Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

Life without the graviton

  1. Nov 15, 2007 #1
    From the Not Even Wrong thread on Garrett Lisi's E8 theory:

    I had some questions in response to this; Woit complained (quite correctly) that my post didn't have anything to do with E8 or Woit's blog and requested the discussion not continue there. So I'm posting it here instead. My response to Marcus is:

    So while I personally find the no graviton approach attractive on several levels, it kind of seems like rejecting the graviton should be considered a dramatic step to take. Not bad-- just dramatic, so it should maybe be done carefully. Do there already exist any other theories of quantum gravity, besides Loll’s, which eschew the graviton or take the “graviton as approximation” approach you describe? Which ones?

    Does LQG, for example, have a graviton? Looking I am finding references to a “graviton propagator” in LQG but it is not immediately obvious whether that’s the same thing.


    Aside from this I find Loll’s argument against the graviton in the paper Marcus links somewhat unconclusive by itself. “Well, we’ve been trying to get useful answers out of this construct for decades and haven’t succeeded, so it’s a good bet we’re doing something wrong” sounds like good strategy to me, but Loll doesn’t seem to actually be putting forth an argument about reality there, only an argument about “how to proceed”. I don’t see any reason that just because we can’t describe the graviton perturbatively, that would mean it doesn’t exist– since, as far as I understand, perturbation theory is supposed to just be an approximation anyway. (And this is of course assuming that perturbatively modeling the graviton is actually impossible, and not just too hard for anyone to manage right now!) Am I missing something about Loll’s argument?
  2. jcsd
  3. Nov 15, 2007 #2


    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    Well all the various approaches that get called LQG---which is quite a bunch. In the GRAVITON PROPAGATOR papers, the graviton that Rovelli gets is in a special approximation that he forces on the theory, forcing it to approximate flat Minkowski space, JUST FOR THE PURPOSE of showing that the theory can act like one expects in that approximation. It is part of showing that the theory has the correct low-energy low-curvature limit.

    The graviton is a mathematical object that arises in approaches called PERTURBATIVE where the metric g is the sum of a fixed metric g' and a small ripple disturbance or perturbation h.
    So g = g' + h
    and the fixed metric is usually the flat (zero gravity, zero matter) solution to Einstein equations called Minkowski. Then there is a series expansion. The perturbative approach does not work where the field is strong, dynamic, lots of curvature---then it cannot be expressed as a static field with a small disturbance added to it. So then gravitons do not arise in the mathematics.

    I have to go, back later

    You may have missed the point. Loll's paper is not about the non-existence of gravitons. that is generally taken for granted in the (non-string) QG community and has been as far back as I have looked. The reason I quoted her was she expressed it succinctly and a side comment in her introduction. I like her writing style.

    You wouldn't pretend that gravitons are real and use them to describe the early universe or the actual collapse of a star to form black hole because those situations don't have an artificial fixed flat background. Graviton is just an effective mathematical construct that is convenient in a limited field theory context, not something that anybody in the community thinks real, as far as I know, or has ever.

    I read Loll's paper as a proud report of progress in her Triangulations approach. It does not revolve around the graviton issue at all!--that is just a side comment which she delivers crisply.

    If interested in QG, I would advise forgetting about gravitons for the moment and reading the introduction section of Loll's paper carefully. It is about the emergence of quantum spacetime from fundamental microscopic dynamical degrees of freedom----the emergence not of one single classical spacetime but a whole slew of them: a pathintegral full of spacetimes. Exploring observables like volume correlation, and dimensionality. I think it is interesting. Don't read it unless you do too. If you like it, it can be a taste of what non-string QG is about.

    There are others one could sample or start with instead, but you've already downloaded Loll's latest paper, so that's one way to go.
    Last edited: Nov 15, 2007
  4. Nov 15, 2007 #3
    I'm (obviously) not Marcus, but I feel compelled to intervene here because I firmly disbelieve in the graviton.

    Have you looked up group field theoretic quantum gravity? It begins with one of its premises being there are no gravitons.

    It was previously discussed on this forum as a matter of fact.

    One of the compelling arguments against perturbative approaches that Thiemann gives in his recent book Modern Canonical Quantum General Relativity (seriously Thiemann, commission is in order for how much I'm plugging your book!):

    --emphasis added

    From pages 7-8 Modern Canonical Quantum General Relativity by Thomas Thiemann (2007).

    The two citations he cites are (from the bibliography):

    [38] N. Marcus and A. Sagnotti. The ultraviolet behavior of N=4 Yang-Mills and the power counting of extended superspace. Nucl. Phys. B256 (1985) 77.
    [39] M. H. Goroff and A. Sagnotti. Quantum Gravity at Two Loops. Phys. Rev. B160 (1985) 81.
  5. Nov 15, 2007 #4


    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    Thanks Angry, you put it like a physicist. I was just saying words.:smile:

    Over at Woit's blog, AGeek put the matter well:
    And it has been known for several decades that gravity can not be described perturbatively because (in perturbative context) it is not renormalizable.
    Last edited: Nov 16, 2007
  6. Nov 16, 2007 #5
    Hi Marcus and AngryPhysicist, you both seem to be basically saying the same thing, which is that the graviton is actually just an artifact of the perturbative approach to quantum gravity and therefore if you can't perturbatively describe the graviton then it DOESN'T exist. AGeek at NEW responded with the same thing:


    So what youall are saying does seem to make sense, I think, although I guess I probably ought to read up on some of this stuff more :) but, in the meantime, there is one thing I am suddenly very confused about.

    My (very vague) understanding is that string theory's claim to be a theory of quantum gravity is based around the presence in the string spectrum of a spin 2 massless boson; we for some reason that I don't quite understand get to assume that the spin 2 massless bosons are all gravitons, and therefore since string theory contains gravitons it is a theory of quantum gravity. But if the graviton is something which does not describe all situations in quantum gravity-- as you all say, it can't be used in a nonperturbative theory (as I understand things string theory is currently used as a perturbative theory, but I have seen discussion implying there was at least one time hope of formulating a nonperturbative version), or when the fluctuations in the background metrics would become "large"-- if this is the case then how can string theory be a theory of quantum gravity, if its notion of quantum gravity is based on gravitons? Am I missing something here? Or does String Theory have some way of handling the large-curvature/large-fluctuation cases, besides the literal usage of spin 2 massless bosons?
  7. Nov 16, 2007 #6
    Gravitons? Uh?

    Well, I´ll do a question for stringy people here. What is the S-dual of the graviton mode of a fundamental string?

    I thnk that the answer could be interesting to understand what string theory has to say about gravity.
  8. Nov 16, 2007 #7


    User Avatar
    Gold Member

    Disbelieving the graviton?

    If you disbelieve the graviton, do you disbelieve in gravitational waves? Or do you believe in gravitational waves and believe that gravitational waves ignore Planck's constant?

    If you believe in gravitational waves and Planck's constant, you are nearly forced into believing in the graviton. Of course, Nobel prize winner Willis Lamb famously disbelieved in the photon.

    However, my understanding is that graviton disbelievers are a very small minority, even among the LQG crowd. Correct me if I'm wrong.

    Instead, I think that some (many?) of them believe the graviton is not elementary, but built up from loops, ribbons, triangulations, (strings), or something else.

    I would be very interested to read a short summary of the position of each of the major groups on the graviton. TIA to anyone who can provide it.

    Best to all.
    Jim Graber
  9. Nov 16, 2007 #8
    I have to wonder if such small amplitude gravity waves can even be detectable. Wouldn't the measuring device have to be larger than the wave itself. And don't small perturbations from flat spacetime mean that those gravity waves would be very long and thus very difficult to detect?
  10. Nov 16, 2007 #9
    Maybe it is obvious, but I already asked myself, how the expansion of the whole universe fits into the spin 2 particle picture of gravity. I guess, the simple answer is, it doesn't.
  11. Nov 17, 2007 #10
    Personally I think this is relatively narrow position (no offense)..."it's either gravitons or else you don't adequately believe in Planck's constant and gravitational waves" is too much of a false dilemma in my opinion.

    What about the geon concept proposed by Wheeler? This is based on the principle of gravitational waves, and quantization presents itself rather readily...all the while, we don't really "need" the graviton.

    Or, as I mentioned earlier I believe, the group field theoretic quantization of gravity? There is Planck's constant involved there and it doesn't contradict gravitational waves.

    Gravity is really a different sort of force since we're dealing with spacetime directly rather than with some force that can be viewed as happening within spacetime. That is where quantum theory sort of...runs into trouble.

    If life were as easy as gravitons, then we'd have had a quantum theory of gravity way back in the 1930s. Fortunately life is interesting, so it's been over 70 years of active research and still we don't have a clue. It's not that we've failed, we've merely succeeded at finding ways that don't work.

    (And by the by, for those that are wondering why I answered four minutes after Marcus, I originally began before him but got distracted by the mail :wink: )
  12. Nov 18, 2007 #11


    User Avatar
    Gold Member

    Hi, Angryphysicist, from a Happy former math student,
    I have several responses, (not necessarily answers).

    Response Number 1:
    Of course, it’s logically possible to believe in the photon and not the graviton, and some physicists have done so, and some still do. Just before the discovery (invention?), of string theory, when many people were despairing of the possibility of a mathematically consistent quantum theory of gravity, this was a more popular idea. But I think it has always been a minority position. I would like your view, and also Marcus’ view on which modern authors uphold this position.

    You seem to suggest Oriti (group field theory quantum gravity) and Marcus quotes Loll.

    However, the quote that Marcus posted merely states that the graviton is not fundamental, not that it doesn’t exist. We used to think the proton was fundamental, now we believe in quarks. But people who believe in quarks mostly don’t believe that the proton doesn’t exist, only that it is not fundamental. The same can be true for gravitons.

    I have scanned some of Oriti’s recent papers on the ArXiv. I didn’t spot any references to gravitons, but there were tons of references to Feynman diagrams and the Feynman propagator for gravitation. The Feynman propagator for gravitation, (actually technically the exchange particle), is the graviton. What makes you think Oriti does not believe in gravitons?

    Of course you can believe in virtual gravitons without believing in real gravitons, but that seems to me to be a rather contorted worldview.

    Response Number 2
    This response is based on the argument that a graviton is necessarily defined as a perturbation on a linear background, and that reality is at least nonlinear, if not background free, thus discrediting the idea of the graviton. Even accepting all the hypotheses in the above argument, some of which are at least arguable if not dubious, it remains true that linearity of spacetime is a very good approximation. Even if the graviton does not exist, it is an excellent approximation. This goes back to “ancient” work by Feynman, Weinberg, Boulware, Deser and others.

    Following this idea, the only places I know in the “real world” where linearity of spacetime is not a good approximation are at the big bang and at the formation/evaporation of a small black hole. In these environments nonlinear spacetime effects can be large. Do you think an evaporating black hole emits Hawking radiation including photons, but not gravitons? (I should probably post that as a separate question over at quantum physics.)

    Enough for now. I may post a third response later.

    Best to all.
    Jim Graber
  13. Nov 18, 2007 #12


    User Avatar
    Science Advisor

    The graviton is really just a simple consequence of using the Einstein Hilbert action and the rules of quantization. So it exists in so far as those things are physical, and whatever theory you output should in principle reduce to those dynamics in the suitable limits.

    By exist of course, we mean in the same way that say a photon or W boson 'exists' (whatever that means). Keep in mind the description of these thigns change as you improve the perturbation series. For instance the form factors of a photon interaction can and does receive nonperturbative corrections in QED.

    Likewise gravitons recieve nonperturbative corrections, and much of the whold business of quantum gravity is trying to figure out what that is. No one is deluded into thinking local weak field limits are the whole story, and indeed they are most assuredly not.

    But yea, it doesnt make much sense to *not* believe in a graviton, and still believe in general relativity and quantum mechanics. They're sorta the same by definition.
  14. Nov 18, 2007 #13
    Just a historical background, String theory was originally invented to solve the problem of Strong force. It turned out to be wrong, and they re-applied it to other problems (the foundation of particle physics as a theory of everything, for example).

    String theory was one of the approaches that was a consistent perturbative approach, but it wasn't the "only" consistent approach as you suggest.

    Other non-perturbative approaches are consistent.

    Marcus gives an argument from Loll, but I cite Oriti's work on group field theoretic quantizations of gravity as one concrete example of a theory of quantum gravity sans graviton.

    There are some serious theoretical problems with the graviton; there is no conclusive "proof" that it exists or it doesn't.

    But as scientists, we're not looking for proofs. We seek evidence.

    Um...I think that the proton is not fundamental, as we now believe it's made out of quarks. As for the photon being made out of quarks, this is news to me.

    Perhaps, if the graviton is massive and fermionic?

    Page 24 from his Loops07 Talk.

    Yes it is an excellent approximation in a weak gravitational field.

    As I quoted Thiemann pointing out, when we break the metric in two [tex]g = \eta + h[/tex] with [tex]g[/tex] being the metric, [tex]\eta[/tex] being the Minkowski metric, and [tex]h[/tex] being the deviation from the minkowski metric, we can expand the Lagrangian with [tex]h[/tex] terms.

    This works when the deviation is within the radius of convergence. Outside of the radius of convergence...like in a black hole, it is utterly useless!

    And places like in a black hole, or the big bang, where the deviation is outside the radius of convergence, the model is completely useless.

    So the only places that we don't know about classically, we can't know about quantum mechanically! That's not good!

    Well, Hawking radiation takes the form of gamma radiation. Unless gamma radiation changes radically due to the existence of gravitons, I don't believe that Hawking radiation includes gravitons.
  15. Nov 18, 2007 #14


    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    An excellent approximation in linear (weak field as Haelfix says) cases, which is most everywhere. I agree of course.

    I just think that it misleads people to give them the idea that gravitons exist like real things whizzing around and that is how you should understand how gravity works. They then imagine that is how it works inside the event horizon of a black hole...

    If someone naively believes gravitons are really how gravity works, not just an extremely useful approx when appropriately applied, then they can apply the idea inapproapriately, as i've seen happen.

    I'd say the graviton exists as a useful analytical tool not as one of God's creatures, and leave it at that.

    so I agree with Loll, but I don't see anything wrong with what you and Haelfix say if it is properly understood

    Jim, thanks for the careful response! I think the point is that outside the event horizon both graviton and photon are useful mathematical tools and I am happy to think with them (to the best of my ability). But inside the event horizon I do not know how to correctly apply the graviton concept to understand gravity. At some point, out of deference to your expertise, I have to take your word about it. I am hardly an authority. I just think gravitons only exist where it is mathematically valid to apply them

    Haelfix, thanks for your carefully worded statement. Properly understood, I think your position actually harmonizes with mine and is articulated in a highly knowledgeable way.

    This is very informative and reinforces for me that the graviton is a mathematical convention which changes description as needed according to the circumstances in which it is serving as an approximate analytical tool.

    I would be happy to believe that none of our fellow PF posters is deluded into thinking that local weak field limits are the whole story. I want them to NOT be deluded and that is why I quoted Loll.

    I disagree with the spirit of your final remark
    I would draw the opposite conclusion from what you just said. But for me GR and QM are not carved in stone. They are still alive and changing: QM is getting extended to cover spacetime dynamics and GR is getting quantized. Maybe I should imagine them as fixed and finalized and then I could accept the spirit of your final remark there.
  16. Nov 18, 2007 #15


    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    Ooops, sorry Angry. I didn't see you were already responding. You are much better qualified to engage in this discussion. Hope your cold is better. I will lay off here and let you prevail.
  17. Nov 18, 2007 #16


    User Avatar
    Science Advisor

    Theres really no great conceptual difference between the various gauge bosons, take the W boson vs the Graviton. The different quantum numbers, spin and mass will distinguish how they behave, but theoretically the jist is identical. Read they're simply quantum point particles that mediate exchanges and appear mathematically in perturbation series.

    These things also live in weak field approximations if you generalize to curved spacetimes. But there too there are ambiguities in description at the global level, which only nonperturbative methods can fully solve. Yet, surely no one disbelieves in the W, since we see them in accelerators.

    Now, if quantum mechanics suddenly stops working near the Planck scale, or if general relativity is wrong near the Planck scale then all bets are off. The graviton may not 'exist' (in the sense that the W 'exists') after all.
  18. Nov 18, 2007 #17

    So I ask above how it is that, if gravitons cannot represent gravity under some circumstances, that String Theory can still be a theory of quantum gravity when it does its gravity by gravitons.

    Is that the explanation, what you say here? I mean, is the idea that the graviton can't be fundamental, but that's okay because the graviton isn't fundamental in string theory, the string is fundamental? Or is something else at work here?
  19. Nov 19, 2007 #18


    User Avatar
    Gold Member

    Hi all,
    Real life intervened and I have not had time to continue this conversation.

    First of all for Coin:
    I believe you have come to the correct logical conclusion:
    Someone who disbelieved in the graviton, or
    who believed that it was only an aproximation,
    would be very likely to conclude that string theory
    is an incomplete or unsatisfactory or only approximate explanation of gravitation.

    And indeed most LQG folks seem to think string theory is unsatisfactory,
    Usually because it is background dependent, in their view.
    (There has been much controversy over this,
    and other people know much more about it than I do,
    so I’ll stop here.)

    Second for Angry,
    Thanks for the reference to page 24 of Oriti’s talk.
    I admit he’s pretty down on the graviton.

    Here is what he says:

    A QFT for Quantum Gravity?
    Immediate questions:
    A QFT on which spacetime?
    a QFT of gravitons on some background doesn’t work
    QG should explain origin and properties of spacetime itself (geometry
    and topology?)
    background independence!
    it can be only be a QFT on some auxiliary or internal space (local
    symmetry group is natural choice)

    a QFT of what? what are the fundamental quanta?
    gravitons do not work
    quanta of space itself! fundamental excitations of space around the
    vacuum (nothing)

    To repeat what I said last post,
    There is a big difference in thinking that the graviton is not fundamental,
    and thinking it does not exist.
    Now what about “gravitons do not work”?

    Not quite the same, but let’s count it as one for the
    Don’t believe in gravitons camp”

    On the other hand here are a few titles suggesting that
    some LQG folks do believe in gravitons or something similar.

    Towards the graviton from spinfoams: higher order corrections in the 3d toy model
    Etera R. Livine, Simone Speziale, Joshua L. Willis gr-qc/0605123 (2006) [Preprint]

    Group Integral Techniques for the Spinfoam Graviton Propagator
    Etera R. Livine, Simone Speziale gr-qc/0608131 (2006) [Preprint]

    Graviton propagator from background-independent quantum gravity
    Carlo Rovelli gr-qc/0508124 (August 2005)

    Towards the graviton from spinfoams: the 3d toy model
    Simone Speziale J. High Energy Phys. 2006 (2006) [Article]

    Finally, I want to ask:

    What do you believe if you don’t believe in the graviton?

    Two possible answers:

    Local moves on graphs (e.g. Fotini Markopoulou)

    An unquantized perfect fluid like gravitational field
    (or electromagnetic field?)

    I’ll post more about these two later, if I get a chance.

    Best to all,
    Bye for now.
    Jim Graber
  20. Nov 19, 2007 #19
    Haha we need to coordinate better is all! :tongue:

    Don't hesitate to give your two cents, I (more than likely) see things far differently than you do so you may have something to add that I overlooked.

    And my cold is getting better, thanks for your concern :smile:

    jimgraber: best of luck to you wading through all of this...
    I tend to keep things simple, and cut out what is unnecessary. It's occam's razor.

    Gravitons don't work, and we're going to perform excitations of spacetime directly? Then...why would we want to use gravitons? It's rather circular reasoning. If there were empirical evidence of gravitons, perhaps you'd have something; but we don't, so you don't either.

    I think the point has been pressed enough that there are serious theoretical problems with the graviton (works as a weak field approximation but not all weak field approximations are good approximations; model becomes totally useless when the fluctuations of spacetime are beyond the radius of convergence for the weak field; perhaps most pressing it tells us nothing new about what we already know, and it tells us nothing about what we don't know, e.g. internals of black holes or the big bang, it's beyond the radius of convergence for the weak field approximation; etc.). I honestly don't see a point rehashing these points again and again. So perhaps we should discuss the "necessity" of the existence of gravitons and the evidence for them.

    Until there is a legitimate argument for the existence of gravitons, we don't need to accept their existence. It's unnecessary to do so, by Occam's razor.

    But, no offense, you appear to be arguing the exact opposite...which is a fallacy, shifting the burden of proof. Like a religious fellow's argument that "God" exists because you can't prove "His" nonexistence!

    I'm still rather confused about your previous remark about electrons not being fundamental and being made out of quarks. This is news to me!

    But if the graviton were not fundamental, were to exist, then doesn't it logically follow that it would be made out of components that we should be able to detect? Why can't we detect the existence of a graviton then? Perhaps more pressingly why should a graviton not be fundamental if it exists?

    I'll be sure to give them all medals.

    It's of marginal interest to me that these people believe in gravitons, graviolis, gravitinos, or the flying spaghetti monster.

    I reiterate my point I previously made: is there any evidence of gravitons? "Give me direct evidence or be quiet!" Isn't that the motto of science? (If not, science has changed a lot since...yesterday...)

    There is evidence of gravitational waves, great, and the argument typically goes "Well, applying standard quantization techniques, you get the graviton."

    But the standard quantization techniques are dubious when applied to gravitational waves, as I previously stated. Quantization techniques deal with waves or fields in spacetime...now what happens when the field is spacetime?!

    The graviton doesn't make any sense anymore. I suppose if you really wanted to you could argue that a graviton "is" spacetime...but this makes no intuitive sense (spacetime is "exchanged" by particles attracting each other via gravitational force in a Feynman diagram? Could I create a "spacetime" ray-gun that attracts stuff to me then?).

    Personally, I think that quantum theory needs to be revisited. It appears to be an approximation to some underlying theory. Just my two cents...

    I know I say this a lot, but how narrow! Only these two options? There are no others?

    I personally believe that there is no graviton, true; I believe that spacetime comes in quantized "chunks" that are generated by the presence of energy/matter. Somewhat a direct quantization of general relativity guided by Einstein's views of his own theory.

    I reiterate that we can't naively apply quantization techniques to general relativity, it's not as simple as that. If it were, then it would have been quantized in the 1930s by people like Bronstein, and we'd be saying "Ah very interesting ... graviton ... blah blah blah ...".

    This sort of necessitates a reconsideration of quantum theory, or general relativity. One or the other...or both!

    I've made my choice with quantum theory first, maybe general relativity needs to be modified too later on because of this. I don't honestly know, hopefully we'll find out in due time.

    But the equivalence principle really allows us to have gravity expressed as the geometry of spacetime; and it's empirically, well, not correct, but it's yet to be falsified.

    So even if you change general relativity (somehow), you would run into the same exact problem. It seems a priori that quantum theory should be the prime candidate for inspection.
  21. Nov 19, 2007 #20


    User Avatar
    Gold Member

    Let's look at the observational evidence so see if theory can approximate what we see. We have theory that tells us that mass is conferred to matter by its interaction with the Higgs field, and that matter's gravitational attraction to other matter is mediated through its interaction with a gravitational field. Is there a problem with this view? I see one. Fields can be polarized/densified/rarified, and they can evolve over time and with location. Everywhere and everywhen we look in our Universe, gravitational behaviors appear to be consistent. If this is to be explained, we must posit that the Higgs field and the gravitational field are exquisitely congruent over all space and time, and this is a coincidence that is unexplained. The observations of gravitational behavior can be most reasonably explained by the explanation that mass, gravitational effects, and inertial effects arise from matter's interaction with a single field. No separate Higgs field and gravitational field, and therefore no mediating particles for these fields. The Higgs boson and the graviton are ruled out by observation unless someone can explain the extraordinary congruence of the the fields that require these mediating particles.
    Last edited: Nov 19, 2007
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook