What is the evidence for GR and against the graviton theory?

[Hurkyl]

I said that in response to, "We also have no theory of gravitons that give the same behaviour as GR," because you seemed to be indicating that the graviton theory was not a valid competing model. I was pointing out that a theory does not need to give the same behavior as GR in order to be a competing theory. The graviton and GR would both be able to explain all observed behavior,

And since (to my knowledge) there aren't any graviton theories that agree with all observed behavior...

(I'm assuming a theory has to be able to make fairly specific predictions before we consider it)

 

[junglebeast]

And since (to my knowledge) there aren't any graviton theories that agree with all observed behavior...

In this framework, the gravitational interaction is mediated by gravitons, instead of being described in terms of curved spacetime as in general relativity. In the classical limit, both approaches give identical results, which are required to conform to Newton's law of gravitation. ... Some proposed theories of quantum gravity[7] (in particular, string theory) attempt to address this issue. In string theory, gravitons (as well as the other particles) are states of strings rather than point particles, and then the infinities do not appear, while the low-energy behavior can still be approximated by a quantum field theory of point particles. In that case, the description in terms of gravitons serves as a low-energy effective theory.

http://en.wikipedia.org/wiki/Graviton

No, what makes GR "correct" is that it matches the observations of every experiment yet devised to challenge it. And finding a graviton doesn't necessarily contradict GR if it is outside the scope of what the math of GR can predict.

By your logic, one could conclude that all theories which explain the same effect can be simultaneously true. That's nonsense! The theories are mutually exclusive

 

[Fredrik]

And since (to my knowledge) there aren't any graviton theories that agree with all observed behavior...

I could be wrong, but I think there's a (non-renormalizable) QFT of gravity (based on the idea of writing the metric as $g=\eta+h$ and only quantizing h) that makes the same predictions as GR if you ignore all the Feynman diagrams that include a loop.

By your logic, one could conclude that all theories which explain the same effect can be simultaneously true. That's nonsense! The theories are mutually exclusive

Don't forget that the only thing experiments can tell us is how accurate a theory's predictions are, or that no theory that's been found so far can possibly be "true".

 

[f95toli]

By your logic, one could conclude that all theories which explain the same effect can be simultaneously true. That's nonsense! The theories are mutually exclusive

There is no such thing as a "true" theory. If two theories give the same result to within the accuracy of an experiment they are both correct in that situation; an obvious example would be SR and classical mechanics. Since most experiments (e.g. all engineering projects except a few notable exceptions such as the GPS) are performed at speeds much lower than c they both agree with experiment (which is why we tend to use Newton's laws most of the time).

Now, SR is obviously valid in situations where classical mechanics is not; but I don't think anyone would argue that Newton's laws are "wrong"; they are simply not valid in all the situations we can test experimentally.

 

[junglebeast]

There is no such thing as a "true" theory. If two theories give the same result to within the accuracy of an experiment they are both correct in that situation...Now, SR is obviously valid in situations where classical mechanics is not; but I don't think anyone would argue that Newton's laws are "wrong"; they are simply not valid in all the situations we can test experimentally.

There is such a thing as a true theory. I think, therefore, I am; therefore the universe is real, and has real laws. It is true that we can never prove one of our made up theories to be exactly in accordance with a true law of the universe, but the goal in science is to keep approaching the truth in the hopes that someday, we actually do know the true laws. Perhaps we are overly ambitious as a species by our pattern of always believing the latest theory up until the point where it is disproven. But once a better theory arises, we can reject an old theory.

Newton's laws are wrong. Undeniably, they are wrong. But they are very good approximations at certain scales, so we still use them as mathematical tools. But it's important to recognize that they are just approximations and we keep in perspective the knowledge of quantum mechanics so that we know where we can and can't use Newton's laws.

GR and the graviton cannot both be true, but one of them might be true. It is logical to continue using GR as a mathematical tool, but it is illogical to believe that GR is correct and can be extrapolated into all circumstances when there are other competing theories sticking their noses out.

Theories of Physics are not only used in the context of the experiments which are used to validate them. They are extrapolated out to situations in which we haven't made observations yet in order to make predictions. They are used to postulate on the origins of our existence in the universe, and they do have philosophical repercussions.

 

[jamesb-uk]

Newton's laws are wrong. Undeniably, they are wrong. But they are very good approximations at certain scales, so we still use them as mathematical tools. But it's important to recognize that they are just approximations and we keep in perspective the knowledge of quantum mechanics so that we know where we can and can't use Newton's laws.

If you believe in quantum mechanics over Newton's laws, then why do you believe in 'the graviton' over GR? Why must quantum mechanics be right, but Newton's laws and GR can't?

I thought the whole point of your argument was that we cannot be certain of any theory, so how can you be certain Newton was wrong, if you are unsure about GR?

 

[Hurkyl]

http://en.wikipedia.org/wiki/Graviton

The next sentence in that article says:

However, attempts to extend the Standard Model with gravitons have run into serious theoretical difficulties​

and this is consistent with my understanding of the situation: we know methods that look like they might give the right answer, but we can't figure out how to make them to work.

(I believe the description there is oversimplified, but I'll let someone more knowledgeable comment on that)

By your logic, one could conclude that all theories which explain the same effect can be simultaneously true. That's nonsense! The theories are mutually exclusive

(I was never talking about "theories which explain the same effect". I was talking about "theories that make agree with observed behavior". There is a difference, y'know: most theories which purport 'explain' some effect get wrong answers)

(1) As the quote said, what makes a theory "correct" is that it matches the observations of every experiment yet devised to challenge it. We use empirical evidence to evaluate the "truth" of our theories, because we don't have some mythical "truth valuation of the universe" which is a function whose inputs are mathematical statements and whose outputs are truth values. Even if such a mythical truth valuation existed, we could never be sure we knew what it was, so it's not a useful concept.

(2) If two theories make exactly the same predictions, then there is nothing to decide between them, no matter how radically different they might appear at first...

(2) But really, you simply aren't being imaginative enough. Go mull over a simple purely mathematical example for a while -- the theory of real number arithmetic and the theory of Euclidean geometry happen to be exactly equivalent mathematical theories.

 

[junglebeast]

If you believe in quantum mechanics over Newton's laws, then why do you believe in 'the graviton' over GR? Why must quantum mechanics be right, but Newton's laws and GR can't?

I thought the whole point of your argument was that we cannot be certain of any theory, so how can you be certain Newton was wrong, if you are unsure about GR?

Newton's laws do not match observations, so they can't be true. Quantum laws do match observations, and there does not exist any other good theory that explains those observations, so to the best of our knowledge quantum theory is true.

GR is one of several theories that matches observations, and the best of our knowledge is not able to rule out other competing theories. Until all but 1 theory can be ruled out, it would be presumptuous to trust any 1 of those contending theories. It would be no different than gambling on which Survivor will win the contest. They aren't all going to win (they aren't all right), but we don't know which.

Whenever there are 2 or more contradictory theories which predict the same results as all experiments (which is the case with GR And the graviton), then none of the experiments can be used as evidence for either theory. It would be like saying, "I heard you talking on the phone today. Therefore, you must have been talking to my sister." No, because it could have been anyone. Likewise, the "evidence" for GR is equally evidence for quantum gravity, and it stops being evidence for either theory over the other. The only evidence that remains for either theory is the fact that the other forces are quantized, which gives a slight edge to the likelihood of quantum gravity.

 

[junglebeast]

The next sentence in that article says:

However, attempts to extend the Standard Model with gravitons have run into serious theoretical difficulties​

and this is consistent with my understanding of the situation: we know methods that look like they might give the right answer, but we can't figure out how to make them to work.

But the next sentence after that proposes a way..

(1) As the quote said, what makes a theory "correct" is that it matches the observations of every experiment yet devised to challenge it. We use empirical evidence to evaluate the "truth" of our theories, because we don't have some mythical "truth valuation of the universe" which is a function whose inputs are mathematical statements and whose outputs are truth values. Even if such a mythical truth valuation existed, we could never be sure we knew what it was, so it's not a useful concept.

(2) If two theories make exactly the same predictions, then there is nothing to decide between them, no matter how radically different they might appear at first...

You are not being logical. Let me try putting this into mathematical terms. A and B are theories, C is observation.

A implies C.
B implies C.
C is true.
You cannot conclude that either A or B is true.

A theory is true under these circumstances only:

A implies C.
Nothing else implies C (so C implies A).
C is true.
Therefore theory A is true.

In reality, we can never prove that "nothing else implies C." We accept a theory as being likely true under this circumstance:

A implies C.
Nothing else that we know of implies C.
C is true.
Therefore theory A is true, to the best of our knowledge.

(2) But really, you simply aren't being imaginative enough. Go mull over a simple purely mathematical example for a while -- the theory of real number arithmetic and the theory of Euclidean geometry happen to be exactly equivalent mathematical theories.

Sorry, but no -- you are confusing theories with theorems. Theories are well-supported hypothesis based on observations, whereas theorems are true by definition. There are no theories in math. A theorem is not accepted until it is proven that it directly follows from other definitions or assumptions under all conditions.

Whereas 2 mathematical theorems may be equivalent, we are not discussing theories which are equivalent. We are discussing theories which are equivalent only up to the presently observed cases. They are not equivalent when extrapolating into all new cases, and are therefore mutually exclusive theories -- NOT just a change in perspective.

 

[Hurkyl]

But the next sentence after that proposes a way..

Sure. But a proposal is merely a proposal.

A theory is true under these circumstances only:

A implies C.
Nothing else implies C (so C implies A).
C is true.
Therefore theory A is true.

In reality, we can never prove that "nothing else implies C."

I've never heard of that logical rule of deduction. Care to provide a reference?

It's irrelevant, though:
(1) "A implies C" cannot be known with certainty
(2) "Nothing else implies C" is trivially easy to disprove. (For example, C implies C)

There are no theories in math.

You lose.

The word is used slightly differently in physics, but none of the differences are relevant to the discussion.

 

[jamesb-uk]

Newton's laws do not match observations, so they can't be true. Quantum laws do match observations, and there does not exist any other good theory that explains those observations, so to the best of our knowledge quantum theory is true.

GR is one of several theories that matches observations, and the best of our knowledge is not able to rule out other competing theories. Until all but 1 theory can be ruled out, it would be presumptuous to trust any 1 of those contending theories. It would be no different than gambling on which Survivor will win the contest. They aren't all going to win (they aren't all right), but we don't know which.

Just because a theory we have is the only one which matches observations, it is not necessarily true- we may very well observe some things in the future which contradict quantum theory- as has happened to many theories in the past. According to your argument, you should feel exactly the same way you do about GR, as all theories which comply with observations.

 

[Tibarn]

The only evidence that remains for either theory is the fact that the other forces are quantized, which gives a slight edge to the likelihood of quantum gravity.

The validity of this argument depends on whether you consider gravity to be a "true" force, which GR does not.

 

[junglebeast]

Sure. But a proposal is merely a proposal.

Well, that's all any theory is -- an idea that was proposed which works to explain observations.

I've never heard of that logical rule of deduction. Care to provide a reference?

If A implies B and B implies A then A = B. I don't remember the name for that logical rule but it's fairly obvious and I shouldn't have to give a reference for it.

The logical fallacy you were committing is called "Affirming the consequent": the antecedent in an indicative conditional is claimed to be true because the consequent is true; if A, then B; B, therefore A.

It's irrelevant, though:
(1) "A implies C" cannot be known with certainty
(2) "Nothing else implies C" is trivially easy to disprove. (For example, C implies C)

(1) A was defined to be a theory which predicts an observation C...so A is chosen by definition to imply C.

(2) It was meant, "no other theory" implies observation "C". An observation is itself not a theory.


You lose.

The word is used slightly differently in physics, but none of the differences are relevant to the discussion.[/QUOTE]

Ok..whatever. There are mathematical theories. But the differences ARE relevant, in the way that I already pointed out, which you ignored.

 

[belliott4488]

junglebeast-

Why do you say that a theory of gravitons (Quantum Gravity) and GR are contradictory? I thought QG is just the proposed quantum field theoretical extension of GR, just as QM is the quantum extension of Classical Mechanics. Not only is QG not contradictory to GR, a fundamental requirement on any graviton theory is that it must match GR in the classical limit, just as QM must reproduce Classical Mech. Do you think that a graviton theory will produce a classical theory other than GR in the appropriate limit? If not, and your preferred graviton theory reproduces all the prediction of GR in all the currently tested energy regimes - how is that a contradictory theory?

I don't know of anyone who says that GR is "true", in the sense of being the only theory of gravity that will ever be needed to predict any conceivable observation. In fact, it's precisely the breakdown of the the theory at the Planck scale that leads us to say that there must be some kind of QFT for gravity - hence the graviton. But GR is the starting point for such theories - they don't contradict it.

 

[Hurkyl]

Well, that's all any theory is -- an idea that was proposed which works to explain observations.

"Proposal" is not synonymous with "an idea that works".

If A implies B and B implies A then A = B. I don't remember the name for that logical rule but it's fairly obvious and I shouldn't have to give a reference for it.

The rule you used was

A implies C.
Nothing else implies C.
C is true.
Therefore theory A is true.​

I decided to analyze it for fun. It turns out that the pair of statements

A => C
For all B: (B => C) => (B <=> A)​

is logically equivalent to the pair of statements

not A
not C​

The logical fallacy you were committing is called "Affirming the consequent": the antecedent in an indicative conditional is claimed to be true because the consequent is true; if A, then B; B, therefore A.

That doesn't resemble anything I said.

(1) A was defined to be a theory which predicts an observation C...so A is chosen by definition to imply C.

Yes, I was wrong -- I was ignoring the degenerate cases. Realistic cases come with uncertainty attached, though.

(2) It was meant, "no other theory" implies observation "C". An observation is itself not a theory.

If your argument is defeated by a degenerate case, it doesn't suddenly become valid if you declare by fiat that we shouldn't consider degenerate cases.

 

[belliott4488]

There is such a thing as a true theory. I think, therefore, I am; therefore the universe is real, and has real laws. It is true that we can never prove one of our made up theories to be exactly in accordance with a true law of the universe, but the goal in science is to keep approaching the truth in the hopes that someday, we actually do know the true laws. Perhaps we are overly ambitious as a species by our pattern of always believing the latest theory up until the point where it is disproven. But once a better theory arises, we can reject an old theory.

Sorry, I think you're conflating a matter of faith with science. A professor I once heard teaching intro-level college Physics (I was a TA, not a student) used to emphatically attempt to drive home the point that no (finite) number of experiments can ever prove a theory true, but it takes only one* experiment to prove it false.

(*Technically, it takes two, since a second one is needed to confirm the results of the first, but that's not really a second experiment; it's a repetition of the first one.)

If you think there are "true" theories floating out in Heaven and that it is our objective to try to approach these ideal theories with our imperfect ones, well - there were certainly a lot of Greek philosophers who would have agreed with you (as would have Newton, it should be pointed out), but I believe that way of thinking is of little relevance to modern science.

All modern scientific theories are conditional, in the sense that they are accepted only as long as they continue to make predictions that agree with observation, but we must always remain open to the possibility that a new experiment will prove even our best theory false. That's why we no longer call the axioms of our theories "Laws", as Newton did in the belief that he was discovering the true laws of God.

GR so far has produced no predictions that are not verified by observation - in that sense it is as "true" as any theory we've ever had. The desire to create a Quantum Field Theory of gravity (and the graviton) comes about because we expect that eventually we will find an experiment whose outcome is not correctly predicted by GR (at the Planck scale, most likely), but no such theory exists yet, in the sense of its being able to make testable predictions. For that reason, it makes no sense to speak of "choosing a graviton theory over GR" - we all hope to have a QFT for gravity some day, but it will necessarily include the predictions of GR in the classical limit.

 

[Vanadium 50]

I've spoken with several physicists on the matter, none of which have been able to point to any evidence that strongly supports GR over the standard model.

What makes you think the standard model has anything to say about gravity at all?

You are drawing-up a conflict that doesn't exist.

Exactly.

 

[Vanadium 50]

My understanding (I'm not a particle physicist) is that the graviton model consists of taking the predictions of GR and determining what properties a graviton would have to match them.

That's essentially correct. What one does is ask what sort of quantum theory would have the the weak-field limit of GR as it's long-distance and weak-field limit.

Because the quantum theory is invalid in the strong-field case, there's no way to match this to the strong-field case of GR.

 

[Vanadium 50]

1) GR was proposed to explain the bending of light by gravity. Although it was a WACKY idea, people chose to accept it because it explained the observable events when nothing else did. Also, people may have been more open to the concept in light of the current political situation.

Virtually none of this is correct. Let me address these points one by one.

GR was not proposed to explain the bending of light by gravity. GR was published in 1916, and the famed Eddington eclipse expedition was in 1919. The deflection of light by gravity in fact is predicted classically: Cavendish discussed the possibility of Newtonian black holes in 1783, as did Laplace in 1796.

2) It was discovered that all of the other fundamental forces are mediated by gauge bosons in the standard model, and all of their associated bosons have been observed. This theory makes intuitive sense.

This is an oversimplifcation. "Mediated by gauge bosons" means that one can take the classical field, and when treated quantum mechanically there emerge discrete excitations of the field. For example, if you start with the discrete electromagnetic fields, you get a photon out. But the existence of a photon doesn't mean that we toss out classical electromagnetism any more than the existence of a graviton means we toss out it's classical field theory: GR.

3) The idea that gravity may also be mediated by a boson is logical, and was proposed, but since we already have a working model for gravity and we can't find a graviton...GR stays. Of course, we have no equipment capable of detecting the graviton, so we have no way to disprove GR.

I challenge you to calculate the perihelion advance of Mercury using gravitons.

The reason people use GR and not a quantum theory of gravity is that they can get answers out of GR, and these answers match observations. We don't have an internally consistent quantum theory of gravity - what you call "graviton theory".

4) The evidence for GR is not actually evidence for GR. It's really just evidence that gravity bends light. But the graviton can also explain that.

Well, that and:

• Perihelion advance of Mercury
• Gravitational time dilation
• Shapiro delay
• Equivalence of free-fall
• Nordtvedt effect nonexistence

So as far as I can tell, both theories explain the observations, and the only reason to believe in GR is it was proposed first.

Are you able to do either of these calculations? I suspect not, because nobody can do calculations in "graviton theory", as the theory is internally inconsistent. What people do is they say "this corresponds to GR in the weak-field long-distance limit" and then do the GR calculation.

If you can't do the calculations yourself, how can you tell that one set is better than the other?