Consistency of the speed of light

[gregory_]

So how did you get the quote if you cannot find the paper? How could you tell us about the content of the paper? How do you know that Gagnon spends 20 pages "analizing other experiments"? as per your post #238.

Because I can still read the abstract online for free through INSPEC.

Aren't you missing a few things?

1. they are using the most general form of the MS spacetime transformation, with all the parameters in place

2. the result of the experiment puts a very hard boundary on the parameters (makes them effectively zero within the experimental error bars)

As I CLEARLY pointed out to you, they do not constrain the parameter \bf{\epsilon}. This is the only parameter that differs between GGT and Lorentz transformations.

In case you missed it , the subject of the discussion is:
- the validity of one way speed measurement (do such experiments exist? yes, they do and they are valid)

No. One way speed measurements are not valid.
If you'd take the time to step back and think about it you'd see this. I really think we should focus on the basics.

In special relativity, the metric is frame independent, so the force law maintains the same form (as expected). However, this is not true for GGT. In GGT the metric is frame dependent ...

You sure about that? This is where you stop calculating and you start talking. Based on the formulation of GGT I would be inclined to believe that the metric is frame independent. Can you prove the above statement mathematically? It should be simple.

Um, are you kidding? Of course the metric is frame dependent in GGT.

In reference 9 of your Gagnon experiment paper, they explicitly work out the metric in GGT. I'm not going to bother working that out for you. Read it yourself if you refuse to think about it and see why the metric has to be frame dependent in GGT (this should be obvious before even doing any calculations).

1. However, this is not true for GGT. In GGT the metric is frame dependent and thus the Lorentz force law changes form when we change frames (the metric is worked out in reference 9, so you can calculate the horrid form of the Lorentz force law using GGT if you so wish).

2. How does this affect Gagnon's paper? It means the boundary conditions they invoke when solving for the fields in the wave-guide are not correct. So their calculations are flawed right at the beginning.

A. Can you prove statement 1 to be true?

B. Can you prove the connection between the two statements (1 and 2)? In my experience many "horrid forms" tend to cancel out when it matters and to produce nice symmetrical results. Look at the "horrid form" of the Mansouri-Sexl transforms...

A] The fact that the metric is frame dependent is not only self evident, but is shown explicitly in reference 9 of Gagnon's experimental paper. Additionally, I have shown that if the metric changes, so too does the Lorentz force law.

B] The Lorentz force law is different. I have shown that.

The boundary conditions on a metal are such that: there is no force on charges in the material, there can only be a force on the surface charges perpendicular to the surface. Because the Lorentz force law normally looks like \vec{F} = q(\vec{E} + \vec{v} \times \vec{B}) this is equivalent to the boundary condition on the fields of \bf{E}_\parallel=0, \bf{B}_\perp=0. Because the force law is not the same in GGT frames, the boundary condition is not the same either.

[in reference to another gagnon paper]
Gregory, how did you get the idea that it might be a retraction?

Because he stated that experiments could not distinguish between the theories at second order. But his previous paper claimed the exact opposite.

------------------------
Summary:

The calculations in the Gagnon experiment have been shown to be wrong on their starting assumptions. They are wrong. Krisher has also been shown to not be relavent to this discussion in regards to GGT. So let's move on.

clj4, you skipped a question:
Are you denying that by definition special relativity and theories invoking GGT agree on the physical laws in one inertial frame?

 

[Aether]

Disprove Gagnon (Phys Rev). With your own calculations.

I still want to do that and learn how to do PDE's in anisotropic space, but that doesn't mean I'm going to let you get away with posting crackpottery of the second kind at PF in the mean time.

If you cannot do it (there has been already three weeks of failed attempts, errors, missteps), then admit it and we move to Kirshner. Same drill: mathematical disproof. Might be easier for you, it is not as terse as gagnon.

I cannot do any more with the PDE's in Gagnon et al. right now. I'll continue to study my new textbooks, gregory's analysis, and the two new papers from T.Chang et al.. I'll also read Krisher.

 

[clj4]

No. One way speed measurements are not valid.








B] The Lorentz force law is different. I have shown that.






The calculations in the Gagnon experiment have been shown to be wrong on their starting assumptions. They are wrong.

Your tensor proof is flawed. you skipped the challenge on the commutation of the product. Therefore your so-called disproof of Gagnon is also flawed. Could you please address it?

 

[clj4]

I still want to do that and learn how to do PDE's in anisotropic space, but that doesn't mean I'm going to let you get away with posting crackpottery of the second kind at PF in the mean time.

No need to resort to personal attacks. You know very well that this thread started elswhere with YOUR claim that there are no valid one-way light speed experiments. I gave you a list of experiments, you started working on disproving Gagnon, we moved it here because of the use of LatEx, in order to facilitate your disproof. Once yoiu get personal it is a sign that you lost the argument

 

[gregory_]

Your tensor proof is flawed. you skipped the challenge on the commutation of the product. Therefore your so-called disproof of Gagnon is also flawed. Could you please address it?

This has already been adressed by another poster. I can show you more explicitly if you wish. But first I need to know something...

I'm not sure how to ask this nicely. And I am not trying to put you down or anything. But, do you understand Einstein summation?

If not, then we can start there as that seems to be your issue.

 

[clj4]

This has already been adressed by another poster. I can show you more explicitly if you wish. But first I need to know something...

I'm not sure how to ask this nicely. And I am not trying to put you down or anything. But, do you understand Einstein summation?

If not, then we can start there as that seems to be your issue.

Thank you, I understand it. So please explain your derivation, contrary to what you say it was not addressed in any other post that we can see.
while you are at it, please explain to all of us how could you apply the Minkowski formalism that is derived for SR to GGT? What makes you think that it is even applicable?

 

[gregory_]

Thank you, I understand it. So please explain your derivation, contrary to what you say it was not addressed in any other post that we can see.

If you understand Einstein summation, then I don't understand your issue here. Each of those components is just a number. Numbers commute. (And yes this WAS explained to you already https://www.physicsforums.com/showpost.php?p=944396&postcount=254".) Furthermore, I can do the summations in any order I choose.

while you are at it, please explain to all of us how could you apply the Minkowski formalism that is derived for SR to GGT? What makes you think that it is even applicable?

Umm... tensor notation is not limited to SR or GR. It can be used for any theory with coordinate systems.

I can't believe I am even having this discussion.
(For your information, Gagnon's results are based on exactly such derivations using tensor notation to derive the electric and magnetic fields in GGT ... again starting with the DEFINITION that GGT agrees with SR on the physical laws in one "special frame".)

 

[clj4]

Question #3] In my explanation of why experiments cannot distinguish between "Generalized Galilean transformations / coordinate systems" and "Lorentz Transformations / Special Relativity's" definition of the one way speed of light, which parts do you disagree with and why?

Looks like we now have THREE Gagnon/Torr/etc papers that fly in the face of your explanation. Plus a few more referenced by the above mentioned papers.
I think your explanation is not an explanation at all, it is pure belief. When the equipment got sensitive enough to detect second (and third) order effects, people started questioning this belief. Now, you countered by questioning the Gagnon paper claiming that they don't set the boundary conditions correctly (questionable demonstration). In the meanwhile, let me remind you that your question 3 is not the main topic of the thread.
The main topic is the validity of one way light speed experiments. So , now that I answered all your questions I have just one for you:

1. Do you question the validity of the one way light speed experiments, especially Kirshner?

 

[clj4]

Because I can still read the abstract online for free through INSPEC.As I CLEARLY pointed out to you, they do not constrain the parameter \bf{\epsilon}. This is the only parameter that differs between GGT and Lorentz transformations.

The above answer is not correct. You know very well that a,b and d can be and also are different. Come on, the authors (and this is the standard approach) assume them to be present and end up constraining them (see bottom of page 733).
They do not constrain \epsilon because of their simplifying choice of clock synchronization. Had they chosen a more complex clock synchro they might just as well derived constrains for \epsilon.

Besides, let me point out to you one more time: the main issue of this thread is the validity of one way light speed experiments.(see my question to you)

No. One way speed measurements are not valid.
If you'd take the time to step back and think about it you'd see this. I really think we should focus on the basics.

Well, with this kind of biased answer I suggest that our discussion is done. You may wish to take your findings on Gagnon and Kirshner and write a report to Phys Review. Be careful how you cook up your math in order to support your point. Good bye,

 

[gregory_]

Now, you countered by questioning the Gagnon paper claiming that they don't set the boundary conditions correctly (questionable demonstration).

Not so fast. Seems like you can never admit you are wrong (or keep your word).

You said you would believe me if I showed you mathematically where Gagnon made a mistake. I was skeptical, but I decided to give you the benefit of the doubt. So I showed you where Gagnon made a mistake, and, even without finding an error you feel I "must" be wrong because of some bizarre unstated reason / internal belief.

It does not work this way.
Gagnon is wrong. I showed you why. Either admit this or point out where you feel I made a mistake.

In the meanwhile, let me remind you that your question 3 is not the main topic of the thread.
The main topic is the validity of one way light speed experiments. So , now that I answered all your questions I have just one for you:

1. Do you question the validity of the one way light speed experiments, especially Kirshner?

Okay, if you wish to word it that way. If we constrain ourselves to inertial frames (which we shall define as frames in which all freely moving object move at a constant velocity), then we can use experiment to constrain the one-way speed of light to either the form according to GGT or the Lorentz transformations. It should be pointed out that while this constrains the form, it doesn't constrain the value of the one-way speed of light (as the "special frame" is arbitrary). Further constraint is not possible (and constraint this far was possible only because we restricted ourselves to inertial frames), because, as you admitted yourself, the one-way speed of light is a coordinate system dependant number (there is no way to define a one-way velocity without defining a coordinate system).

As I CLEARLY pointed out to you, they do not constrain the parameter \bf{\epsilon}. This is the only parameter that differs between GGT and Lorentz transformations.

The above answer is not correct. You know very well that a,b and d can be and also are different.

No. Read the definition of the "generalized galilean transformations". Only the \bf{\epsilon} parameter differs between them. Krishner does not distiguish between them.


Because theories invoking GGT agree with SR on the physical laws in one special frame BY DEFINITION, they cannot be distinuighed by experiment BY DEFINITION (unless we find some physical law that is not lorentz invarient, which we both know hasn't happenned as of yet).

Are you denying that by definition special relativity and theories invoking GGT agree on the physical laws in one inertial frame?

 

[clj4]

Gagnon is wrong. I showed you why. Either admit this or point out where you feel I made a mistake.

You made several mistakes, not one. I'll get back with a full list. Ciao.

 

[clj4]

You conveniently "forgot" this one:

No. One way speed measurements are not valid.
If you'd take the time to step back and think about it you'd see this. I really think we should focus on the basics.

Well, with this kind of biased answer I suggest that our discussion is done. You may wish to take your findings on Gagnon and Kirshner and write a report to Phys Review. Be careful how you cook up your math in order to support your point. Good bye,

 

[gregory_]

You seemed to have missed this question:

Because theories invoking GGT agree with SR on the physical laws in one special frame BY DEFINITION, they cannot be distinuighed by experiment BY DEFINITION (unless we find some physical law that is not lorentz invarient, which we both know hasn't happenned as of yet).

Are you denying that by definition special relativity and theories invoking GGT agree on the physical laws in one inertial frame?

 

[clj4]

See post 282.

In the meanwhile, if I were you I would look over the Gagnon "disproof".

 

[gregory_]

You made several mistakes, not one. I'll get back with a full list. Ciao.

I look forward to seeing your "complaints". (This better not be on par with your "but how can you commute numbers?" complaint.)

I hope in researching this that you finally realize your mistakes.


Also, considering that I've been nice enough to answer all of your questions, I don't understand why you always insist on not answering mine. Please answer the following:

Are you denying that by definition special relativity and theories invoking GGT agree on the physical laws in one inertial frame?

 

[clj4]

Please answer the following:

Are you denying that by definition special relativity and theories invoking GGT agree on the physical laws in one inertial frame?

I agree that SR and GGT agree in the preferential frame of GGT.

OK, so answer this :

No. One way speed measurements are not valid.
If you'd take the time to step back and think about it you'd see this. I really think we should focus on the basics.

Well, with this kind of biased answer I suggest that our discussion is done. You may wish to take your findings on Gagnon and Kirshner and write a report to Phys Review. Be careful how you cook up your math in order to support your point.

While you are at it, ponder on this:

http://prola.aps.org/abstract/PRA/v34/i3/p1708_1

You may have to add it to the growing refutations of Gagnon/Torr/Kirshner...

Try to be less arrogant, your counter on Gagnon might not hold. Use your time to double check it...

 

[Aether]

Would you agree that Eq. (8) of (Krisher et al., 1990) is the ultimate prediction for this paper, and that a null result is predicted when \alpha=-\frac{1}{2}, \beta=\frac{1}{2} and \delta=0? That is exatly what I mean (e.g., \alpha=-\frac{1}{2}, \beta=\frac{1}{2} and \delta=0) when I say that we're assuming perfect Lorentz symmetry for the purposes of this discussion. Note number 14 goes to what we're talking about here: "Notice that the result is independent of the synchronization procedure embodied in the vector \epsilon.[14]" And note 14 says this: "This had to be the case, since the experiment contains only two clocks. Because we look only for a variation in the relative phase with angle, the relative syncronization of the two clocks at an initial time is completely arbitrary."

 

[clj4]

I think that you are reading (8) wrong. The whole idea is to derive (8) assuming that the theory in cause is NOT SR (the authors tell you right above (8) that "in SR \alpha=-1/2,...") but a DIFFERENT one (i.e. Mansouri Sexl with a simplified clock synchro). The authors proceed with constraining 1+2\alpha, etc through the proposed experiment. This is standard procedure in test theories.

 

[Aether]

I think that you are reading (8) wrong. The whole idea is to derive (8) assuming that the theory in cause is NOT SR (the authors tell you right above (8) that "in SR \alpha=-1/2,...") but a DIFFERENT one (i.e. Mansouri Sexl with a simplified clock synchro). The authors proceed with constraining 1+2\alpha, etc through the proposed experiment. This is standard procedure in test theories.

One of us is wrong about this, and our basic disagreement probably stems from that. Here's how I interpret this: the general Mansouri-Sexl transform (by design) reduces to SR when the four parameters are \alpha=-1/2,\ \beta=1/2,\ \delta=0,\ \epsilon=-v/c^2; and it reduces to GGT when the four parameters are \alpha=-1/2,\ \beta=1/2,\ \delta=0,\ \epsilon=0. This (at least the part about \epsilon) is explained very clearly at the bottom of page 355 of T.Chang et al.. The only difference is in the parameter \epsilon which is not subject to empirical measurement (e.g., it is conventional, and entirely coordinate-system dependent). The other parameters are measureable in a coordinate-system independent way, but they are exactly the same for SR and GGT.

 

[clj4]

One of us is wrong about this, and our disagreement probably stems from that. Here's how I interpret this: the general Mansouri-Sexl transform (by design) reduces to SR when the four parameters are \alpha=-1/2,\ \beta=1/2,\ \delta=0,\ \epsilon=-v/c^2; and it reduces to GGT when the four parameters are \alpha=-1/2,\ \beta=1/2,\ \delta=0,\ \epsilon=0. This is explained very clearly on page 355 of T.Chang et al..

No argument with the above.
The Krisher (thank you for correction) experiment uses neither of the above.

 

[Hans de Vries]

One way speed measurement.

Maybe everybody can agree on this:


It's possible to do a one way speed measurement if you
can establish a reference coordinate system, that is,
assign a (t,x,y,z) to each event in space time.

-We can establish SR reference frames (think GPS)
-We can not (currently) establish a GGT reference frame
since we need to know the preferred frame.

The result of measurements in an SR frame may be know a priory
because of the way the reference frame was established.

It becomes different when we do high precision measurements
to test aberrations, non-linearities, violations. These aberrations
may prevent us to establish a sufficiently exact coordinate system
and thus prevent meaningful one way speed of light measurements
in the context of high precision measurement.


Regards, Hans.

 

[pervect]

My $.02 on isotropy (i.e. the one-way speed of light).

There is a real, testable prediction to be made. This is the prediction that when coordinates are chosen such that the speed of light is isotropic, so is "everything else". This is the prediction SR makes.

Example - to clarify the above vague statement by example. Consider a 100 Mev electron beam that travels very close to the speed of light. SR predicts that a choice of coordinates that make light isotropic also make this 100 Mev electron beam isotropic, i.e. light moves at 'c' in all directions, the electron beam also moves at a uniform velocity in all directions, at just a hair under 'c'. SR also makes the prediction that 1 ev electron beams should also move isotropically (at some very low velocity) - the exact energy of the beam isn't really relevant. There's nothing special about electrons, either - if one can prepare a beam of any sort of particle of known specific energy, SR predicts that this beam will be isotropic (i.e. have isotropic velocities) when light is isotropic.

Beams of a specific energy are not even the only way to define isotropy, one could look at specific momentum (momentum / unit mass) rather than energy, and make the statement that uniform beams of constant specific momentum also have isotopic velocities.

This is the prediction of relativity.

Other theories might make the prediction that something will be anisotropic even when light is isotropic. It's a bit ugly, but it's testable. So far experiment upholds relativity. Not only light, but everything else as well, appears to behave isotropically when the correct coordinates are used.

When looked at with this viewpoint, the isotropy of light is being used as a definition to make sure that velocities are being measured properly, and what is being measured is not the isotropy of light, but the isotropy of "something else", exactly which something else may depend on the exact experiment.

This seems to me to be not only the simplest way to formulate the problem, but one which has firm historical roots. Apparently, though, not everyone views things from this viewpoint.

[add]And I don't mean just as evidenced in this thread, either, I've read a number of papers which do not take the viewpoint I advocate either. I would still like to immodestly promote my viewpoint, though, because I think that it's reasonable and avoids extended discussions of a lot of non-issues.

 

[Aether]

Maybe everybody can agree on this:


It's possible to do a one way speed measurement if you
can establish a reference coordinate system, that is,
assign a (t,x,y,z) to each event in space time.

Only if you're using the term "measurement" very loosely, because all real "measurements" are coordinate-system independent, although the result of a measurement can be interpreted with respect to a reference coordinate system. For example, a Doppler shift (e.g., \frac{\lambda}{\lambda_0} is a real (e.g., coordinate-system independent) measurement. The relativistic Doppler equation (e.g., \frac{\lambda}{\lambda_0}=\gamma (1\pm \frac{v}{c})) can be used to solve for \frac{v}{c} which is also a real measurement. However, solving for v by assuming that c=c_0 necessarily requires one to choose a coordinate system, and v does not represent a real measurement; it is a coordinate-system dependent interpretation of a real measurement.

-We can establish SR reference frames (think GPS)
-We can not (currently) establish a GGT reference frame
since we need to know the preferred frame.

Even in the absense of a violation of local Lorentz symmetry, we can still establish (currently) an arbitrarily preferred GGT frame. In the presence of any violation of local Lorentz symmetry (which we can not detect currently), we could futher establish a locally preferred Mansouri-Sexl frame, but that's a different issue.

The result of measurements in an SR frame may be know a priory
because of the way the reference frame was established.

It becomes different when we do high precision measurements
to test aberrations, non-linearities, violations. These aberrations
may prevent us to establish a sufficiently exact coordinate system
and thus prevent meaningful one way speed of light measurements
in the context of high precision measurement.

We can measure local Lorentz symmetry to ever increasing precision, but we need to keep the physical meaning of real measurements separate from the coordinate-system dependent interpretations of the measurements. GGT is important because it allows us to distinguish the coordinate-system dependent content of SR from it's physical content. There is a clear and present propensity for people to wrongly attribute physical significance to the coordinate-system dependent content of SR, and the more people strain against this distinction the more convinced I become that SR is misleading when it isn't viewed in light of GGT. Of course, Mansouri-Sexl goes beyond both SR & GGT and can be considered in terms of the search for violations of local Lorentz symmetry, but that's a different issue.

This seems to me to be not only the simplest way to formulate the problem, but one which has firm historical roots. Apparently, though, not everyone views things from this viewpoint.

I agree with you that this is an excellent way to formulate most practical problems.

 

[Aether]

No argument with the above.
The Krisher (thank you for correction) experiment uses neither of the above.

Immediately before Eq. (7) they say "The observable quantity is the variation in the phase differences as \theta changes...". How does a series of such observations translate into a "test of the isotropy of the one-way speed of light" unless you pick a value for \epsilon (see Eq. (3))? Theoretically at least, this appears to be an attempt to constrain both the rotational invariance component of Lorentz symmetry (e.g., (1/2+\delta-\beta)) as well as a first-order effect (e.g., (1+2\alpha)), and not really an attempt to measure the one-way speed of light per se.

 

[clj4]

Immediately before Eq. (7) they say "The observable quantity is the variation in the phase differences as \theta changes...". How does a series of such observations translate into a "test of the isotropy of the one-way speed of light" unless you pick a value for \epsilon (see Eq. (3))? Theoretically at least, this appears to be an attempt to constrain both the rotational invariance component of Lorentz symmetry (e.g., (1/2+\delta-\beta)) as well as a first-order effect (e.g., (1+2\alpha)), and not really an attempt to measure the one-way speed of light per se.

They are doing what everyone else does (did) , they are constraining the light speed anisotropy to within a few hundreds of m/s.

 

[Aether]

They are doing what everyone else does (did) , they are constraining the light speed anisotropy to within a few hundreds of m/s.

Assuming that \epsilon=-v/c_0^2, sure. Otherwise, no.

 

[clj4]

Assuming that \epsilon=-v/c_0^2, sure. Otherwise, no.

The correct statement is that each experiment measures one way light speed isotropy within the adopted clock synchro scheme. Makes sense?

 

[gregory_]

The correct statement is that each experiment measures one way light speed isotropy within the adopted clock synchro scheme. Makes sense?

So you are saying: using Einstein synchronization we find the light speed to be isotropic? OF COURSE, since you defined the light speed to be isotropic to setup the coordinate system.

The point is that the Krishner experiment does not distinguish between synchronization schemes. We can use the GGT synchronization scheme and it will agree with their experiment.

Read this:

As you already admitted yourself, one way velocities are a coordinate system dependant thing. Krishner himself states that they can't distinguish between coordinate systems that differ only in clock synchronization. Therefore they did not distinguish between the GGT or SR coordinate systems which have different one-way speeds of light.

Do you agree with that?
If not, please state specifically what you disagree with and why.

I agree that SR and GGT agree in the preferential frame of GGT.

Good.

Now answer this:
How can you claim SR and GGT make different predictions for any experiment then? Are you trying to claim GGT is mathematically inconsistent?


I am also still waiting for your "disproof" of the fact that Gagnon forgot to use the GGT version of the Lorentz force which made his calculations incorrect.

 

[clj4]

No. One way speed measurements are not valid.
If you'd take the time to step back and think about it you'd see this.

With such an embarassing biased statement all you'll get is the disproof for your Gagnon disproof. It will take me some time, in the meanwhile you may want to double check your disproof.

 

[clj4]

How can you claim SR and GGT make different predictions for any experiment then? Are you trying to claim GGT is mathematically inconsistent?

Read post 286. If you have difficulties with comprehension, read it again.

I am also still waiting for your "disproof" of the fact that Gagnon forgot to use the GGT version of the Lorentz force which made his calculations incorrect.

You'll have to wait. In the meanwhile double check your "disproof".