David Gross Admits String Theory is in Trouble

The latest issue of New Scientist has an article entitled Nobel Laureate Admits String Theory is in Trouble. It describes remarks by David Gross at the recent Solvay conference in Brussels, mentioned here earlier.

Gross described the current state of string theory as “We don’t know what we are talking about”, and also admitted:

“Many of us believed that string theory was a very dramatic break with our previous notions of quantum theory,” he said. “But now we learn that string theory, well, is not that much of a break.”

He said the field was in “a period of utter confusion”, and compared the current situation to that at in 1911, at the time of the first Solvay conference, when no one had any idea what was causing radioactivity.

“They were missing something absolutely fundamental,” he said. “We are missing perhaps something as profound as they were back then.”

The same issue of New Scientist also has an editorial on the subject entitled Physics’ greatest endeavour is grinding to a halt, which ends as follows:

For decades, string theorists have been excused from testing their ideas against experimental results. When astronomers discovered the accelerating expansion of the universe, which string theory fails to account for, many string theorists took shelter in a remarkable excuse: that their equations describe all possible universes and should not be tied to matching data in just one of them.

But when the theory does not match the one data set we have, is it science? There is a joke circulating on physics blogs: that we can, after all, call our universe unique. Why? Because it is the only one that string theory cannot describe. Should we laugh or cry?

There is a growing feeling that string theory has run into the sand. Gross thinks we are missing something fundamental. We need a leap in understanding, though where it will come from is not clear. Many of the greatest minds in physics were there at last week’s conference, and none had an answer.

We are approaching the end of Einstein’s centennial year – a celebration of physics. While some lesser-known areas of the subject are flourishing, the search for a theory of everything is in a sorry state. Unless string theory gets a radical shake-up, gifted but frustrated minds will begin to drift into other areas of science. And if that is what makes biology the subject of the century, it will be depressing reason indeed.

Update: Lubos Motl has some comments on this. He compares the current devastation of string theory to the effects of hurricane Katrina and me to an Islamic extremist, while arguing against the terrible danger to physics if all this leads to study of a “diversity of approaches” other than string theory.

Update: Gross now claims his words were misinterpreted.

This entry was posted in Uncategorized. Bookmark the permalink.

41 Responses to David Gross Admits String Theory is in Trouble

  1. blank says:

    Ouch!

    Now is Lubos going to tell us that Gross has s**t for brains?

  2. Who says:

    for me the big question is whether David Gross acknowledged the existence of other interesting approaches to quantum spacetime or quantum gravity that might reasonably be pursued.

    there has been a consistent tendency for string leaders to say things “our one best hope” and “the only consistent quantum theory of gravity”. Steve Shenker used that phrase at an embarrassing moment in Strings ’05 when someone in the audience asked why study string theory and the panel members didn’t have a ready answer.

    It is a kind of monopoly delusion—-of being God’s chosen research program—-which is maintained by repetition, like a creed. The only time I heard any string notable break from it was in what Andy S said during the same discussion. He outright said there were other interesting approaches that you could reasonably investigate. No sneer or any hint of condescension. Apparently it takes courage and honesty to say that.

    For me the question is if David Gross went that far. If his message was merely “Oh my we are the only game in town and look how confused we are!!! This must be an exciting historical moment, like 1911, when relativity and quantum mechanics were waiting in the wings.”
    If his message was merely that, then it lacked integrity.

    But if he said “we are confused right now and there are alternatives—we aren’t sure of being the one right approach”, then I would put that on par with the Andy S statement at Strings ’05. It is the kind of healthy realism that leadership has to show before things can start getting better.

  3. Who says:

    a propos “physics greatest endeavor” grinding to a halt, I seem to remember that in past years Michael Peskin’s Topcite Review of previous year has been out by around September.

    but if you look here
    http://www.slac.stanford.edu/library/topcites/
    you see that it promises 4 items of information for 2004 but has delivered only two, not including Peskin’s review

    —-quote—
    2004 Edition

    (The 2004 edition covers all HEP papers from January 1, 2004 and December 31, 2004. The following articles are available for this edition.)

    Review of top cited HEP articles 2004
    List of top cited HEP articles in 2004
    List of all-time top cited HEP articles (as of 12/31/2004)

    List of top cited articles from the E-print archives 2004

    —-endquote—

    only the two bolded items appear to be actually available

    Is Michael Peskin going to write the review this year, in fact, or is he reluctant to continue performing that ceremonial function?

    has the annual Rose Bowl of cite-ranking become so embarrassing to string theorists that the Stanford librarians have tactfully decided to let the practice lapse?

  4. Doug says:

    It would be really interesting to know what Gell Mann, t’Hooft, Atiyah, Weinberg and the rest had to say. Does someone know of any plans to make the proceedings available?

  5. woit says:

    Who,

    I doubt it’s the Stanford librarians deciding this, but suspect that, given the discouraging situation of virtually no new ideas in particle theory to report, Peskin has been avoiding the task of sitting down to write up the kind of thing he’s done in past years. Maybe he’ll get to it at some point, maybe he’ll give up on it for good.

  6. Is it really a consensus in the string theory community that this enormous so-called landscape of vacua is in fact an unavoidable result from the “theory”? Or there are people that do not really accept this result as a final word and attempt to find other possibilities (within the “theory”, I mean)?

    Now one more thing concerning this apparent dead-end for string theory. This question is ashamedly basic and possibly a nonsense, but I will ask anyway.

    – This landscape thing. Has anyone thought about some “meta”(?)-variational principle or something analogous so that a unique vacuum could be selected from such “principle”? I have not the slightest idea if that makes any sense, but I appreciate any comments on this.

    Christine

  7. Aaron Bergman says:

    Is it really a consensus in the string theory community that this enormous so-called landscape of vacua is in fact an unavoidable result from the “theory”?

    Yes and no. There seem to definitely be an infinite number of vacua — toroidal compactifications at various radii, for example. The new idea is the production of vacua that have no massless scalars. These are all AdS vacua. The techniques involved in the production of these vacua aren’t completely rigorous, but persuasive.

    Because they’re AdS, none of these vacua are realistic. One needs to lift them to de Sitter vacua with a positive cosmological constant. This step is much more sketchy.

    Even when you accept all that, one has to ask how many vacua are there that have a reasonable matter content. For any useful definition of reasonable, that’s a difficult question. It has been joked by various string theorists that people have extrapolated the existence of a huge number of realistic vacua when string theory hasn’t yet produced a single one.

    One scenario for predicitivity of string theory is that there turns out to be very few vacua consistent with current observations. If we can pin down the precise vacuum that we live in, then the theory becomes predictive.

    Lots of people have thought about selection mechanisms for vacua, but — in that we don’t actually understand string theory — it’s all pretty speculative.

  8. A. nonymous says:

    I think the variational principle involved is that the background should be a (meta-)stable vacuum. This will ensue that you have a discrete set of candidate backgrounds. The problem is that, although there are discrete vacua, there are a very large number of them. You need some principle which selects a good element from a very large and discrete set, and those kinds of principles are much harder to come by than principles which let you choose a special element from a continuous set.

  9. Who says:

    christine, before people start discussing the “landscape” I want to make one point of information about the 23rd Solvay CONFERENCE. It was myopically organized, to a disappointing extent I think. Ostensibly, the conference was about research areas where the non-string quantum gravity community has been especially successful recently—and by contrast little of significance has been happening in string.

    When this was realized, it may well have left a residue of embarrassment—-either in David Gross himself or in some of the others. It would have been a better conference had it been less parochial. I will illustrate how 23rd Solvay shot itself in the foot.

    To see the program follow links at:

    http://tena4.vub.ac.be/23Solvay/qsst/

    The subject was the
    QUANTUM STRUCTURE OF SPACE AND TIME
    broken down into several main foci of discussion
    each with a “rapporteur” and a panel. I will list the rapporteur
    for each of several main topics:

    A. singularities (Gibbons)
    B. mathematical structures (Dijkgraaf)
    C. emergent spacetime (Seiberg)
    D. cosmology (Polchinski)

    the page I mentioned has links to lecture notes for three of the four rapporteur talks—-A. C. and D. Gibbons, Seiberg, Polchinski
    I was impressed by how insubstantial these lecture notes were, and how much they missed by way of active research and recent results in these areas

    In each of these three cases, significant advances are being made outside string, to which the rapporteur does not refer.

    A. in the case of singularities the removal of the black hole and cosmological singularities in the Loop context, by Martin Bojowald and others should have been highlighted (in any but a narrowly string-centered context) and apparently was not even mentioned

    C. as for emergent spacetime, that has been exactly the point of recent work by Loll and others—the emergence of a 4D spacetime in Causal Dynamical Triangulations. In fact it was surprising that Loll was not on hand.

    D. as for cosmology, I assume this means Quantum Cosmology since the overall subject of the conference was the quantum structure of space and time. Abhay Ashtekar, who was at 23rd Solvay, could well have pointed out the rapid growth in activity in Loop Quantum Cosmology, with connections beginning to be made with observational astronomy. But quite possibily he would not have done so in public, since the organization made it awkward. Ashtekar was not included on the cosmology panel. In fact there was no LQC representation on the cosmology panel at all! It consisted of Banks, Guth, Kachru, Kallosh, Linde, Steinhardt, Weinberg! I see that not as a recipe for quantum cosmology discussion, but instead for speculative talk about eternal inflation, bubbles, multiverses, and the landscape.

    maybe some discouraging noises in the Solvay wake is due to unwise organization, I’m thinking.

  10. Not a Nobel Laureate says:

    Radioactivity was an experimental observation that no one knew how to explain in 1911.

    What experimental data requires a string theory explanation?

    Physics is still an experimental science.

    Deduction by pure reason is the domain of meta-physics or, in the case of strings, Ubian pata-physics.

  11. secret milkshake says:

    M-theory => Marshland Monstrosity

  12. Aaron Bergman says:

    Physics is still an experimental science.

    Would that we could do experiments that probe quantum gravity.

  13. D R Lunsford says:

    Would that we could do experiments that probe quantum gravity.

    What would we learn? I don’t think anyone can realistically formulate what you would measure.

    Sometimes I’m amazed at how bright people such as you can defend these labyrinthine thought constructions. No successful theory of anything was ever so complicated. On the contrary, the good theories like Dirac and GR are simple and involve a bare minimum of structure. Isn’t it just intuitively obvious that when things get so complicated, it’s time to try something else? Even if you choked some kind of answer, even a wrong one, from such a theory, who would be happy?

    -drl

  14. Thank you all for the instructive comments on my landscape question an to “Who” for the important comments on the 23rd Solvay conference.

    All the best,
    Christine

  15. FP says:

    > we could do experiments that probe quantum gravity

    Let’s assume that somebody can show that the Pioneer anomaly is real and
    measure it exactly.
    How would this help string theorists? (And how would this help LQG?)
    As far as I know “the only consistent theory of quantum gravity” cannot estimate
    deviations from Newton’s potential.

  16. Not a Nobel Laureate says:

    NNL: Physics is still an experimental science.

    AB: Would that we could do experiments that probe quantum gravity.

    NNL: Since we can’t then what’s the point of constructing theories that can’t be falsified, be they stringy or loopy.

  17. Aaron Bergman says:

    Since we can’t then what’s the point of constructing theories that can’t be falsified, be they stringy or loopy.

    Because they might be right? I don’t know about loopiness, but we’ve also learned a lot of things about susy gauge theories and mathematics from the study of string theory, too.

  18. Anonymous says:

    Peter,

    Lubos Motl is a clown, and a rather pathetic one at that. That much is obvious to anyone who’s had a glance at his blog. What I don’t understand is why you are so concerned about his silly tirades and opinions (e.g., the update to this post). If he has’nt anything substantiative to say, why not just let him rot in this own dimensions?

    Just curious.

    -Anon.

  19. woit says:

    Anonymous,

    Sometimes I think Lubos’s antics are just hilarious, and don’t want anyone needing entertainment to miss them, thus the update to the post. OK, this is not very high-minded behavior on my part, I admit it.

  20. Anonymous says:

    I thought I would share this link demonstrating that even if the older members of the string community are becoming more pessimistic about the current incarnation of string theory, younger people certainly aren’t. The website:
    http://www.interactions.org
    Has an article written by a grad student describing the field to a lay person:
    “String theory a very exciting prospect for physicists”
    http://www.theithacajournal.com/apps/pbcs.dll/article?AID=/20051207/LIFESTYLE18/512070307/1025

  21. D R Lunsford says:

    In the day, we usually had (or taught ourselves) a course in theoretical physics (Sommerfeld, Landau, etc.) which included not only QM and relativity, but things like hydrodynamics, elasticity, kinetic theory of gases, optics, etc. etc. I doubt this student has much acquaintence with these things, and therefore he is not in a position to know what he’s missing with ST.

    -drl

  22. Aaron Bergman says:

    younger people certainly aren’t

    Like older people, there’s a wide array of opinions among younger people.

  23. Chris Oakley says:

    Another interesting point you might have heard about is that string theory predicts that there are really ten dimensions — the three big spatial ones and time, which we are used to, and six more tiny dimensions.

    This is a bit like saying, “I’m a qualified brain surgeon: all I need to do is to get into medical school and study for seven years”

  24. Not a Nobel Laureate says:

    NNL: Since we can’t then what’s the point of constructing theories that can’t be falsified, be they stringy or loopy.

    AB: Because they might be right?

    NNL: The odds of finding the right theory be it string or loopy or something entirely different in the space of all possible theories is vanishingly small – the history of pure reason is the history of dead ends.

    If some people want to pursue this fine – they should, but it is inappropriate that it should come to dominate theoretical HEP.

    AB: I don’t know about loopiness, but we’ve also learned a lot of things about susy gauge theories

    NNL: Great. Now what’s the experiment evidence for SUSY?

    AB: and mathematics from the study of string theory, too.

    NNL: That’s also great. And it reinforces why strings belong in the math dept not the physics dept – until they can predict some physical phenomena.

  25. Aaron Bergman says:

    The odds of finding the right theory be it string or loopy or something entirely different in the space of all possible theories is vanishingly small – the history of pure reason is the history of dead ends.

    Like GR?

    If some people want to pursue this fine – they should, but it is inappropriate that it should come to dominate theoretical HEP.

    It doesn’t. There are more posts on hep-ph than on hep-th. String theory, for better or worse, just gets most of the press. In the future, there’s going to be even more of a trend towards hep-ph jobs. You can see that happening right now.

    Great. Now what’s the experiment evidence for SUSY?

    SUSY for years was impressive in that it provided a dark matter candidate, unified the couplings and made things more natural. It’s not looking as pretty as it once did these days, but we’ll see in a few years.

    That’s also great. And it reinforces why strings belong in the math dept not the physics dept – until they can predict some physical phenomena.

    Who cares what department it’s in?

  26. Not a Nobel Laureate says:

    NNL: The odds of finding the right theory be it string or loopy or something entirely different in the space of all possible theories is vanishingly small – the history of pure reason is the history of dead ends.

    AB: Like GR?

    NNL: Thanks. I was expecting this reply.

    1. The validityof inertial and gravitational mass as physical concepts and the WEP were already established by experiment – although Einstein states that he was unaware of the Eotvos experiement until later.

    The concept of strings is not established experimentally.

    2. Even if we allow that you’re right, GR is atypical of the way the progress in physics is usually made:

    Technical advance in instrumentation -> observation of new and unexplained phenomena -> candidate theories -> new predictions -> experimental tests.

  27. dan says:

    before we write off string theory shouldn’t we wait for LHC to come online?

    I think SUSY ans SUSY-m-theory are important research programs for *physics* in that they help
    1- encourage investments into particle accelerators, including LHC
    2- tell experimentalists what sort of particles to look for (Higgs, extra dimensions, SUSY-particles)

    if LHC does find SUSY-particles such as the neutralino, such a finding would strongly support the string theory research program.

    as difficult as the landscape problem presents, m-theory may be worth pursuing if LHC finds evidence of SUSY particles and/or higher dimensions.

    if LHC does not find SUSY-particles or extra dimensions, i would agree with Woit we should more aggressirvely pursue other research programs, including LQG.

  28. Chris W. says:

    (Taking off from NNL’s comment —)

    Both special relativity (SR) and general relativity (GR) are atypical, inasmuch as they were motivated in Einstein’s mind by facts about certain relations of existing theories to well-established experimental observations and to each other.

    In the case of GR the relevant fact was the apparent equivalence of gravitational and inertial mass, which was relied upon but not explained in the formulation and application of Newton’s theory of gravitation. The significance of this fact was amplified for Einstein by Ernst Mach’s acute analysis of the difficulties that universal gravitation created for any effort to give operational meaning to the idea of inertial motion.

    In the case of SR the relevant fact was the equivalence of the currents induced in a wire (eg, in the form of a coil) when (1) the wire was held stationary relative to the laboratory and a magnet was moved near it, and when (2) the magnet was held stationary and the wire was moved such that the relative velocity of wire and magnet was the same in the two situations. The measured and calculated currents (in Maxwell’s electrodynamics) were known to be the same in the two situations, but the description of the two situations in Maxwell’s theory as it was understood at the time differed markedly. The significance of this fact was amplified in Einstein’s mind by the apparent theoretical possibility—given Newtonian kinematics—of overtaking an electromagnetic wave and rendering its oscillation stationary in the observer’s frame of reference.

    Other physicists of the time could have responded to both of these concerns this way: “Well, so what? What observation is being contradicted here? I don’t see any real difficulty, certainly not one that justifies reconsidering the foundations of our science, or the assumptions behind our current investigations into the properties of the luminiferous aether.”

    The point in the case of SR is that the theoretical issue restated here was as much or more of a preoccupation for Einstein than any of the specific observational oddities that were perplexing his contemporaries. They were concerned with specific discrepancies between theoretical expectations and observation. He was concerned with what he saw as a deficiency in explanatory power in situations where no specific discrepancy between calculated consequences and corresponding observations was at issue. His genius lay in the fact that he had latched onto genuine problems within the overall theoretical framework of classical physics that had not been clearly seen before, proceeded to resolve them, and in so doing showed that observational anomalies already known could be understood and resolved.

    This echoes our current predicament in essential respects. The fact that string theory in its early formulations did not follow directly from supporting observations is beside the point. (We should be so lucky…) We do have fundamental inconsistencies within the overall structure of late 20th century theoretical physics, and we do have observations and experimentally determined quantities which play a fundamental role in our theories but which have no deeper explanation. Our problem, as it was for Einstein, is to understand these difficulties, see how they can be resolved, and see how the resolution can be tested, ie, how it manifests itself in the world we now observe or might observe if we are sufficiently clever.

    There is no guarantee that this can be done in a way that doesn’t amount to an epistemological capitulation, an abandonment of essential methodology of science—that we find things out by having ideas and trying very hard to see where they might fail as well as where they succeed. To do this we must have ideas that are not designed at the outset to be immunized against empirical failure. Such immunization must itself be regarded as a failure; this point is what this weblog has become largely about. Einstein had a profound faith that one could adopt this attitude and yet not be hopelessly stymied in understanding the world.

  29. D R Lunsford says:

    Chris W makes excellent points, and just to emphasize, GR was no more (or less) a product of “pure reason” than was the Dirac equation, and it fit in nicely with the historical development in its time, no matter how surprising the form*. Antimatter and horizons are concepts that emerged from the two formalisms – they were not inputs. In contrast, string theory puts in by hand an assumed form of matter. To make an analogy, one cannot derive the Dirac equation from the existence of antimatter, because the Klein-Gordon equation also leads to the same idea. String theory is always trying to derive what we already know with an assumed form of matter, which is backward – the traditional way is to model what we don’t know as competently as possible and see what new ideas emerge, which inevitably correspond to real facts.

    (markus: that is an intriguing idea 🙂

  30. Aaron Bergman says:

    The program of quantum gravity — not just string theory — is certainly an act of hubris. That nobody has made it very far in twenty-five years is humbling. The hope of everyone in string theory, for years, was that it would uniquely predict our world. At best, in now seems more likely that string theory has a huge multitude of vacua, and the best hope of obtaining predictions would be to overconstrain the particular vacuum we live in or discover some way to probe the universal aspects of the space of vacua.

    I don’t know if that’s possible. One can argue aesthetics all night, but string theory’s still the only idea out there that has derived a black hole entropy. Until we’ve got some experimental data, we’re stuck with our hubris. It’s either that or go work on something with the prospect for data — which a lot of people are doing these days.

  31. FP says:

    Aaron,

    > Until we’ve got some experimental data, we’re stuck with our hubris.

    I have asked this question before. How would experimental data help you (or any other string theorist) ?

    E.g. If the Pioneer anomaly is real and we could measure it exactly,
    how would this help with superstrings (or LQG) ?
    You cannot calculate corrections to the Newton potential and thus you cannot decide if the Pioneer anomaly confirms or falsifies string theory.

    I am just using this as an example, the same would be true for any other (gravitation) experiment.
    What if Gravity probe B data show a small (or large?!) deviation from GR? How would a superstring theorist react to such news?

  32. Arun says:

    Chris W., thank you!

    FP, if the Pioneer anomaly is real, and cannot be accomodated in Einstein’s GR or similar geometric idea, then the idea of gravity as the effect of the geometry of space-time may be overthrown. Who knows what will happen to blackholes and their entropy? The idea of space-time fluctuations at the Planck scale, etc., may be discarded and so on. The correct classical theory of gravity is an essential ingredient to a theory of quantum gravity and thereby to string theory as well. For instance, whatever the corrected theory of gravity turns out to be, it had better show up in the string theory derivation of the consistency condition on the background space-time that looks like Einstein’s equations.

  33. Aaron Bergman says:

    I have asked this question before. How would experimental data help you (or any other string theorist) ?

    It depends on the type of the experiment. If one could (unlikely) directly probe high energy quantum gravity, that would be immensely useful.

    Low energy things, on the other hand, are much more difficult to exploit. The cosmological constant, for example, is a bit of low energy experimental data that has drived people to despair anthropic explanations.

    Nonetheless, it’s still better than nothing. And the more stuff we have, the more likely it is that someone might be able to guess some unifying theory.

  34. FP says:

    Arun,

    > The correct classical theory of gravity is an essential ingredient to a theory of quantum gravity and thereby to string theory as well

    I always understood it the other way around. The quantum theory of strings, or
    M-theory if you will, contains necessarily a theory of gravitation and there is no freedom to begin with different ingredients …

  35. theon says:

    Did anybody question whether we really have to quantize gravity?
    If not, then a lot of the current problems need no answer.

  36. Aaron Bergman says:

    Did anybody question whether we really have to quantize gravity?

    Yes. It’s very hard to write down anything that makes sense that way. Jacques talked a bit about it here.

  37. Question especially for Peter and Aaron (or anyone else who knows): I have been hearing for many years some variation on “we don’t really know what String Theory (or sometimes M-Theory) is.” Clearly string theorists know how to compute a lot of things, so what do they mean by that? If it’s not the theory of these little string like things, why not?

    Is some key conceptual or dynamical piece still missing? If so, isn’t it likely that discovering it might simplify landscape, swamp, etc.?

  38. woit says:

    CIP,

    Basically what is understood is “perturbative string theory”, which is conjectured to be a series expansion of some underlying “non-perturbative string theory, AKA M-theory”. Many string theorists certainly hope that when M-theory is finally understood, it will solve the problems currently plaguing string theory. But some things are known about M-theory, and what is known gives no evidence at all for this to be true.

  39. Aaron Bergman says:

    If it’s not the theory of these little string like things, why not?

    Damned if I know.

    Really, it’s not at all clear that we understand what the fundamental degrees of freedom actually are in string theory. We have control over a few things like AdS/CFT and a perturbation expansion, but beyond that, it’s a lot of patchwork approximations.

    It’d be cool if understanding all this stuff got rid of the multitude of vacua — Tom Banks has argued that many of the approximations break down for subtle reasons having to with asymptotics in cosmology — but it’s hard to say that it’s likely.

    You ask if some key conceptual piece is still missing. I think, even after who knows how many ‘revolutions’, we still understand next to nothing. The anthropic and string phenomenology types get a lot of the press, but there’s still a whole bunch of us who would prefer to understand just what string theory is before trying something as audacious as applying it to the real world.

    Of course, that’s hard, so I’m just working on trying to understand what a quiver gauge theory on a D-brane is….

  40. Who says:

    Theo asked a good question.

    *Did anybody question whether we really have to quantize gravity?*

    there is a thread at PF about this question, in case anyone has ideas about why it could be useful to quantize gravity.

    http://www.physicsforums.com/showthread.php?p=856677#post856677

  41. Wolfgang says:

    > Did anybody question whether we really have to quantize gravity?
    Yes of course. And the answer is, yes we do have to quantize gravity,
    as explained e.g. by Jacques Distler (the part about ): http://golem.ph.utexas.edu/~distler/blog/archives/000639.html

Comments are closed.