Scientists Find a Practical Test for String Theory

This sort of thing seemed to be dying down (2013 required a record low number of “This Week’s Hype” postings), but 2014 is starting off with the usual promotion by physicists of nonsense about how they have “found a test for string theory”. This time the news that Scientists find a practical test for string theory comes from a group at Towson University, who are basing their claims on this paper, published here. I’m not sure where got this, but it reads like a university press release, and they credit “Provided by Towson University”.

What’s actually in the paper is a proposal for a test (and not a very good one, as far as I can tell…) of the equivalence principle. The claim is then that a violation of the equivalence principle would be evidence for string theory. I’ve written about this kind of claim before (see here), pointing out that string theorists sometimes argue that the equivalence principle is a prediction of string theory. So, string theory can be tested, and the test is even “practical”, but since the prediction is that either the equivalence principle will be violated or not, it’s pretty likely to pass the test.

Update: Another source for the press release is here.

Update: Matt Strassler weighs in, a week later:

Baloney. Hogwash. Garbage.

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

46 Responses to Scientists Find a Practical Test for String Theory

  1. CIP says:

    If I understand correctly, the question here is really the validity of the so-called “strong” or Einstein Equivalence Principle (EEP), which insists that fields that couple directly to matter (like gravitation), couple through the metric. Brans-Dicke and other fifth force variants don’t.

    The same may be true of the dilaton and some other string theoretic fields, but that doesn’t seem to be iron clad. Still, any fifth force type effects seen would seem to give string theorists and other alternative gravity people something to work with.

  2. Jim Akerlund says:


    Here is a different test for equivalence principle, but no mention of String Theory.

  3. Peter Woit says:

    One thing that I don’t doubt is that the day after evidence for a fifth force shows up, the headlines will be “Evidence found for string theory!”….

  4. Jim Akerlund says:

    Here is the missing link. Sorry.

    Three Star Test

  5. David says:

    How do they know that there isn’t a fourth component in the system, too dark to see, but with enough mass to cause deviations from the computed 3-body solution?

  6. Allan Rosenberg says:

    You have to give them credit for honesty. How many people proposing this sort of test have the cojones to say that “[s]tring theory is infamous as an eloquent theoretical framework to understand all forces in the universe….”

  7. Paul says:

    One of the studies authors had this to say:

    “Hi, I’m one of the authors of the paper. It’s crazy to see my undergraduate research on Reddit!
    Basically, some versions of string theory propose that elements will “fall” at different rates (i.e. violate the equivalence principle). Well, planets are “falling” around the Sun, along with asteroids in the same orbit. Moons are falling around planets as well.
    We have a pretty good idea of exactly where everything is supposed to be. When we calculate these things, we assume that everything falls at exactly the same rate (that is to say, inertial mass is the same as gravitational mass). And within margins of error, they ARE right where we expect them to be.
    So, if these versions of string theory that say things can fall at different rates are true, we should see some kind of discrepancy between the models and what we really observe. We don’t see these discrepancies. This places limits on the parameters of those versions of string theory (and other “beyond the standard model” theories that don’t technically involve strings).
    Of course, we don’t know where everything is to an infinite number of decimal places. Perhaps Jupiter is 0.00001 arcseconds away from where we claim it is. We can’t measure that precisely. So if there is an effect due to something from string theory, it would have to be hiding in that uncertainty. In our paper, we take the limits of that uncertainty and turn it into limits on how much something can violate the equivalence principle. From there, if you had a string theory model, you could calculate how much it would violate the equivalence principle, and if it was bigger than our numbers, you might want to be skeptical about your model (especially since our numbers are quite conservative and aren’t even as strong as other methods).
    Hope that helps.”

    Interested to hear your response as I am not nearly smart enough :)

  8. Bee says:

    They propose to test for some kind of dilaton coupling. I actually read the paper last year. You only have to look at the first page to see that they are not actually claiming that they test string theory. They propose to look for generic violations of the equivalence principle that could hint at physics beyond the standard model. I think it’s a reasonable proposal, but the headline is more than just misleading. It is a) wrong and b) most likely deliberately so.

  9. Bee says:

    PS: I said reasonable – I don’t think it’s terribly exciting because I can see no good reason the effect should appear in the parameter range that the experiments are sensitive to. That isn’t to say one shouldn’t look though.

  10. Zhiming Wang says:

    Wow, I like that “Three Star Test” referenced by @Jim Akerlund. A great example of how to spice up some regular scientific discovery/progress so that it sounds revolutionary.

  11. Shantanu says:

    I asked matt strassler about whether string theory violates equivalence principle
    or not (since as Peter mentioned I have heard conflicting claims on this issue) and his reply was
    (although you will have to sift through lots of comments to see his reply inline)

    Regarding the actual paper, it doesn’t cite the limits from binary pulsar(pulsar/white
    dwarf systems) which I think they are equally stringent compared to solar system limits.

  12. Peter Woit says:

    For a serious discussion of the physics here, see Thibault Damour’s recent
    especially page 12. I think a fair reading of Damour is that “the current string landscape prediction is no equivalence principle violation, but if equivalence principle violation is found, that just means string theorists need to look at other currently less popular string theory models”.

    So, usual story for string theory “predictions”.

  13. chris says:

    if a violation of the strong equivalence principle would indeed be found, the one theory that raises in popularity would probably be MOND.

  14. Curious Wavefunction says:

    That’s a little weird and it almost sounds like a tautology. It would be like claiming that the existence of friction is evidence for string theory (since string theory is supposed to be true).

  15. BobDastro says:

    What’s actually in the paper is a proposal for a test (and not a very good one, as far as I can tell…) of the equivalence principle

    The authors are not proposing a test of the equivalence principle. Rather, they actually determine new (and significantly improved) upper limits to deviations from the equivalence of inertial and gravitational mass based on existing Solar System observations.

    I would also say the paper makes no real claims about testing string theory per se. The authors cite in passing one speculative stringy model, but it is mentioned as only one example among several other classes of theoretical proposals that allow for composition-based differences between inertial and gravitationally-derived masses.
    IMHO this is a clearly written paper that is well worth reading for those of us interested in observational data that bear on the fundamental properties of gravitation.

  16. Eli Rabett says:

    Wife’s comment on seeing the press release about a test for string theory: Oh, they found the Big Ball

  17. MathPhys says:

    From a quick reading of the paper, I agree with BobDastro’s comment above. It’s a paper on theoretical astronomy that basically says “If you observe this, then that means that”, which (assuming it’s technically correct) is fine and educational.

    Now if someone has decided to blow the string-theory-can-be-tested trumpet to attract attention, that’s a different (and sad) issue.

  18. Peter Woit says:


    This is following the usual pattern: published article includes only minor references to string theory, since no referee would allow the author to claim that this was a “test of string theory” (since it isn’t). On publication of the article, the author has their university press office issue a press release about how they have discovered a “test of string theory” (I don’t believe in claims that university press offices issue press releases about their faculty’s work without the faculty member’s agreement). The press release then gets spread through various media outlets, often with the outrageousness of the claims increasing as it spreads. Finally, you end up with lots of news stories like

    There are by now dozens of examples of this. You can argue about who is responsible for the public getting misled here, my vote would be for the physicists who allow or encourage such press releases to go out (together with their colleagues who raise no objection or sometimes provide supporting quotes for the stories).

  19. Zathras says:

    It seems to be almost a contradiction in terms to have universities put out non-peer-reviewed press releases on peer-reviewed results. Maybe these press releases should be peer reviewed as well?

  20. chris says:

    the peers of university press people are other universities press staff, right? that would be a nice peer review.

  21. DaveC says:

    My university has put out press releases on our work (in condensed matter; not at our suggestion but at the journal’s). Our press guys take great pains to represent faithfully what we talk about in the interviews and incorporate all our own proof edits. I find it hard to believe this is the exception rather than the rule.

  22. Chris W. says:

    Press release = Marketing blurb

    Is that too strong? There is an incentive for both the researcher and the institution to claim too much. (And yes, universities are operating more and more like for-profit entities.) Unless there are clear negative consequences in terms that truly matter to them, this stuff will tend to become standard practice.

  23. I think a fair reading of Damour is that “the current string landscape prediction is no equivalence principle violation, but if equivalence principle violation is found, that just means string theorists need to look at other currently less popular string theory models

    I don’t think so. As far as I understand, Damour says that EP violation is a generic prediction of string theory, but a widespread assumption is to say that there might be a mechanism that protects the EP. Though, it is just a “convenient” assumption that is used by “model builders” . In general, Damour’s work (originally with Polyakov [1]) shows that even a massless dilaton (which has a long range dynamic that should imply a long range violation of the EP) can lead to a very small EP violation only, because the dilaton field’s cosmological equation can have an attractor at late cosmic times (hence leading to a very small violation of the EP at current epoch, that would thus be compatible with current experiments). (Note that it is said that they “can” and not that they “would”, see also [2]).

    In any case, although string theory generically predicts EP violation, observing one of its various representations is indeed not a test of string theory since string theory is not the unique theory to propose EP violation anyway. But it would certainly be a small argument in favor of string theory. Also, this “prediction” is qualitative only, which is not quite the standard of “checkability” in modern science…

    Otherwise, as far as I understand, there might be a way to put a quantitative constraint on the string theory effective action. Indeed, the full loop effective action of string theory in the gravitational sector should predict a unique specific coupling between the dilaton field and material fields. But it has recently been shown that depending on the such a coupling, the so-called post-Newtonian parameter gamma can be either more than one or less than one [3] (or even equal to one for a very specific coupling [4]). Hence, as explained here, the sign of 1-gamma (if measured) may put a strong constraint on the non-perturbative effective action of string theory, if someone is able to compute it, of course…..

  24. Peter Woit says:

    You’re ignoring the “string landscape” issue completely. What “string vacuum” are you planning to use to evaluate the effective action of string theory? Damour explains clearly on page 12 that the standard string landscape scenario implies a prediction of no observable EP violations:
    “Though such a mechanism might entail observable short-range modifications of
    gravity [30], it predicts the absence of any long-range EP violations.”
    This same argument has often been given by Douglas in recent years as a response to those like me who say string theory predicts nothing. According to him, the string theory landscape does make a prediction: no observable EP violations (because un-fixed moduli will contribute to the CC). According to you, string theory predicts EP violations.

    I do think you’re both right: string theory predicts anything you want, either EP violation or no EP violation.

  25. Dear Peter, Damour also says p.12 (a little after your quote)

    “This family likeness between the dilaton φ(x) and the metric g_{μν}(x) (which entails a correlated likeness, say in heterotic string theory, between g_{μν}(x) and the gauge couplings g^2_a(x), as well as the string-frame gravitational coupling G(x)) suggests that there might exist consistent string vacua where some of the moduli fields are not stabilized, but retain their long-range, spacetime-dependent character. As recalled above, such a situation would entail long-range violations of the EP. How come such violations have not yet been observed, given the exquisite accuracy of current tests of the universality of free fall (at the 10^{-13}level [26]) and of current tests of the variability of coupling constants [32]? A possible mechanism for reconciling a long-range, spacetime varying dilaton (or, more generally, moduli) field φ(x) with the strong current constraints on the time or space variability of coupling constants is the cosmological attractor mechanism [33, 17,34]“.

    As far as I understand, it says that one can expect some moduli fields to be long-range, just as the metric is, and in particular the dilaton field. In that case, one should expect an EP violation, at least form the dilaton field.

    Hence, although a violation of the EP wouldn’t prove string theory of course (nor would it be a striking evidence that string theory might be correct), it still seems to me that it would be a small argument in favor, since one should expect such a violation from string theory.

    Otherwise, in the last part of my last comment, I assumed that the dilaton-matter coupling in the effective action does not depend on the string vacuum considered (for instance, at tree level, the dilaton coupling is universal no-matter what). But I’d love to have a comment from an actual string theorist on this issue.

  26. Peter Woit says:

    It still seems to me that Douglas’s argument is that no unstabilized moduli is the generic expectation in the landscape picture, and thus no EP violation. Damour is saying it is worth thinking about alternatives and “suggests that there might exist” ones that violate the EP. This is different than your “one can expect some moduli fields to be long-range”.

    I don’t disagree at all that one can find string theory models with or without long-range moduli and thus EP violation. The question is which alternative to expect, with Douglas saying he expects no EP violation from string theory, you saying you expect some. String theory here has no predictive power. Put differently, if we see EP violation, Douglas’s claims implies this is evidence against string theory, while you and the authors of this press release want to claim it would be evidence for string theory.

  27. Dear Peter,

    it is not because several people claim opposite things that it means that the theory actually allow opposite things to exist within its paradigm. It could simply be that one of the two arguments is wrong. Only a formal proof would tell which argument is correct, but as far as I can tell, formal proofs are rather difficult to get in string theory. I don’t know Douglas’s argument that you are refereeing to. Could you send me a link please? I guess you have my email address. Thanks a lot :)

    Otherwise, if the effective dilaton-matter coupling(s) in the effective 4D action is(are) indeed string vacuum independent, if the dilaton field is proven to be massless (or light), and if one can compute the full loop effective dilaton-matter coupling(s), then a measurement of the value of 1-gamma could be a strong constraint on string theory, because its sign depends on such a coupling [1]. That’s a lot of ifs, but it’s better than nothing…

  28. Jeff M says:

    What, physicists do formal proofs, really :-)

  29. Peter Woit says:

    For a paper by Douglas relevant to this, see
    But the reason I refered to the Damour paper is that he makes the basic argument explicitly (end of page 11 to page 12). For another similar Damour claims, see
    where on page 8 he writes “EP tests are important because they test an assumption commonly made in string theory, and could refute it.”
    I don’t see any reason why dilaton matter couplings should be string vacuum independent.

    In any case, Damour clearly is claiming that an assumption commonly made in string theory implies no EP violation. This is inconsistent with any claim that string theory generically will imply EP violation.

  30. Dear Peter,

    thank you very much for the links.

    Still, the way I understand Damour is rather that: in general, string theory predicts EP violation. However, one can use an assumption that forces string vacua to respect the (long-range) EP in order to get specific models one can play with. May it be widely used or not, as far as I understand, it is still only an assumption. Not a prediction.

    Regarding Banks, Dine and Douglas paper, as far as I understand, it only argues that it wouldn’t seem natural to have a variation of the dimensionless couplings, according to our current understanding of QFT. Though it probably shows the state of the art regarding the QFT’s point of view on the problem of the variation of the dimensionless constant, I’m not quite sure one can deduce that it means that string theory predicts no EP violation, as you seem to argue. But maybe have you linked the wrong paper? Or maybe have I missed something?


  31. Peter Woit says:

    I’ve heard the argument a couple times directly from Douglas that string theory predicts no variation in moduli. For an example in print
    page 18
    “This is perhaps the simplest testable prediction of string/M theory for which contrary evidence has ever been reported”
    (referring to claimed observations of variation of the fine structure constant). Here he refers to the Banks/Dine/Douglas paper. You’re right though that I don’t think it’s just a naturalness argument, what I remember hearing from Douglas was an argument that variation of moduli would imply a huge variation in the vacuum energy, but I don’t immediately know of a source where that appeared.

    Of course the occurrence of large number of moduli fields in string theory means that generically you have lots of different sizable long-range forces, so that is some sort of “generic” (and completely wrong) prediction of string theory. The question is whether some such things survive the standard landscape picture. My reading of both Damour and Douglas is that this is not supposed to happen in this picture. Douglas is quite definite about there being a “prediction of string theory” here, and it’s the opposite of a prediction of observable moduli.

  32. Rafael says:

    Dear Peter, do you believe it is true ?

  33. Peter Woit says:

    Seems unlikely to be true (solution to Millenium problem), but I know little about this subject.

  34. Bernhard says:

    And it will be quite hard to verify it since the guy decided to publish the solution in a Kazakh journal.

  35. Dear Peter,

    thanks again for this new link! :)

    Thus, as far as I understand, according to the reference to Banks/Dine/Douglas paper in Douglas/Kachru paper, no EP violation is a “prediction” of QFT alone, not string theory in particular. (It seems to me that it is actually not a prediction, but more an expectation based on our current understanding of QFT). Therefore, it appears to me that Douglas’s claim that no EP violation is a prediction of string theory is exaggerated (if not wrong), unless I have missed something.

    If the EP is to be expected from QFT, as Banks/Dine/Douglas paper seems to tell, then it would mean that any theory with (measurable) EP violations would be “unnatural” from the QFT point of view, may this theory be one of the vacua of string theory with no EP violation or anything else.

    It is very interesting. However, one should remain cautious with QFT’s predictions since, as far as I know, QFT is still far from being completely understood, not to mention the possible loopholes that may hide behind Banks/Dine/Douglas argumentation.

    Anyway, thanks for the discussion :)

  36. vmarko says:


    “no EP violation is a “prediction” of QFT alone, not string theory in particular”

    The equivalence principle is a statement describing how gravity couples to matter. As such, it cannot be predicted by any theory which does not incorporate gravity at some deeper level. String theory notwithstanding, there are no standard QFT’s out there that contain gravity in any shape or form, and therefore no ordinary QFT can ever predict the EP.

    In order to seriously discuss the theoretical prediction/violation of EP, you need a theory of quantum gravity. And today it is common wisdom that QG cannot be described with the formalism of QFT (except in some very unusual hypotetical constructions involving the existence of a nonperturbative fixed point etc.).

    So QFT in general has nothing to say about EP.

    HTH, :-)

  37. Peter Woit says:

    So, according to you, Douglas/Kachru are just completely wrong to in their claim about a “prediction” of the string theory landscape. Fine, I’m definitely not going to be the one to spend time defending string theorist’s claims for their “predictions”….

  38. David Brown says: Otelbaev’s publication in Russian with English abstract at the end

  39. Marko,

    It might be a matter of nomenclature. By QFT, I mean a set of tools one applies on actual theories originally defined classically in order to get their quantum behavior. Now, as far as I understand, Banks/Dine/Douglas suggest that a coupling of the form “scalar times Lagrangian of material field (=\Phi L_m)” would be “unnatural” from the QFT point of view for a light scalar field (actually they restrict their attention to L_m=F^2). Such a type of coupling could indeed come from string theory but from many thing else as well. For instance, one can simply postulate such a coupling to begin with, without invoking string theory. I don’t quite see how Banks/Dine/Douglas argumentation can be transformed into a “no Ep violation is a prediction of string theory”. As far as I can see, it has nothing to do with string theory, but depends on QFT arguments alone.


    at least from what I read, I don’t understand how such a claim can be made. But I could easily miss something. However, as far as I know, the string dilaton is massless at first loops and its coupling is of the form of the one considered by Banks/Dine/Douglas (see (2.1) and (2.2) in [Damour and Polyakov]). Besides, Damour/Piazza/Veneziano say

    A striking prediction of all string theory models is the existence of a scalar partner of the spin 2 graviton: the dilaton φ, whose vacuum expectation value (VEV) determines the string coupling constant g_s=e^{φ/2} [1]. At tree level, the dilaton is massless and has gravitational-strength couplings to matter which violate the equivalence principle [2]. This is in violent conflict with present experimental tests of general relativity. It is generally assumed that this conflict is avoided because, after supersymmetry breaking, the dilaton might acquire a (large enough) mass.

    That would be nice to have Douglas explaining from where the claim that “EP satisfaction” is a prediction of string theory is coming from because, as far as I understand, the reference Douglas/Kachru use in their paper doesn’t seem to say so. But I am certainly not the best person to talk about that.

  40. vmarko says:


    “By QFT, I mean a set of tools one applies on actual theories originally defined classically in order to get their quantum behavior.”

    Correct. And if your classical theory contains gravity (say GR), the QFT formalism breaks down (loses predictive power) since the theory is nonrenormalizable.

    “scalar times Lagrangian of material field (=\Phi L_m)”

    In the context of string theory, the “scalar” is the dilaton, which (one could argue) is a part of gravitational sector. The above coupling then violates the EP, unless one breaks supersymmetry high enough etc.

    Outside of the context of string theory, there is absolutely no justification to claim that the “scalar” is part of the gravitational sector. From the point of view of QFT, it is just another matter field, and the above coupling has nothing to do with EP. You could also argue, for example, that this scalar is the piece of the Dicke-Brans-Jordan gravity (the scalar-tensor theory), but in that case there should also be the curvature term in the theory, again rendering it nonrenormalizable. Consequently, QFT breaks down and can make no statements about EP.

    Don’t get me wrong — I agree that the formalism of QFT is made use of in discussing the EP violation. But this can be done only if that QFT is considered to be an effective low-energy limit of string theory. The argumentation about EP violation simply does not hold outside the context of string theory, because you don’t know if that scalar field is part of “gravity” or “matter”.

    HTH, :-)

  41. Marko,

    The low-energy effective action of string theory is a scalar-tensor theory (Brans-Dicke-like) with a non-minimal scalar-matter coupling (see (2.1) in Gasperini/Piazza/Veneziano. Therefore, I don’t quite understand when you seem to argue that with a scalar-tensor theory with non-minimal scalar-matter coupling, one cannot make statements about EP; while one could(?) with string theory. Please specify what you have in mind.

    Otherwise, you don’t need string theory to postulate this kind of coupling in an otherwise usual scalar-tensor theory. I.e. you don’t need string theory to postulate the action (2.1) of Gasperini/Piazza/Veneziano. Other principles than the string theory ones can lead to such an action.

  42. vmarko says:


    “Therefore, I don’t quite understand when you seem to argue that with a scalar-tensor theory with non-minimal scalar-matter coupling, one cannot make statements about EP; while one could(?) with string theory.”

    Ok, suppose that you have the scalar-tensor theory with nonminimal coupling to matter fields as your classical theory, without any mention of string theory. If you try to quantize this classical action (which you must if it is to be fundamental), you will find that it is nonrenormalizable — you cannot eliminate divergences consistently, and thus the corresponding QFT is not well-defined. Consequently, that QFT does not have any predictions at all, and in particular no predictions regarding the EP.

    Now consider that same action as a low-energy effective action of the string theory. The effective action (by its nature) is not something that needs to be quantized, and consequently there are no renormalization issues. This is because string theory has some form of UV completion (which makes it perturbatively finite), and because your effective action is valid only at low energies — at higher scales it will receive nontrivial correction terms from string-UV-completion, and these terms are supposed to cure any and all divergences. But at low energies those corrections are supposed to be small, and the effective action should be approximately correct. Consequently, you can analyse its form and find out that EP is violated, unless the dilaton gets a large mass from supersymmetry breaking (in which case EP is not violated), etc.

    It goes without saying that the string theory has its own share of problems regarding the prediction of EP violation. Peter’s argument is basically that due to all those problems (essentially lack of uniqueness of the formulation of M-theory), string theory also doesn’t make any predictions regarding the status of EP, but for different reasons and despite the knowledge of the low-energy effective action.

    So it’s the same action, but in two different contexts — one where you don’t have a well-defined UV completion (and consequently you don’t have any theory to speak of), and one where string theory does attempt to give you a UV completion (and if successful, you consequently do have a well-defined effective theory, which can be studied).

    The general moral of the story: QG =/= QFT. ;-)

    HTH, :-)

  43. Marko,

    are you saying that the somehow very general argumentation in Banks/Dine/Douglas paper is actually restricted to string theory because the discussion wouldn’t make sense otherwise? Am I correct to assume that it is what you are saying?

    Also, as far as I can tell, Banks/Dine/Douglas paper is about the violation of the Eintein equivalence principle, while you seem to talk about the weak equivalence principle. Indeed, they are interested in the variation of the fine structure constant only. But a variation of the fine structure constant is a violation of the Einstein (and the strong) equivalence principle — and in that case, the scalar-field (φ) that couples to the material sector (eg. φ F^2) is not necessarily part of the gravitational sector: it could be any moduli field. On the contrary, if the scalar-field is part of the gravitational sector as in your exemple, and as it is for the dilaton for instance, then indeed it would lead to a violation of the weak equivalence principle in addition to the strong and Einstein EPs. But Banks/Dine/Douglas don’t seem to consider this situation in particular.

  44. vmarko says:


    “are you saying that the somehow very general argumentation in Banks/Dine/Douglas paper is actually restricted to string theory because the discussion wouldn’t make sense otherwise?”

    Essentially yes, that’s right, with a caveat that their argumentation could be also restricted to some other, non-string theory of QG. So yes, outside of a context of some theory of QG, their argument wouldn’t make sense. Moreover, the authors actually say that themselves in the first sentence of the second paragraph of their paper:

    “Within both the contexts of effective field theory and M-theory, it is natural to
    model the change of the fine structure constant by coupling a dynamical scalar field φ to the photon kinetic term in the low energy effective action.”

    Note the phrases “M-theory”, “effective field theory” and “low energy effective action”. Without M-theory (or some other theory of QG), the latter two phrases don’t actually have any meaning. One can discuss them only in a QG context, never in the QFT context alone. As I said earlier, QG has proven to be extremely hard (if at all possible) to construct within just the QFT framework.

    “Banks/Dine/Douglas paper is about the violation of the Eintein equivalence principle, while you seem to talk about the weak equivalence principle.”

    The various flavors and formulations of the EP basically depend on how you make a split of all relevant fields into “gravity” and “matter” sectors. One version of the (strong) EP which is particularly suitable says that the coupling between gravity and matter is such that when one writes down the laws of matter fields in a locally-inertial coordinate frame, they reduce to the form that they would have if the gravity sector was “turned off” completely (i.e. laws of physics would then take the so-called “special-relativistic” form). The enforcement of the EP dictates the “minimal coupling” between gravity and matter.

    So if the scalar field in the BDD paper is considered to be a “matter” field, then their analysis says absolutely nothing about the EP (any flavor). If, OTOH, the scalar is considered to be a part of “gravity”, then its coupling to EM-field clearly violates the EP. But considering the scalar field to be part of gravity can be done only in the context of some QG theory, never in QFT alone, because no QFT can describe QG.

    Note also that the BDD paper is all about estimating a cutoff scale to fix the upper limit on the variation of the fine structure constant. The very presence of a cutoff scale is an indication of the fact that they work in some (unspecified) QG context, where above that cutoff QFT fails to be a valid description of nature. IOW, the violation of EP (as embodied in the variation of the fine structure constant) depends on the scale where QFT doesn’t work anymore. This is precisely inside the domain of QG and outside the domain of QFT. That is just the same statement formulated in different language — QFT alone cannot say anything about the EP.

    Btw, we are getting sort of off-topic here, and it seems that I am repeating the same statement in four posts already…

    Best, :-)

  45. Marko,

    “with a caveat that their argumentation could be also restricted to some other, non-string theory of QG.”

    Well, the caveat here changes everything on whether EP is a string theory prediction, as claimed in Douglas/Kachru (mentioning Banks/Dine/Douglas), or a generic prediction that comes from QFT arguments only — which was my original point — not mentioning that it is actually not a prediction but rather an expectation.

    Of course the argumentation is restricted to a theory that might have a chance to be proven well defined down to the quantum level. The point if that, as far as I know, we currently do not have the nonperturbative tools in order to (even simply) know whether or not a theory might be correct down to the quantum level.

    “So if the scalar field in the BDD paper is considered to be a “matter” field, then their analysis says absolutely nothing about the EP (any flavor).”

    As far as I can tell, this is not correct. If the fine structure “constant” is a function of any non-gravitational scalar field, then it violates the local position invariance and therefore Einstein EP. Indeed, then the outcome of a local non-gravitational experiment in a freely falling laboratory is NOT independent of its location in spacetime (see Will’s Living review). Anyways, Banks/Dine/Douglas actually don’t even mention the equivalence principle but only argue about the possibility of a variation of the fine structure constant :P

    So the question here really is: does Banks/Dine/Douglas paper show that string theory in particular predicts the invariance of the fine structure constant? or rather that QFT seems to put a strong constraint on string theory (among other theories)?

    Otherwise, I don’t think we are off-topic here. But maybe Peter would indeed prefer if we were continuing our discussion privately. Peter?

    Best :)


  46. Peter Woit says:

    I don’t think you’re off-topic, but I do think this has gotten to the point where it would be better if the two of you continue the discussion privately. If either of you needs help figuring out how to contact the other, let me know.