- #1

- 16

- 0

You are using an out of date browser. It may not display this or other websites correctly.

You should upgrade or use an alternative browser.

You should upgrade or use an alternative browser.

- Thread starter BluMuun
- Start date

- #1

- 16

- 0

- #2

ghwellsjr

Science Advisor

Gold Member

- 5,122

- 147

Nope, no one can help. Light's one-way speed of exactly c is Einstein's second postulate on which Special Relativity is based. It's consistent with all experiments but it cannot come from an experiment.

- #3

Nugatory

Mentor

- 13,339

- 6,271

Can you turn this around? Give us a scenario in which the experimental results would not be consistent with the speed of light being exactly c in all inertial coordinate systems?

If you're saying that you can't come up with an experiment that shows that one-way speed of light is c in all inertial coordinate systems, you're absolutely right. The one-way speed of light is a postulate and cannot be proven by experiment (of course it could be disproven, bu that hasn't happened).

- #4

- 31,217

- 7,908

http://www.edu-observatory.org/physics-faq/Relativity/SR/experiments.html

- #5

- 16

- 0

Thanks for the replies, guys.

Nugatory asked about the opposite scenario, and luckily for me Einstein had already been there and done that, as follows:

During the creation of relativity (SR), Einstein noted that prior to SR it was possible for light's one-way speed to be c for one observer but c - v for another. It was this scenario that gave Einstein a major pain and caused him to create SR on the spot. http://www.bartleby.com/173/7.html

Einstein's cause for concern was his notion that the above one-way results conflicted with the principle of relativity. (Einstein's w = c - v is not a closing velocity - such a velocity would not even apparently conflict with the PR.)

As for the postulate part, it seems to me that both Nugatory and Wells are slightly confusing a math axiom or postulate with a physical or scientific postulate.

A scientific postulate is a guess, a supposition, a hunch, or a hypothesis about the nature of nature.

Indeed, as Nugatory noted, all scientific postulates must be experimentally testable.

So this leads to the question How can SR's one-way invariance be tested?

- #6

- 31,217

- 7,908

See the reference in my post above.

- #7

- 16

- 0

SR says that light's one-way, two-clock speed (with the caveats above) must be c in all inertial frames, but I am still unable to see how it can be c in any inertial frame.

Indeed, according to John A. Wheeler (in his book _Spacetime Physics_), clocks are simply set by definition to get the chosen value "c." There is really no way to falsify a definition or to even test one.

Color me confused.

- #8

ghwellsjr

Science Advisor

Gold Member

- 5,122

- 147

If you are "unable to see how it can be c in any inertial frame", then it should be easy for you to falsify the definition, shouldn't it?

SR says that light's one-way, two-clock speed (with the caveats above) must be c in all inertial frames, but I am still unable to see how it can be c in any inertial frame.

Indeed, according to John A. Wheeler (in his book _Spacetime Physics_), clocks are simply set by definition to get the chosen value "c." There is really no way to falsify a definition or to even test one.

Color me confused.

If light propagated at c only in a single absolute ether rest state, and if the Principle of Relativity were not true, then it would be very easy to falsify Einstein's second postulate, wouldn't it?

- #9

Vanadium 50

Staff Emeritus

Science Advisor

Education Advisor

- 26,813

- 10,479

Professionals don't worry about this because it's an underconstrained system. There are two things you need to know - the clock synchronization convention and the one-way speed of light. You can't determine two things with a single measurement.

A similar example: the sun is directly overhead. What is your longitude?

- #10

- 31,217

- 7,908

How would you suggest synchronizing the clocks in such an experiment?Oddly enough, given all the info at that site that you cited and I sighted, one would think that the direct experiment would be mentioned somewhere. By "direct" I mean using two mutually-at-rest clocks that are not rotating and have not been transported (because such clocks run slow).

- #11

Nugatory

Mentor

- 13,339

- 6,271

I asked a while back in this thread, and I'm asking again... What is a situation in which you find this assumption problematic? That is, can you show us an example of how it's hard to see that the one-way speed of light could be c in all inertial frames? (It's easy to find situations in which if you assume that the one-way speed of lightSR says that light's one-way, two-clock speed (with the caveats above) must be c in all inertial frames, but I am still unable to see how it can be c in any inertial frame.

That's true. We do that because if we set clocks any other way, we get various horrible complications, or internal contradictions, or results that don't match experiment. Thus, even though it is not possible to measure the one-way speed of light (dig deep into any experiment that appears to measure the constancy of the one-way speed of light and you'll find that the clocks used to measure the travel time were synchronized using some procedure that assumes what we're trying to prove) it is very convenient to assume that the one-way speed of light is constant.Indeed, according to John A. Wheeler (in his book _Spacetime Physics_), clocks are simply set by definition to get the chosen value "c."

If the definition leads to a contradiction it's falsified. If the definition leads to predictions that don't match experimental results, it's falsified. The one-way speed of light assumption has held up just fine against challenges of this sort and there are no alternatives that do so as well.There is really no way to falsify a definition or to even test one.

Last edited:

- #12

ghwellsjr

Science Advisor

Gold Member

- 5,122

- 147

Could you please supply the page number from which you took this quote?Indeed, according to John A. Wheeler (in his book _Spacetime Physics_), clocks are simply set by definition to get the chosen value "c."

- #13

- 3,872

- 90

A simple, direct test is completely useless, insofar as two-way light speed has been tested already. That's due to the synchronization convention which makes the one-way speed of light equal to the (average) two-way speed of light. As a matter of fact, that's already implied in the first section of Einstein's 1905 paper : http://www.fourmilab.ch/etexts/einstein/specrel/www/[..]

A scientific postulate is a guess, a supposition, a hunch, or a hypothesis about the nature of nature.

Indeed, as Nugatory noted, all scientific postulates must be experimentally testable.

So this leads to the question How can SR's one-way invariance be tested?

Thus the one-way speed is made equal to the two-way speed.

The self-consistency can be tested by indirect measurements, such as for example a measurement of the speed of light after clock transport. For example one synchronizes two clocks that one next moves apart very slowly or at equal speeds relative to the lab. A light signal that is next sent one way should be recorded with those clocks as taking the same time as a signal that is sent the other way. Similarly, one could rotate the setup (or wait for the lab to rotate), and again the measured speeds should be c in both directions.

ADDENDUM: those are tests of the constancy of the one-speed of light - that is, using the same reference system. For such a test one doesn't need to re-synchronize the clocks, as they should remain in sync with each other wrt the lab. However, in order to find that the one-way speed of light is invariant (between inertial frames that are accelerated to different velocities) one must in general first re-synchronize each pair of clocks after acceleration.

Last edited:

- #14

- 8,679

- 2,742

Could you please supply the page number from which you took this quote?

I don't have Wheeler's book, but I think that the original poster is talking about the Einstein synchronization convention: You set two distant, comoving clocks so that the one-way time for a light-signal is half the round-trip time. This makes light-speed isotropic by definition.

This doesn't make SR into a tautology, because there is more than one way to synchronize clocks. For example, slow clock transport: You synchronize two clocks, then slowly (at a speed much less than the speed of light) separate them. SR says that two clocks synchronized by this definition will also be synchronized by the Einstein synchronization convention.

- #15

- 8,679

- 2,742

Conceptually, the empirical meaning of the constancy of the one-way speed of light is given by:

- Pick an inertial reference frame F. Assume that we have a large, flat platform at rest in F to do our experiments on.
- Draw a straight line on the platform, and mark two spots on the line a distance L apart (as measured by meter sticks at rest in F).
- Take two identical clocks, bring them together at one of the marked spots and synchronize them.
- Move one of the clocks at constant speed to the other spot, and then bring it back to rest in frame F. Let [itex]v[/itex] be the speed that the clock was transported (as computed by [itex]v=L/\delta t[/itex], where [itex]\delta t[/itex] is the elapsed time on the transported clock.
- At time [itex]t_0[/itex] according to the first clock, send a light signal toward the second clock.
- Let [itex]t_1[/itex] be the time the signal arrives, according to the second clock.
- In the limit of slow clock transport ([itex]v \rightarrow 0[/itex]), the ratio [itex]L/(t_1 - t_0)[/itex] will approach a constant c.
- This constant is found to be the same in every inertial reference frame F.

I think it's clear how it could work this way in a single frame. Then the question is: How could work this way in EVERY frame?

Well, the Lorentz transformations relating inertial coordinates in different rest frames show how it could work in every frame.

- #16

- 8,679

- 2,742

Conceptually, the empirical meaning of the constancy of the one-way speed of light is given by:

- Pick an inertial reference frame F. Assume that we have a large, flat platform at rest in F to do our experiments on.
- Draw a straight line on the platform, and mark two spots on the line a distance L apart (as measured by meter sticks at rest in F).
- Take two identical clocks, bring them together at one of the marked spots and synchronize them.
- Move one of the clocks at constant speed to the other spot, and then bring it back to rest in frame F. Let [itex]v[/itex] be the speed that the clock was transported (as computed by [itex]v=L/\delta t[/itex], where [itex]\delta t[/itex] is the elapsed time on the transported clock.
- At time [itex]t_0[/itex] according to the first clock, send a light signal toward the second clock.
- Let [itex]t_1[/itex] be the time the signal arrives, according to the second clock.
- In the limit of slow clock transport ([itex]v \rightarrow 0[/itex]), the ratio [itex]L/(t_1 - t_0)[/itex] will approach a constant c.
- This constant is found to be the same in every inertial reference frame F.

I think it's clear how it could work this way in a single frame. Then the question is: How could work this way in EVERY frame?

Well, the Lorentz transformations relating inertial coordinates in different rest frames show how it could work in every frame.

Note that this operational meaning of lightspeed does NOT make it true by definition that light has one-way speed c.

- #17

- 31,217

- 7,908

If you synchronize the two clocks using Einstein's synchronization convention then it is guaranteed to be c. Do you understand why?SR says that light's one-way, two-clock speed (with the caveats above) must be c in all inertial frames, but I am still unable to see how it can be c in any inertial frame.

I agree. In the reference I posted above, that is what they are talking about when they say:Indeed, according to John A. Wheeler (in his book _Spacetime Physics_), clocks are simply set by definition to get the chosen value "c." There is really no way to falsify a definition or to even test one.

Experimental Basis of SR (emphasis added) said:Note that while these experiments clearly use a one-way light path and find isotropy, they are inherently unable to rule out a large class of theories in which the one-way speed of light is anisotropic. These theories share the property that the round-trip speed of light is isotropic in any inertial frame, but the one-way speed is isotropic only in an aether frame. In all of these theories the effects of slow clock transport exactly offset the effects of the anisotropic one-way speed of light (in any inertial frame), and all areexperimentally indistinguishable from SR. All of these theories predict null results for these experiments. See Test Theories above, especially Zhang (in which these theories are called “Edwards frames”).

- #18

- 16

- 0

>If you are "unable to see how it can be c in any inertial frame", then it should be easy >for you to falsify the definition, shouldn't it?

>If light propagated at c only in a single absolute ether rest state, and if the Principle of >Relativity were not true, then it would be very easy to falsify Einstein's second >postulate, wouldn't it?

You cannot falsify a definition because it is a given (given by man, not nature). All two-clock measurements in relativity are given by definition (the synchronization definition), including light's one-way speed and the transformation equations. As long as we are given Einstein's clocks, we are stuck with his "c invariance."

Also note that the principle of relativity does not call for c invariance, but only for the same laws. It does not say that frames are indistinguishable, but that their laws are indistinguishable. For example, if the one-way, two-clock light speed law is c ± v (as in Einstein's own example ref'd above), then all frames share the same law, but they are distinguishable (and can detect their absolute motion).

But not only is c invariance "given" by definition, it cannot happen even on paper as far as I can see as long as two mutually-at-rest clocks are used. (No transported or rotated clocks allowed because they run slow.)

For Wells: Wheeler's stuff came from Wheeler & Taylor's book

_Spacetime Physics_, 1963 edition, page 18.

Dalespam asked:

"How would you suggest synchronizing the clocks in such an experiment?"

The same way clock were set on paper for 100's of years prior to Einstein. Can you tell me why Einstein decided to get rid of such clocks?

Nugatory asked re no experiment showing c invariance:

"I asked a while back in this thread, and I'm asking again... What is a situation in which you find this assumption problematic? That is, can you show us an example of how it's hard to see that the one-way speed of light could be c in all inertial frames? (It's easy to find situations in which if you assume that the one-way speed of light is not c in all inertial frames you get horrible complications)."

To put the ball back in your court, please show on paper any frame getting c for the one-way, two-clock speed of light (not using slowed clocks as in clock transport or in the case of rotated clocks). As I see it, this must be possible or the c invariance of SR is not scientific.

And I am very interested in the "horrible complications" of not getting c? Can you cite one please?

Nug also wrote:

"Thus, even though it is not possible to measure the one-way speed of light (dig deep into any experiment that appears to measure the constancy of the one-way speed of light and you'll find that the clocks used to measure the travel time were synchronized using some procedure that assumes what we're trying to prove) it is very convenient to assume that the one-way speed of light is constant."

What does "the one-way light speed is constant" mean if it is not an experimental (or a possible experimental) result? Of what use is it to physics?

Why is it convenient to assume that which cannot happen?

DaleSpam quoted:

"In all of these theories the effects of slow clock transport exactly offset the effects of the anisotropic one-way speed of light (in any inertial frame), and all are experimentally indistinguishable from SR."

Sorry, but this is not true. As Einstein said, a theory that contains truly synchronous clocks predicts a variable one-way, two-clock light speed, unlike SR.

http://www.bartleby.com/173/7.html

For all of those who mentioned clock transport: I said up front that this is not allowed because moving clocks run slow, no matter how slowly they move. (I also disallowed rotating clocks for the same reason.)

In other words, I am looking for a legit experiment that shows a two-clock one-way light speed of c in any inertial frame, much less many.

DaleSpam wrote:

>If you synchronize the two clocks using Einstein's synchronization convention then it is >guaranteed to be c. Do you understand why?

But it is not via experiment; the clocks are merely manually forced to get "c," as Wheeler said.

Nugatory noted:

"If the definition leads to a contradiction it's falsified. If the definition leads to predictions that don't match experimental results, it's falsified. The one-way speed of light assumption has held up just fine against challenges of this sort and there are no alternatives that do so as well."

How can a definition lead to any contradictions if it has been assumed that any and all results of the definition are acceptable? For example, if two same-frame SR clocks are

compared with a passing clock, then the result is that "the passing clock ran slow." This may or may not be true when it comes to the intrinsic atomic rhythm of the clock, but it is accepted in SR as a valid and meaningful result. (However, it does not seem to be all that meaningful when you consider the fact that SR also has the other frame's clock running slower so that two clocks both run slower than each other. Is this really physics?)

I am not trying to be argumentative, but just trying to see how any frame's observers could use two mutually-at-rest clocks to get a one-way light speed of c experimentally. Of course, one of you said this:

stevendaryl noted:

"This doesn't make SR into a tautology, because there is more than one way to synchronize clocks. For example, slow clock transport: You synchronize two clocks, then slowly (at a speed much less than the speed of light) separate them. SR says that two clocks synchronized by this definition will also be synchronized by the Einstein synchronization convention."

Everyone agrees that using slowed clocks can result in approximate c invariance, but what I am wondering about is what happens using two mutually-at-rest clocks because SR says that this will also result in c (exactly c) experimentally. To me, these are two entirely different things (slow transport & true c invariance).

Harrylin wrote:

>A simple, direct test is completely useless, insofar as two-way light speed has been >tested already.

But SR did not predict two-way invariance - this was given prior to SR via a direct experiment. This left only the one-way case for SR to consider, and SR claims that light's one-way speed is c in all frames. Is this via experiment or merely given by definition? Can it even happen on paper?

If c invariance cannot happen, then it would seem to be of no real use to physics. This is the only reasonable conclusion that I see.

- #19

Nugatory

Mentor

- 13,339

- 6,271

And I am very interested in the "horrible complications" of not getting c? Can you cite one please?

One such problem is that we can

Rejecting Maxwell's equations doesn't work. There's just too much experimental evidence from the past few centuries of studying electricity and magnetism.

Different laws of electricity and magnetism for people moving at different speeds would imply that our observations on earth would change with the seasons because of the earth's motion around the sun. They don't, to the limits of accuracy of some very precise experiments.

Some history: Maxwell's laws were discovered in 1861. Reconciling them with the intuitive notion that if you're moving at speed ##u## relative to me and I measure the speed of a light signal moving in the same direction as ##c##, you will measure the speed as ##c-u## was one of the great challenges of the second half of the 19th century. That's why when Einstein published his classic paper in 1905 resolving this problem by proposing a consistent and experimentally confirmed theory that assumed a constant one-way speed of light he titled it "On the electrodynamics of moving bodies".

- #20

ghwellsjr

Science Advisor

Gold Member

- 5,122

- 147

Could you please supply the page number from which you took this quote?Indeed, according to John A. Wheeler (in his book _Spacetime Physics_), clocks are simply set by definition to get the chosen value "c."

I have the 1966 edition which you can see the first part of here.For Wells: Wheeler's stuff came from Wheeler & Taylor's book

_Spacetime Physics_, 1963 edition, page 18.

I cannot find anything even remotely resembling your quote. Could you please find about where it got edited out?

- #21

PAllen

Science Advisor

- 8,280

- 1,542

Dalespam asked:

"How would you suggest synchronizing the clocks in such an experiment?"

The same way clock were set on paper for 100's of years prior to Einstein. Can you tell me why Einstein decided to get rid of such clocks?

That's a total non-answer. Again: how do you think this was done? Tell us how you think clocks were synchronized between e.g. London and Glasgow in the 1800s?

- #22

ghwellsjr

Science Advisor

Gold Member

- 5,122

- 147

Suppose instead of Einstein's convention, we use one that says that the outbound propagation of light is twice as fast as the inbound. Don't you think that would easily be falsifiable?GHWells noted:

>If you are "unable to see how it can be c in any inertial frame", then it should be easy >for you to falsify the definition, shouldn't it?

>If light propagated at c only in a single absolute ether rest state, and if the Principle of >Relativity were not true, then it would be very easy to falsify Einstein's second >postulate, wouldn't it?

You cannot falsify a definition because it is a given (given by man, not nature).

I don't understand why you say it cannot happen even on paper. I draw spacetime diagrams all the time that illustrate how light propagates at c in any Inertial Reference Frame (IRF) and how the coordinates of all the events can be transformed to any other IRF moving with respect to the original one and yet light still propagates at c and all observers continue to see and measure exactly the same things in all IRF's.All two-clock measurements in relativity are given by definition (the synchronization definition), including light's one-way speed and the transformation equations. As long as we are given Einstein's clocks, we are stuck with his "c invariance."

Also note that the principle of relativity does not call for c invariance, but only for the same laws. It does not say that frames are indistinguishable, but that their laws are indistinguishable. For example, if the one-way, two-clock light speed law is c ± v (as in Einstein's own example ref'd above), then all frames share the same law, but they are distinguishable (and can detect their absolute motion).

But not only is c invariance "given" by definition, it cannot happen even on paper as far as I can see as long as two mutually-at-rest clocks are used. (No transported or rotated clocks allowed because they run slow.)

Is it that you have never seen how this happens on paper because you don't understand Special Relativity or is it that you thoroughly understand Special Relativity but still find fault with it?

- #23

- 3,872

- 90

[..]

Harrylin wrote:

>A simple, direct test is completely useless, insofar as two-way light speed has been tested already.

But SR did not predict two-way invariance - this was given prior to SR via a direct experiment. This left only the one-way case for SR to consider, and SR claims that light's one-way speed is c in all frames. Is this via experiment or merely given by definition? Can it even happen on paper?

If c invariance cannot happen, then it would seem to be of no real use to physics. This is the only reasonable conclusion that I see.

I think to see a slight but perhaps important misunderstanding here. The purpose of SR was to match two apparently contradictory facts that were concluded from observations:

1. The principle of relativity is also valid for light propagation (=>invariance of the two-way speed of light)

2. Maxwell's model of light propagation is valid (=>the speed of light is independent of that of the source)

Those two constraints together led to strong predictions for the two-way speed of light and many other things such as electrons and clocks. If the two-way speed of light is not found to be the same c in different inertial frames then also the one-way speed of light can't be c both ways. As by definition the one one-way speed is made equal to the two-way speed, c invariance means experimentally the invariance of the two-way speed. And that is obviously very useful to physics.

In short, one-way speed can distract from the experimentally verifiable predictions of SR.

Note also, I'm surprised by the following remark:

"what I am wondering about is what happens using two mutually-at-rest clocks because SR says that this will also result in c (exactly c) experimentally. To me, these are two entirely different things (slow transport & true c invariance)."

I somewhat answered that question of yours in my earlier post, before you asked it. However I provided different answers, depending on the exact test conditions. I even indicated how SR predicts that with a certain procedure one can find a value different from c. Either my answers were not clear, or your question was not clear, but I don't know which. Also your remark that "No transported or rotated clocks allowed because they run slow" seems to contradict your earlier question and my earlier answers. If you like, we can elaborate on that.

Last edited:

- #24

- 31,217

- 7,908

Hi BluMuun, please use the quote and multi-quote features, particularly in a long post like your previous one where you are responding to multiple people.

Mansouri, Sexl. A test theory of special relativity: I. Simultaneity and clock synchronization. General Relativity and Gravitation. July 1977, Volume 8, Issue 7, pp 497-513

If you don't have access to that then there is this arxiv reference, but it is a little math-heavy

http://arxiv.org/abs/physics/0510260

It is even mentioned on Wikipedia: http://en.wikipedia.org/wiki/Test_theories_of_special_relativity

You seem to think that coordinates are somehow out there in nature, waiting to be experimentally determined. They are not. Coordinates are human-made inventions. You have to "manually force" them to be something since nature doesn't provide them. So you have to choose some simultaneity convention. It is just a convention, so there is no "right" or "wrong" way to do it as long as you are consistent.

Einstein's convention is popular because if you use it then Maxwell's equations take their "textbook" form, and otherwise you have to re-write them. The one way speed of light is governed by Maxwells equations, so it is determined by the synchronization convention chosen.

What way was that? I am unaware of anyone addressing clock synchronization issues prior to Einstein.Dalespam asked:

"How would you suggest synchronizing the clocks in such an experiment?"

The same way clock were set on paper for 100's of years prior to Einstein.

It is true and is well known in the literature. See for example:DaleSpam quoted:

"In all of these theories the effects of slow clock transport exactly offset the effects of the anisotropic one-way speed of light (in any inertial frame), and all are experimentally indistinguishable from SR."

Sorry, but this is not true.

Mansouri, Sexl. A test theory of special relativity: I. Simultaneity and clock synchronization. General Relativity and Gravitation. July 1977, Volume 8, Issue 7, pp 497-513

If you don't have access to that then there is this arxiv reference, but it is a little math-heavy

http://arxiv.org/abs/physics/0510260

It is even mentioned on Wikipedia: http://en.wikipedia.org/wiki/Test_theories_of_special_relativity

Good, so you do, in fact, see how the one way speed of light can be c in a frame, you simply "manually force" it to be c.>If you synchronize the two clocks using Einstein's synchronization convention then it is >guaranteed to be c. Do you understand why?

But it is not via experiment; the clocks are merely manually forced to get "c," as Wheeler said.

You seem to think that coordinates are somehow out there in nature, waiting to be experimentally determined. They are not. Coordinates are human-made inventions. You have to "manually force" them to be something since nature doesn't provide them. So you have to choose some simultaneity convention. It is just a convention, so there is no "right" or "wrong" way to do it as long as you are consistent.

Einstein's convention is popular because if you use it then Maxwell's equations take their "textbook" form, and otherwise you have to re-write them. The one way speed of light is governed by Maxwells equations, so it is determined by the synchronization convention chosen.

Last edited:

- #25

- 16

- 0

One such problem is that we cancalculatethe speed of light from Maxwell's laws of electricity and magnetism, and Maxwell's equations make no allowance whatsoever for the velocity of the observer.

Different laws of electricity and magnetism for people moving at different speeds would imply that our observations on earth would change with the seasons because of the earth's motion around the sun.

Note that Einstein made no mention of Maxwell's equations during his creation of SR. He talked only about the "apparent" incompatibility of the simple light law with the principle of relativity. This tells us that Maxwell's equations had nothing to do with SR.

Why is this?

It's because the equations do not refer to a coordinate light speed, but merely to light's propagational speed in space. Maxwell used no clocks to measure light's speed. As Einstein said, if the clocks of classical physics were used to measure light's one-way speed, then it would vary with observer velocity. And, again, he made no mention of this being a problem re Maxwell's equations. http://www.bartleby.com/173/7.html

Also note that the laws of mag&elect must necessarily NOT reflect observer velocity because they depend ONLY on *relative* motions of magnets and wires, so they must be the same for all observers.

Can you list a real complication?

Thanks.

Share: