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

Gravitons and infinite regress.

  1. Feb 8, 2008 #1
    Can anyone tell me what is wrong with the follow observation (if anything)?

    (1) The graviton (putatively) mediates the force of gravity. The phenomenology of masses moving toward each other is due to an exchange of gravitons. Call these gravitons G1.

    (2) The G1 are themselves, as massless particles, subject to the gravitational force. Their paths through spacetime are bent by the presence of masses. But if the gravitational force is exerted on G1, it must be exerted via gravitons (call these gravitons G2).

    (3) But then the G2 are subject to gravitational effects, and G3 must be introduced. G4 for G3, G5 for G4, and so on.

    Thanks in advance.
  2. jcsd
  3. Feb 8, 2008 #2

    Whether or not the graviton becomes an accurate description of gravitation- I will not consider- but the what you are describing basically (I think) found within the standard model already.

    Because the gluon carries color charge, it has a 3 gluon and 4 gluon vertex in the perturbative version of QCD. This leads to the possibility of a color neutral ensemble of gluons called a glueball.

    I hope this helps.
  4. Feb 8, 2008 #3
    Well, it seems like if the graviton account of gravitation is correct, then there must be an infinite number of gravitons (G1, G2, . . ., Ginf), which is absurd.
  5. Feb 9, 2008 #4
    Why can't one graviton mediate the force between two other gravitons? Perhaps they can, perhaps they cannot. Hasn't really been solved yet right? Maybe gravitons only mediate interaction between its originating mass/energy and space(its field), and its the field that interacts with other particles. Therefor no need for multi-graviton interaction.

    Make your model, if it works collect your nobel prize.

    Also there is nothing absurd about there being an infinite (or near) number of possibilities in an interaction. Thats just how it is, with less and less probability.
  6. Feb 9, 2008 #5


    User Avatar
    Science Advisor

    Thats how quantum field theory works. Its the same for QED. Loops within loops within loops all the way to infinity.

    There is nothing absurd about it, but keep in mind thats just a *picture* based on perturbation theory that so happens to be mathematically consistent with how you might solve the action..
  7. Feb 9, 2008 #6
    This wouldn't matter, because you'd still have an infinite number of gravitons.

    Yes, that's a possibility, but then the graviton would have to have some special property that makes it immune to the effects of gravitation (unlike, say, the photon).

    Not what I said. You're right that there's nothing wrong with "an infinite . . . number of possibilities". There is something wrong with an infinite number of particles.

    It is obviously absurd to posit an infinite number of gravitons. And indeed, not just an infinite number of gravitons, but an infinitely dense number of gravitons (i.e. even if the universe is infinite, given any finite volume V there are an infinite number of gravitons in V. That's what the graviton account says.
  8. Feb 9, 2008 #7

    Vanadium 50

    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor
    2017 Award

    Consider the electric field from a static charge. You're not in an eigenstate of photon number (virtual or otherwise), so talking about how many there are - finite or infinite - makes no sense.

    Same thing with gravitons from a static mass.
  9. Feb 9, 2008 #8


    User Avatar

    If we step back and talk about a simple QED process, let's say. Would you still have the same concerns? consider the scattering of an electron from another electron.

    Now, we can not calculate this exactly in QED so we use perturbation theory. Then we introduce this cute little mathematical trick of an expansion in Feynman diagrams and so-called virtual particles. It's just a mathematical trick, really, forced upon use because we can't solve the problem exactly. Now, even in that simple case, in principle there are an infininite number of diagrams to consider with an ever increasing number of virtual photons, electrons, etc. It never stops. But again, it's just a formal trick.

    The picture is not any different with hypothetical gravitons except that the interaction is nonlinear so the gravitons couple to each other (like gluons). The fact that they interact with one another does not change th ebasi c picture of an infinit enumber of processes with more and more virtual particles. But again, this is just a formal, mathematical, trick to do the calculation. If we were smart enough to solve exactly, we would never even talk about Feynman diagrams.
  10. Feb 9, 2008 #9
    My concern with that response is that gravitons are (alleged to be) real, and not virtual, particles. Even in a nonperturbative treatment we'd be facing an infinitely dense sea of gravitons, wouldn't we?

    And speaking more generally, aren't gravitons irreconcilably at odds with background independence?
  11. Feb 10, 2008 #10

    Vanadium 50

    User Avatar
    Staff Emeritus
    Science Advisor
    Education Advisor
    2017 Award

    Not so. You can have real and virtual gravitons, much like you have real and virtual photons.

    At the risk of repeating things that have been said before, if you're not in an eigenstate of particle number, counting particles makes no sense.
  12. Feb 11, 2008 #11
    I may be wrong, but I thought gravitation could be described either by gravitons or by general relativity (curvature of space-time).
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook