This Week’s Hype

This week’s contribution to the long tradition of universities issuing press releases hyping non-existent “experimental tests of string theory” by their employees is from Duke University, which advertises “String Theory in a Lab“. This is based on a paper that just appeared in Science describing measurements of the viscosity of a Fermi gas. The paper explains the relationship of the measurements to string theory as:

The measurement of the viscosity is of particular interest in the context of a recent conjecture, derived using string theory methods, which defines a perfect normal fluid.

referring to this paper which first suggested that gauge/gravity duality implied a value of 1/4π for the ratio of shear viscosity to entropy density.

In the press release, this connection to string theory has been promoted to a headline, as well as to the claim that:

The results may also allow experimental tests of string theory in the future.

which I suppose is better than the usual claim in these press releases that what is being promoted is already an experimental test of string theory. It seems likely that one reason this isn’t yet an “experimental test” is that the data comes out 4 to 5 times higher than the string theory value.

This entry was posted in This Week's Hype. Bookmark the permalink.

30 Responses to This Week’s Hype

  1. Amused says:

    Hi Peter,

    I`m just amused seeing You jump head on at every headline appearing somewhere containing the S-word, independent of what the context is …

    That`s so predictive and sure as death and taxes !

  2. Peter woit says:


    For years I’ve had a Google alert set up that lists new news stories involving the terms “string theory”. The long term pattern is that fewer and fewer of the things listed have anything to do with string theory itself. These days the majority come from music: “string theory” is very popular as the name of a musical group or event.

    Besides the music ones though, many of the rest are the all-too-familiar press releases promoting some kind of hype about a bogus “experimental test of string theory” (and the varying numbers of news stories that pick up the press release). I’m well aware that my pointing these out is all too predictable and kind of tedious by now, since I’m pretty bored with this myself. However, it still seems to me that someone should try and do something to slow down this phenomenon, and I’m not seeing anyone else interested in the job…

  3. Derek Teaney says:


    I have not read the article by Schafer et al. (I’m at home
    and its not free) but certainly the 1/4\pi result is not to be made
    light of. The 1/4\pi result does provide an “exact” example
    of a strongly coupled theory close to the quantum limit where
    the shear viscosity can be computed. Of course the 1/4\pi
    result does rely on
    the AdS/CFT and certain properties of black holes which haven’t
    been proven. Measuring the shear viscosity in fermions system
    close to unitarity does not prove the exsistence of strings or the validity of string theory. But, it does strongly suggest the inherent value in these methods as a tool of theoretical physics.

  4. Peter Woit says:


    I’m not making light of gauge-gravity duality, or its possible applications to computations in strongly-coupled quantum systems.

    I am making light of the endless less-than-honest attempts to use this to generate misleading press releases about “experimental test of string theory finally found!” Or, actually, I’m not really making light of it, I think it’s kind of disgraceful…

  5. Amused says:

    Ok Peter,

    the Google alert explains it …

    Perhaps nobody else does what You are doing because it`s really enough now and it`s the job of present and future experiments to decide what is right and what is not?

    Instead of trying to get any research on ST in particular or on all of the other approaches to quantum gravity which You seem not to be able to accept neither eliminated, it would be much more interesting to read about what You are working on and what You appreciate as good and fascinating news in physics …

    Or perhaps Your intention is to merit a special chapter in the new
    german popular science book “Kampfhähne der Wissenschaft” by
    Heinrich Zankel ;-))) ?


    Seriously, I`d really prefere reading more about fascinating in interesting things You surely have to write about than seing this neverending picking on others …

    At the present state of affairs, reading some physics text, on QFT for example, is much more promising for me to understand what is going on in modern physics.

    Best wishes

  6. New says:

    Is the press release by string theorists? It looked like it was the experiementers at Duke who said (correctly) that it might be interesting from a string theory perspective. The Duke string theory group is into much more formal string theory, like CFTs for heterotic models etc.

    In any event, I am not sure why this is hype. They say very clearly that the string theory computation is a lower bound and that the experiment is five times the bound or so. The actual claim is that “The results may also allow experimental tests of string theory in the future” which is hardly hype.

    The reason why this is interesting is that standard approaches other than string theory give vastly bigger results, while the lower bound from string theory is within an order of magnitude of the experimental value.

    What is so upsetting to Woit here is anybody’s guess.

  7. Peter Woit says:


    The point is that this has nothing at all to do with experimentally testing string theory, either now or in the future, and those promoting this kind of hype know this, whether they’re condensed matter physicists or string theorists.

    What one can’t now, but in principle might be able to test this way is the combination of some conjectured strongly coupled model for the system together with some conjectures from gauge/gravity duality about how solutions to the model behave. If such an experimental test fails it means either the model is wrong, or solutions to the model have different properties than conjectured.

    The first alternative has nothing to do with string theory, and the second one is a mathematical question. Whatever the correct answer to the mathematical question is, there’s no “experimental test of string theory” there. If experiment disagrees with the gauge/gravity duality “prediction”, I doubt anyone is going to give up on gauge/gravity duality, instead they’ll give up on the particular strongly-coupled model used, or decide that the application of gauge/gravity duality in this situation is too naive.

  8. Chris Austin says:

    Hi Derek Teaney,

    Would you be able to give the authors or title of the paper with page reference


    mentioned by luny in response to your comment on the same topic here? I can’t find it on SPIRES just from the page reference.

  9. thomas says:

    Neither the heavy ion experiments nor the experiments involving dilute Fermi gases provided quantitative test of string theory right now. The heavy ion experiments do not produce an N=4 SUSY plasma at large N_c, and the Fermi gas experiments involve a scale invariant non-relativistic fluid, but not one of the systems for which a precise correspondence has been formulated.

    The experiments nevertheless provide a path for testing the string theory framework. Ideally, this will happen when holographic dualities are extended to more realistic systems. Even if this turns out to be difficult, we can use the experiments to verify numerical calculations of transport properties (in lattice QCD or non-relativistic quantum Monte Carlo), and then check the dependence on the matter content numerically. If string theory correctly predicts the N=4 large N_c limit, and if at strong coupling the dependence on the precise matter content is indeed weak, then this would constitute a pretty non-trivial test of string theory as a quantum gravity framework, and a test of the holographic dualities.

    Of course this does not directly tell us much about attempts at deriving physics beyond the standard model from string theory, but the Duke press release is actually fairly honest (on the scale of these things) in not claiming anything like this. The press release also points out connections to the physics of neutron stars. Here, the situation is not so different — some additional theory is required to extrapolate from cold atoms to cold neutrons.

    Finally, as Derek points out, nobody can seriously doubt that string theory has been extraordinarily fruitful in thinking about strongly coupled fluids.

  10. Peter Woit says:


    You’re just exploiting ambiguities in language when you write about “testing the string theory framework”. Most people when they read that would assume that you are talking about the conjectured use of strings as elements of a fundamental theory, which is the way string theory has been advertised. You’re referring to something completely different, a string-theory-inspired mathematical conjecture that may or may not be right, and may or may not some day have a use as an approximate calculational method.

    People have noticed that after 25 years the heavily promoted “string theory” predicts nothing and instead of admitting the reason for this failure, there’s an active campaign to avoid acknowledging this by misleading people with bogus claims of new “experimental tests of string theory”. I don’t think trying to justify these kinds of press releases is a good idea.

  11. Chris Austin says:

    With reference to my comment above, perhaps I should have pointed out that it would be interesting to see the arguments in the paper mentioned by luny, which luny states finds a limit that is surprisingly similar to 1/(4 pi), using back-of-the-envelope arguments stemming from the uncertainty principle, because Kovtun, Son, and Starinets find on page 5 that the uncertainty principle leads to a lower bound of about 1, rather than about 1/(4 pi), on the ratio of viscosity to entropy density.

  12. Derek Teaney says:

    Hi Chris,

    The paper is a classic of our field

    P.~Danielewicz and M.~Gyulassy,
    %“Dissipative Phenomena In Quark Gluon Plasmas,”
    Phys.\ Rev.\ D {\bf 31}, 53 (1985).
    %%CITATION = PHRVA,D31,53;%%

    The point is though, that its based on extrapolation from weakly coupled theories to a regime where they are not valid, rather than
    a valid scheme.

  13. Giotis says:

    Peter I’m curious; do you have any theoretical reasons to believe that String theory is wrong as a fundamental theory or the fact that it can’t be currently tested is an adequate reason for you to dismiss it altogether?

  14. Peter woit says:


    I wrote a whole book about this…

    In brief, the mathematical structure of the standard model is quite marvelous and uses very deep mathematics. My guess is that going beyond the Standard Model will involve going deeper in those mathematical directions.

    I never was a fan of the idea of throwing most of this out, with the idea that completely different objects (strings and extra dimensions) were what was really fundamental. When people first tried to get a unified theory from strings it was a reasonable (if not especially compelling) thing to try. It soon became clear though that this didn’t work. The lack of any predictions is just a symptom though of the underlying disease, which is that you have to choose more and more complicated backgrounds just to avoid contradiction with experiment. You never get more out than you put in, which is the signature of a bad idea.

  15. critic says:

    Peter wrote: For years I’ve had a Google alert set up that lists new news stories involving the terms “string theory”.

    That’s quite an occupation you have invented for yourself. Is this the world’s newest profession?! If this is your occupation, of course you will always criticize these press releases, no matter what they say. Who cares anyway: you don’t have any working knowledge of string theory. It is easy to dismiss things you have no clue about!

    Obviously, what you are doing is not stopping all the string theory chatter whether it makes sense or not. Quite the opposite, it brings it to the attention of a wider audience. And anyway, complaining is `un-American’ and rarely brings intended results.

    You seem like a smart guy: find something better to do with your blog!

  16. Peter Woit says:

    Hi “critic” from Princeton, New Jersey,

    You don’t have any substantive argument, but you seem to think an anonymous personal attack will help your cause. Thanks for the encouragement to keep devoting a small fraction of my time to this activity. If it’s causing behavior like yours, it must be having an effect…

  17. Peter Woit says:

    Many commenters claim to be annoyed that I’m spending too much of my time on this blog complaining about string theory. This seemed odd to me, since I couldn’t remember thinking or writing at all about the topic for quite a while until spending the fifteen minutes or so it took to write this post after reading the press release. So, I took at look at the last ten posts on the blog. No complaints about string theory in any of them, the only real reference to the subject was some links to postings about ALICE results in one of 5 parts of one multi-part post.

    It’s pretty clear that as far as certain people are concerned, the problem is not that I’m spending too much of my time criticizing string theory, it’s that I’m doing it at all.

  18. luny says:

    To Derek and Chris, of course the limit in the paper is a “back of the envelope estimate”, since it uses an approximation beyond its range of validity.
    However, while the “AdS/CFT limit” is certainly more accomplished, its physical foundations are exactly as shaky. We do not know to what extent does the strongly coupled theory with 4 supersymmetries and an infinite number of colors approximate any physical system. We do not know how stable is the “limit” against perturbative corrections (see, for example,
    Heck, we dont really know weather AdS/CFT is really “right”,its a conjecture.

    So both approaches are based on assumptions and extrapolations.
    If anything, the most interesting aspect is how close is the AdS/CFT result come to the “back of the envelope” result of 1985.

    And Peter, keep ’em coming. Its true that the eta/s result is not to be made fun of, but its also true that a press release describing Fermi Gas measurements as “string theory in a lab” deserves as much making fun of as it can get.
    I kind of disagree that this is exclusive to string theory though
    (This whopper
    is as hilarious as anything in your blog).
    Unfortunately, in this day and age, scientists are under pressure of spinning anything they do into something fashionable, and these press releases are a byproduct of this

  19. New says:

    Woit: “If experiment disagrees with the gauge/gravity duality “prediction”, I doubt anyone is going to give up on gauge/gravity duality, instead they’ll give up on the particular strongly-coupled model used, or decide that the application of gauge/gravity duality in this situation is too naive.”

    This is exactly the same as *not* throwing out quantum field theory because not all QFTs are experimentally relevant.

    General predictions of QFT (like scalings of scattering amplitudes near fixed points etc.) still have value. Thomas is absolutely right that string theory as we underatand it now is a framwork and not a model.

  20. Peter Woit says:


    Encouraged by thomas, you’re mixing up two completely different things:

    1. string theory as a “framework” for producing TOEs that include the SM + GR in the low energy limit. The question of the value of such a “framework” is whether it predicts anything useful at all, allowing it to be tested, but that’s a different issue than the one addressed in this posting.

    2. string theory as a “framework” for producing dualities between strongly interacting qfts and gravity theories in another dimension. Assuming this works as advertised, it can give a very useful approximate calculational method. That’s what this press release is about.

  21. Chris Austin says:

    Hi Derek, thanks for the reference.

  22. Mean and Anomalous says:

    I certainly do not see what the problem is with exposing over-hyped ‘scientific’ claims for what they are. Keep up the good work Peter.

  23. New says:

    Sorry Peter, I mixed up nothing. The press release said nothing about whether the potential predictivity of string theory was about Unification or not. You dragged that in, as your handy strawman.

  24. seth thatcher says:

    Peter, thanks for continuing to point out the hype associated with ‘tests’ of string theory. I’m as anxious as anyone to see progress in this area and have no animus toward string theory. Rather, as a layman, I prefer to read both sides of an argument so I can make an informed opinion of the science. As far as I can tell, you are the only one providing a reasoned alternative to string theory. It is a service to me. Of course, I also have no problem with those wanting to attack Peter’s viewpoint as long as it is based on logical objections of his analysis of the science. But most of these objections are not logic based.

  25. AL says:

    Very interesting……I have to agree with you in this case, since I published some work on the non-relativistic gas in question and calculated the ratio of viscosity to entropy-density and obtained about 6 times the conjectured “bound”, in agreement with experiments.
    Though my calculation uses “standard” methods, I haven’t
    checked, but I am pretty sure it wasn’t cited.

  26. AL says:

    Actually I just checked my paper again and obtained something between
    4.5 and 5 times the bound, which is even better.

  27. New says:

    There are some calculations I have seen where people use the QCD Lagangian directly and try to do manipulations with it (or do naive manipulations with it inside the path integral) to get the eta/s ratio for quark-gluon plasma.

    I have heard statements defending these approaches, like “at least these people are using the correct (i.e., QCD) Lagrangian, whereas AdS/CFT is using a conformal, supersymmetric theory”. This misses the point completely.

    None of these computations are meangful. This is because asymptotic freedom and confinement make QCD an intrinsically quantum theory. The Lagrangian is an utterly meaningless thing to work with (outisde of lattice-based approaches) when the coupling is strong. Said another way, there are no classical, propagating Yang-Mills waves, unlike electromagnetic waves, so classical physics of this kind is useless. AdS/CFT is useful because it gives a direct handle on quantum QCD via classical gravity. At finite temperature, both supersymmetry and conformal invariance are broken and the theory is deconfined. So it is a far better candidate for comparison against the deconfined phase of QCD than any ill-conceived approach using the QCD Lagrangian.

    Similar statements hold for any approach that claims to make predictions using “classical” approaches for strongly coupled quantum theories.


  28. AL says:

    NEW, I agree, for gauge theories like QCD the Ads/CFT correspondence
    has been very useful in offering new insights and even quantitative results. Although it remains a conjecture, there is a lot of evidence
    that gauge theories are related to gravity in one higher dimension. But this post is about non-relativistic quantum gases, i.e. cold atoms, and it is fair to say that the attempts to use Ads/CFt ideas here are just not as successful.

  29. chris says:

    Dear Apple-Anny,

    why don’t you sell oranges?

    isn’t that kind of ridiculous?

    so why on earth do people feel the need to pressurize an anti-string blogger into blogging about advances in QFT?

    seriously people, if you are so bored, go somewhere else, read something different.

    Oh and New,

    “The Lagrangian is an utterly meaningless thing to work with (outisde of lattice-based approaches) when the coupling is strong.”

    wow. why not tell this to Mr. Leutwyler? or to Mr. Goldstone? sure the QCD Lagrangean is entirely worthless at low energies. what a statement.

  30. AL says:

    There is excessive hype in the press release, but one has to distinguish between the media writers and the physicists. I looked at this paper, and it is a solid paper reporting interesting and difficult measurements in cold atoms. The paper itself does not try to promote string theory. I don’t see anything wrong with comparing with a conjecture on the lower bound of this quantity……it’s just interesting. Ads/CFT has certainly provided unique insights into Yang-Mills. Though the connection with relativistic quantum critical points of condensed matter is more speculative, it has nevertheless provided interesting perspectives on transport phenomena. This post was about cold atoms, which are non-relativistic and at LOW energy…..the success here has been nothing to write home about,
    but the papers on the subject are honest and worthwhile attempts. Where things seem to go wrong is when the non-scientific media tries to make a splash.

Comments are closed.