Some notes on LET for reference

[robphy]

[Moderator's note: Thread spun off due to topic change, for reference only.]

(I'll ask my questions here [hoping for short answers].. and quote the earlier discussion below.)

• Does Lorentz Ether Theory have its own predictive power?
  Or is it essentially a direct and straightforward (as opposed to convoluted) alternative-interpretation after Special-Relativity has predicted and established the result? ( McDonald's vs McDowell's ? )
• Does this extend to General Relativity?

Do you refer to the Lorentz Ether Theory?

No, and I don't believe that theory is correct. It doesn't fit observation.

It's final version does fit observation:

In the absence of any way to experimentally distinguish between LET and SR, SR is widely preferred over LET, due to the superfluous assumption of an undetectable aether in LET, and the validity of the relativity principle in LET seeming ad hoc or coincidental.

Source:
https://en.wikipedia.org/wiki/Lorentz_ether_theory

See the related PF policy:
https://www.physicsforums.com/insights/pfs-policy-on-lorentz-ether-theory-and-block-universe/

 

[PeterDonis]

Does Lorentz Ether Theory have its own predictive power?
Or is it essentially a direct and straightforward (as opposed to convoluted) alternative-interpretation after Special-Relativity has predicted and established the result?

LET makes all of the same predictions as standard SR; it's just a different interpretation (and one we don't discuss here since there's no way to test any claims about LET vs. standard SR by experiment).

Does this extend to General Relativity?

GR has its own set of interpretations: spacetime curvature is one, "spin-2 field on a flat spacetime background" is another. (The latter is technically limited to spacetimes with topologies that can have a flat background put on them.) The second is kinda sorta LET-ish, I suppose (the flat background is in principle unobservable), but it has no historical lineage in common with LET that I'm aware of.

 

[Sagittarius A-Star]

• Does Lorentz Ether Theory have its own predictive power?
  Or is it essentially a direct and straightforward (as opposed to convoluted) alternative-interpretation after Special-Relativity has predicted and established the result? ( McDonald's vs McDowell's ? )

There were different versions of LET.

• Early versions (between 1892 and 1895) explained the failure of the negative ether drift experiments (i.e. MM) to first order in v/c. So they did have limited predictive power.
• The final version (1904, with corrections in 1906), which is today usually referred to as "LET", is a convoluted alternative-interpretation to SR. Unlike SR, it does not follow Ockham's razor and does not provide additional predictive power.

• Does this extend to General Relativity?

Wikipedia refers to a related 2012 paper by Schmelzer, which I don't know.
There is no universally accepted "Lorentz Ether Theory for GR."

Source:
https://en.wikipedia.org/wiki/Lorentz_ether_theory

 

[PeterDonis]

Wikipedia refers to a related 2012 paper by Schmelzer

Whose claims were repeatedly and thoroughly discussed here at PF years ago. They didn't get any real traction here, and AFAIK they have not in the general physics community either.

 

[PeterDonis]

Schmelzer

Note, btw, that his claimed "ether theory" including gravity does not make all of the same predictions as GR, so it is not just an interpretation of GR, it's a different theory.

 

[FactChecker]

The final version (1904, with corrections in 1906), which is today usually referred to as "LET", is a convoluted alternative-interpretation to SR. Unlike SR, it does not follow Ockham's razor and does not provide additional predictive power.

To me, this sounds like SR just accepts the results of the speed of light in very precise experiments and sees what those results lead to, whereas LET tries to explain why those same experimental results happen. Therefore, LET is more complicated and leads to the same results.
Mathematically, with very simple and desirable assumptions, there are only two possible versions of relativity, Euclidean and Einsteinian (see @Ibix 's link in post#2). I don't see why one mathematical version needs some justification, beyond the experimental results. It is what it is. Would Euclidean relativity have needed justification?

 

[Sagittarius A-Star]

Euclidean and Einsteinian

I think it is "Galilean and Einsteinian".

 

[PeterDonis]

SR just accepts the results of the speed of light in very precise experiments and sees what those results lead to, whereas LET tries to explain why those same experimental results happen.

I'm not sure I would put it that way. SR explains the speed of light experimental results by saying they are a consequence of the geometry of spacetime. That's not the same as not explaining them at all.

 

[robphy]

The final version (1904, with corrections in 1906), which is today usually referred to as "LET", is a convoluted alternative-interpretation to SR. Unlike SR, it does not follow Ockham's razor and does not provide additional predictive power.

...SR just accepts the results of the speed of light in very precise experiments and sees what those results lead to, whereas LET tries to explain why those same experimental results happen.

I'm not sure I would put it that way. SR explains the speed of light experimental results by saying they are a consequence of the geometry of spacetime. That's not the same as not explaining them at all.

So, this goes back to the heart of my initial query.
Does LET make predictions on its own, using its own principles (whatever they may be) [without looking over the fence as to what SR is doing]?
What does LET say about causality and light-cones? Can LET get [on its own] to the spacetime geometry of GR and all of its solutions to the field equations and various theorems (positive energy, initial value, causal-structure, singularity, etc.) ?

 

[Dale]

Does LET make predictions on its own, using its own principles (whatever they may be) [without looking over the fence as to what SR is doing]?

Yes. It just postulates the Lorentz transform directly. Then it makes all the usual predictions.

Can LET get [on its own] to the spacetime geometry of GR

No. LET is equivalent to standard SR and fails to replicate GR in the same way that standard SR does.

 

[Sagittarius A-Star]

Does LET make predictions on its own, using its own principles (whatever they may be) [without looking over the fence as to what SR is doing]?

Yes, the principles of LET are different than those of SR. But the resulting predictions coincide.
Principles:

• SR: There is no preferred inertial frame and no absolute time. Time is, what a clock measures.
• LET: There is an undetectable preferred frame and and undetectable absolute time.
  • In a rest-frame moving relative to the ether:
    • Clocks are physically time-dilated (without further explanation)
    • Bodies are physically length-contraced (without further explanation)
    • Clock synchronization: Light speed is physically anisotropic, but measured as isotropic

EDIT:
Insights were taken over from Einsteins SR to LET, for example time-dilation.

Source:
Rindler: Relativity: Special, General, and Cosmological (2nd edition), chapter 1.8 "Lorentz's ether theory"
https://www.amazon.com/Relativity-S...fgang-Rindler/dp/0198567324?tag=pfamazon01-20

 

[PeterDonis]

Does LET make predictions on its own, using its own principles

What would that even mean? You don't make predictions using "principles". You make them using a well-defined mathematical model. LET's well-defined mathematical model is the same as that of standard SR.

 

[robphy]

Does LET make predictions on its own, using its own principles

What would that even mean? You don't make predictions using "principles". You make them using a well-defined mathematical model. LET's well-defined mathematical model is the same as that of standard SR.

From my understanding,

• LET is based on the Lorentz Transformations--- somehow, maybe just as a set of formulas or maybe an element of a group.
• But, special relativity "SR" (in the spirit of Minkowski spacetime) is more than the Lorentz Transformations...
  SR has the spacetime geometrical structure, has the light-cone structure, has the causal structure, etc.... It features and emphasizes structures that can and have been weakened to allow generalizations to say non-R^4 manifolds, pseudo-riemannian curvature, spacetimes with various types of causality violations.
  
  In other words, would an LET-theorist or experimentalist actually be led to a route or something parallel to the route that SR followed, especially to get to GR. (One might argue that LET-folk could follow the route because it was mathematical consequence of the Lorentz Transformation, but would they?
  Would LET-folk focused on "length-contraction" think about
  the spacetime-geometric separation between two parallel inertial worldlines and the rest of what could be uncovered using spacetime geometry?)

The question in the back of my mind is "Why does LET matter today?"
It seems like a historically influential heuristic that may have influenced how Special Relativity developed.
But in light of General Relativity and its successes, why is LET still "a thing"?
[Maybe it's not a perfect analogy.... although we acknowledge the Bohr model in the historical development of quantum physics, we don't patch it up and prop it up alongside the equations of modern quantum mechanics (Schrodinger et al.).]

(update: I have similar issues with Loedel diagrams, Brehme diagrams, Epstein diagrams.)

 

[Matterwave]

LET is equivalent to standard SR and fails to replicate GR in the same way that standard SR does.

What does it mean to "replicate" a theory? I would say SR is a special (non gravitating) case of GR. Not the other way around.

(I feel like we are really hijacking this thread though )

 

[robphy]

(I feel like we are really hijacking this thread though )

I agree... and I was hesitant to follow up on that early comment mentioning LET...
but this LET-thing has always been puzzling to me.

I would suggest that the LET side-discussion be split-off into another thread. [Moderator's note: And now it has been.]
To me, this side-discussion has been helpful... and it might be for others as well.