The graviton pisses me off.. .

[KingOrdo]

Well, not really, but I am really struggling with understanding the motivation for its inclusion in the Standard Model; I'm hoping someone here might provide some guidance.

My understanding of General Relativity is that gravitation is a purely local phenomenon; i.e. two massive entities in some sense 'appear' to be attracted to each other because the manifold upon which they reside is in fact curved. So it is really not a force at all but only an apparent one, and one which can be roughly described in most cases by Newtonian mechanics.

But the graviton seems to be just an analogue of the photon; a mediating particle that (I assume) travels at the speed of light and is therefore not local at all, and in fact suggests that the gravitational 'force' would not be instantaneously 'felt', as it would be if it were a phenomenological manifestation of a local property.

What gives here? Why introduce a particle without cause? William of Occam warned us against this sort of thing, didn't he? I'm just confused. Thanks!

 

[sirwebber]

I'm not an expert by any means, but I think your wrong about one thing. Einsten's Theory of General Relativity does not assert that the presence of matter instantaneously effects the manifold around it. Instead, if all of a sudden the sun would disappear, we would still feel the force of its pull for 8 mins (the time it takes light to travel to the earth) because nothing, including information about the manifold, can travel faster than the speed of light. Is this reasoning correct? Someone please correct me if I am wrong.

 

[coalquay404]

Well, not really, but I am really struggling with understanding the motivation for its inclusion in the Standard Model; I'm hoping someone here might provide some guidance.

My understanding of General Relativity is that gravitation is a purely local phenomenon; i.e. two massive entities in some sense 'appear' to be attracted to each other because the manifold upon which they reside is in fact curved. So it is really not a force at all but only an apparent one, and one which can be roughly described in most cases by Newtonian mechanics.

The first thing to point out is that general relativity makes no claims about gravitation being a local phenomenon. There are certain situations in which you can restrict your attention to the local behaviour of the gravitational field around an object (such as considering an asymptotically flat spacetime with an interior boundary to represent, say, an event horizon), but Einstein's field equations are essentially statements about the global behaviour of spacetime and, hence, of the gravitational field. In fact, a really active area of research in mathematical relativity at the moment is the study of existence theorems and global dynamics of solutions.

But the graviton seems to be just an analogue of the photon; a mediating particle that (I assume) travels at the speed of light and is therefore not local at all, and in fact suggests that the gravitational 'force' would not be instantaneously 'felt', as it would be if it were a phenomenological manifestation of a local property.

There is, in a sense, a real philosophical gap between particle theorists (and high energy physicists more generally) and relativists. People who are interested in high energy physics usually take their lead from quantum field theory and have an innate desire to express everything they see, including the fundamental forces, in terms of particles. Most probably, this is because when they are introduced to quantum field theory they are told over and over again that it's a good thing to look for particle interpretations of QFTs. Given that, the high energy crowd have been trying for years to, as you say, explain away gravity by means of the graviton. A graviton is a particle which mediates the force of gravity and is, by necessity, a massless spin-two particle (it's believed to be massless because of the proposed equivalence between the speed of propagation of the gravitational field and that of the electromagnetic field, and it's spin-two because this is the lowest spin particle possible which produces only attractive forces). String theorists, for example, believe that string theories can explain gravity because they all contain massless spin-two states, states which can be interpreted as particles which carry the gravitational force. (Hopefully this answers your query as to why people introduced the idea of a graviton.)

On the other hand, relativists usually frown on the graviton idea for a number of reasons. Firstly, high energy physicists who want to explain gravity in terms of a graviton usually propose that gravitons are particles which exist on a flat spacetime background. To a relativist, picking out a flat spacetime and claiming that it is somehow "fundamental" is anathema: the whole content of general relativity is that it requires non-flat spacetimes to describe non-trivial gravitational processes, so why bother claiming that Minkowski space is somehow special?

Secondly, there's a *huge* question which hangs over quantum field theory that's often brushed under the carpet by high energy physicists. The essential idea is this: you can interpret quantum field theories in terms of particles and particle excitations only when the quantum field theory is defined on a flat manifold. In general, if you attempt to do quantum field theory in a curved spacetime a particle interpretation of the theory does not exist! Relativists (rightly, in my opinion) claim that if you can't interpret QFTs in curved spacetimes in terms of particles, then what's the point in proposing a graviton? You can do QFT in these curved spacetimes, but if the particle interpretation goes out the window, how do you explain the curvature of the spacetime in terms of a graviton?

It's all a bit complex, and discussions about this eventually end in brawls about the importance of background independence, but I guess the basic answer to your question is simply that people who think that the SM is more fundamental than GR will always want to explain *everything*, including gravity, in terms of particles.

 

[Haelfix]

Err particle physicists are interested in *fields*, not particles. Particles have always been somewhat mysterious and by assumption, purely local measurement phenomena (say something that heats up a glass of water in an experiment).

The Graviton *field* otoh reproduces Einsteins equations of motion and his field equations, so it by default has the global properties required of General relativity in the semiclassical limit. The problem is the quantum limit, where things behave extremely bad. QFT with gravity has no problem with GR so to speak, rather it has issues with quantum mechanics.

The language QFT uses is sometimes clashy with GR, but if you are careful you can make things work (see for instance textbooks by Wald and so forth). In any event the problem is more technical than fundamental. The quantum problem otoh *is* fundamental, which is why things need to be changed at high energy scales.

 

[Hurkyl]

Einstein's field equations are essentially statements about the global behaviour of spacetime

The EFE's are differential equations -- they are the epitome of a local constraint!

 

[ohwilleke]

Short version:

In the standard model, the other three forces can be described as mediated by force carriers electro-magnetism (photon), weak force (Ws and Zs), strong force (gluons). Gravity seems like a force.

If there is such a force carrier for gravity, it ought to look like a graviton which if as theorized would produce something, at least, quite similar to, and perhaps even identical to, GR. A spin 2 particle would be necessary because if it weren't an integer spin it would behave like a matter particle, and if it were spin 1 then gravity would have features (perhaps polarity or positive v. negative charge) that it doesn't appear to have.

Hence, the graviton is a quite plausible idea.

Of course, the basic issue (and a key divide between stringy people and loop quantum gravity people) is the question of whether you could have a GR that doesn't flow from the nature of time-space itself. String theory basically leans towards saying yes. But, LQG people say that gravity is more than a force, so a force model from the standard model of particle physics can't possibly work.

 

[bananan]

which research program do you think is more promising? if gravity is just a field embedded in space-time, then what is space-time?

Short version:

In the standard model, the other three forces can be described as mediated by force carriers electro-magnetism (photon), weak force (Ws and Zs), strong force (gluons). Gravity seems like a force.

If there is such a force carrier for gravity, it ought to look like a graviton which if as theorized would produce something, at least, quite similar to, and perhaps even identical to, GR. A spin 2 particle would be necessary because if it weren't an integer spin it would behave like a matter particle, and if it were spin 1 then gravity would have features (perhaps polarity or positive v. negative charge) that it doesn't appear to have.

Hence, the graviton is a quite plausible idea.

Of course, the basic issue (and a key divide between stringy people and loop quantum gravity people) is the question of whether you could have a GR that doesn't flow from the nature of time-space itself. String theory basically leans towards saying yes. But, LQG people say that gravity is more than a force, so a force model from the standard model of particle physics can't possibly work.

 

[bananan]

how do gravitons, which cause attraction, cause time dilation, in a way that the other 3 forces apparently do not.

Err particle physicists are interested in *fields*, not particles. Particles have always been somewhat mysterious and by assumption, purely local measurement phenomena (say something that heats up a glass of water in an experiment).

The Graviton *field* otoh reproduces Einsteins equations of motion and his field equations, so it by default has the global properties required of General relativity in the semiclassical limit. The problem is the quantum limit, where things behave extremely bad. QFT with gravity has no problem with GR so to speak, rather it has issues with quantum mechanics.

The language QFT uses is sometimes clashy with GR, but if you are careful you can make things work (see for instance textbooks by Wald and so forth). In any event the problem is more technical than fundamental. The quantum problem otoh *is* fundamental, which is why things need to be changed at high energy scales.

 

[Farsight]

What I don't understand is how gravitons, virtual or real, can account for black hole gravity. Surely nothing can escape the event horizon. I'd be grateful if anybody could explain this.

 

[Careful]

***
On the other hand, relativists usually frown on the graviton idea for a number of reasons. Firstly, high energy physicists who want to explain gravity in terms of a graviton usually propose that gravitons are particles which exist on a flat spacetime background. To a relativist, picking out a flat spacetime and claiming that it is somehow "fundamental" is anathema: the whole content of general relativity is that it requires non-flat spacetimes to describe non-trivial gravitational processes, so why bother claiming that Minkowski space is somehow special? ***

You probably know that one can derive GR from building a consistent theory of graviton-graviton interaction on Minkowski ? Moreover, the latter is the only global vacuum solution (without cosmological constant) which does not contain singularities (which would have to be interpreted as point particles). So, even for a relativist such solutions are damn special ...

***
Secondly, there's a *huge* question which hangs over quantum field theory that's often brushed under the carpet by high energy physicists. The essential idea is this: you can interpret quantum field theories in terms of particles and particle excitations only when the quantum field theory is defined on a flat manifold. In general, if you attempt to do quantum field theory in a curved spacetime a particle interpretation of the theory does not exist! ***

Hmmmm ... can you not simply construct a particle interpretation by considering localized self interacting (bullet like) wave packages as particles ?? I mean, I know of the``problem'' that in generic curved spacetime the particle notion is unstable (a gravitational decay time if you want to) but that could simply be expected when quantum waves interact with a background field.


***
Relativists (rightly, in my opinion) claim that if you can't interpret QFTs in curved spacetimes in terms of particles, then what's the point in proposing a graviton? You can do QFT in these curved spacetimes, but if the particle interpretation goes out the window, how do you explain the curvature of the spacetime in terms of a graviton? ***

Well, again, just consider localized ``bubbles'' ...

 

[KingOrdo]

Thanks for all the fine, insightful replies.

Perhaps a better way of putting it is: what is the Universe really like?

Is gravitation a force like electromagnetism, mediated by particles flying through spacetime? Or, is it just a phenomenological manifestation of the curvature of spacetime itself? It seems to me that it cannot be both; the two notions are physically incommensurable.

Of course, we might adopt scientific antirealism and simply claim something like 'it doesn't matter, it's meaningless: the equations work out', and in fact I am not opposed to such a thing. But the physicists I know are generally scientific realists, and it is still not clear to me how this seemingly foundational problem is handled.

 

[pervect]

The way I understand the particle-field issue is that in field theory, particles are viewed as irreducible representations of the Poincare group. This defintion is due to Wigner.

If one has a general curved space-time, one doesn't have a Poincare group, and it turns out one can't apply Wigner's defintion. AFAIK there aren't any alternate competing defintions that one can apply either.

This defintional issue of "what is a particle" is also related to the the Unruh effect. An accelerating observer sees "particles", a non-accelerating observer does not. This demonstrates that the notion of what constitutes a "particle" depends on the observer, that the notion of exactly "what constitutes a particle" is not an observer indepedent notion. At the very least, the notion of "real" particles is shown to be observer dependent by the Unruh effect.

As an aside, I suspect one might be able to describe the Unruh effect as virtual particles being converted into real particles, and vica-versa, but I haven't seen anyone actually offer this descritption in print.

This probably gets off the main point, which is that from the field POV, particles are the consequence of certain symmetries of space-time, and that if you don't have such a symmetry, you can't define particles in terms of fields.

 

[CarlB]

The way I understand the particle-field issue is that in field theory, particles are viewed as irreducible representations of the Poincare group. This defintion is due to Wigner.

If one has a general curved space-time, one doesn't have a Poincare group, and it turns out one can't apply Wigner's defintion. AFAIK there aren't any alternate competing defintions that one can apply either.

I believe that this is correct, an impression I got from Weinberg's famous for being difficult to understand text (and therefore my understanding is likely to be faulty). But a problem with the program is that there is a sort of circular reasoning going on here. If gravitation gets replaced with a particle operating on flat space, it does at least some damage to the special theory of relativity in that one can suppose that after one chooses the flat space to use it takes on the character of a preferred reference frame in QFT. But if the special theory of relativity is iffy, then it weakens the reasoning behind picking out the Poincare group.

In other words, we have some reason to suspect that both the foundations are faulty. QM implies the specialness of a flat space that is denied by GR. A problem with GR implies a problem with SR. And a problem with SR weakens the foundations of QM.

I had to drive from Seattle to Spokane and back the day before yesterday and as always, I had lots of time to think about physics. (I'm the guy the rest of you honk your horns at.) My agnosticism on SR is decaying into atheism, and it comes from the anthropic principle.

M&M looked for an effect due to the Earth's movement through the ether. Because they were running the experiment before much was known about the galactic and larger environment, they were only assuming a speed through the ether of the Earth's speed around the sun. But if we ran the experiment today, we'd assume a much larger speed, about 0.0001 c, if I recall correctly.

With that much of a speed through the ether, it would have been that much easier to detect the speed of light. But the complete effect of a detectable ether would have been far worse than that. Our bodies are made out of very delicately balanced molecules. With a .01% change in the speed of light, we would probably have difficulty surviving a rotation of our bodies. The chemical properties would depend on what direction we were facing.

Life requires stability that lasts for long enough for evolution to do its slow work. Having different frames of reference share the same laws of physics helps contribute to that stability. So we can say that SR follows not because of simplicity, but because of the anthropic principle, and we should have known that the result of the experiment would be close to null before we ran it.

It's easy to get in trouble with laws that follow from anthropic principles and the reason is because anthropic principles cannot place requirements on physical laws that are exact. They can only place limits, and those limits can be rather vague.

As an example of this, the fact that life requires a stable environment implies that the eccentricity of the Earth's orbit is limited. This means that our orbit is close to a circle. Early astronomers saw that the orbit was close to a circle, and then concluded that the orbit was exactly a circle (with the sun going around the earth). Why? The circle was the simplest and most elegant orbit and a lot could be easily derived from the assumption. Why believe in perfect SR? Because it is a simple and elegant theory and a lot can be easily derived from the assumption.

The Ptolemaic model of the universe brought baggage with it, namely the circles within circles that were required to give retrograde motion. The baggage that SR brought mostly seems to be causality issues.

But let me get back to the problem with the fact that the particles are modeled as symmetries of the Poincare group. In modeling particles this way we are making an assumption about what part of nature is simple and elegant. It could be that we are wrong.

I see physics as having three parts, any combination of which may be simple and elegant. At the highest level are principles such as Einstein's SR. At the mid level are the conserved quantities which are associated with symmetries such as the Poincare group. And at the lowest level are the equations of motion. Newton's physics was fairly simple at all three levels. But if you look at the conserved quantities for gravitation, you will see that they are not quite as simple as the equations of motion. I think that we should think more carefully about what we are assuming to be simple.

A program for modeling particles in curved space could be to take the equations of motion in flat space, convert them into equations of motion for curved space, and then look for the symmetries in curved space. If you can do this, the mathematics might work out to be a practical example of David Finkelstein's no-field theory.
https://www.physicsforums.com/showthread.php?t=129527

Carl

 

[Careful]

**The way I understand the particle-field issue is that in field theory, particles are viewed as irreducible representations of the Poincare group. This defintion is due to Wigner.

If one has a general curved space-time, one doesn't have a Poincare group, and it turns out one can't apply Wigner's defintion. AFAIK there aren't any alternate competing defintions that one can apply either. **

Hmm, one can do free quantum field theory for any stationary spacetime since we also have a conserved current T^{ab} V_b (with V^b Killing), and therefore a conserved Hilbert space inner product (apart from the conserved symplectic form). Also here, one can speak of creation and annihilation operators (with the ordinary commutation relations), conserved particle numbers and so on... Just work it out: take any metric of the form -dt^2 + h(x^{\alpha})_ab dx^a dx^b, look for the solutions of constant energy (with respect to t) and you will see ordinary mode decomposition proceeds as usual.

One merely wants Killing properties in order to have stable and unique particle notions, in de Sitter for example there will be some steady particle creation rate from the vacuum (due to the expansion of the universe).

***
As an aside, I suspect one might be able to describe the Unruh effect as virtual particles being converted into real particles, and vica-versa, but I haven't seen anyone actually offer this descritption in print. ****

I guess Robert Wald has done that in detail in the 80 ties, including how the adsorbtion of a Rindler particle by an accelerated detector corresponds to the emission of Minkowski particle (negative and postive frequencies get mixed up in the Unruh effect due to the nonlinear coordinate transformation).

**
This probably gets off the main point, which is that from the field POV, particles are the consequence of certain symmetries of space-time, and that if you don't have such a symmetry, you can't define particles in terms of fields.**

I disagree here, your particles will just decay after a while, that's all. There is no problem in thinking about particle notions in QFT, it is just that they are generically decaying as well as non-unique.

Careful

 

[Haelfix]

Yes particles as irreducible symmetries of the Poincare group is out for physics in curved spacetimes, it no longer is a sensible definition in that regime.

Which is fine, the field dynamics still work but you are left with an observable problem of QG, eg what to measure that makes sense in an intrinsic invariant way.

For the moment, (and its slightly unsatisfying) the SMatrix is really the only thing that quantum gravity people can make sense off, and that works only in very special conditions (say with asymptotic flatness). Again it might be more of a technical problem with the language we are using, rather than some deep physical issue.

 

[turbo]

The first thing to point out is that general relativity makes no claims about gravitation being a local phenomenon.

GR does not claim that gravitation is local, but if you will read Einstein's 1924 treatise "On the Ether", you will see that he realized that GR was woefully incomplete, and that he needed an "ether" or interactive medium with which matter could interact to provide gravitational attraction and inertial effects. The medium in which matter is imbedded would endow matter with these effects, and the matter would in turn fundamentally effect the properties of the medium. What's more, he insisted that gravitation and inertial effects had to arise from matter's interaction with local ether - he rejected Machian "action at a distance" completely. Einstein never successfully made this extension to GR, but Sakharov hinted at the path in the 1960's when he hypothesized that matter's interaction with the quantum vacuum was responsible for mass, gravitation, and inertia. Recently, Padmanabhan has revived this idea, but he has neglected to consider that changes in the bulk properties in the vacuum can have effects on the strength of the gravitational and inertial effects that the vacuum confers to matter. This denial of dynamical characteristics cannot possibly lead to a good theory of quantum gravitation - we're not there yet.

 

[Farsight]

Sympathising with Kingordo, I was looking on Google for explanations of gravitons (real or virtual) escaping black holes to cause gravity. But I couldn't find anything that I thought was convincing. It was stuff like this:

http://www.astronomycafe.net/qadir/q756.html

Can anybody supply a better link?

 

[wolram]

I suspect that maths is the problem with any of this theory, it is thought that maths can solve any problem, i would say that only statistical maths
is worth following, maths is a double edged sword, it can generate unreal
solutions, as well as 1+1 =2, i am far from an understanding of maths, but
if you look in the arxives maths can prove anything with a 95% cl, and now you are all going to howl (we are nothing without maths) well can you say what the nothing means? Poincare (spelling) is up for an award, but his maths pinches out black holes to make the topology (smooth) is that real?
( my memory is bad it may not be poincare), any way the fact is there are no known rulers in cosmology only light, a fish in water?

 

[Hurkyl]

GR does not claim that gravitation is local

How do you figure? The EFE's are differential equations, as are the equations of motion. You can't get more local than that!

 

[turbo]

How do you figure? The EFE's are differential equations, as are the equations of motion. You can't get more local than that!

Yes, but GR is commonly interpreted as if gravitational attraction is a force acting over a distance - a misconception that Einstein tried to clear up later. If you can get hold of Saunders' book "The Philosophy of Vacuum" and read chapter One (Einstein's essay "On the Ether"), you will see that he is making a subtle distinction that is not properly appreciated today.

 

[Careful]

Yes particles as irreducible symmetries of the Poincare group is out for physics in curved spacetimes, it no longer is a sensible definition in that regime.

Which is fine, the field dynamics still work but you are left with an observable problem of QG, eg what to measure that makes sense in an intrinsic invariant way.

For the moment, (and its slightly unsatisfying) the SMatrix is really the only thing that quantum gravity people can make sense off, and that works only in very special conditions (say with asymptotic flatness). Again it might be more of a technical problem with the language we are using, rather than some deep physical issue.

Here I disagree (with your last statement), there is clearly a further issue to be considered : that is how we can *observe* correlations in QFT *inside* the universe (and not on some conformal boundary added to spacetime). This is not just a problem of QG, it is a problem of QFT (even on flat spacetime).

Careful

 

[nitin]

String theory graviton

I am currently reading Prof. http://www.astro.virginia.edu/people/faculty/txt/""). A few pages ago, I read something which still confuses me. I have read, and understood, on several occasions and at several locations on the web that, even if supersymmetry does not show up at the LHC (or at any present/future accelerator for that matter), this does not mean that (super)string theory is wrong. However, in the book, I read that the graviton appears in the theory of strings only when supersymmetry is included. Is the author talking about one of the different string theories, or is it true in general? Because then, to my understanding, without supersymmetry, there is no such thing as unification by string theory. I think the author might be wrong there, right? Or maybe has omitted some essential element?

 

[selfAdjoint]

I am currently reading Prof. http://www.astro.virginia.edu/people/faculty/txt/""). A few pages ago, I read something which still confuses me. I have read, and understood, on several occasions and at several locations on the web that, even if supersymmetry does not show up at the LHC (or at any present/future accelerator for that matter), this does not mean that (super)string theory is wrong. However, in the book, I read that the graviton appears in the theory of strings only when supersymmetry is included. Is the author talking about one of the different string theories, or is it true in general? Because then, to my understanding, without supersymmetry, there is no such thing as unification by string theory. I think the author might be wrong there, right? Or maybe has omitted some essential element?

You can get the graviton out of non-supersymmetric "bosonic" string theory. The problem then is that at the simple-minded level that doesn't give you much other physics, specifically fermions. Superstring theory was discovered in the effort to find a stringy description of fermions.

As I understand it (which isn't much ) there are now some very non-obvious constructions which do fermions in non-supersymetric string theory, but someone else is going to have to tell you about those.

But what you read is not true in general and must have been describing some particular case.

 

[Careful]

***
As I understand it (which isn't much ) there are now some very non-obvious constructions which do fermions in non-supersymetric string theory, but someone else is going to have to tell you about those.
**

Don't worry, fermions are very strange creatures (and I do not understand them myself well yet, bosons are much easier). Some useful points of view on a single spinor wave are
(a) the Feynman checkerboard model: here the spinor wavefunction can be reproduced by a local stochastic process using complex amplitudes associated to Zitterbewegung
(b) some classical models based upon zitterbewegung.

If there would be some nontrivial model for fermions in string theory I would be definately interested to hear about it.

Careful

 

[nitin]

Self-Adjoint:
I must say that author is not explicit in the book which theory he is talking about. Anyway, thanks for your effort. I might grab Polchinski (or Zwiebach?) and check. I know that the graviton was already part of the theory before supersymmetry was included in the picture. Anyway, thanks for the effort. Maybe I should try to understand supersymmetry.