# The speed of gravity

Discussion in 'Astronomy, Exobiology, & Cosmology' started by Aloysius, May 30, 1999.

1. ### AloysiusGuest

When we calculate the orbit of the earth due to the sun, do we use the instantaneous position of the sun, or do we use the position of the sun 8 minutes ago? (which is the length of time it takes for light to reach us from the sun).

I hope that at least Boris will address this.

3. ### PlatoGuest

We calculate our position from the sun with the simple laws of Newton. So at each instant in time, we know where we are at the ellips (which is the orbit).
If something would happen to the sun though, we woudn't notice it until the eight minutes have past ! Suppose the sun suddenly (by a stroke of magic) vanished, the earth would continue in it's orbit until the eight minutes have passed. You see for us everything that happens on the sun is eight minutes old, this is the consequence of the local nature of the fundamental forces.
So if we want, by experiment (this means looking at the sun), to know what the distance is of the sun to the earth, we are measuring this distance to the eight minute old sun because that is the only real sun for us.

------------------
greetings,
Plato

5. ### OOM-9Guest

Gravity travels at the speed of light. Gravitational waves, like those which are being searched for by the LIGO project, travel outward from binary black holes or neutron stars at the speed of light. If the sun were to instantaneously disappear, the earth would continue in its orbit for 8 minutes before moving off in a straight line. So in this sense the answer to your question is that yes, we use the position of the Sun 8 minutes ago in our orbit calculations. of course, as we see it, that's the correct time anyway, it would be more of a pain to forward-calculate where the Sun REALLY is now, when we won't know that for another 9 minutes. Hope this helps.

7. ### BorisGuest

I agree with both Plato & OOM-9. (Are these names getting stranger by the day, or what?

)

The current force Earth feels is due to the Sun's position 8 minutes ago. The answer is definitely 'yes'. Although, to be a pain in Plato's behind, I'll mention that our orbit is not an ellipse. There are relativistic effects to consider (precession), as well as multiple-body effects (the Sun is also wabbling instead of staying still). But to be fair, those effects for Earth are so tiny, that Newtonian approximation works extremely well.

However!! Nobody has yet proven that gravity indeed propagates at speed of light; that's only a result stemming from Einsteins General Relativity. (although, again to be fair, Einstein hasn't been wrong yet! (at least not with GR)

) Among other things, LIGO will either confirm or disprove this particular result. So, the final answer is, "probably".

------------------
I am; therefore I think.

8. ### AloysiusGuest

Well, I am informed otherwise, to my consternation, by the physicist Tom van Flandern.

On http://www.vigier.yorku.ca/VigierII/VANFLAND.html
he writes:

THE SPEED OF GRAVITY -- WHAT THE EXPERIMENTS SAY

T. Van Flandern
University of Maryland, USA.

If gravity from the Sun propagated outward at the speed of light, the transmission delay would progressively increase the angular momentum of bodies orbiting the Sun at so great a rate that orbital radii would
double in about 1000 revolutions. Direct measurements of the directions of bodies and their accelerations (using, e.g., planetary radar ranging) show that, while light of any wavelength undergoes aberration as
an immediate consequence of its finite speed, gravity has no such aberration at a detectable level. Dynamical studies of binary pulsars show that not only the position and velocity of a source of gravity are
anticipated without light-time delay but accelerations of the source are anticipated as well, in contrast with the behavior of electromagnetic forces. Indeed, Newton's universal law of gravity, to which general
relativity is supposed to reduce in the low-velocity, weak-field limit, unconditionally requires infinite propagation speed for gravity. These paradoxes are supposed to be explained by general relativity's curved
spacetime interpretation of gravity. Yet that interpretation leads to new, equally unresolvable paradoxes, especially acute in the case of binary black holes. Moreover, that interpretation is in conflict with
results from neutron interferometer experiments (D.M. Greenberger and A.W. Overhauser, Sci.Amer. 242, May, pp. 66-76, 1980). One resolution of the paradoxes is to interpret the experiments literally,
and from them deduce that the speed of propagation of gravitational force is no less than 1010 c. Although this is inconsistent with the Einstein interpretation of special relativity, it is perfectly consistent with
Lorentzian relativity, a variant espoused by Lorentz until late in his career, which behaves like Einstein relativity in most particulars, but permits faster-than-light propagation in forward time. This subtle
alteration in our thinking about what is allowed under the laws of physics has several immediate consequences, all beneficial, occurring in areas that are also facing paradoxes or internal contradictions, such as
the locality dilemma of quantum mechanics or the existence of singularities in nature ("black holes"). The "speed of gravity" issue may be pointing us to the "wrong turn" physics made to get into these
difficulties.

His home page, with a much fuller account of this assertion, can be found at:

http://www.ldolphin.org/vanFlandern/gravityspeed.html

9. ### BorisGuest

This is amazing! Basically, I've been trying to contemplate a 'refracting medium' interpretation all on my own; I never knew there was so much activity already around the issue. Many thanks for bringing this to my (our

) attention!

Plato and others who have read my posts in "With regard to space access" thread, will probably see that I've been having very similar problems with the standard formulations of QED and relativity. I've been looking for an 'underlying medium', or in a way, a preferred frame of reference, and it seems I'm not alone! (nor the first

!)

Anyway, a lot of what Dr. Van Flandern says makes perfect sense to me. Although some of his statements concerning acceleration within the geometrical framework of GR are a bit off, I also see how an underlying medium with varying density can be equivalent to warped geometry -- heck, lately I've been thinking along some very similar lines... The aberration argument also seems to be pretty solid; at least I can't find a flaw in it. Though of course I'm not neccessarily saying that the theory is sound. After all, I'm not a world-renown authority in cosmology (or math for that matter) -- so at this point I'm very cautious of what my intuition is trying to tell me.

The few points in the paper I'm not sure are correct:

1) There is a reference to stretching a taut line between two bodies on the same orbit, and claiming that this is a straight line in local spacetime. Well, in fact it's not a straight line (geodesic) as far as GR is concerned; although the line may achieve the shortest distance, it does not represent the shortest Minkowsky interval between the orbiters.

2) A claim is made that gravitational fields somehow transfer momentum, and therefore cannot be static (as in non-regenerating). Well, this can still be accomplished if we remember that it's not just the 'source' that emanates a field, but that the target does so as well; so whereas the source's field affects the target, the target's field affects the source. So, there is really no communication necessary between the two in order for momentum transfer to occur.

Hey Plato, do you think it couldn't be just gravity that goes faster than light, but also electroweak and strong forces? (basically, anything except EM?) Could be a neat explanation for the EPR effects, and why universe-wide wave functions are a good approximation to experiment!

But anyway, it's pretty promising stuff (I think). I look forward to what may develop henceforth.

------------------
I am; therefore I think.

10. ### PlatoGuest

I hate to break the expectations but I would ask for some caution in this matter. For one thing, I did a little search on other things this Van Flandern guy (this is stange because I am actually of Flanders) published and he is kind of notorious of putting forward controversial ideas. Like claiming that Mercury is a former moon of Venus or discediting the Big Bang theory.
I'm not saying there is anything wrong with that because science needs people like him to pour out new ideas but what I am saying is that his theories might leap ahead on experimental evidence.
He claims that there is experimental evidence that gravity travels at 10^10 c. On the other hand under "Does gravity travel at the speed of light? [98-05-06]" at http://src.doc.ic.ac.uk/usenet/news-faqs/sci.astro/%5bsci.astro%5d_Astrophysics_(Astronomy_Frequently_Asked_Questions)_(4_8).
There they talk about the observations of binary neutron stars where the orbit is gradually decaying. This can only be because of gravitational radiation, just like in electromagnatic theory. This automatically implicates that gravitation has a finite speed. If one agrees with the general relativity theory, the dampening effect leads to a speed of gravity as being the same as the speed of light, with an error of 1%.
There are other experiments on the way to further refine these findings but a speed of 10^10 c is a bit more than 1% deviation I would think...

------------------
greetings,
Plato

11. ### AloysiusGuest

Plato, you mention gravity waves travelling at light speed, and I (and Van Flandern) agree with this. It's the speed of assertion of the gravitational potential, not gravity waves (which require acceleration to be produced) that is under discussion here.

To be more concise, "What's the speed of gravity?" refers not to gravity waves, but to the gravitational field. I wonder if that's causing confusion here.

As a footnote, I would have thought an experiment could be quite easily set up to check van Flandern's assertion. After all, gravitational fields are much stronger than waves (for normal events around here, else we'd be dead

) so one could precalculate the two possible orbits for a given set-up (one for instantaneous and one for lightspeed-delayed interaction) and just go measure the darned thing.

I would have thought that this had already been done. Van Flandern asserts, using as his set-up the earth-sun system, that he has an open-and-shut case. I'd like to take a closer look at that I think, because that's the lynchpin of his assertion right there - and checkable against hard experimental evidence with some not-too-hard maths.

12. ### PlatoGuest

Aloysius,

I might have been on the wrong subject there but have you read the URL link ? It also talks about the field itself. If one does calculations of the retarded field (in electromagnetics as in GR) there it seems as if the force still works instantanious while the speed limit of the virtual gravitons is still limited to the lightspeed, as is the speed of the virtual photons who give way to the electromagnetic force.
Now I do know that regarding GR I shouln't mention the gravitons because they arise in quantum gravitational theories but I was merely using them to get the analogy with photons right.
You see, gravitation waves consist of real gravitons, these are the same physical particles as their virtual counterparts who give rise to the gravitational field.

------------------
greetings,
Plato

13. ### BorisGuest

Plato:

As I said, I'm not all that sure that Dr. Van Flandern (VF) has found a true problem with GR. However, even after reading the posts you referenced, I still think there might be something to it.

a) to be calculating retarded potentials, and in general to be treating geometry as if it was a field, you need some kind of a restoring force that will 'untwist' the geometry after a massive object has passed. There's nothing like that inherent in GR, as far as I am aware. Also, I find it amazing that the restoring force, even if it exists, is perfectly dampened (i.e. it doesn't oscillate the geometry back and forth, but simply restores it directly to the most relaxed state.)

b) VF has a point about reference frames. (Did you read the paper on his website?) In fact, I have a thought experiment to practically demonstrate that a universal preferred reference frame exists and is measurable. (Although I might be missing something, and may be totally wrong...) The preferred reference frame would be the same one in which gravity wells exist and light propagates. Here's my thought experiment:

Assume for now that nothing exists in the entire universe except for the following setup:
Imagine you have a really massive object (e.g. a neutron star) being orbited by two very light objects (the weight of ping-pong balls). The two orbiters are actually on the same orbit, but separated from each other. Now, in this system the neutron star is stationary to astonishing accuracy, so I'll treat it as stationary (and call it N). Also assume that the two orbiters (call them A and B, and assume that A is ahead of B in direction of orbit) are on a perfectly circular orbit. Now, imagine N emits a flash of light. The spherical wavefront propagates out and strikes A and B simultaneously. At this point, each orbiter emits its own flash of light (of a different frequency maybe). Keep in mind that throughout this, both A and B are continuing forward on their orbits. The wavefront from A propagates out and strikes B, and likewise B's wavefront will eventually reach A. However, A's wavefront will strike B first because B will in effect 'run into' it. A, on the other hand, is 'running away' from B's wavefront, and will receive B's flash later. Now, as soon as B receives A's flash, it sends a signal back to N; likewise A sends a signal to N as soon as it receives B's flash. The two signals take exactly the same time to propagate back to N since the orbit is perfectly circular. Therefore, N will receive a signal from B first, and only later will it receive a signal from A. From N's point of view, this looks like the only plausible sequence of events. At least I can think of no other way this system could work.

However, what I just showed is that light propagates within N's reference frame! That includes light from both A and B. However, A and B are both in freefall, and N's reference frame should not be preferred over theirs! In fact, this way A and B can use N as an intermediary to measure their velocities with respect to this absolute reference frame -- using none other than light itself!! Notice, however, that in my perfectly empty universe (and assuming N doesn't have any distinguishing marks on its surface), A and B originally have no way of telling in which direction they are moving (i.e. which is ahead, A or B?) However, simply by sending and receiving a few light signals, they are able to make just such a determination, which goes against the principle of equivalence!

Now, imagine that this three-body system is orbiting yet another object, still far more massive. Now, by similar arguments, we will see that the velocity of the three-body system (essentially of N) will also influence its reception of signals from A and B. However, this influence will manifest itself differently when A and B are moving in N's direction of motion rather than opposite to N's direction of motion. Suppose that the super-massive object around which N orbits is a black hole, H. Then, using its two orbiters, N can measure its velocity around H! Also, even if N can't see H, it should be able to deduce in which direction H (i.e. the center of mass) is located! And not only that, A and B can measure their absolute velocities now as well...

By extension, we can now add the rest of the universe -- and see that even with all the complex gravitational contributions to the orbit of N, A and B can still measure their absolute velocity with respect to the underlying medium -- that medium which contains the gravity gradient as well as propagates light. In fact, if we now take the existence of such a medium for granted, and return to the three-body system, we will see that even if the system simply flies along a straight line in Eucledian space, it will be able to determine in which direction, and at what speed, it's flying -- by using nothing else other than light signals.

So, it seems to me that although without gravity special relativity makes sense, once gravity can be used in an experimental setup, special relativity breaks down.

In fact, with non-Euclidean spacetime you don't even need any orbiters to determine your absolute velocity with respect to the gravity gradient. All you need is a gyroscope to define a fixed orientation in space, and a meter stick. Then, you can orient the meter stick in a fixed direction, and measure tidal distortion on the stick as you move along on your orbit. Granted, in most cases the tidal force is miniscule, and your measurement would have to be really accurate. However, it would enable you to directly detect gravity gradient, as well as direction of the center of mass.

Then I also have this quantum idea. Since we are supposed to have particle pairs popping up and disappearing all over the place throughout the vacuum, then what if we somehow found a way to make them interact with a laboratory setup? And if so, would they not provide an absolute velocity measurement, by simply being blue-shifted in one direction and red-shifted in the opposite direction? (Though I'm not so sure about this particular idea, since I'm not exactly a QED guru).

The whole point is: there seem to be ways of detecting a preferred inertial reference frame. Which is what I've been talking about all along. Existence of any such frame comes into direct contradiction with any relativistic theory.

Now, I may be totally wrong on all counts, and may simply be an ignorant ass professing my misunderstanding of the theory. So, if there are any other relativists out there, how about taking a shot at my 'thought experiments'?

------------------
I am; therefore I think.

14. ### AloysiusGuest

Boris, that's a delicious thought experiment. I don't have the time right now to say anything deep about it, except two immediate hunches which might gainsay it:

1. You'll get a null result if the orbiters lie precisely 180 degrees apart. Because then they are indistinguishable, of course.

2. Mossbauer effect plays a role, rendering to Caesar all that is Caesar's? (erm I mean Albert

)

15. ### PlatoGuest

Boris,

First of all, I will take a look at your thought experiment a little closer this evening but for the moment I have one comment that kind of is in the line of Aloysius. I don't know if you have any experiance with the three body problem ? It is a very tricky thing which has real solutions only in certain cases (I'll tell you later which since I don't have the formula's with me here at work). One of the cases was indeed if one of the three is a very massive object but I think it was shown that the only solution of stable orbits in that case was that the two objects would be on opposite sides of the heavy one.
I promise I'll get back to this later...

------------------
greetings,
Plato

16. ### AloysiusGuest

Plato, I read the Alcubierre Mark II paper with interest. Unless I'm mistaken, the author has actually proposed the Tardis! For non-cognoscenti of Doctor Who, this is a fictional time machine which is larger inside than it is outside.

Funky.

17. ### BorisGuest

Plato, Aloysius:

I see where you guys are coming from -- trying to ruin the perfect setup with technical difficulties

But no matter; it is only important that the system be semi-stable for a short time so that the measurements can be made. That's why I made the central body so massive and the orbiters so puny. Also, Plato: although the particular three-body system will not be stable as I described it, the problem is easily fixable by making the three-body system into a symmetric n-body system, where now we have n bodies orbiting the neutron star, all of them separated by exactly the same distance. E.g. a simple construct would be three bodies orbiting in an equilateral triangle configuration (with 60 degrees between any pair). Though the equilibrium in such a system would still be precarious. But no matter -- it's the *principle* that's important

Aloysius: what does Mossbauer effect have to do with anything? Is there humor in there too subtle for my thick skull? (It's been a looong day...)

----------------------------
I'm too tired to think...

18. ### PlatoGuest

Ok Boris, I think I got it know.
De way to make you see that it only 'seems' that N is your preferential reference frame is to do your same argument in the reference frame of A and B.

First, you will agree that A and B have the same reference frame as they completely are at rest to each other. Let's take our origan of this frame also on their orbit and as far from A as from B. Thus AOB is in effect a two legged triangle with | |AO| | = | | OB | |.
In this reference frame N is seen to revolve around O in the opposite direction as in the N-frame.
Let's say N starts his flash when he crosses the vertical axe (time t). Since the veritical axe is the set of all points at equal distance of A and B, the lightsignal will indead arrive at the same time (t_0) in A as in B.
Then A signals to B and vice versa. Since they are at rest to each other, they to will get each others signal at exactly the same time (t_1). So at t_1 A and B send their signal to N, this however has moved further along his orbit, since a time t_1 - t has passed. So now N is actually closer to B then to A, this is why B's flash reaches N before A's.

You see that the result is completly the same but in a totally different referential system. So I have to conclude that there is no preference for one frame above the other to explain the same effect.

------------------
we are midgets standing on the backs of giants,
Plato

19. ### BorisGuest

Plato:

Nice try, but I'm afraid it's more complex than that... One very simple thing you're missing: N never gets any closer to B than it is to A -- in fact it remains at the center as A and B orbit it -- and they can verify that using radar ranging (i.e. verify that the distance to N is not changing). Also, A and B are not exactly at rest with respect to each other. For example, if they both start out facing in the same direction, so that B stares at A's back, then half a revolution later it'll be A staring at B's back! So indeed A and B have reasonable means to deduce that it is they who are orbiting, and not N. In fact, due to this same rotational effect, if you put an 'origin' between A and B, then to this origin it will appear that all 3 bodies orbit it, not just N. And from those orbits, it will be clear that the distances of NA and NB remain constant. Finally, it is clear who orbits whom simply from taking a look at the surrounding gravitational gradient. And logically, it is nonsensical for a neutron star to orbit a ping-pong ball. So I'm afraid the 'absolute reference frame' is quite a tenacious little beast in this scenario...

Rotation is tricky. In fact, it is the one thing that doesn't seem to be relative, but indeed absolute. For example, you can't argue that it's the rest of the universe rotating around you instead of you rotating on your own -- simply because it is you who feels the centrifugal force. That's why gyroscopes can act as beacons of direction -- yet another indication of an ultimate underlying reference frame.

Well, ok, maybe I don't have the justification for pushing the line _that_ strongly... But there's something out there... somewhere.

20. ### PlatoGuest

I admit that I was a little hasty in my previous mail. It appears I applied two different transformations for A,B and N.

I have put in more thought about it now and have the following claim.
There will be no difference in time what so ever between the two signals because there is no relative velocity between A, B or N if there are no fixed points in the universe to orient your 'fixed' frame in. You can argue that they have to turn around N because that is the massive one and in order not to fall to N they have to be in orbit. But that is only when you can proof there is something as gravity. In the system that you described there is no way these three object can proof that, the way they are orbiting. Only if A is on a different distance from N then B there is going to be relative velocity between them and one can deduce the fact of turning around N but then your little scheme won't work anymore.

I guess what you have done is assume there is a fixed reference frame in which A and B are turning around N and then after some elaboration you proved that there is indeed a fixed reference frame in which A and B are turning around N. This is a trivial proof but not a valid one because it is as meaningless as saying statement A equals statement A.

------------------
we are midgets standing on the backs of giants,
Plato

21. ### AloysiusGuest

Boris, you're a bright young sonofabitch.

RE. Mossbauer effect, I originally meant it as serious (so there!

) but there is a humorous side in my case.

In my _viva voce_ at Oxford for my physics degree, I was quizzed by Nobel Laureate Sir Rudolph Peierls. It appears that the question on the Mossbauer effect had run afoul, in the fear and loathing of my tortured adolescent brain, of a certain paregoric (this was 1969) which I had simply transferred from one piece of silver paper to another. Unfortunately, I licked my finger...can you spell Sunshine?... and this was to be my undoing in the finals paper on the morrow in Nuclear Physics. I calculated, in my euphoria that next day that the increase in weight of a cosmic meson hitting the earth would result in an effective mass of ten TONS when striking the surface of the planet. It seemed to make sense at the time.

Fortunately, at the viva I got it right, and maintained my honours degree status.

So, I always giggle when I say Mossbauer.

You were not to know.
I apologise for the wealth of personal information, rather akin to Lori's confession of the hicks that beat up their spouses and rev their engines.

It will all end in tears, I'm telling you.

22. ### AloysiusGuest

OK, since I started this, and since Boris is homing in on where I'd hoped someone would home in - i.e. Machian philosophical quandaries about the nature of absolute versus relative rotation - I'd like to throw a second link into the soup. God I'm so manipulative.

Actually, you guys are great.

http://chaos.fullerton.edu/Woodward.html

23. ### PlatoGuest

Aloysius,

First of all I remember the Mossbauer effect as being one of the most boring experiments we did in my student time. We had the choise between electron-positron annihilation experiments or Mossbauer and since everyone wanted to do the first one I thought it was cool to take Mossbauer... not !

Anyway I kind of suspected you were manipulating us, this question in the beginning was far to innocently put as if it was to lure us into some tricky discussion about absolutism >< relativism.

------------------
we are midgets standing on the backs of giants,
Plato