The speed of gravity revisited (nueva revisión de la velocidad de la gravedad)

Posted by Albert Zotkin en abril 10, 2013

Hola amable lector, hoy me gustaría traerte un thread de usenet que inicé hace ya algunos años. En el grupo de usenet sci.physics.relativity creé un thread con el título de “The speed of gravity revisited”. Dicho thread llegó a alcanzar los 368 posts, y entre los numerosos comentaristas podemos encontrar nombres como Tom Van Flandern, Steve Carlip, Tom Roberts, ó Juan R. González-Álvarez.

Una breve presentación de estos cuatro relevantes comentaristas científicos:
1. Tom Van Flandern, era un prestigioso astrónomo americano especializado en mecánica celeste, y entre otras muchas cosas, contribuyó notablemente a mejorar el GPS.
2. Steve Carlip es un prestigiso profesor de física en la Universidad de California, Davis. Son destacables sus papers en gravedad cuántica (2+1) dimensional, fundamentos gravitacionales cuánticos de la termodinámica de los agujeros negros, o en triangulaciones dinámicas causales.
3. Tom Roberts, PhD en fisica, prolífico comentarista en usenet, y acérrimo defensor de la relativdad Einsteniana, trabaja en Fermilab, es autor, entre otros trabajos, de “What is the experimental basis of Special Relativity?”

4. Juan R. González-Álvarez. Cientifico español, con una sólida base académica. Estudió físca y quimica en la Universidad de Vigo. Trabajó en temas científicos en el Ilustre Colegio de Químicos de Galicia y fue investigador asistente de bioquímica de las Rias en el CSIC, participó en varios simposios, conferencias e informes. Puedes encontrar algunos de sus papers en FQXi Community, por ejemplo este.

Lo que sigue son los primeros posts de ese histórico thread en sci.relativity.

 1. Albertito: 6 mar 2008, 22:41 There are evidences showing that in Solar system, the speed of gravity is many orders of magnitude higher than the speed of light. But, what must we understand by speed of gravity?. Aetherists often claim that gravity are longitudinal waves, whereas light are transverse waves through the aether. We know that in any medium longitudinal waves travel faster than transverse waves. We can find that longitudinal speed, cL, and transverse cS, in a medium, with Young’s modules E, Poison’s ratio v and mass density d0, are $\displaystyle c_L^2 = \left (\cfrac{E}{d_0(1+v)}\right ) \cfrac{1-v}{1-2v} \\ \\ \\ c_S^2 = \left (\cfrac{E}{d_0(1+v)}\right ) \cfrac{1}{2}$ We also know there exists a relation between those elastic constants, as $\displaystyle E=2G(1+v)=3K(1-2v),$ where G is shear modulus and K is bulk modulus. So, we have $\displaystyle c_L^2 = \left (\cfrac{2G}{d_0}\right ) \cfrac{1-v}{1-2v} \\ \\ \\ c_S^2 = \cfrac{G}{d_0}$ Therefore, for a Poison’s ratio of v=1/2, it would result an infinite longitudinal speed. In general we have $\displaystyle c_L^2 +c_S^2 = \cfrac{G}{d_0} \cfrac{2(1-v)}{1-2v}+1 \\ \\ \\$ This quadratic relation suggests it is a universal constant for vacuum. This suggests $\displaystyle \cfrac{G}{d_0} \cfrac{2(1-v)}{1-2v}+1 = \cfrac{R^2}{t_p^2}$ where R is a scale parameter and tp is Planck time, or $\displaystyle \cfrac{G}{d_0} \cfrac{2(1-v)}{1-2v}+1 = \cfrac{c^2}{l_p^2}$ where lp is Planck length $\displaystyle c_L^2 + c_S^2 = c^2 \cfrac{R^2}{l_p^2}$ So, for a speed of light being cS=c, it would yield $\displaystyle c_L^2 + c_S^2 = c^2 \cfrac{R^2}{l_p^2} \\ \\ c_L = c \sqrt{\frac{R^2}{l_p^2} -1},$ which is roughly $\displaystyle c_L = \cfrac{c L}{l_p}$ if R is meaningfully larger than lp. If we define R = R_h (Hubble radius), then the speed of gravity, there where the local speed of light is c, would be $\displaystyle c_L = \cfrac{c R_h}{l_p}$ it is saying it would be a very superluminal speed (i.e. infinite velocity, for practical purposes).

 2. Tom Roberts: 7 mar 2008, 17:44 Albertito wrote: > There are evidences showing that in Solar system, > the speed of gravity is many orders of magnitude higher > than the speed of light. Sure. But this is MODEL DEPENDENT. In the model of Newtonian gravitation, gravity propagates INSTANTLY (i.e. with infinite speed). In the model of GR, gravity does not propagate at all, but changes in gravity propagate with speed c. The GR model agrees with all these “evidences”, and indeed it accounts MUCH more accurately than the Newtonian model for measurements in the solar system (including the perihelions of Mercury and other planets, the Shapiro time delay, the bending of EM radiation by the sun, the operation of the GPS, the frame dragging measured by the LAGEOS satellites, etc.). Bottom line: it is MUCH better to discuss models and their agreement with experiments than to discuss MODEl-DEPENDENT quantities like “speed of gravity”. That is, discuss science (experiments) rather than engineering (measurements), and avoid unacknowledged puns (such as model-dependent meanings of words that are treated as if they had a single meaning) like “speed of gravity”. > [… further nonsense based on unrealistic models (“aetherists”)…] Tom Roberts

 3. Juan R. González-Álvarez: 7 mar 2008, 20:17 Tom Roberts wrote on Fri, 07 Mar 2008 15:44:38 +0000: > Albertito wrote: >> There are evidences showing that in Solar system, the speed of gravity >> is many orders of magnitude higher than the speed of light. > Sure. But this is MODEL DEPENDENT. In the model of Newtonian > gravitation, gravity propagates INSTANTLY (i.e. with infinite speed). Being a AAAD theory, nothing propagates in Newtonian gravitation. speaking about infinite speed is misleading also. Infinite speed of what? > In > the model of GR, gravity does not propagate at all, Gravitational waves travel at c like changes in spacetime geometry do. >> but changes in > gravity propagate with speed c. The GR model agrees with all these > “evidences”, and indeed it accounts MUCH more accurately than the > Newtonian model for measurements in the solar system (including the > perihelions of Mercury and other planets, the Shapiro time delay, the > bending of EM radiation by the sun, the operation of the GPS, the frame > dragging measured by the LAGEOS satellites, etc.). GR gives better results (i would not say “MUCH”) for purely relativistic effects. Since NG is non-relativistic, this is not kind of surprising. The problem with NG is that lacks an adequate Newtonian limit. GR literature is incorrect at this point. Moreover, NG is free from several difficulties affecting GR: energy problem, systems of reference problems, unphysical boundaries, quantization, N-body theory… — I apply http://canonicalscience.org/en/miscellaneouszone/guidelines.txt
4. Tom Roberts: 8 mar 2008, 04:11
Juan R. González-Álvarez wrote:
> Tom Roberts wrote on Fri, 07 Mar 2008 15:44:38 +0000:
>> In the model of Newtonian
>> gravitation, gravity propagates INSTANTLY (i.e. with infinite speed).
> Being a AAAD theory, nothing propagates in Newtonian gravitation.
> speaking about infinite speed is misleading also. Infinite speed of what?

Infinite speed of gravity, of course. You are just saying the same thing
using different words (AAAD == infinite speed of propagation of influence).
>> In
>> the model of GR, gravity does not propagate at all,
> Gravitational waves travel at c like changes in spacetime geometry do.

Of course — gravitational waves _ARE_ changes in spacetime geometry.
>> The GR model agrees with all these
>> “evidences”, and indeed it accounts MUCH more accurately than the
>> Newtonian model for measurements in the solar system (including the
>v perihelions of Mercury and other planets, the Shapiro time delay, the
>v bending of EM radiation by the sun, the operation of the GPS, the frame
>v dragging measured by the LAGEOS satellites, etc.).
> GR gives better results (i would not say “MUCH”) for purely relativistic
> effects. Since NG is non-relativistic, this is not kind of surprising.

Hmmm. If you mean NG is accurate in the non-relativistic regime, then
sure. But such a statement carries no information. And the usual meaning
of “relativistic effects” does not apply to any of the measurements I
mentioned. In any case, my “MUCH” is certainly justified — NG fails to
predict ANY of them anywhere close to correctly (why else do you suppose
I chose them?):

 Measurement NG GR Perih. of Mercury et al zero correct Shapiro time delay zero * correct Bending of EM radiation zero * correct operation of GPS hopeless correct frame dragging zero correct

Where “correct” means within the appropriate experimental resolution.
* For NG applied to EM waves, I use the fact that
such waves are massless in making the NG prediction.
> The problem with NG is that lacks an adequate Newtonian limit. GR
> literature is incorrect at this point.

If this is not a typo it makes no sense. If it is a typo, writing “NG”
when you meant “GR”, then you are wrong — there is nothing “inadequate”
about the Newtonian limit of GR.
> Moreover, NG is free from several difficulties affecting GR: energy
> problem, systems of reference problems, unphysical boundaries,
> quantization, N-body theory…

Some of those “difficulties” are merely complications that are
inescapable: energy problem, systems of reference problems. Some are (as
best I can tell) figments of your imagination: unphysical boundaries,
N-body problem. Yes, quantization is a problem for GR and severely
limits its domain of applicability, but NG has much worse problems
(disagreement with numerous experiments within its domain of applicability).
Tom Roberts

 5. Juan R. González-Álvarez: 8 mar 2008, 14:51 Tom Roberts wrote on Sat, 08 Mar 2008 02:11:31 +0000: > Juan R. González-Álvarez wrote: >> Tom Roberts wrote on Fri, 07 Mar 2008 15:44:38 +0000: >>> In the model of Newtonian >>> gravitation, gravity propagates INSTANTLY (i.e. with infinite speed). >> Being a AAAD theory, nothing propagates in Newtonian gravitation. >> speaking about infinite speed is misleading also. Infinite speed of >> what? > Infinite speed of gravity, of course. Gravity in AAAD has not a property called “speed”, of course. > You are just saying the same thing > using different words (AAAD == infinite speed of propagation of > influence). No, i am just saying the contrary: in AAAD nothing propagates including “influences”. You would not confound AAAD models with field-metric models. >>> In >>> the model of GR, gravity does not propagate at all, >> Gravitational waves travel at c like changes in spacetime geometry do. > Of course — gravitational waves _ARE_ changes in spacetime geometry. Of course “like… do” could be emphasized as “_LIKE_… _DO_”. >> GR gives better results (i would not say “MUCH”) for purely >> relativistic effects. Since NG is non-relativistic, this is not kind of >> surprising. > Hmmm. If you mean NG is accurate in the non-relativistic regime, then > sure. But such a statement carries no information. Hmmm. Crizing a non-relativistic theory because fails on relativistic regimes is very old relativistic tactic but is clearly unfair. > And the usual meaning > of “relativistic effects” does not apply to any of the measurements I > mentioned. Sure perihelions for Mercury, the Shapiro time delay, bending of EM radiation by the sun, and GPS operation contain relativistic effects, if one takes the general meaning not just a kinematic meaning. “Relativistic effects” had certain restricted meaning in 1908 because then only SR was known… > In any case, my “MUCH” is certainly justified — NG fails to > predict ANY of them anywhere close to correctly (why else do you suppose > I chose them?): As explained before NG does not exactly fail to explain relativistic effects. That is wrong claim. NG does not apply to relativistic phenomena because is a non-relativistic theory. Nobody would imagine one can apply NG *outside* its range of validity waiting adequate answer, unless that person does not understand SCIENCE. But that is another point… > Measurement NG GR ———————- > ———– ——— Perih. of Mercury et al zero > correct Shapiro time delay zero * correct Bending > of EM radiation zero * correct operation of GPS > hopeless correct frame dragging zero > correct > Where “correct” means within the appropriate experimental resolution. > * For NG applied to EM waves, I use the fact that > such waves are massless in making the NG prediction. This table has been clearly done to confound readers. Computes total values for entries making *sense* and try next ratio NG value _________________________________________ NG value + relativistic correction You will find most of ratios are very small. Rest is so unfair as a table comparing quantum gravity with GR. >> The problem with NG is that lacks an adequate Newtonian limit. GR >> literature is incorrect at this point. > If this is not a typo it makes no sense. If it is a typo, writing “NG” Only a genious could see it is a typo, thanks by kindly correction! “The problem with GR is that lacks an adequate Newtonian limit.” > when you meant “GR”, then you are wrong — there is nothing “inadequate” > about the Newtonian limit of GR. You are wrong. The NG limit does not exist and the several Newtonian-like limits tried on relativistic literature are not actually working (lacking mathematical rigor, unphysical boundaries,…). >> Moreover, NG is free from several difficulties affecting GR: energy >> problem, systems of reference problems, unphysical boundaries, >> quantization, N-body theory… > Some of those “difficulties” are merely complications that are > inescapable: energy problem, systems of reference problems. They they are “inescapable” when you decide to introduces it on physics, i.e. when you insist on a geometrical interpretation of gravity. > Some are (as > best I can tell) figments of your imagination: unphysical boundaries, > N-body problem. Those problems are well-known and studied on literature. Several proposals are done to correct eliminate them. Yes, you are not aware of them but as is known from sci.physics.research “Yours is a statement of profound ignorance in all of its parts.” — Uncle Al to Tom Roberts. Feb 2008 > Yes, quantization is a problem for GR and severely > limits its domain of applicability But NG can be quantized without the further problems of GR! — I apply http://canonicalscience.org/en/miscellaneouszone/guidelines.txt

 6. Tom Roberts: 8 mar 2008, 22:26 Juan R. González-Álvarez wrote: > Tom Roberts wrote on Sat, 08 Mar 2008 02:11:31 +0000: > Hmmm. Crizing a non-relativistic theory because fails on relativistic > regimes is very old relativistic tactic but is clearly unfair. Criticizing a non-relativistic theory for disagreeing with experiments is not “unfair” at all. This is supposed to be science, and YOU are the one pushing NG. >> And the usual meaning >> of “relativistic effects” does not apply to any of the measurements I >> mentioned. > Sure perihelions for Mercury, the Shapiro time delay, bending of EM > radiation by the sun, and GPS operation contain relativistic effects, if > one takes the general meaning not just a kinematic meaning. Hmm. The usual meaning of “relativistic effects” is that they are important only for speeds approaching c (SR) or very strong fields (GR). None of the experiments I mentioned have either. You seem to mean “relativistic effects” when the non-relativistic theory fails. That’s silly, and useless — for good enough measurement resolution the non-relativistic theory is completely useless. Several of the experiments I mentioned have extraordinarily good resolutions, and _that_ is why they are important. > As explained before NG does not exactly fail to explain relativistic > effects. That is wrong claim. NG does not apply to relativistic phenomena > because is a non-relativistic theory. There is no “relativistic phenomena” involved in ANY of the experiments I mentioned, unless one uses your silly meaning. > Nobody would imagine one can apply NG *outside* its range of validity > waiting adequate answer, unless that person does not understand SCIENCE. Ok. I’m not the one pushing NG, you are. Note its “range of validity” depends on one’s measurement accuracy, and for good enough accuracy its “range” is essentially empty. Certainly such accuracy is common today (a \$200 GPS receiver), and will be even more common in the future as measurement techniques improve. >> [my list of experiments, totally corrupted and now unreadable] > This table has been clearly done to confound readers. The “confounding” is all yours. > Computes total values for entries making *sense* and try next ratio > NG value > _________________________________________ > NG value + relativistic correction That is a very silly way to do this. And your denominator is outrageous — it should at least be “GR value”. Note, however, the CORRECT way to do this is to compare the theories via these two ratios: |NGvalue – Experiment| / sigma_experiment |GRvalue – Experiment| / sigma_experiment [sigma_experiment is the experimental resolution.] When one does that, one finds that for EVERY ONE of the experiments I mentioned the NGvalue is so different from the experimental value that NG is soundly refuted; the GR value is quite reasonable for all of them. > “Yours is a statement of profound ignorance in all of its parts.” > — Uncle Al to Tom Roberts. Feb 2008 I merely remark that neither Uncle Al nor you have ever responded to my followup — in the physics community it is quite common to consider one’s understanding of a subject to be measured by the ability to explain it to a graduate student or postdoc not expert in the field. You both fail that criterion, and instead rely on “dense spews of jargon indistinguishable from nonsense” [Tom Roberts to Uncle Al, in the thread you quoted]. > But NG can be quantized without the further problems of GR! Whyever would that matter? — who cares about a demonstrably incorrect and soundly refuted theory like NG? That’s like claiming 2+2=5 can be generalized without the “problems” of number theory. Tom Roberts

 7. Juan R. González-Álvarez: 9 mar 2008, 18:13 Tom Roberts wrote on Sat, 08 Mar 2008 20:26:21 +0000: > Juan R. González-Álvarez wrote: >> Tom Roberts wrote on Sat, 08 Mar 2008 02:11:31 +0000: Hmmm. Crizing a >> non-relativistic theory because fails on relativistic regimes is very >> old relativistic tactic but is clearly unfair. > Criticizing a non-relativistic theory for disagreeing with experiments > is not “unfair” at all. English may be not your natural language. I said Criticizing a non-relativistic theory for disagreeing with relativistic experiments is “unfair” at all. >> Sure perihelions for Mercury, the Shapiro time delay, bending of EM >> radiation by the sun, and GPS operation contain relativistic effects, >> if one takes the general meaning not just a kinematic meaning. > Hmm. The usual meaning of “relativistic effects” is that they are > important only for speeds approaching c (SR) or very strong fields (GR). > None of the experiments I mentioned have either. Completely wrong. E.g. anomaly Mercury perihelion is explained by two relativistic corrections. But since you only look to “how” instead “why” you lack understanding. >> Nobody would imagine one can apply NG *outside* its range of validity >> waiting adequate answer, unless that person does not understand >> SCIENCE. > Ok. I’m not the one pushing NG, you are. One of your usual FALSE accusations tactics. Read i exactly said. >>> [my list of experiments, totally corrupted and unreadable] > When one does that, one finds that for EVERY ONE of the experiments I > mentioned the NGvalue is so different from the experimental value that > NG is soundly refuted; the GR value is quite reasonable for all of them. No SERIOUS scientist would apply a theory outside its range of applicability waiting meaningful answers. Tom, that is not how science works. And no HONEST scientist would use those answers to attack that theory he DISLIKE/HATES. Science is a dialog with Nature Tom. > in the physics community it is quite common to consider > one’s understanding of a subject to be measured by the ability to In the physics community it is rather common to provide detailed replies when one is sure the other can understand it. One aloso usually ignores unfair queries That is because you received that reply in sci.physics.research. > Whyever would that matter? — who cares about a demonstrably incorrect > and soundly refuted theory like NG? “Yours is a statement of profound ignorance in all of its parts.” — Uncle Al to Tom Roberts. Feb 2008 — I apply http://canonicalscience.org/en/miscellaneouszone/guidelines.txt

 8. Tom Van Flandern: 1 abr 2008, 19:17 Tom Roberts” writes: > [Roberts]: In the model of GR, gravity does not propagate at all, but > changes in gravity propagate with speed c. That is directly in contradiction to experiment and observations. Binary pulsars are an obvious example, as I demonstrated (without any dissent) in Reference B below. But even the simplest orbit computation program can show the same thing. If you use light-time-retarded positions of bodies to compute orbits, the computed orbits are open spirals, in contradiction to observations. But you've obviously never done the experiment yourself, or have used only propagation delays in the potential field, which are irrelevant for orbit computation. See Reference (C). There is no way known to any person on this planet to avoid the conclusion that gravitational force propagates >> c without invoking some kind of physical miracle, such as an effect without a cause or the creation of new momentum out of nothingness. Mathematical relativists don't seem bothered by such miracles. Meanwhile, real world physicists know they must not invoke miracles in their theories because that makes them non-falsifiable, and therefore unscientific. [See Reference E.] > [Roberts]: The GR model agrees with all these "evidences", and indeed it > accounts MUCH more accurately than the Newtonian model for measurements in > the solar system (including the perihelia of Mercury and other planets, > the Shapiro time delay, the bending of EM radiation by the sun, the > operation of the GPS, the frame dragging measured by the LAGEOS > satellites, etc.). True but irrelevant because GR is a field theory and describes only the field. The gravitational potential field causes all the effects on your list. But it does not cause ordinary orbital motion. Nor do the field equations describe ordinary orbital motion. To get that, one must take a gradient of the potential (or its equivalent) to get what you like to call an "approximation" theory. In simple, classical physics lingo, that process develops an expression for the 3-space (Euclidean) acceleration of bodies in coordinate time, which gives the orbital motion, which is then compared against astronomical observations made in Euclidean 3-space using proper time clocks. Try computing an orbit with GR just once in any system with at least two significant masses, and you will discover that you cannot do it without adopting near-infinite gravitational force propagation speed between bodies applying forces to one another. Then the dawn will come, and you will finally understand what the "speed of gravity" issue is about. > [Roberts]: it is MUCH better to discuss models and their agreement with > experiments than to discuss MODEL-DEPENDENT quantities like "speed of > gravity". The "speed of gravity" is not a model-dependent concept except at the level of parts per 100 million, any more than "perihelion motion" is model-dependent. Its simple meaning is: When a source mass accelerates, the speed of gravity is the ratio of the distance of a target body to the time elapsed before the target body responds. And every known experiment measures that elapsed interval to be zero within experimental error, making the speed of gravity >> c and approximately infinite. Relativists like to redefine the concept to refer to the speed of changes in the gravitational potential field, which everyone agrees is c. But that refers to gravitational waves, and avoids the issue of the propagation speed of gravitational force for determining the ordinary orbital motion of two masses around a common center of mass. One must either give up the causal link to a source mass, or agree that the force propagates from the source mass to the target body faster than c. > Tom Van Flandern does not understand the real issues, and uses egregious > PUNs to promulgate his claims. In particular, what he calls "speed" is not > what anybody else would call "speed". The experiments he cites do NOT > measure speed (usual meaning), and their actual measurements are fully > consistent with GR, in which nothing propagates faster than c. Quit making up nonsense. The published papers are in references (A), (B), (C), and (D) below. "Speed" has its unambiguous, classical meaning in all of them, as the editors, reviewers, and readers have all understood. Where are your publications on the subject? >> [Juan]: For calculations of orbits we have to use the actual positions of >> bodies and not the perceived locations. > [Roberts]: True in Newtonian mechanics; irrelevant in GR. The comparison of theory with observations is not relevant? How absurd! You are disconnected from reality. References: ** (A) "Possible new properties of gravity", Astrophys.&SpaceSci. 244:249-261 (1996); http://metaresearch.org/cosmology/gravity/possiblenewpropertiesofgrav… ** (B) "The speed of gravity – What the experiments say", Phys.Lett.A 250:1-11 (1998); http://metaresearch.org/cosmology/speed_of_gravity.asp ** (C) "Reply to comments on 'The speed of gravity'", Phys.Lett.A 262:261-263 (1999). ** (D) "Experimental Repeal of the Speed Limit for Gravitational, Electrodynamic, and Quantum Field Interactions", T. Van Flandern and J.P. Vigier, Found.Phys. 32:1031-1068 (2002); preprint under title "The speed of gravity – Repeal of the speed limit" at http://metaresearch.org/cosmology/gravity/speed_limit.asp ** (E) "Physics has its principles", in Gravitation, Electromagnetism and Cosmology, K. Rudnicki, ed., C. Roy Keys Inc., Montreal, 87-101 (2001); http://metaresearch.org/cosmology/PhysicsHasItsPrinciples.asp Tom Van Flandern – Sequim, WA – see our web site on frontier astronomy research at http://metaresearch.org

Anuncios