Is the Graviton Fundamental in Quantum Gravity Theories?

[friend]

Marcus,

Slightly more of what Loll wrote:

“
The
failure of the perturbative approach to quantum gravity in terms of linear fluctuations around
a fixed background metric implies that the fundamental dynamical degrees of freedom of
quantum gravity at the Planck scale are definitely not gravitons. At this stage, we do
not yet know what they are. Neither do we have the luxury of hints from experiment or
observation of what they might be, …”

So if gravitons are particles associated with weak gravitational fields, then at what level of gravitational curvature do gravitons cease to be relevant? What do they turn into and at what level of gravitation? Are they supposed to decay at some level of gravity - into what? I would think that the impossibility of something slowly disappearing proves they don't exist to begin with, right?

 

[jimgraber]

Hi Friend,
you wrote:
"
So if gravitons are particles associated with weak gravitational fields, then at what level of gravitational curvature do gravitons cease to be relevant? What do they turn into and at what level of gravitation? Are they supposed to decay at some level of gravity - into what? I would think that the impossibility of something slowly disappearing proves they don't exist to begin with, right?"

Think more and more gravitons as the curvature gets stronger, not less and less.
Also think interactions with other kinds of particles (real and virtual) as well as gravitons.

The "failures" only happen down near the Planck scale if they happen at all.
The theoretical failures, which string theory claims to cure, come in the form of uncontrollable or "nonrenormalizable" infinities in the calculations as you go to higher orders.
So take the word "failure" with a big grain of salt. The vast majority of physicists still believe that the zero mass spin two graviton is our best current understanding of gravitation. (I include string theorists in this group. Their graviton is a massless spin two closed string.) Second place goes to good old fashioned unquantized general relativity. Even some of the most developed Loop Quantum Gravity approaches contain gravitons or near equivalents. Only a very small percent of physicists investigate theories with no graviton, or no equivalent. I could be wrong, but I do not know of any well worked out approach that makes significantly different predictions from the graviton theory. However, some people are making valiant attempts along these lines. Some day they may succeed. But I'm not holding my breath.
Best.
Jim Graber

 

[notknowing]

I posted a message already on a more recent tread about the graviton (see 12-06-2007: Could the quanta of gravity be something other than spin 2 particles) but it seems more appropriate to post it here.
I'm also convinced that gravitons do not exist and it is easy to prove by a simple thought experiment:

Describing gravitation by means of virtual particles quickly leads to a contradiction:
Let us suppose indeed that they exist. Even if they are supposed to be (rest)massless spin 2 particles, they should contribute to the total energy of the system, just like the virtual (massless) photons of the electromagnetic field somehow contribute to the mass of the system. The nonlinear character of the Einstein equations also seems to indicate that the "gravitational field" itself gravitates (I know this is strictly not correct, I use this only to guide the intuition). So, if the gravitons themselves are subject to the gravitational field, one should have virtual gravitions exchanging between these gravitons and other gravitons or other masses - AND - you can not stop this process: you just have to keep on adding gravitons to include all the reactions. So, you get an infinite number of gravitons per unit volume resulting in an infinite energy density and the whole of space would be just a massive "block" of gravitons. A ridiculous situation. This proofs that gravitons can not exist.

So, if gravitons do not exist, what then? Well, one can describe gravity as an induced force resulting from all other forces in nature. This was originally proposed by Sakharov. In this case, one does not need gravitons. I have also been publishing a paper about this in 1992 (see http://home.online.no/~avannieu/darkmatter/ first paper)

Rudi Van Nieuwenhove

 

[friend]

Think more and more gravitons as the curvature gets stronger, not less and less.
Also think interactions with other kinds of particles (real and virtual) as well as gravitons.

The "failures" only happen down near the Planck scale if they happen at all. ...

The vast majority of physicists still believe that the zero mass spin two graviton is our best current understanding of gravitation. (I include string theorists in this group. ...

Only a very small percent of physicists investigate theories with no graviton, or no equivalent. I could be wrong, but I do not know of any well worked out approach that makes significantly different predictions from the graviton theory. However, some people are making valiant attempts along these lines. Some day they may succeed. But I'm not holding my breath.

From a previous thread I wrote:

In "QFT in curved spacetime", Robert Wald states that in highly curved or fast changing gravitational field there is no particle interpretation of QFT; it's all just fields. So strictly speaking, in high curvature or fast changing gravitational fields, there are no particles at all - not even gravitons.

Describing gravitation by means of virtual particles quickly leads to a contradiction:
Let us suppose indeed that they exist. Even if they are supposed to be (rest)massless spin 2 particles, they should contribute to the total energy of the system, just like the virtual (massless) photons of the electromagnetic field somehow contribute to the mass of the system. The nonlinear character of the Einstein equations also seems to indicate that the "gravitational field" itself gravitates (I know this is strictly not correct, I use this only to guide the intuition). So, if the gravitons themselves are subject to the gravitational field, one should have virtual gravitions exchanging between these gravitons and other gravitons or other masses - AND - you can not stop this process: you just have to keep on adding gravitons to include all the reactions. So, you get an infinite number of gravitons per unit volume resulting in an infinite energy density and the whole of space would be just a massive "block" of gravitons. A ridiculous situation. This proofs that gravitons can not exist.

There are iterative processes that don't necessarily result in infinity. Self-gravitation may be one of those. In my opinion, it might even be possible that self-gravitation may result in the gravitational effects of dark matter, who knows?

 

[jimgraber]

Disclosure: The giant graviton industry is not paying me a huge bribe to write this post. (Drats!)

I think there are five main categories of alternatives seriously studied or used at the present time:

1. Do not quantize gravity. Keep it continuous.

2. The “old fashioned” spin 2 point particle graviton approach. You can modify this slightly by giving the graviton an extremely tiny rest mass.

3. The “new-fangled” string theory closed string graviton, which is also a spin 2 zero rest mass object.

4. Mixing a spin 2 zero rest mass graviton with a little bit of something else. (The classic thing is a tiny bit of one or more scalar i.e. spin 0 fields, but a little bit of a massive spin 2 graviton has also been looked at. String theory can also generate lots of scalar fields, often called moduli. The new TeVeS theory, related to MOND, also mixes in some vector, or spin 1, contribution.) Scalar-tensor gravity is a big industry, once in eclipse, but now revived by the string theory connection, and also by attempts to work on the dark energy and the dark matter.

5. Replacing gravitons with elementary moves OF or ON some lattice (Dynamic triangulation, spin foam, or what have you. If you want to get fancy, make your lattice out of ribbons rather than lines.) In this case the graviton is not fundamental, but something that acts like a zero rest mass spin two graviton can usually be constructed from these elementary lattice moves. (Otherwise, the theory is likely to predict wrong answers.)

To me, there is very little difference between a fundamental graviton and one that is composite or derived. Of course, if these lattices help build a unified theory, or do some other useful work, more power to them. But that’s different from saying the graviton does not exist, or there is no graviton.

Also, as far as I know, there is no fully worked out version of anything in class 5 that does not have a graviton equivalent. Please correct me if I’m wrong.

Now about quantum theory in curved spacetime. Despite what you might think from the name, it falls in class 1.

I quote the closing lines of Bob Wald’s recent review:

arXiv:gr-qc/0608018v1 3 Aug 2006
The History and Present Status of Quantum Field
Theory in Curved Spacetime
Robert M. Wald


“All of the above results have
been obtained without any appeal to a notion of “vacuum” or “particles”.
These and other results of the past decade have demonstrated that quantum field
theory in curved spacetime has a mathematical structure that is comparable in depth to
such theories as classical general relativity. In particular, it is highly nontrivial that quan-
tum field theory in curved spacetime appears to be mathematically consistent. Although
quantum field theory in curved spacetime cannot be a fundamental description of nature
since gravity itself is treated classically, it seems hard to believe that it is not capturing
some fundamental properties of nature.
The above results suffice to define interacting quantum field theory in curved spacetime
at a perturbative level. However, it remains very much an open issue as to how to provide
a non-perturbative formulation of interacting quantum field theory in curved spacetime.
It is my hope that significant progress will be made on this issue in the coming years.”

It is very true that particles are not appealed to in this theory, but gravity is not quantized.
This also means photons and electrons are not appealed to. There is a very old way of thinking and speaking which says in effect that “photons and electrons do not exist, they are just knots in the electromagnetic field”. You can obviously do the same with gravitons, if you wish. To me it is not clear that anything is gained by this way of thinking or talking, but YMMV.

As I wrote earlier, the way to cut to the chase is to ask yourself, “Do you believe Planck’s constant is relevant for the other forces, but not for gravity?” For me, the answer is “No” and that rules out all approaches in category 1.

Finally, I give below a reference for the work, initiated by Freeman Dyson, but completed by Rothman and Boughn, that shows how difficult it is to detect a single graviton, as opposed to a wave composed of “billions and billions” of gravitons.


arXiv:gr-qc/0601043 [ps, pdf, other]
Title: Can Gravitons Be Detected?
Authors: Tony Rothman, Stephen Boughn
Comments: This version as appeared in Foundations of Physics
Journal-ref: Found.Phys. 36 (2006) 1801-1825
Subjects: General Relativity and Quantum Cosmology (gr-qc); Astrophysics (astro-ph)

Best to all
Jim Graber

 

[Coin]

Hi jimgraber, that's an extremely interesting overview. To get offtopic for a moment though I'm very curious about this option:

1. Do not quantize gravity. Keep it continuous.

What would be an example of this, besides the "QFT in curved spacetime" people you cite? I am not aware of any work that reconciles gravity with quantum theory without quantizing gravity (and if I understand your quote it sounds like the curved QFT people only have half the equation so to speak, knowing how to perform QFT on surfaces where gravity exists but not knowing how quantum "things" can impact the gravitational field).

 

[Haelfix]

I think Jim means that one proposal is that you have general relativity and field theory, and the two don't mix at all, and they are taken as separate formalisms. Eg three forces are described by field theory, and gravity is done via general relativity and stays classical through all energy scales. Some people still believe this (a minority).

Its important to note that field theory in curved space vs classical general relavitiy have different physical predictions. If you do field theory in curved space, gravitational interactions will receive quantum corrections not unlike how say a photon receives radiative corrections and leads to the Lamb shift. We cannot measure such things yet in the lab or from the cosmos, so no one knows what the answer is, but for instance things like the black hole area-entropy law makes sense if you work with field theory, but doesn't at all with just classical GR.

Now, there is yet another point of view that says.. Ok, well we will take field theory in curved space as fundamental. Who cares if its nonrenormalizable and ceases to be predictive around the Planck scale, that's just a technical problem due to human incompetence. Boom, done, QG is solved! Some people also believe in this (also a minority)

String theory basically takes the last point of view, except that it UV completes the field theory so that it becomes predictive. It 'completes' and finishes the last few orders of magnitude not well described by qft in curved space and limits it to within a larger theory.

 

[Coin]

Haelfix, thanks, that's really interesting. By the way, did you notice that was your 666th post?

Kind of a tangential question:

Now, there is yet another point of view that says.. Ok, well we will take field theory in curved space as fundamental. Who cares if its nonrenormalizable and ceases to be predictive around the Planck scale, that's just a technical problem due to human incompetence. Boom, done, QG is solved!

So if we do this... exactly what kind of predictions are we "losing"? That is to say, is there anything happening around the Planck scale which, if we lost predictivity around the Planck scale, it would impact our ability to solve non-planck-scale problems? I am suddenly, to my slight embarrassment, realizing I cannot think of any real-world problems that we need quantum gravity to solve that do not involve black holes. Are there any good examples of what kind of non-black-hole questions we are limiting ourselves from by taking this "whatever" approach to curved qft?

 

[Haelfix]

Its an open question. To my knowledge the only glimmer of a chance is big bang physics. For instance inflation.

Inflation is an incomplete physical model. What you do is sort of guess the form of some quantum field (something that would naively be involved with QG), and then evolve it (by that point the universe can be described classically via GR). Now we really are interested in the exact nature and shape of that guess, and it would be nice if we could derive it from first principles (no one has yet).

The problem here is that by the very nature of inflation and efolding, it washes away most of the information contained in that initial guess. For instance if you take the slow roll approximation, and inflate the universe through ~60 efolds its really the only last 2 or 3 efolds that have any observational consequence, everything else gets swamped.

So yea, even that might be lost to us.

 

[friend]

Its an open question. To my knowledge the only glimmer of a chance is big bang physics. For instance inflation.

Inflation is an incomplete physical model. What you do is sort of guess the form of some quantum field (something that would naively be involved with QG)...

So if we at present don't have any experimental evidence to support it, why the interest in quantizing gravity? I mean do we suppose GR should be quantized simply because it is just another field? Or are there more substantial reasons that necessitate quantizing gravity? Thanks

 

[Haelfix]

Well, there are plenty of purely theoretical and aesthetic reasons. But for instance the success of inflation begs the question already at an observational level.

It requires some sort of quantum fluctuation in the early universe to plant the seeds necessary for the dynamics to take over. Now these fluctuations occur close enough to the Planck scale that they should in principle have some elements of QG dynamics in them, and its not clear how you would do away with them.

Without inflation you run into the usual problems faced by cosmology in the 80s like
1) Monopole abundance
2) flatness problems
3) Horizon problems

And people view these as serious enough to accept the paradigm (or some slight modification thereof).