Gravitons in Loop Quantum Gravity

  1. How do you understand about Gravitons in Loop Quantum Gravity? All I know about it is from what I heard that "LQG hopes that its predictions for experiments occuring far below the Planck scale will be almost identical to that of gravitons on flat spacetime".

    So do you consider it a pure graviton at all or some kind of pseudo-graviton (in LQG)?

    I just found a paper about gravitons in LQG but may not understand it 100%

    Do you think it's correct? Please share other papers about gravitons in LQG if there is. Thanks.
  2. jcsd
  3. For example someday the LHC were able to detect gravitons. Can we tell whether it is a String graviton or the pseudo-graviton of Loop Quantum Gravity? What are the differences? Maybe the latter involves some kind of geometric based graviton or something?
  4. atyy

    atyy 10,098
    Science Advisor

    Here's a more recent paper that references the one in the OP.
    Euclidean three-point function in loop and perturbative gravity
    Carlo Rovelli, Mingyi Zhang
    "In particular, the low-energy limit of the two-point function (the “graviton propagator”) obtained in this way from the improved-Barrett-Crane spin foam dynamics [7–12] (sometime denoted the EPRL/FK model) correctly matches the graviton propagator of pure gravity in a transverse radial gauge (harmonic gauge) [13, 14]. .......... The obvious next step is to compute the three-point function. In this paper we begin the three-point function analysis. We compute the three-point function from the non-perturbative theory."

    BTW, Rovelli and Zhang's calculations are based on a different spin foam proposal from that used in Bianchi et al's paper referenced in the OP. Bianchi et al base their calculations on the Barrett-Crane proposal, which is now thought to be wrong, because it did not match the graviton propagator. The Rovelli and Zhang paper is based on the EPRL/FK/KKL proposal which is reviewed in Rovelli's Zakopane lectures.
    Last edited: Feb 28, 2012
  5. What's odd in the two papers above is they mentioned about "Graviton propagator" instead of just "Graviton". Why is that. In your own understanding, why do they call it "Graviton Propagator"? Why "Propagator"?
  6. atyy

    atyy 10,098
    Science Advisor

    In quantum theory, a propagator contains information about the probability that a particle will take a certain path.
  7. marcus

    marcus 24,316
    Science Advisor
    Gold Member
    2014 Award
    LQG propagator from the new spin foams
    Eugenio Bianchi, Elena Magliaro, Claudio Perini
    (Submitted on 25 May 2009)
    We compute metric correlations in loop quantum gravity with the dynamics defined by the new spin foam models. The analysis is done at the lowest order in a vertex expansion and at the leading order in a large spin expansion. The result is compared to the graviton propagator of perturbative quantum gravity.
    Comments: 28 pages
    Lorentzian spinfoam propagator
    Eugenio Bianchi, You Ding
    (Submitted on 29 Sep 2011)
    The two-point correlation function is calculated in the Lorentzian EPRL spinfoam model, and shown to match with the one in Regge calculus in a proper limit: large boundary spins, and small Barbero-Immirzi parameter, keeping the size of the quantum geometry finite and fixed. Compared to the Euclidean case, the definition of a Lorentzian boundary state involves a new feature: the notion of past- and future-pointing intertwiners. The semiclassical correlation function is obtained for a time-oriented semiclassical boundary state.
    Comments: 13 pages
    In this limit, the two-point function we obtain exactly matches the one obtained from Lorentzian Regge calculus [38]. We therefore extend to Lorentzian signature the results of [13].​
    [13] E. Bianchi, E. Magliaro, and C. Perini, Nucl. Phys.
    B822, 245 (2009), arXiv:0905.4082 [gr-qc].

    So Bianchi Ding is a sequel to Bianchi Magilaro Perini. It completes that initiative by doing the earlier analysis in the Lorentzian setting
    Last edited: Feb 28, 2012
  8. About LQG vs String Theory. Let me share an answer to a question I asked 2 weeks ago. I was asking if loop quantum gravity was also about spin-2 on flat spacetime. The answer is no, because LQG starts from curved spacetime and is simply a canonical quantization thing and the curvature is primary. While String Theory is actually spin-2 on flat spacetime with the spacetime curvature as secondary effect. Although one can say LQG can be modelled as spin-2 on flat spacetime by taking the GR part in LQG and doing it. But for full fledged LQG. It has spacetime curvature a priori and there is no flat spacetime underneath it. I think everybody agrees now and I assume there are no objections from anyone.
    David Horgan likes this.
  9. atyy

    atyy 10,098
    Science Advisor

    Although perturbative string theory has flat spacetime (or some other vacuum solution of the Einstein equation) as background, in some versions of non-perturbative string theory such as AdS/CFT the bulk space has no background and is completely free to fluctuate, so it is similar to LQG in this respect. In fact it is more radical, since LQG assumes space as a fundamental entity that fluctuates, whereas the bulk space is not a fundamental entity in AdS/CFT.
    Last edited: Feb 29, 2012
  10. tom.stoer

    tom.stoer 5,489
    Science Advisor

    In string theory and in most treatments of QFTs one starts with quantized excitations on top of a classically fixed background. The excitations are the quanta of the associated fields (photons, gravitons, ...). This approach has some limitations and LQG tries to get rid of them.

    LQG never introduces a background and excitations living on this background, so LQG does not use gravitons as building blocks. Instead one expects that one may recover a kind of semiclassical limit or weak field limit where something like "gravitons" will show up again.

    So in contrast to any other QFT where the "...ons" are the fundamental (mathematical and physical) entities in LQG the gravitons are not fundamental but only to be considered in a certain limited approximation.
    Last edited: Feb 29, 2012
    John M. Carr likes this.
  11. marcus

    marcus 24,316
    Science Advisor
    Gold Member
    2014 Award

    This is an excellent concise statement! Offhand I can't think of any place where these basic facts have been expressed more clearly.

    Just to amplify, unnecessarily I think, any approach which starts with a classical fixed geometry and lays quantized excitations on top of it is called "perturbative". The primary aim of LQG, and so to speak its "claim to fame", is to strive for a "non-perturbative" quantum geometry.

    In the loop approach there is no mathematical "thing" representing space or space time. There are only relationships among measurements. These are the quantum states of geometry (a state is a web of interrelated geometric measurements). It's a minimalist approach, in a sense, and makes for challenging mathematical problems.
    Last edited: Feb 29, 2012
  12. tom.stoer

    tom.stoer 5,489
    Science Advisor

  13. Well I believe the story is roughly like follows. Strings, in the usual worldsheet formulation, assume some fixed background. Eg for the flat empty space, take this to be the Minkowski space described by a metric eta_mn, plus small fluctuations delta around it:

    g_mn = eta_mn+ delta_mn

    Essentially these delta describe gravitational waves whose quanta are gravitons. Other, curved backgrounds g_mn are equally possible, like eg. black holes, and one can expand around them analogously.

    In LQG the “expansion point” is more like g_mn=0, so no spacetime is there. In order to recover GR as we know it, one needs to specify a backgound around which one wants to expand, essentially by putting some non-zero g_mn in by hand. Only then one can try to see what a graviton propagator etc looks like, namely by expanding around this ad-hoc background.

    It is (for me) an open question what the admissible choices are, certainly flat space should be an allowed possibility. AFAIK is has not yet been proven that flat space is a solution to LQG at all. As often said, showing that that graviton propagator comes out right it is a necessary, but by no means a sufficient condition, as the 2-point function just captures the free theory.

    Some ppl seem to claim the LQG ought to be background independent and thus be able to “dynamically decide itself” what the backgound is supposed to be; but how can this ever be possible without additional input. Namely in particular our whole universe should be an allowed solution, including the gravitational fields of us sitting in front of our computers, residing on earth orbiting the sun; etc. Roughly the whole solution space of GR (plus whatever is necessary to make the theory quantum mechanically consistent) must be allowed vacua of any theory of gravity incl LQG.

    The theory can’t know by itself what solution to choose, so it must be told by specifying boundary conditions or a boundary state in LQG. So roughly the necessary, fixed choice of g_mn in string theory is replaced in LQG by a choice of boundary conditions that induce the desired background g_mn in the low-energy limit. But what are then the rules that determine or constrain the admissible boundary conditions?

    This is the landscape “problem” in disguise, since one has to specfify as extra data what semiclassical long distance limit g_mn one wants to talk about. This landscape “problem” is actually not a problem and never was. The multitude of possible solutions must be a property of any theory of gravity. So the question is what LQG can possibly add here. One thing LQG might be able to do in principle at some point (and which is not possible in the standard world-sheet formulation of string theory), is to compute transition amplitudes between certain such boundary conditions. However, not all boundary conditions, or space-times g_mn seem to be allowed, which restricts the usefulness of this kind of ideas. See the review by Rozali on background independence for further details.
  14. Are you saying the AdS/CFT is an alternative formulation of quantum gravity other than string theory and LQG, but AdS/CFT has anti-desitter spacetime with negative curvature which doesn't describe our universe at all. Unless you are saying that perhaps a future version of Ads/CFT with the right positive curvature can describe our universe?
  15. atyy

    atyy 10,098
    Science Advisor

    AdS/CFT is non-perturbative string theory. I don't know if a future version of something like AdS/CFT can describe our universe, but I hope there will be.
  16. I know AsD/CFT is non-perturbative string theory. But it describes negative curvature which doesn't describe our universe at all. So you are saying that if in the future we can't find any AsD/CFT that doesn't describe our universe, then the dual idea is refuted. Also an actual AsD/CFT would be sD/CFT but then the analogy or duality no longer holds because the antidesitter thing is done. So why is there high hopes reserved for the AsD/CFT programme?
  17. atyy

    atyy 10,098
    Science Advisor

    Well, you study what you have, and learn from it. The first relativistic theory of gravity was Nordstrom's. Einstein studied Nordstrom's scalar gravity on flat spacetime, reformulated it as a curved spacetime equation, and formulated the correct tensor theory. Similarly, AdS/CFT is the first complete theory of quantum gravity, so maybe by studying it, we can learn what the correct theory is.
  18. You are kinda saying that in the end, our universe may be described by something like "Michael Talbot's The Holographic Universe?

    It's bizarre but this seems to be what our quantum gravity programme is pointing to... because if we can find the right version of AsD/CFT, it would be "The Holographic Universe" as described Talbot.
  19. tom.stoer

    tom.stoer 5,489
    Science Advisor

    I think you are right to approx. 99%, so it's unnecessary to comment on these 99%, and it's better to avoid any comments on the 1% in order not to confuse the reader with minor details.
  20. We have been focusing on the idea of spin-2 fields plus flat spacetime = curved spacetime and applying this even to String theory. But have we forgotten that the compactified dimensions in String Theory are really 6 dimensions? Unless you are saying the 6 dimensions are flat? How does one resolve this?

  21. tom.stoer

    tom.stoer 5,489
    Science Advisor

    In the standard formulation in ST one introduces a spacetime metric. Some decades ago this was typically "4-dim. Minkowski spacetime" * "6-dim. compactified Calabi-Yau space"; in the meantime other geometries have been discovered and are studied extensively.

    From ST one can derive a consistency condition for the spacetime on which strings are propagating. This consistency conditions requires Ricci-flatness and forbids arbitrary spacetimes and arbitrary compactified dimensions; Minkowski spacetime * Calabi-Yau is a typical solution, but as I said, others are possible, and in the meantime string theorists were able to relax these conditions.

    The problem I see with string theory is that "spin-2 fields plus flat spacetime = curved spacetime" does not really work; you have to chose are curved spacetime in the very beginning and study propagation of strings as "weak distortions" on top of it". But these strings do never change the whole spacetime dynamically, it always stays in some fixed subsector which does not change dynamically. This holds afaik for other approaches (e.g. branes, fluxes, ...) as well.

    This is what is called background dependence and is basically due to the approach "fix a background and then quantize small distortions". LQG tries to get rid of this problem and avoids to fix a classical background. But as suprised explained, one still has to introduce some kind of boundary condition or background when doing physics; it's not required for the definition of the theory (and that's a major step forward), but it's required for detailed calculations (e.g. when studying black holes in LQG one has to define an isolated horizon classically; a full dynamical setup w/o any input like boundary conditions or background is not possible).
Know someone interested in this topic? Share a link to this question via email, Google+, Twitter, or Facebook

Have something to add?