I am not sure about the answer yet. 😉

  21. Lubos Motl says:

    So far, no one knows anything about it. I’ve asked a couple of phenomenologists.

    All of them say that they did the 1loop calculation themselves, and they believe that the full 2loop result is even better – and the nice intersecting graphs we usually draw are the full 2-loop results.

    At any rate, you are on the same boat with Gordon Kane about this ;-). On the other hand, it may be helpful if you found the scientific references containing this actual calculation – because once again, what two of you say sounds bizarre.

  22. Lubos Motl says:

    Thanks! I did not know that – and of course, I will recheck elsewhere.

    We should be careful about two-loop results. Several years ago, Brookhaven claimed some deviation for the muon anomalous magnetic moment, or something like that, and it was used as evidence for new physics etc.

    Finally it turned out to be a sign error in the theoretical 2-loop calculations.

    I would still prefer to trust the 1-loop results more than the 2-loop results because they can be quite easily verified.

    I agree that a 10% error in such a calculation is pretty large. Nevertheless, the running is better than in the pure SM, so in this case, it is still experimentally suggested that GUT prefers to get SUSY as its ally.

  23. Peter says:

    I could find a better reference, but here’s a quote from Gordon Kane (hep-ph/0202185) (I seem to have been too kind to quote 5-10%)

    “The supersymmetric gauge coupling unification misses by about 10%. More precisely, the experimental value of the strong coupling \alpha_3 is about 10-15% lower than the value computed by running down theoretically from the the point where the SU(2) and U(1) couplings meet. The details are interesting here — the one-loop result is somewhat small because of a cancellation, and the two-loop contribution therefore not negligible. If one only took into account the one-loop effect the theoretical value would be close to the experimental one, but the two-loop effect increases the separation.”

    I’ll let my comments about the arguments for supersymmetry stand. I think all the arguments are very weak, the strongest is coupling constant unification, and even there, being off by 10-15% is not impressive.

    If Witten really is quoting one-loops and ignoring two-loops, he wouldn’t be the first one, but I don’t think that’s an honest thing to do.

  24. Lubo Motl says:

    It looks like a good talk by Witten.

    I agree with Witten’s summary of the arguments for SUSY – and it even looks like the phenomenologists would agree, too.

    Peter, could you give us some references about the 5-10%-off two-loop results on unification? I’ve never heard something like that, and it seems to be your only argument against SUSY (and GUT), except for totally unjustified nervous and hostile comments.

    I understand your point of view “if I don’t believe anything, I don’t need to believe anything else”. I just hope that you don’t expect me to think that this set of arguments of yours goes too much beyond the thinking of the average monkeys. 😉

    It’s much easier to agree that the anthropic predictions of the SUSY breaking scale are vague hand-waving. But once again, being negative about everything does not solve anything, and you seem to be negative about everything.

  25. Peter says:

    I don’t think there’s been anything new about neutrino masses in quite a while. You can easily extend the standard model to incorporate neutrino masses in at least two ways:

    1. Just add right-handed neutrinos with no charge under the standard model gauge group and then, just like the other leptons, Yukawa Higgs couplings give neutrinos mass. These couplings have to be much smaller than the couplings to the other leptons, but then again, no one knows why such couplings have such a huge range of values already in the standard model.

    2. Add a right-handed neutrino with a lepton-number violating Majorana mass term. If you make this large enough you can generate small masses for the observed neutrinos through the”seesaw” mechanism.

    SO(10) GUTs naturally contain a right-handed neutrino state with no SU(3)xSU(2)xU(1) charge.

  26. dolt says:

    Peter thank you for your summary of the colloquim and for stating the obvious!

    But one definite pro for SUSY and string theory (“unique/parameter free”), which is left unstated is that they are great fields for publishing lots of papers! A definite pro when it comes to careers in academia these days and explains the herd mentality.

    BTW, what is the latest on incorporating neutrino masses?


  27. serenus zeitblom says:

    Very often a resort to anthropy is just a sign of lack of imagination. Even the landscape may not need it:
    and similarly the famed “cosmic coincidences” don’t require it either:

Comments are closed.