I Understanding the historical shift away from absolute simultaneity

[Iskandarani]

Hello everyone,

I'm trying to strengthen my understanding of the foundational principles of Special Relativity. My intuition, like that of many classical physicists, leans towards a concept of absolute simultaneity, where we could determine a universal "now" by accounting for the travel time of a signal, much like we do with sound.

I understand that the constancy of the speed of light for all observers makes this simple picture impossible. Could someone help walk me through the key experimental evidence (like the Michelson-Morley experiment) and the logical steps that forced the physics community to abandon the idea of a universal "now" (and the related concept of a luminiferous aether) in favor of the relativity of simultaneity?

I'm particularly interested in understanding how physicists first grappled with the counter-intuitive consequences of c being constant. Thank you!

 

[phinds]

what part of the Michelson-Morley experiment do you not understand ?

 

[FactChecker]

In general, you should separate the goals of clear understanding of physics versus knowing the history. There were many confused people in history.
That being said, this might be one exception. How can the speed of a light pulse be measured as identical in two different inertial reference frames? It's impossible if there is absolute simultaneity. There must be something about how the clocks, synchronized in their respective frames, measure speed that explains it.

 

[A.T.]

I understand that the constancy of the speed of light for all observers makes this simple picture impossible. Could someone help walk me through the key experimental evidence (like the Michelson-Morley experiment) and the logical steps that forced the physics community to abandon the idea of a universal "now" (and the related concept of a luminiferous aether) in favor of the relativity of simultaneity?

The video below explains that pretty well.

 

[martinbn]

Could someone help walk me through the key experimental evidence (like the Michelson-Morley experiment) and the logical steps that forced the physics community to abandon the idea of a universal "now" (and the related concept of a luminiferous aether) in favor of the relativity of simultaneity?

I'm particularly interested in understanding how physicists first grappled with the counter-intuitive consequences of c being constant. Thank you!

You want some one to teach you, in a forum post, the theory of relativity and its history!

 

[Ibix]

For me, the simplest argument is to propose that if global inertial frames exist at all then the transformation between them must map straight lines ($x=vt+x_0$) to straight lines ($x'=v't'+x'_0$), so the transformation must be representable as a 2×2 matrix whose components can only depend on the relative velocity of the frames. By considering a few obvious cases (where (vt,t) must map to, for example) and insisting that the composition of two transforms must have the same form as a single transform you can narrow things down to the Galilean transforms or the Lorentz transforms (with an unknown constant $c$ with dimension of velocity). Then you can look at Bertozzi's experiment (Bertozzi himself recorded it, and it's posted on YouTube) to see that the Galilean transforms are incorrect. And you could work out what that $c$ is at that point too.

 

[BWV]

My intuition, ..

you do realize your intuition is worthless? Intuition comes from past learning and experiences. I am sure trained physicists develop a good intuition over the years, but raw untrained human intuition about physics simply sucks

 

[Iskandarani]

Thank you all for the replies and resources, this is very helpful. My apologies for the overly broad nature of my original question.

@A.T. - I've watched the video you linked, and it did a great job of visualizing the core concepts. Thank you for that.

@phinds - To take up your challenge and be more specific about the Michelson-Morley experiment: I understand that the experiment's null result was a major blow to the idea of a static luminiferous aether. My specific point of confusion is this: How did the physics community distinguish between the conclusion "there is no aether" and the alternative hypothesis of "full aether drag" (where the Earth drags the aether along with it), which would also seem to explain the null result?

Was the aether drag hypothesis falsified by a different experiment, like the Fizeau experiment or observations of stellar aberration?

@Ibix - Your mathematical argument is fascinating. I'll need some time to fully digest that, but it seems to be the most rigorous way to approach the problem.

 

[Herman Trivilino]

As you probably know the MM experiment was done in 1889 and Einstein's theory was published in 1905. Einstein claimed that his notion of the invariance of the speed of light came from Maxwell's Equations (ca. 1860's) not from MM. During those decades physicists were concerned with unifying Maxwell's theory with newtonian physics. The predominant idea was that Maxwell's Equations needed to be modified to account for an absolute motion through a medium, the aether. See eg. Lorentz Theory.

Einstein's revolutionary idea was to leave Maxwell's Equations unchanged and to replace newtonian physics with a new theory.

It was by no means accepted by physicists, for the most part, even though others such as Poincare had written and spoken about the need to replace the newtonian foundation based on absolute time.

It took a lot of experimental evidence over the next decades for Einstein's notion to be widely accepted. Read Clifford M Will's writings for summaries of the overwhelming evidence of the validity of Einstein's theories of relativity.

 

[Ibix]

How did the physics community distinguish between the conclusion "there is no aether" and the alternative hypothesis of "full aether drag"

Other experiments. The pre-1905 section of the SR experimental basis FAQ (linked in this sticky thread at the top of the forum) describes some of it - in particular the aberration of starlight is inconsistent with a dragged ether.

 

[A.T.]

@phinds - To take up your challenge and be more specific about the Michelson-Morley experiment: I understand that the experiment's null result was a major blow to the idea of a static luminiferous aether. My specific point of confusion is this: How did the physics community distinguish between the conclusion "there is no aether" and the alternative hypothesis of "full aether drag" (where the Earth drags the aether along with it), which would also seem to explain the null result?

The MM-Experiment by itself doesn't imply SR. You need other observations to rule out alternative theories. See for example:
https://en.wikipedia.org/wiki/Emission_theory_(relativity)

 

[Iskandarani]

Thank you, @Herman Trivilino, for that insightfull historical summary. It clarifies that the physics community was at a genuine fork in the road. (i dont know if fork is the right term, but as a programmer Im used to it)

It seems one path, Lorentz's, was to propose new physical dynamics—real length contraction, for instance—to explain how matter interacts with a preferred reference frame (the æther) to produce the observed results. The other path, Einstein's, was to propose new principles of spacetime itself, which made the aether concept unnecessary.

My question is this: Given that both a Lorentz-style theory and Special Relativity were constructed to be mathematically consistent with the experiments of the time (like M-M), was the widespread acceptance of Einstein's interpretation over Lorentz's driven more by a philosophical principle like Occam's Razor (eliminating the "unobservable" aether), rather than a direct experimental falsification of all possible preferred-frame theories?
--------------------------------------------

And I want to thank you both, @Ibix and @A.T. This is a crucial point that I was missing. It's clear now that the Michelson-Morley experiment on its own was not the final word. The combination of its null result with observations of stellar aberration and the Fizeau experiment seems to definitively rule out both a simple static æther and a fully dragged æther.

This leads me to a deeper question about what was actually left on the table. What about a more sophisticated model, like a Lorentz Ether Theory, where:

1. A preferred frame (an æther) does exist.
2. Any measuring device (clocks, rulers) moving through this frame undergoes real, physical changes (time dilation, length contraction).

In such a theory, the physical effects would conspire to make the speed of light measure the same for all inertial observers, making the preferred frame itself hidden from kinematic experiments.

My question is: Is this type of Lorentz Ether Theory considered experimentally falsified, or is it generally accepted as being kinematically equivalent to Special Relativity, with the choice between them being a matter of convention and interpretation?

 

[PeterDonis]

Is this type of Lorentz Ether Theory considered experimentally falsified

It can't be since it makes all of the same experimental predictions as standard SR.

is it generally accepted as being kinematically equivalent to Special Relativity, with the choice between them being a matter of convention and interpretation?

I would say the choice isn't really a matter of "convention and interpretation" but simplicity. LET is basically SR plus the claim that there's some undetectable absolute reference frame. But the latter claim does zero work, so it gets scraped right off by Occam's Razor.

 

[Iskandarani]

that's a very clear and powerful argument. I understand completely.

If the only job of a theory is to describe relativistic kinematics, then adding an undetectable preferred frame on top of SR is a textbook case for Occam's Razor. The additional entity does no work in explaining the phenomena SR already describes, so it should be discarded for being less simple.

This leads me to a final question on the matter. The argument that the preferred frame "does zero work" seems to depend on the scope of the problem we're trying to solve.

But what if that preferred frame, while being hidden in kinematic experiments, was actually essential for explaining other fundamental phenomena that lie outside the scope of Special Relativity? For example, what if the physical properties of that frame were the key to a mechanical explanation for the origin of particle mass or the quantization of charge—mysteries that SR doesn't address?

In such a hypothetical case, would Occam's Razor still cleanly shave away the preferred frame? Or would that frame no longer be an extraneous entity, but rather the foundational postulate required to build a bridge between relativity and particle physics? It seems the razor is only decisive if the entity truly does no work across all domains of physics.

 

[Herman Trivilino]

But what if that preferred frame, while being hidden in kinematic experiments, was actually essential for explaining other fundamental phenomena that lie outside the scope of Special Relativity?

Every theory has limits of validity. So it's always possible that a theory can be replaced with another that has less restrictive limits, but all physical theories require experimental verification.

LET has neither larger limits of validity nor any experimental verification. Aether theories and the MM experiment are well over a century old. Science and engineering have advanced so far in the intervening years, improving not only the experimental evidence supporting Einstein's relativity, and perhaps more importantly its utility, that their significance is purely historical and heuristic.

 

[PeterDonis]

what if that preferred frame, while being hidden in kinematic experiments, was actually essential for explaining other fundamental phenomena that lie outside the scope of Special Relativity?

If someone ever comes up with a theoretical model that includes such a testable prediction, then we can test it and see.

But LET is not such a model. So what you say is not relevant to LET.

 

[robphy]

But what if that preferred frame, while being hidden in kinematic experiments, was actually essential for explaining other fundamental phenomena that lie outside the scope of Special Relativity?

(I'm new to the whole LET discussion.... and maybe this is off-topic...
while LET appears to be primarily an interpretive alternative to special relativity,
does it have any predictive power on its own, from purely LET-foundations?
Could an LET-theorist, from first LET-principles alone and experimental data
(without peeking at what SR is doing),
formulate/develop the same results from special relativity and, presumably later, general relativity?

 

[PeterDonis]

from first LET-principles alone and experimental data
(without peeking at what SR is doing)

There is no such thing. LET, as I've said, is SR, plus the claim that there's an unobservable absolute reference frame. It has to use all of the standard SR machinery for making predictions. There are no "first LET-principles" independent of the ones that ground standard SR.

Historically, there were attempts to construct "ether theories" from independent foundations--and all of them were falsified. That's why LET ended up being what I described above.

 

[Filip Larsen]

It can't be since it makes all of the same experimental predictions as standard SR.

I was under the impression that the historical proposed ether theory in principle always can be falsified by measuring the speed of light at enough different relative velocities to exclude possibility of a single absolute and rigid (i.e. homogeneous) ether. If that is correct, I assume then from your description that you are just saying that any "modern" ether theory which proposes to use an ether that is non-rigid (i.e. parts of the ether will move relative to other parts of it) in just the right way to avoid being disproved by any light speed measurement will, per Occam's Razor, have no physical meaning?

 

[PeterDonis]

the historical proposed ether theory

There wasn't just one. There were various proposals for ether theories, assigning different properties to the ether (for example, whether or not "ether drag" would exist).

a single absolute and rigid (i.e. homogeneous) ether

That was just one of the proposals.

any "modern" ether theory which proposes to use an ether that is non-rigid (i.e. parts of the ether will move relative to other parts of it)

I don't know that any "modern" ether theory makes any such claim; AFAIK historical "non-rigid" ether models were also falsified (they made predictions that contradict standard SR, which has now been extensively confirmed within its domain of validity).

That's why I described "modern" LET the way I did: it is standard SR, plus the claim that there is some one inertial reference frame that's "absolute", but we have no way of telling which one. That claim is indeed scraped off by Occam's Razor, but even leaving that aside it's an extremely weak claim, since it assigns no properties whatever to any "ether" defined by the unobservable "absolute" frame.

 

[ersmith]

It should perhaps be noted that while LET can reproduce all of the predictions of SR, it does not naturally extend to a theory of gravity (general relativity) in the same way that the geometric interpretation of SR does. This is also a significant reason to favor the geometric interpretation.

 

[ersmith]

Putting aside the history for a moment, the significance of universal simultaneity in Newtonian mechanics comes from there being no maximum speed: this makes the causal past expand to fill all of spacetime in the coordinate past. In a universe in which there is a maximum speed this is not the case. Instead, causality becomes local, and simultaneity is just a coordinate convention with no particular physical significance.

Edited to add: I didn't really express myself clearly; what I'm trying to say is that simultaneity is often confused with causality, and one of the things physicists learned in the 20th century is the difference between the two. Grafting a notion of absolute simultaneity onto SR (as LET does) buys you nothing in terms of causality.

 

[phinds]

ETA I didn't really express myself clearly;

ETA normally means "estimated time of arrival". What do you mean by it?

 

[ersmith]

ETA normally means "estimated time of arrival". What do you mean by it?

Edited to add... I'll change it in the comment

 

[Iskandarani]

Thank you all for the excellent and challenging responses. I think I have a much clearer picture of the consensus view now. If I can try to summarize the main points I've gathered from your posts:

• Modern Lorentz Ether Theory (LET) is defined as being kinematically identical to SR, but with an added, experimentally undetectable preferred frame.
• Because it is kinematically identical, it is unfalsifiable by experiment and, by Occam's Razor, less simple than SR.
• Crucially, this version of LET offers no new predictions, has no independent first principles from which to derive results without "peeking" at SR, and doesn't extend naturally to a theory of gravity the way SR's geometric interpretation does.
• Therefore, its core feature—a universal 'now'—is seen as physically insignificant compared to the concept of causality.

I hope this is correctly summarized, because this leads me to a final thought experiment to ensure I understand the boundaries. It seems all these powerful objections apply to a theory that is merely a philosophical reinterpretation of SR. But what if a different kind of preferred-frame theory were proposed? Let's call it a 'Constructive Theory' for the sake of discussion.

Imagine a theory with the following properties:

1. It has Independent First Principles: It doesn't start with the postulates of relativity. Instead, its foundations might be, for example, hydrodynamic or topological. It begins with a physical medium with concrete, defined properties.
2. It Derives, Not Assumes: From these physical principles, it derives the Lorentz transformations as an effective, low-energy description of how physical objects (rulers, clocks) and fields behave within this medium. It could therefore meet @robphy's challenge of building up results from its own foundation.
3. It Makes New, Falsifiable Predictions: The primary purpose of the preferred frame isn't just to exist. Its physical properties are used to make new, testable predictions in a domain outside of kinematics, for example, by deriving the particle mass spectrum from the topology of stable excitations (like vortices) in the medium. This would give the preferred frame real "work" to do, satisfying the Occam's Razor challenge from @PeterDonis.
4. It Proposes an Alternative to GR: It wouldn't need to "extend to GR." It would propose its own alternative and emergent theory of gravity based on the collective dynamics of the medium, as mentioned by @ersmith.

My question is: would such a 'Constructive Theory'—one that is not just a philosophical layer on top of SR, but a completely separate, predictive, and potentially unifying framework—still be subject to the same historical objections? Or would it represent a genuinely new path that must be evaluated on its own predictive merits?

 

[PeterDonis]

Imagine a theory with the following properties

That's out of bounds here. We can't discuss imagined theories. We can only discuss theories that have actually been published in the peer-reviewed literature, so that we have an explicit presentation of the theory that can be used to make predictions. You can't make predictions from an imagined theory. Nor can you say what its implications would be.

And with that, this thread is closed since the OP question has been sufficiently addressed.