Thread on Lorentz Invariance Violation

[Russell E]

The frequency of the interference filter is specified very clearly in Ref [14], which is available at http://archiv.ub.uni-heidelberg.de/...xte/2005/5934/pdf/doktorarbeit_sreinhardt.pdf . In particular, the caption of Figure 6.7 of Ref [14] states “… and an interference filter (transmission 548 nm, half-width 10 nm (s.[Mer00])) are placed”. This is the transition frequency $$\eta_o$$ in the laboratory frame as seen in Figure 3.1 of Ref [14]. .

Thanks for the link. It confirms what I previousloy surmised, namely, that the filter is irrelevant. Look, the filter is centered on 548 nm with half-wide of 10nm, so it passes quite well anything between 538 nm and 558 nm. Now, they are using a maximum beta of 0.064, so the transverse Doppler shifted wavelength is about 549 nm, which is still more or less at the center of the +-10 nm band pass. So the flouresence they observed is perfectly suitable for establishing the laser tuning at the Lamb dip.

The point you should keep in mind is that the experiment is not measuring the frequency of the ion emissions, it is simply using the intensity of those emissions at various laser settings to identify the resonant condition. A filter of half-width 10 nm centered on 548 nm will obviously work just as well as one centered on 549 nm to identify the distinctive Lamb dip. Also, note that they made adjustments until they maximized the flouresence, which ensures they are looking at the main resonance, and not some incidental side resonance, as you seem to suppose.

Your fundamental confusion is that you were misled into thinking the filter they used would somehow cause them to mis-identify the resonant condition, and you compounded this misconception with the idea that the Stark and/or Zeeman effects would somehow conspire to fool the experimenters into thinking they were looking at emissions from nominal ions, and then you RE-compounded your confusion by convincing yourself that the issue had something to do with interpretation of special relativity, when in fact you were claiming gross experimental error (falsely, as it turns out).

You should also note that many other experiments of this general kind have been performed with no filter at all, so your whole thesis is based on a misunderstanding of a very particular experimental detail of one particular experiment.

I do not think that there is an error in the experiments. For example, if the photons emitted at $$\eta_o$$ (in the moving ion-frame) were to be measured then the interference filter would have to be at $$\eta_o/\gamma$$ in the laboratory frame. However, this would require a-priori knowledge of the time-dilation factor $$\gamma$$, which is what the experiments are trying to evaluate (at high speeds).

You completely misunderstand the situation, because you mistakenly imagined that a filter (used in one particular experiment) had a narrow enough bandwidth to exclude the photons from the ions at resonance, which led you to believe the experimenters must have tuned the lasers to a sub-population of ions, whose resonant frequency you fantasized must have been shifted by the Stark and/or Zeeman effects by an amount that exactly matched the transverse Doppler shift between ions and lab (so it could get through the infinitely narrow filter that you mistakenly imagined). Hopefully you now realize that the filter they used can pass 549 nm photons with no difficulty, so the flouresence profile from the ions (at 548 nm in their rest frame) will be clearly apparent, and hence the lasers will be tuned correctly, and hence the test is valid and np*na/no^2 = 1 for Lorentz invariance.

(Actually they evaluated np and na at two different ion speeds, 0.03c and 0.064c, and then showed that (np1*na1)/(np2*na2) = 1 to extremely high precision.)

Therefore, placing the interference filters at $$\eta_o$$ (in the laboratory frame) is quite reasonable in the experiments and not an error.

It is not an error, but only because the bandpass of the filter is perfectly well suited to pass the resonant emissions. If the filter had the effect that you fantasize it had (i.e., blocking the main resonant emissions from contributing to the flouresence, leading them to mis-identify the primary resonance condition and laser settings), it WOULD have been a gross and totally inexplicable experimental error. Experimental physicists are not nearly as stupid as you imagine.

However, the time-dilation effect between the emitted and observed photons needs to be included in the SR calculations with the use of the pre-filter --- since the observed lamb-dip is for photons at a specific frequency$$\eta_o$$ in the laboratory frame.

Hopefully by now you see that your claim is completely erroneous, based on a total misunderstanding of the experiment, combined with numerous logical fallacies and non-sequiturs. The filter used in the experiment is quite suitable for correctly identifying the Lamb dip for the main ion resonance of 548 nm in the rest frame of the ions (about 549 in the lab frame). The appropriate value to use for nu_a* in the definition of R is nu_0, i.e., the frequency corresponding to 548nm, and this gives R = 1 is Lorentz invariance is true. Also, as mentioned above, the paper actually bases its test on two non-zero velocities, so nu0 doesn't even enter into the final calculations.

 

[Dale]

OK, so here is my work. I used the wave four-vector with c=1 and with the timelike component first:
$$\mathbf f = (\nu,\nu \mathbf n)$$

where $\nu$ is the frequency and $\mathbf n$ is a unit vector in the direction of propagation. An observer moving with a four-velocity
$$\mathbf v = (\gamma, \gamma v, 0, 0)$$

will detect or emit a frequency
$$\nu = \mathbf v \cdot \mathbf f$$

Where the dot product represents the Minkowski inner product. The advantage of this formalism is that the detected frequency automatically includes all relativistic effects including time dilation, Doppler shift, and aberration.

So, in the lab frame we have
$$\mathbf f_a = (\nu_a,-\nu_a,0,0)$$
$$\mathbf f_p = (\nu_p,\nu_p,0,0)$$
$$\mathbf f_e = (\nu_e,0,\nu_e,0)$$

Thus, for a particle with four-velocity $\mathbf v$ the detected/emitted frequencies are:
$$\nu_a* = \mathbf v \cdot \mathbf f_a = (1+v) \gamma \nu_a$$
$$\nu_p* = \mathbf v \cdot \mathbf f_p = (1-v) \gamma \nu_p$$
$$\nu_e* = \mathbf v \cdot \mathbf f_e = \gamma \nu_e$$

The resonant condition gives:
$$\nu_a*=\nu_p*=\nu_e*=\nu_0$$

Substituting the resonant condition into the above equations for the particle-frame frequencies and solving for the lab-frame frequencies we get:
$$\nu_a=\frac{\nu_0}{(1+v)\gamma}$$
$$\nu_p=\frac{\nu_0}{(1-v)\gamma}$$
$$\nu_e=\frac{\nu_0}{\gamma}$$

Then by substitution and simplification we obtain:
$$R = \frac{\nu_a \nu_p}{\nu_0^2} = 1$$

 

[bcrowell]

Is it a policy of this forum to ban all these ignorant, unwilling to learn, "trolls, fools, kooks, and/or uninformed loudmouths"?

Here are the rules of the forum: https://www.physicsforums.com/showthread.php?t=414380

 

[bcrowell]

The anti-paper states that in those experiments the transverse Doppler effect gets canceled by time dilatation and length contraction in the ion's RF. Where can it be observed and why it is not canceled in those cases?

See the sticky at the top of this forum titled "FAQ: Experimental Basis of Special Relativity."

 

[Weber2]

OK, so here is my work. I used the wave four-vector with c=1 and with the timelike component first:
$$\mathbf f = (\nu,\nu \mathbf n)$$

where $\nu$ is the frequency and $\mathbf n$ is a unit vector in the direction of propagation. An observer moving with a four-velocity
$$\mathbf v = (\gamma, \gamma v, 0, 0)$$

will detect or emit a frequency
$$\nu = \mathbf v \cdot \mathbf f$$

Where the dot product represents the Minkowski inner product. The advantage of this formalism is that the detected frequency automatically includes all relativistic effects including time dilation, Doppler shift, and aberration.

So, in the lab frame we have
$$\mathbf f_a = (\nu_a,-\nu_a,0,0)$$
$$\mathbf f_p = (\nu_p,\nu_p,0,0)$$
$$\mathbf f_e = (\nu_e,0,\nu_e,0)$$

Thus, for a particle with four-velocity $\mathbf v$ the detected/emitted frequencies are:
$$\nu_a* = \mathbf v \cdot \mathbf f_a = (1+v) \gamma \nu_a$$
$$\nu_p* = \mathbf v \cdot \mathbf f_p = (1-v) \gamma \nu_p$$
$$\nu_e* = \mathbf v \cdot \mathbf f_e = \gamma \nu_e$$

The resonant condition gives:
$$\nu_a*=\nu_p*=\nu_e*=\nu_0$$

Substituting the resonant condition into the above equations for the particle-frame frequencies and solving for the lab-frame frequencies we get:
$$\nu_a=\frac{\nu_0}{(1+v)\gamma}$$
$$\nu_p=\frac{\nu_0}{(1-v)\gamma}$$
$$\nu_e=\frac{\nu_0}{\gamma}$$

Then by substitution and simplification we obtain:
$$R = \frac{\nu_a \nu_p}{\nu_0^2} = 1$$

Nice analysis --- however, the above analysis does not include the effect of the prefilter in the observations (which is the main point of the EPJ C article).

If only photons at $$\nu_o$$ (in the laboratory frame) are being measured due to the narrow prefilter, then the resonant condition is being measured for a specific set of photons, i.e., of frequency $$\nu_e = \nu_o$$ (in the laboratory frame) which corresponds to photons emitted at $$\nu_e^* = \gamma\nu_e = \gamma\nu_o$$ (in the moving ion frame).

This along with the resonant conditions (for the photons being measured) $$\nu_a*=\nu_p*=\nu_e*$$

leads to
$$R = \frac{\nu_a \nu_p}{\nu_0^2} = \gamma^2 \ne 1$$

 

[Dale]

Nice analysis --- however, the above analysis does not include the effect of the prefilter in the observations (which is the main point of the EPJ C article).

Filters don't change frequencies, so they can't change R (even in principle) which is just a ratio of frequencies.

There is nothing wrong with the experiments at all.

If only photons at $$\nu_o$$ (in the laboratory frame) are being measured due to the narrow prefilter

You cannot have it both ways. Either the experimenters are all stupid and all set up their experiments completely wrong and therefore did not actually measure R or the experimenters are not all stupid and they actually did measure R. You need to pick a position on this issue, because you are contradicting yourself.

In either case, the SR prediction is clearly that R=1. If the experimenters were all stupid and did not measure R then their results are invalid and therefore not evidence of Lorentz violation but rather evidence that even a lot of smart people are stupid. If the experimenters did measure R then their results are also not indications of Lorentz violation since R=1 is the predicted and measured result (within experimental error).

 

[Russell E]

Filters don't change frequencies, so they can't change R (even in principle) which is just a ratio of frequencies.

He isn't claiming that the filter changes the frequency. He's claiming (mistakenly, of course) that the filter passes *only* the wavelength 548 nm, and that it excludes light of 549 nm, which is the wavelength we would expect (after accounting for the transverse Doppler shift) for light emitted from the ions at their resonant wavelength of 548 nm. He is mistaken about this, because the filter in question had a half-width of 10 nm, so it passed 549 nm light just fine. Hence the "reasoning" in that EJPC letter is based on a completely false premise.

Based on the mistaken belief that only the precise wavelength 548 nm can pass through the filter, he infers that we must be observing light emitted from ions with a frequency of 547 in their rest frame (accounting for transverse Doppler). He argues that this is possible, because some of the ions may have their resonant frequencies shifted by the Stark and/or Zeeman effects, due to some uncontrolled external fields in the apparatus. He doesn't actually analyze these effects to see if this is even a realistic possibility (which it isn't), he just offers it as conjectural explanation for why the actual resonant frequency of the ions observed through the filter (and hence tuned to the lasers) is really 547 nm instead of 548 nm.

On the basis of these factual errors, conceptuals misunderstandings, and unfounded imaginings, he arrives at the conclusion that the relevant ratio condition for Lorentz invariance is (np)(na)/(ne*)^2 = 1 where ne* is the frequency corresponding to 547 nm. This of course is just the usual condition, the only difference being that he thinks the resonant frequency of the ions in their rest frame is different than it is measured to be. However, to increase the level of obfuscation, he chooses to state his conclusion slightly differently, by saying that, if we define n0 as the frequency for 548 nm (the emission frequency in the lab frame, rather than the ion frame), then the condition for Lorentz invariance is (np)(na)/(n0)^2 = 1 + v^2. This is a stupid way of expressing it (because the accepted definition of the denominator is the ion frequency in the ion frame, not the lab frame), but that's how he chooses to express it. He then tries to pretend that he is making a statement about relativistic effects, whereas in fact he is claiming the experiment was botched by allow uncontrolled Stark and Zeeman effects to offset the result (by an amount that just by coincidence gave the expected ratio for the wrong frequency!)

His position is totally absurd. His errors have been clearly explained to him multiple times, but he appears determined to persist in error, and to disregard the explanations.

One more comment: The paper he cites doesn't actually even use the ion rest frame frequency in their ratio calculation. They perform the experiment for two different velocities, v1 = 0.064c, and v2 = 0.030c, and then they compute that the ratio

np(v1)na(v1)/[np(v2)na(v2)] = 1

Of course, if v2 = 0, this formula reduces to np(v1)na(v1)/[ne*^2] = 1, because ne* is the resonant frequency we would measure if the ions have v = 0, by definition. This just emphasizes even more how totally clueless is the EJPC letter.

 

[Dale]

He isn't claiming that the filter changes the frequency. He's claiming (mistakenly, of course) that the filter passes *only* the wavelength 548 nm, and that it excludes light of 549 nm, which is the wavelength we would expect (after accounting for the transverse Doppler shift) for light emitted from the ions at their resonant wavelength of 548 nm.

Yes, in which case all the previous experimenters were not actually measuring R. But R=1 is unambiguously the prediction of relativity and is not at all affected by the presence or absence of filters.

In any case such a sharp filter would be very impressive.

 

[bcrowell]

There is nothing wrong with the experiments at all.

If only photons nuo (in the laboratory frame) are being measured due to the narrow prefilter

You cannot have it both ways. Either the experimenters are all stupid and all set up their experiments completely wrong and therefore did not actually measure R or the experimenters are not all stupid and they actually did measure R. You need to pick a position on this issue, because you are contradicting yourself.

This is a typical situation when trying to have a logical debate with creationists, anti-relativists, etc. They don't commit themselves to any specific theory that makes predictions, because then their theory would be proved wrong by experiments. Here it may be helpful to understand something about the belief system of Assis, Devasia, et al. They want to believe that Weber's electrodynamics provides a Machian theory that overturns relativity. Assis has built a nice Machian toy model called relational dynamics, which is predicated on a notion of absolute time, but is compatible with something resembling a Lorentz contraction of lengths, as well as dynamical effects such as corrections of order v^2/c^2 to momenta. The model is cute, but it can't be applied to our universe, because it is contradicted by all the experimental evidence on Lorentz invariance, and it also requires the "tired light" hypothesis in order to explain cosmological redshifts.

This may make it clearer why Weber2 is contradicting himself, and why Devasia wants to claim that experiments on Lorentz invariance for the past century have all been wrong. Devasia's claims about transverse Doppler shifts don't make any sense, but from the point of view of the Assis types it absolutely makes sense to focus on transverse Doppler shifts, because transverse Doppler shifts are basically interpreted as kinematic time dilation, and their theory is not compatible with the idea that time is not absolute. On the other hand, Assis and Devasia both know that they will get nowhere if they admit too forthrightly that they think SR is wrong and time is absolute. For example, Devasia clearly never would have been able to slip his EPJ paper past a sloppy referee if the abstract had said, "SR is wrong." This is why we get all the waffling about how the experiment are wrong, the experiments are not wrong, SR is wrong, SR is right, etc.

 

[Tantalos]

Thanks for the link. It confirms what I previousloy surmised, namely, that the filter is irrelevant. Look, the filter is centered on 548 nm with half-wide of 10nm, so it passes quite well anything between 538 nm and 558 nm. Now, they are using a maximum beta of 0.064, so the transverse Doppler shifted wavelength is about 549 nm, which is still more or less at the center of the +-10 nm band pass. So the flouresence they observed is perfectly suitable for establishing the laser tuning at the Lamb dip.

The point you should keep in mind is that the experiment is not measuring the frequency of the ion emissions, it is simply using the intensity of those emissions at various laser settings to identify the resonant condition. A filter of half-width 10 nm centered on 548 nm will obviously work just as well as one centered on 549 nm to identify the distinctive Lamb dip. Also, note that they made adjustments until they maximized the flouresence, which ensures they are looking at the main resonance, and not some incidental side resonance, as you seem to suppose.


(Actually they evaluated np and na at two different ion speeds, 0.03c and 0.064c, and then showed that (np1*na1)/(np2*na2) = 1 to extremely high precision.)

So how they knew the frequency was 548nm and not 549nm?
Measuring the frequency would disperse any confusion.

 

[Doc Al]

This is a typical situation when trying to have a logical debate with creationists, anti-relativists, etc. They don't commit themselves to any specific theory that makes predictions, because then their theory would be proved wrong by experiments. Here it may be helpful to understand something about the belief system of Assis, Devasia, et al. They want to believe that Weber's electrodynamics provides a Machian theory that overturns relativity. Assis has built a nice Machian toy model called relational dynamics, which is predicated on a notion of absolute time, but is compatible with something resembling a Lorentz contraction of lengths, as well as dynamical effects such as corrections of order v^2/c^2 to momenta. The model is cute, but it can't be applied to our universe, because it is contradicted by all the experimental evidence on Lorentz invariance, and it also requires the "tired light" hypothesis in order to explain cosmological redshifts.

This may make it clearer why Weber2 is contradicting himself, and why Devasia wants to claim that experiments on Lorentz invariance for the past century have all been wrong. Devasia's claims about transverse Doppler shifts don't make any sense, but from the point of view of the Assis types it absolutely makes sense to focus on transverse Doppler shifts, because transverse Doppler shifts are basically interpreted as kinematic time dilation, and their theory is not compatible with the idea that time is not absolute. On the other hand, Assis and Devasia both know that they will get nowhere if they admit too forthrightly that they think SR is wrong and time is absolute. For example, Devasia clearly never would have been able to slip his EPJ paper past a sloppy referee if the abstract had said, "SR is wrong." This is why we get all the waffling about how the experiment are wrong, the experiments are not wrong, SR is wrong, SR is right, etc.

Moderator's Note: With these words from bcrowell, I think it is time to close this thread. I would like to thank everyone for their contributions.