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

What is the behavior of gravitons

  1. May 23, 2003 #1
    i think i asked a question here a long time ago and nobody fully answered it so here goes:
    what is the behavior of gravitons. i know they are virtual particles and cannot be observed, but is there a way to manipulate them such as we can with charged particles? also, have we any observations (or speculations) on an antiparticle for a graviton and if so does it repell? cause that would just be cool...
    also, what exactly happens when objects are gravitationally attracted. i believe it is the exchange of gravitrons but if so, a)do they therefore travel faster that a photon, b) does the "production" of a graviton reduce an objects mass or alter its energy level like it would during a chemical exchange of electrons?
    i suppose a graviton can't have any mass or else we'de had a new particle creating gravity for them!
    ohhh, and also, what relationship does the physical properties of a graviton have in the space-time distortion=gravity theory?
    Last edited by a moderator: Feb 5, 2013
  2. jcsd
  3. May 23, 2003 #2


    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    who says gravitons exist?

    I dont know if there is room for gravitons in what I suspect is the
    most promising theory of gravity.
    Ive been reading Smolin's survey comparing Loop Q. Gravity with
    Strings as two competing attempts to build a quantum theory of
    Smolin accepts gravitons as part of the "semiclassical approximation" but they do not seem to be fundamental to LQG.

    Personally dont want to try answering questions about them
    if they are some imperfect idea in some temporary kludge
    attempt to make quantum gravity in the context of QFT---something that does not fit General Relativity because the
    spacetime of QFT is fixed, rigid, undynamic. So graviton may
    well just be part of an approximation that will get downgraded
    in the course of time.

    Other people may think differently, and may believe gravitons are very interesting and valuable ideas, which they can tell you about.
    But I would advise you to look at Smolin's survey of the current situation in quantum gravity


    you can download a few pages in pdf and check it out.
    Im looking for people who have read this paper and would be
    willing to discuss it with me.
    Last edited: May 23, 2003
  4. May 23, 2003 #3
    Gravitron? is that something you seen while watching the didney film Tron?
  5. May 23, 2003 #4


    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    gravitons are in Smolin survey paper

    maximus, I overlooked some stuff on first reading
    and have been going back over Smolin's survey
    of loop QG more carefully and see that he does
    have room for gravitons in the discussion

    the theory isnt built out of them, as presented in
    but they can occur as "excitations of the theory"

    (point 5. on page 22)

    Another thing I just notice which astonishes me is
    that the dark energy density is quantized in integer steps.

    I recently calculated it in natural units (which these theories
    are based on) and found that it was 1.3E-123

    Smolin says that the reciprocal times 6pi is an integer.

    No the man is not insane he is a topnotch physicist
    maybe currently the most creative theoretical physicist in North America. However the idea of the dark energy density (cosmological constant) being quantized like that is very hard
    for me to swallow. Temporarily boggled.

    It is point 15. on page 24.

    It will turn out to be some whole number around 14.5E123.
    He is referencing somebody elses work here, which I will have
    to look up. Anyway gravitons are part of it! cheers
  6. May 24, 2003 #5
    Re: gravitrons

    Graviton is a real particle at least as much as photon, gluon, and W, Z bosons. Graviton is massless and has spin 2 (all others above has spin 1. Also, elusive Higgs suppose to have spin 0) It means graviton mediate long range force like photon - this is why gravity falls like 1/ distant^2. Spin 2 means graviton has two polarization state, in other words two degrees of freedom. Direct observation of gravity wave hasn't been made yet, but there are ample indirect evidence of it (Taylo and Hulse were awarded Nobel for it.). Whenever there is a "oscillating" massive body gravitational wave is produced. It is just too weak to detect it yet, but work is currently undergoing. Like photon gravitons are their own anti particle. (Gluon case is different since they have colour charges.)

    Being massless graviton travels with speed of light (at lwast in theory.) Yes, production of graviton reduces objects mass. For example, if two bodies are rotating around each other they produce gravitational wave (graviton) They lose their orbital energy and falling into each other. The rate of which they fall exactly matched the amount of energy carried out by gravitational wave. This was measured at least for one well known example (Hulse-Taylor pulsa in Crab nebula. As I said earlier they got Nobel for this.) Also, theoretically it is well known that when massive body collapse gravitationally they loose ample amount of energy through gravitational wave (This was calculated in 60s and 70's by many people especially Richard Price in Utah.)

    Gratitons carries energy even though its proper mass is zero (exactly like photons.) Unlike photons though they interact each other (obviously because gravity is there whenever there is energy. Phtons doesn't interact each other because they don't carries charges.) . This is why Einstein's equation is non-linear. Graviton is a linear perturbation of background spacetime geometry. (so sometimes it is called "ripples on geometry")

    Last edited by a moderator: May 24, 2003
  7. May 24, 2003 #6
    ^^^ Instanton knows his stuff, listen to him. :smile: You can think of gravitons as the 'minimum possible ripple in spacetime geometry' if you want.

    They are their own antiparticles.

    Gravitons will show up in any fundamental theory; they are really a 'semiclassical approximation' in a sense. Under very basic conditions (notably Lorentz invariance) you can show that any fundamental theory will have a low-energy/long-distance approximation that is a quantum field theory. ('Long' here is likely only around a few hundred Planck lengths, so it's not really very long.) Because we know gravity is attractive, that approximation will necessary involve a spin-2 particle which we call a graviton.
  8. May 28, 2003 #7


    User Avatar
    Science Advisor

    Re: who says gravitons exist?

    Why do you suspect that LQG is the most promising quantum theory of gravity?
  9. May 28, 2003 #8


    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    Re: Re: who says gravitons exist?

    Hello steinitz, since you quoted me accurately you most likely realize that I am not making a declaration that it is "most promising" or offering a reasoned argument. I simply suspect that it is.

    You are asking me to explain my subjective mental state---to give some personal history as to how I came to have this hunch.

    It has to do with simplicity and completeness and testability.
    I have a personal bias towards theories that are simple and
    comparatively verifiable (disprovable by experiment).

    Also it has to do with mathematical taste. I find the mathematics upon which LQG more spare, great, and tasteful than the mathematics of strings and branes with all the rolled-up extra dimensions.

    Also it has to do with a sense of humor. I find it humorous that strenuous highly publicized efforts have been made for 4 decades in strings and there are still huge gaps----whereas by comparison the less publicized approach of LQG has covered pretty much all the bases and is closer to testability. It has moved faster with less hype.

    I decline to offer arguments why YOU should suspect one field to be more promising than the other. This is my subjective take. You ask me to introspect and try to understand why my hunch is this rather than that.

    If you have different tastes and hunches, feel free to express them. I will not argue with you since these are subjective matters.

    It would be especially amusing to hear what you have to say if you think String theory is the most promising approach to quantum gravity!
  10. May 28, 2003 #9


    User Avatar
    Science Advisor

    Re: Re: Re: who says gravitons exist?

    How well do you understand the math?
  11. May 28, 2003 #10


    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    Re: Re: Re: Re: who says gravitons exist?

    You keep asking questions. Why dont you lay your cards on the table? Can you think of anything hopeful to say about string theory and its offspring? Have you a spark of enthusiasm for it?
    Speak your mind.

    As for me, I rely largely on surveys and overviews by other people and my own judgement as to which are intelligently written. I have offered some arXiv references here---and invite
    you to reciprocate with some survey articles that accord with
    your viewpoint whatever it is.

  12. May 28, 2003 #11


    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    Loops are hot and strings are cold [<:)]

    You really arent being very communicative, steinitz. You just ask these guarded questions. If you are a diehard String fan it would be interesting and entertaining if you would come out in the open and lay out why you favor that approach.

    We've all seen areas of research become hot----we know the signs----it is a general phenomenon in academic research in many fields (not even limited to physics).

    The signs, as I read them, are that Strings are cold and LQG is hot.

    Here is one of the signs-----it doesnt take a historian of science to read this one:

    The seminal results in QG are the mid70s results of Bekenstein Hawking and Unruh. these sparked the development we see in progress. The attempt to understand black holes (and acceleration itself in Unruh's case) both quantum mechanically and thermodynamically has been central----because it's an arena where quantum mechanics, relativity, and classical physics meet.

    Right now black holes are LQG turf. Theyve proven that the horizon area is quantized and calculated the area step.
    An impressive collective effort which Baez summarized in a recent Nature article.


    It is short and easy to follow. I invite you to read it and comment.
    Love to hear what you have to say.
  13. May 28, 2003 #12


    User Avatar
    Science Advisor


    As you know, LQG is the result of employing some novel techniques to construct a QGT that encodes certain preconceived notions of what such theories should look like. The hope is that among these theories there will be one that has the correct low energy limit. However, whether the LQG group has achieved this remains an open question.

    On the other hand, it appears that any consistent string theory yields a QGT that does reproduce GR in the low energy limit.

    There is also a sense among string theorists that they're involved not in a program of construction, but of discovery. Viewed from this perspective, the results of LQG seem contrived.

    However, unless you have a technical understanding of string theory - and I do - it's difficult to convey why string theorists feel this way.

    I also share with most high energy theorists a basic belief in supersymmetry.
    Last edited: May 28, 2003
  14. May 28, 2003 #13
    marcus, I think you've been reading too much stuff by Baez and Smolin -- their prejudices are starting to rub off on you. ;)
  15. May 28, 2003 #14
    Even though I generally agree on your comment regarding Marcus I do have some comments on LQG. Because your post was too short to justify both approach I worry that some misunderstanding is due. LQG is a fairly straightforward application of well known quantization technique to GR. Due to technical reason they started not from real GR, but a complexified GR called Plebanski action. Classically, there is no ambiguity to recorver GR from it, but as Steinitz pointed out, no one knows yet whether current form of LQG has a semi-classical limit that looks like GR. The difficulty lies largely on solving Hamiltonian constraint. Readers who are interested in the recent development should go read Thomas Thiemann's review in LIVING REVIEW. (The review is very good, but not for a fainted heart.). But, despite of problems LQG still has virtues too. Unlike stringy calculation they can compute entropy of physical blackholes. Also, recently Ashetekar and his collaborator were able to compute generalized entropy in the case that matter fields are not minimally coupled to curvature. (It has been shown in classical GR that , in non-minimally coupled case , the entropy of BH is not just proportional to the area, but alsodepends on fields )

    On the other extreme I will say that the recovery of GR from string is not what they say it is. I never personally understood why they are claiming they recorverd the GR. (Of course string theory has massless spin 2 boson. Using the theorem by Weinberg you can argue that the theory couple to the energy momentum tensor universally. Therefore it must be GR. But......) I hope someone here can answer some of my naive questions regarding this.

    1. it is not GR, but rather a Brans-Dicke type theory on higher dimension. You may argue that after moduli fixing, and somehow all those extra exotic fields like 3-form fields go away we recorver GR. But, as we all know, it is a thorny question.

    2. String theory doesn't even have a well defined theory on DeSitter space. Unless we get a good idea how to break SUSY it doesn't seem promising to explain current data in cosmology.

    3. This is little bit more personal. To me it is not surprising that string recovers GR (or GR-like theory) in low energy limit. After all if you write down effective action of matter fields on curved spacetime you will get something like GR with cosmological constant. (In heat kernel expansion of determinant of propagator we have Ricci scalar. In leading order that is going to be the Einstein-Hilbert looking term in effective action.). So, it is strange to saying that recovering GR in low energy limit is a virtue of string theory.

    4. Do we really have a controllable non-perturbative string theory? Closest one I know of is AdS/CFT. But, I don't think the program is reaching its maturity to adress real quantum gravity questions like what is observable. (I mean non-perturbative definition of them.)

    At the end I feel like I should give some credit to string theory. Personally, I think string is probably the most developed quantum theory of gravity we have now. But, despite many claims (especially in popular venues - you don't hear these outrageous claims in professional venues.) I don't think any of these theories are mature enough to claim victory, yet.

    Last edited by a moderator: May 28, 2003
  16. May 28, 2003 #15


    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    Perhaps so, but you see I have elicited a declaration of faith in string theory and supersymmetry from steinitz---and an interesting one I may say!

    In fact one engages in certain kinds of discussion so that one's beliefs and assumptions can be questioned. Unquestioned ones are not worth having. Smolin has done respectable work in both strings and loops and his side-by-side comparison of the two approaches is, IMHO, the strongest challenge yet faced by the string people. It's a discussion analogous in some ways to the Ptolemaic versus Copernican debate and worth following.

    Hope you are getting something out of it too,
  17. May 28, 2003 #16


    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    On your suggestion I will go back to the LIVING REVIEW article and see if I can get more out of it. I have already spent some time on it and it is----as you say----not for the faint of heart.

    But there is a problem here---I was earlier reading ROVELLI'S review of LQG in LivingReviews. Could you have meant to say
    Rovelli instead of Thiemann? There is no Thiemann LQG survey in LivingReview that I could find.

    Could you give a brief overview of how Loop Quantum Gravity is constructed? Just list the main steps---a thumbnail sketch.

    One striking feature is that it starts with a smooth manifold &Sigma; x R with no metric. Like starting with a blank slate.

    And many of the objects studied are equivalence classes under
    diffeomorphism (smooth 1-1 onto mappings). Makes good mathematical sense and also

    seems intuitively appealing, since the metric is what one hopes to quantize! Other (eg stringy) theories start with a fixed arbitrarily chosen metric which they must then perturb in a quantum fashion. This seems to prejudice the whole proceedure at the outset.
    Last edited: May 28, 2003
  18. May 28, 2003 #17

    Sorry. Here


    Well I guess I can try. The first step is understanding classical GR in the form we can apply usual quantization technique. It is called canonical formalism. In this form the phase varibles are metric on (spatial) 3-surface and its conjugate momenta (extrinsic curvature - roughly time derivative of 3-metric). The action of GR is not like usual theory, but it is a set of constraints. I means it is a set of conditions that phase variables should satisfy. There are 4 constraints at each points on the manifold. 3 of them are called momentum constraints, and the other is called Hamiltonian constraints. Roguhly speaking they are manifestation of diffeomorphism invariance of action. Momentum constraints says only the phase variable that is diffeo invariant on 3 -surface are physical ones. (It is exactly analogous to Gauss's law in EM. Only those E and B satisfy Gauss's law at each moment in time are the physical fields.) Canonical quantum gravity usually means you quantize these constraints. That is, you choose out wave functions that satisfy operator version of constrains out of some Hilbert space. Now, it turns out that these constraints are very difficult to handle usually because they are not polynomials.

    Here, Ashtekar's idea comes in. It turns out we can change old phase variables into new ones, so that the action looks like that of gauge theory. Then, with these new variables constraints becomes polynomial and a lot easier to handle. Especially, Rovelli and Smolin discovered that momentum constraints has a complete set of solutions on the space of functionals of knot invariants. It is not really surprising in a sense that knots are only diffeo invariant quantities. So, now we can build a Hilbert space out of L^2 functional over Wilson loops. These works , by many people, especially, Ashetekar, Lewandowski, Baez, Smolin, Rovelli, Pullin, and Gambini, basically defines kinematic state space for loop gravity as a L^2 functional over soem special class of graph called spin network.

    Now, one thorny problem is solving Hamiltonian constraint. Thomas Thiemann suggested a solution, which is only well known solution as far as I know. But, as Stenitz suggested earlier, it is not known whether the choice will recorver GR in some (not-yet known) semi-classical limit. Curretly, there are lots of effort in this direction is undergone especially by AEI group, Penn State group, and Warsaw group.

    There are some neat applications of the formalism like computation of BH entropy, and singularity free quantum cosmology, but I gotta leave soon, so may be next time.

  19. May 28, 2003 #18


    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    What a pleasure! thanks. I had not seen Thiemann's excellent 2001 review and printed out portions of it immediately, to read this afternoon.

    I'm looking forward to hearing more about the things you mentioned.

    In the meanwhile I will try to formulate my own overview
    (outrageously oversimplified, but have to try for some perspective).

    Thiemann's first 20 pages are highly accessible and give a clear motivation for the development of the theory. Hope that more of us at PF read it.
  20. May 29, 2003 #19


    User Avatar
    Science Advisor
    Gold Member
    Dearly Missed

    I will echo back some of what you said as a check of my understanding. Tell me if I misinterpret.

    Roughly speaking this corresponds to focussing on the connection as the primary object (carrying the geometry) rather than on the metric.

    A metric determines a connection (parallel transport mechanism) and the metric can be recovered from its connection. So one is going to quantize the space of connections rather than the metrics.

    This is an exciting idea. The behavior of a connection can be told by seeing how it does parallel transport around loops. Loops detect the characteristics of the geometry.

    Knots are diffeo equivalence classes of loop. Certain kinds of knots apparently can be picked out to form an orthonormal basis for the Hilbert space.

    What is meant by "L^2" functionals. In some treatments there is the function space L2 (A/G) where A is the connections and G is a gauge group-----so some measure has been constructed (for integration purposes) on the equivalence classes of connections. And we have a Hilbert space of the square-integrable functions. I am guessing that this is what its meant here.

    Narrowing down from Wilson loops to spin networks seems to be because there are too many Wilson loops. To have a linearly independent basis for the space one must get rid of redundancy. So the spin network business is just a way of systematically picking out the basis.

    This is a great thumbnail sketch! Thanks much!
    If I made any gross mistakes in interpreting your sketch please tell me.

  21. May 29, 2003 #20


    User Avatar
    Science Advisor

    1. As you know, massless spin-2 excitations are produced as a direct result of quantizing a theory of objects extended in a single spatial dimension. Further - as you more or less intimated - the corresponding symmetric tensor particle must couple to a conserved symmetric source, of which the only one available (aside from some special but irrelevent cases) is the energy-momentum tensor, and the consistency of this coupling requires the theory to have spacetime coordinate invariance. This means that for string theory to be consistent, these excitations must be gravitons and string theory must contain GR, but a priori, there's no obvious reason why it should. The "miracle" is that it does. Also surprising is the manner in which this fact emerges: The gravitational field equations of GR - a theory of spacetime - are recovered not from an explicitly spacetime formulation, but indirectly through imposition of consistency on a 2-dimensional quantum theory formulated as a nonlinear sigma model for which spacetime is simply it's target space. Specifically, the world-sheet theory must be weyl-invariant, a condition which is enforced by requiring that the relevent beta functions vanish. Working in the low energy limit of the massless bosonic closed string sector and computing the beta functions yields background field equations which include the gravitational field equations with a source term that in this kinematical sector is a function of the dilaton (which is not the scalar field of brans-dicke theory) and the field strength associated with the antisymmetric potential. But let me add for emphasis that despite the need to make choices - like how to compactify the spacetime manifold, or of backgrounds - to obtain realistic phenomenologies from these equations, we still have GR.

    2. The manifold ways string theorists break SUSY in attempts to obtain "reasonable looking" phenomenologies notwithstanding, the mechanism underlying the current accelerating expansion of our universe remains unknown and so may not require a vacuum energy to explain.

    4. We describe particular formulations of a theory and not the theory itself as being perturbative or non-perturbative. Now, assuming you meant "perturbative" - and in the preceding sense - the answer depends on what you mean by "controllable". If you mean borel summable, than no. However, none of the standard gauge theories of particle physics are borel summable. It's just that unlike string theory, these latter theories are not meant to be TOEs, and this embarrassment is what lead in the early 90s to the formulation of the non-perturbative methods used over the last decade to obtain so many remarkable exact - i.e. non-perturbative - results. Btw, AdS/CFT is not a string theory, it's a conjectured duality between CFT in AdS and YM on it's boundary.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook