Following up on this question on whether it's possible to construct a physical theory with invariant space-time and variable speed of light.

I am looking for a authoritative and more definitive answer to the hopefully more precise question:

Are Lorentz ether theory and special relativity fully compatible/interchangeable?

Interchangeable in the sense that, at least in principle,there is either no different prediction the other doesn't make or if there is it could be settled experimentally (or has been) in favor of one of the two?

From the wikipedia page I gather that they are indistinguishable, but SR is preferred over LET either for practical reasons (elegance, convenient) or because it's assumption are somewhat less arbitrary or don't require somewhat problematic entities such as ether.

  • $\begingroup$ This would be better for History of Science and Mathematics $\endgroup$ – user4552 Jan 9 '20 at 19:17
  • $\begingroup$ No, the author mentions ether in passing to... SR treatment is stated clearly by the inventor and suggests a symmetric and drastically simplified electro/magento-dynamics and concludes there is no absolute rest, or absolute anything, except the constant speed of light in vacuum, which is amazingly just a postulate equivalent to the conclusion, and no suggestion is needed. $\endgroup$ – user192234 Jan 9 '20 at 19:34
  • $\begingroup$ @safesphere as you can see from answers and comments this is not entirely clear/accepted. It surely is not in the world of non physicists, but perhaps also among physicists. Maybe state this as an answer? $\endgroup$ – Three Diag Jan 12 '20 at 1:00
  • $\begingroup$ A major reason for the ambiguity in the answers it that "Lorentz ether theory" can be taken to mean a lot of things, ranging from long-disproven options, to options which are manifestly indistinguishable from ordinary relativity. $\endgroup$ – knzhou Jan 12 '20 at 23:23
  • $\begingroup$ @ThreeDiag There are Lorentz's comments on Einstein's theory and there are Einstein's comments on ether theory. $\endgroup$ – Alex Trounev Jan 14 '20 at 11:21

Are Lorentz ether theory and special relativity fully compatible/interchangeable? Interchangeable in the sense that, at least in principle,there is either no different prediction the other doesn't make or if there is it could be settled experimentally (or has been) in favor of one of the two?

Special relativity (SR) uses the Lorentz transform to make all of its experimental predictions. Lorentz aether theory (LET) also uses the Lorentz transform to make all of its experimental predictions. Furthermore, the reading of a clock or a ruler would be mapped to the same variable in both and also a given reference frame would either be determined to be inertial or non-inertial for both.

As a result, there is simply no possible way to distinguish between the two experimentally. Any result that is predicted by one is predicted by the other also. Their disagreements are entirely philosophical or metaphysical. Due to that fact it is sometimes considered that both LET and SR are simply different interpretations of the same theory (usually confusingly called SR).

Here is a solid reference describing the experimental equivalence and the philosophical distinction between the two:


  • 1
    $\begingroup$ @AgniusVasiliauskas Just using the word "ether" doesn't make a theory invalid. It depends entirely on how they are using it. If it agrees with experiments, that's really all that truly matters. The words they use to describe that prediction is a secondary concern, and never enough to invalidate a theory. It mostly only affects popularity. $\endgroup$ – Arthur Feb 26 '20 at 8:50
  • 2
    $\begingroup$ @AgniusVasiliauskas Ether in the particular form usually referenced in popular science was disproven by the MM experiment. The idea of ether in general wasn't. $\endgroup$ – Arthur Feb 26 '20 at 8:53
  • 1
    $\begingroup$ @AgniusVasiliauskas They can't be differentiated, though, as the ether in LET is undetectable. Just because one theory has a concept that the other does not, that doesn't mean they can be differentiated experimentally. Any experiment will give the same predicted result in LET as in SR. The fact that one theory uses a word you don't like to reach said result doesn't change anything. $\endgroup$ – Arthur Feb 26 '20 at 11:06
  • 2
    $\begingroup$ @AgniusVasiliauskas It's the exact same story: A mathematical construct that can't be measured or observed, used to facilitate calculations so that you can make predictions. Those predictions, be it the movement of light, or the probability density of positions and momenta of particles, can be measured (well... the probability density can't really be measured either, just sampled from). But what exactly is going on behind the scenes can't. I fail to see the distinction here. At any rate, I don't see either of us convincing the other, so there is no point in carrying on this discussion. $\endgroup$ – Arthur Feb 26 '20 at 13:55
  • 3
    $\begingroup$ @Arthur just for the record, I agree with AgniusVasiliauskas characterization of the Lorentz aether as "silly" and as violating Occham's razor. However, the question was not about that, it was about experimental equivalence. So I answered the factual question on its own terms without injecting my opinions. Regarding the wave function, if someone were able to create a mathematical framework that removed the need for the wavefunction then I am sure that the wavefunction would also go the way of Lorentz' aether. Occham's razor requires two alternatives, and we don't have one for the wavefunction $\endgroup$ – Dale Feb 26 '20 at 14:09

Lorentz ether theory describes a world in which light moves through a medium called ether, and observers that are not at rest with respect to this ether see everything Lorentz transformed. In some philosophical sense, there is a preferred reference frame: that in which the ether is at rest.

But the funny thing about Lorentz transformations is precisely that they relate reference frames in which light moves in the same way; the speed of light does not change upon applying your favorite Lorentz transformation. For this reason, it is not possible even in principle to determine experimentally what the ether rest frame is, since even light (the stuff that the ether was invented for) moves in the same way in different inertial reference frames.

The theory of special relativity explains all of this more naturally by postulating that all inertial reference frames are similar and that there is no ether, at least not one that sets one specific reference frame apart. One could try to argue that the electromagnetic field as found in quantum electrodynamics can be called an 'ether'; however, since the electromagnetic field does not have one specific inertial frame it likes to live in, this is mostly semantics.


No. Such equivalence is inconsistent with Galilean principle: inertial observers are equivalent, and there's no physical processes lest you difference between beying at rest or moving with constant speed.

  • 1
    $\begingroup$ en.wikipedia.org/wiki/Lorentz_ether_theory: Today LET is often treated as some sort of "Lorentzian" or "neo-Lorentzian" interpretation of special relativity. The introduction of length contraction and time dilation for all phenomena in a "preferred" frame of reference, which plays the role of Lorentz's immobile aether, leads to the complete Lorentz transformation (see the Robertson–Mansouri–Sexl test theory as an example). Because the same mathematical formalism occurs in both, it is not possible to distinguish between LET and SR by experiment. $\endgroup$ – Albert Jan 9 '20 at 17:59
  • $\begingroup$ There is no asumption about constant speed of light it stw. It is only a trick used in high school when pupils don't know about Maxwell equations. Here you have proof starting from Galilean principle, and you get maximal speed as a parameter of the theory: jakubw.pl/faq/fizyka/node31.html It is not an accident that Einstein works about this, was published as papers about Maxwell equations. $\endgroup$ – kakaz Jan 9 '20 at 18:04
  • 1
    $\begingroup$ mpiwg-berlin.mpg.de/litserv/diss/janssen_diss/Chapter3.pdf "In an important paper published in 1899, both in a Dutch and a slightly different English version (Lorentz 1899a, 1899b), Lorentz introduced a new transformation under which the source free Maxwell equations are exactly invariant. Up to an undetermined constant, which Lorentz would eventually set equal to unity (Lorentz 1904b), this new transformation is what we now recognize as the exact Lorentz transformation. $\endgroup$ – Albert Jan 9 '20 at 18:12
  • $\begingroup$ Sorry, I don't have time to read something which is incompatible with the facts $\endgroup$ – kakaz Jan 9 '20 at 18:15
  • 1
    $\begingroup$ Hi @kakaz. I just went through the link you posted, but since I don't understand Polish (?) I might have misunderstood the point. It seems to me that if you assume linear relations, you can derive the Lorentz Transformations. However, when you actually work out the velocity addition formula, you find that $$u^\prime = \frac{u-v}{1-uv/k^2},$$ where $k$ is a parameter in the problem (related to $C$ in your link). Of course, if $k=c$, it's STR, and $k=\infty$ gives Galilean Relativity, so I think the argument needs either an experimental reason to have a frame independent speed, or a postulate. $\endgroup$ – Philip Jan 9 '20 at 19:04

Ether – less Special Relativity cannot be sound alternative to Lorentz theory.

Special Relativity assumes isotropy of one – way speed of light in all relatively moving frames.This isotropy follows from Einstein synchrony convention, that makes the one-way speed equal to the two-way speed.

Michelson - Morley experiment confirms, that in moving laboratory round trip speed of light appears to be isotropic, but this experiment says nothing about one - way speed of light. Measured isotropy of two - way speed of light in moving laboratory is due to distortion of interferometer (Lorentz - contraction).

It should be noted, that one – way speed of light is anisotropic relatively to a rim of rotating ring (Sagnac Effect). The ring can be of arbitrarily large diameter. It is experimentally confirmed fact. Hence, speed of light is anisotropic relatively to any laboratory on the Earth surface.

The only correct answer to this question is that if we assume that the speed of light is isotropic in one reference frame (and synchronize clocks Einstein – way), it is necessarily anisotropic in the other, relatively moving one and one must employ anisotropic (Reichenbach's) synchrony convention which is also self- consistent.

My answer to this question demonstrates, that time dilation is not reciprocal and only absurd synchronization leads to reciprocal time dilation.

If you consider which theory is more elegant and simple, just keep in mind that resolution of twin paradox in frameworks of Lorentz Theory and the ether – less Special Relativity demonstrates ugliness of the latter. Resolution of the Twin paradox in Lorentz Theory fits into one page, while raving fans of ether – less relativity still put forward hypotheses, each his own one.

The same goes to Bell’s spaceship paradox. Explanation why the thread breaks (or spaceships scatter from each other) is very simple in the framework of Lorentz theory.

Champeney and Moon time dilation test demonstrates absence of reciprocal time dilation for two observers that are placed on opposite sides of rotating ring. It is clear why – because their clocks dilate at the same magnitude and this dilation chancel each other. Neither rotating observers, nor purely inertial ones that coincide with rotating at the moment of measurement would not measure any reciprocal time dilation ( See fig. 4 in this article)

Minkowski formalism cannot be sound alterative to the Ether; it simply turns simple things into complex ones: it’s just a matter of time when this annoying concept will just be dumped, I believe that the following comment here indicates that this moment is getting closer:

‚@Dale If one fills all space time with particle fields on which operators work, and posits that this is the underlying reality, as most theoretically inclined people do, what is it but a reemergence of an aether? It just fulfills the function of an aether, for signals, in this case particles.‘


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.