Talk:Special relativity/Archive 6

Archive 1Archive 4Archive 5Archive 6Archive 7Archive 8Archive 10

Relativistically Rotating objects.

Suppose you take a large spoked wheel and spin it up so that the rim is moving at a high relativistic speed (say .9c). What would you see if you looked down at it from above? The rim would appear length contracted, but the spokes would still have the same lengths and angles to each other. What would happen here? 12.37.33.3 23:04, 17 March 2006 (UTC)

Because the parts of the wheel aren't moving in interial frames (free-fall), but are instead undergoing acceleration, you have to use general relativity to figure out what happens. --Christopher Thomas 18:04, 22 March 2006 (UTC)
No GRT is needed for this analysis: SRT has no problems with describing the behaviour of moving objects in a single inertial frame! What happens in practice is that the rim expands due to inertial forces. But if we forget about those forces, the rim would very slightly Lorentz contract and push against the spokes. As the spokes push back with a negligible force (compared to the rim), this relativistic effect is really very easy to calculate.
But note that such questions are more properly forwarded in a newsgroup, such as sci.physics.relativity. Harald88 20:20, 22 March 2006 (UTC)
Harald, the rim is being accelerated radially so that it can follow a curved path. This causes its time dilation to be other than what SR predicts, as any given point in the rim is not in an inertial frame. If memory serves, the Einstein lectures explicitly cover this case as an example of where it's hard to apply SR. --Christopher Thomas 22:13, 22 March 2006 (UTC)
Well, as I remember it, Einstein uses the SR relations to show that accleration changes geometry (or "curves space"), the ratio of radius to circumfernce (pi) is changed. So you can use the SR relations on these accelerating objects in some sense. I will look for a reference where Lorentz deals with this using GR. E4mmacro 22:26, 22 March 2006 (UTC)
A good example of time dilation with a curved path can be found in Einstein's 1905 SRT paper; that paper is online and linked from this article. BTW, the first verified prediction of SRT was on accelerating electrons...
Using GRT/SRT, Lorentz provided a more complex calculation (because it doesn't have a simplification as with spokes) about a rotating disc, in Nature vol.106, p.793-795 (1921). Harald88 11:03, 23 March 2006 (UTC)

Ives, H. E. "Theory fo the double Fizeau toothed wheel" Journal of the optical Society of America, v29, p472-478 thinks the disk must bend (curve into a cup shape) when it rotates. The cruvature of the disk depends on the speed. The ratio of radial distance along the curved shape to the circumference depends on the rotation speed. It appears he has not considered (on top of that) the deformation due to the stresses which supply the centripetal forces. (real disks have a limited rotation speed before the material yiled stress is reached). Ives's analysis is all from an interial frame POV using the length contractions. No GR involved. E4mmacro 23:46, 22 March 2006 (UTC)

Two Small Questions

One: Could someone show me where mass is mentioned in any original papers on special relativity?

Look at section 10 of "On the electrodynamics of moving bodies".

Two: Could someone tell me why, when two bodys are experiencing motion relative to one another, time compresses for one clock, but not for another? That is, why, after the bodies become at rest with respect to one another, is one clock objectively slower than the other? SJCstudent 01:23, 6 April 2006 (UTC)

Actually, if both bodies decelerate identically until there is no relative motion between them, their clocks will be synchronized (assuming they were synchronized when they were instantaneously co-located). A moving clock appears to run slow. Accelerated clocks actually do run slow. The twin 'paradox' occurs because one twin experiences acceleration (the one in the spaceship) while the other does not. In GR terminology, geodesics of spacetime are paths of maximum elapsed proper time. Alfred Centauri 02:12, 6 April 2006 (UTC)

But this is exactly the source of my confusion. In Einstein's 1905 paper, he never mentions acceleration to account for this difference. Yet people did not reject that theory outright. What, within Einstein's SR, backs this up? SJCstudent 03:40, 6 April 2006 (UTC)

See Twin paradox, it's "all" there (you'll notice that you're right that acceleration doesn't account for the difference). Harald88 22:16, 6 April 2006 (UTC)
You sound like a smart person so do your own research. There's plenty of scientific, pseudo-scientific, and non-scientific (historical) articles written about SR on the web. Have fun. Alfred Centauri 03:53, 6 April 2006 (UTC)

Uh, ok... Thanks Harald88. This helps, but I'm still not sure where anything about inertial reference frames is mentioned in SR proposed by Einstein. Look, I'm not interested in joining the list of whackos attempting to disprove relativity. I am simply interested in moving past a particular impasse in the course of my studies. I understand, Alfred, that I ought to do research, and that is exactly what I am doing by asking these questions to those that are so knowledgable that they wrote this article. Sorry for all the trouble. SJCstudent 03:16, 10 April 2006 (UTC)

2nd postulate update

The current version of the special relativity "Postulates" section contains the following statement: "An observer attempting to measure the speed of light's propagation will get the same answer no matter how the observer or the system's components are moving."

Apparently, that particular sentence was given by User:Christopher Thomas, who is currently on intermittent sabbatical. (As far as I could see, here is when he added said sentence: Thomas link - the page as of 16:44, 23 November 2005)

Here is the problem with the given sentence: "Measure" means "To ascertain the quantity of, using standard measurement instruments." Since the round-trip light speed case was closed prior to special relativity, this left only the one-way case for postulation; however, it is currently impossible to measure the one-way speed of light. Specifically, it's not possible (currently) to measure light's speed between two clocks.

Here is a suggested replacement for the second postulate section:

2. Second postulate - In empty space, light's round-trip, one-clock speed is c per experiment, but no one has yet measured light's one-way, two-clock speed due to the lack of absolute clock synchronization; therefore, the one-way speed of light was simply defined to be c to follow Einstein's assumption that inertial frames should not be distinguishable.

Since even Einstein agreed that light's one-way speed would vary given (absolute) synchronization, his other (hidden) assumption was that (absolute) synchronization is impossible, although he did not prove this. Indeed, since one cannot prove such a negative, the one-way case remains open, and can only be closed if and when (truly) synchronous clocks are used to actually measure light's one-way speed. Cadwgan Gedrych 18:31, 10 April 2006 (UTC)

In the "Status" section it is mentioned that the speed of light is understood two be the two-way speed. I think there should be no change in the "Postlates" section, because only two-way speed can be measured, there is no way to conduct "absolute synchronization". Icek 18:16, 11 April 2006 (UTC)

As I mentioned, at the time of relativity's creation, there could not have been any postulation regarding light's round-trip speed (because that case was closed via experiment); therefore, the second postulate could not apply to the round-trip speed of light, but could only apply to the one-way speed.

As I also mentioned, the one-way case remains open due to the possibility of absolute clock synchronization. (It remains possible because no one can prove that it is impossible.)

To reiterate, Einstein did not postulate regarding the round-trip case, but simply accepted the null result as an experimental fact. This left only the one-way case, and his postulate in that case was "Since my [Einstein's] rule is null results always, I postulate that clocks should be set to obtain a one-way null result, even if it means that they are not absolutely synchronous." Cadwgan Gedrych 18:48, 12 April 2006 (UTC)

Being experimentally tested and not falsified does not contradict being used as a theory's postulate - the experimental evidence was the basis of the theory. The first postulate is also well-supported by experiments. In Einstein's original publication, simultaneity is actually defined by synchronization by light signals (emitted in the same reference frame, as I understand it). So I agree partly, the definition of simultaneity should be in the article. But I don't think of it as a hidden assumption but rather as a mere definition and as such cannot be tested. If someone finds a way for instantaneous communication and thus absolute synchronization, it would, if one accepts Einstein's definition, allow for communication backward in time in certain reference frames. Icek 23:49, 12 April 2006 (UTC)

Thank you, Icek, for your comments. Just for the sake of the argument, let's say that you are correct about the round-trip case, i.e., that it was a part of the 2nd postulate at the creation of special relativity; this still leaves the problem of the one-way case.

By "the one-way case," I mean (as did Einstein) the specific case where light's one-way speed is measured by two clocks which are neither rotated nor transported.

In other words, the part of the second postulate that addresses light's one-way speed says that whenever two mutually-at-rest clocks are used to measure this speed, it will be c (in all frames and in all directions).

Icek, would you be so kind as to tell us how this one-way experiment could be done. You can even use the ideal clocks and rulers of theoretical physics. Please show each step, starting with two unstarted clocks. 66.147.55.213 18:39, 13 April 2006 (UTC)

The relevant sentence from Einstein's paper:

We have not defined a common "time" for A and B, for the latter cannot be defined at all unless we establish by definition that the "time" required by light to travel from A to B equals the "time" it requires to travel from B to A.

Einstein did not explicitely state that the light should be emitted in the rest frame, but I think this is implied since many physicists in 1905 still thought that velocities simply add up.

Taking into account the definition of simultaneity, a one-way experiment can actually be done:

  • Leave one clock (A) at the coordinate origin (which shall be at the starting point of the clocks). Take the other clock (B) to distance x, as measured by ruler.
  • Now both clocks are at rest. Start clock A and at the same time send a light signal toward clock B. When it arives at clock B, set clock B to time x/c and start it. By definition, the clocks are now synchronized. It should be noted that the light signals where emitted in the rest frame.
  • Now let a spaceship fly by clock A with velocity v. When it is at the position of clock A, it shall send a light signal toward clock B. Record the sending and arrival times and subtract one from the other.
  • Divide x by the time difference to get c(v), the one-way light time.

Icek 00:56, 14 April 2006 (UTC)

Now let's click our heels, and go back to Kansas! ;-) Nice try, Icek! (Seriously!) But now that we are back in Kansas, let's focus on your word "experiment"; by definition, an experiment is an attempt to discover the nature of Nature, so there can be no rigging of the result by man; however, in the case you cited above, man rigged the result ("c invariance/isotropy") by forcing the clocks to obtain "c" (in the form of using the rigged "time" "x/c").
Yes, I know that you used a different source for the second light ray, but since experiment says that light is source-independent, this matters not.
It also matters not what the 2nd postulate says about the round-trip case because that case was closed experimentally prior to special relativity. (That is, special relativity has no unique position re this case because every theory must have a null result for it.) It is only the one-way case that really matters. It is only regarding that case that special relativity (SR) can take a unique position.
And that unique SR position is exactly (and only) what the Wiki SR story should tell.
What, specifically, does SR say about the only case that matters?
What, specifically, does SR say about the case where a light ray's speed is experimentally measured between two same-frame clocks sans rigging?
For example, does the Wiki SR article say that the one-way case is still open?
For another example, does the article state that no one has ever actually used two same-frame clocks to measure light's one-way speed?
For yet another example, does the Wiki article say that Galileo and Newton and Lorentz have not been proved wrong in the only case that matters?
And here is yet another example: Does the article state that one-way light speed invariance has not been shown, and cannot be shown (experimentally)?
Cadwgan Gedrych 01:53, 15 April 2006 (UTC)

I agree that the article should be changed.

There is indeed the hidden assumption that light propagation is isotropic and c is invariant in at least one reference frame when the light is emitted in this same reference frame. Else the definition of time could lead to contradictory results. But if we accept that, time is no longer "rigged", it's just a definition. Maybe we should include the hidden assumption in the article. Icek 15:13, 15 April 2006 (UTC)

You've confused me with your word "invariant"; normally, "invariant" means "the same for all frames," and yet you mentioned only one frame.
You also confused me by reverting to your prior position of bringing up whichever light source emits the light, as if that makes any difference.
You have further confused me by your statement that "the definition of time could lead to contradictory results"; please explain this.
If Einstein's time is just a definition, then what has it got to do with the physics of space and time? Why not explicitly state that it is merely a definition given by man in lieu of absolute synchronization, which we (Einstein) cannot obtain.
Can Einstein claim that the results a mere definition are laws of physics? (I am talking about the standard SR results such as c invariance, the transformation equations, and the composition of velocities theorem.)
Why does the Wiki 2nd postulate section not say that it is currently impossible to experimentally measure light's one-way speed?
Why does the Wiki 2nd postulate section not say that the only way to correctly measure light's one-way speed is to use absolutely synchronous clocks (but Einstein does not have such clocks)?
Cadwgan Gedrych 19:20, 17 April 2006 (UTC)

I should clarify a few things:

  • When I answered at 18:16, 11 April 2006, I had not taken into acount a proper definition of time, so let's ignore that answer, since I have thought more about the problem afterwards.
  • "Invariant" is probably not the right word in my last answer, a better word would be constant. The statement means that there exists (at least) one reference frame (which I will subsequently call the rest frame) for which the two-way light trip time only depends on the distance, provided that the light is emitted in the rest frame.
  • It should be clear now that if the two-way light trip time does e. g. also depend on the direction of the light beam there could be a different times at clock B depending on whether you synchronize it directly with a light signal from A to B or you synchronize it via a "relais" clock C which does not lie one the line passing through A and B.
  • Every physical concept is "just a definition". Definitions aid us in describing the world. The definition of a quantity should include a method for measuring that quantity. How would one define "absolute time" or "absolutely synchronous clocks"?
  • Definitions may rely on assumptions as I showed above for the case of relativistic time.
  • Definitions cannot be disproved, but underlying assumptions can be, and this would make the definition unuseable.
  • In relativity, only synchronization is defined, but not local time (we are speaking of "clocks" without specifying how they measure time). In physics, (local) time is currently defined by a resonant frequency of the caesium atom.
  • The definition of time in relativity allows us to measure one-way light trip times for light emitted in other reference frames than the rest frame.
  • In contrast to the definition of time, the postulates of special relativity can be tested.

Icek 12:10, 18 April 2006 (UTC)

Let's touch on three of your just-given points.
  • The definition of time in relativity allows us to measure one-way light trip times for light emitted in other reference frames than the rest frame.
Please remember that "measure" does not include rigging the result, and as we all know, Einstein's definition of "time" certainly does that. (It does it by presetting (rigging) the clocks to read the prechosen time "x/c" - as I have already mentioned.)
  • In contrast to the definition of time, the postulates of special relativity can be tested.
So tell us how we can test Einstein's second postulate. (That is, how can we experimentally test whether or not light's one-way speed between two same-frame clocks is invariant? Please show all steps, starting with two unstarted clocks.)
  • Every physical concept is "just a definition". Definitions aid us in describing the world. The definition of a quantity should include a method for measuring that quantity. How would one define "absolute time" or "absolutely synchronous clocks"?
You seem to think that there is no definition of absolute time or absolute clock synchronization. Einstein himself gave two very good definitions of absolute synchronization, to wit:

[Quoting Einstein:] "This is what is meant when we say that the time of classical physics is absolute: The simultaneity of two definite events with reference to one inertial system involves the simultaneity of these events in reference to all inertial systems." [Appendix V of Relativity] [sentence order reversed and colon used]

[Quoting Einstein:] "w is the required velocity of light with respect to the carriage, and we have

w = c - v.

The velocity of propagation of a ray of light relative to the carriage thus comes out smaller than c.

But this result comes into conflict with the principle of relativity...." http://www.bartleby.com/173/7.html

Sure, Einstein (baselessly) rejected the result as being wrong, but the point is, he did derive it, so it could happen, at least on paper, and by this simple example Einstein told us that absolutely synchronous clocks would find that light's one-way speed varies with frame velocity, thereby giving us a definition of absolutely synchronous clocks.

Cadwgan Gedrych 18:45, 18 April 2006 (UTC)

On absolute synchronization: I don't see a definition, i. e. how would you theoretically "absolutely synchronize" two clocks which are not at the same location?

On the other two points you picked: I've answered these questions above. If there is a logical flaw in my answers, I don't see it, so please tell me exactly why you are not happy with my answers. Icek 20:13, 18 April 2006 (UTC)

Re absolute synchronization: You're confusing a definition with a procedure.
I see nothing wrong with either of Einstein's definitions. Do you?
The logical flaws were given above, but it seems that I must now repeat them.

You wrote:

  • The definition of time in relativity allows us to measure one-way light trip times for light emitted in other reference frames than the rest frame.
Please cite the date and persons involved when light's one-way speed was measured between two same-frame clocks. I maintain that this experiment not only has not been done, but cannot be done today.

You wrote:

  • In contrast to the definition of time, the postulates of special relativity can be tested.
Please back this claim by showing step-by-step how the 2nd postulate can be tested. As we know, the 2nd postulate pertains to the one-way speed of light (saying that light's one-way speed per two same-frame clocks is invariant and isotropic), so this is essentially the same problem as above where I asked you to supply the date and experimenters.
Tell us how the 2nd postulate applies to anything in physics.

Cadwgan Gedrych 20:44, 19 April 2006 (UTC)

I partly disagree: as Einstein pointed out, both postulates were based on experience. Moreover, the one-way speed of light can indeed not be determined except by convention, while physics theories are about what can be determined experimentally. Thus the second postulate is primarily concerned with the (experimental) round speed of light. It sets the constant c in the LT. Harald88 20:56, 19 April 2006 (UTC)
The elapsed time required for light to propagate along a closed path can be measured by a single clock where synchronization is not an issue. It is my understanding that experiments show that the length of the closed light path divided by the elapsed time is the constant speed c. Now, consider an experiment whereby light is sent along a path towards a reflector oriented to reflect the light back along the incident path. Is it not true that the electromagnetic traveling waves that are the incident and reflected light superpose to produce a familiar standing wave pattern of stationary nodes and anti-nodes? But, if the incident light propagates at a different speed than the reflected light, the nodes and anti-nodes will not be stationary but will instead move in the direction of the 'faster' light path, right? The speed with which the nodes move, in conjuction with the result that the closed path average light speed is c, can be used to calculate the one-way speed of light for the incident and reflected paths. Of course, if the nodes are stationary for all orientations and states of uniform motion of the apparatus, then one-way light speed must be c which would imply that the notion of absolute synchronization of spatially separated clocks is without meaning. Alfred Centauri 00:47, 20 April 2006 (UTC)
Alfred, I also thought so once, but after thorough calculation it turned out to be not so. I don't remember clearly what the error in your "nodes" picture, but probably it is that the long and short waves together reproduce the same result. One can't invalidate the Lorentz transformations with such simple tricks. ;-) Harald88 11:15, 25 April 2006 (UTC)
What turns out not be so - that one cannot calculate the OWSL in this way? Long and short waves from what perspective? Alfred Centauri 13:48, 25 April 2006 (UTC)
One can of course calculate it it, but not more than with any other method: you get out what you first put in. What look like equally long waves in the co-moving frame, look like short and long waves in all other frames - but they add up in a way that everyone agrees on what the light pattern is. Just try it for yourself (I did). Or (to get back to what this page is supposed to be about!), give a reference to a peer reviewed paper that found another result. 20:45, 25 April 2006 (UTC)

Thanks for the input, Harald88 and Alfred, but we are losing focus here, and I would hate to see this simple discussion drag on forever!

Surely you admit that special relativity (SR) pertains to the case of the one-way light speed per two same-frame clocks. Surely you admit that SR states that this speed must be experimentally measured as c in all frames. Surely you admit that if this is not the case, then SR falls, regardless of round-trip nullness or node nullness or whatever.

However, as I have tried to get across, it is currently physically impossible to experimentally measure light's one-way speed between two same-frame clocks. (This is why the one-way Michelson-Morley experiment has never been performed.)

Also (as I have also tried to get across), even Einstein admitted that two absolutely synchronous clocks would disprove SR by finding a variable one-way light speed, regardless of round-trip nullness or node nullness.

[RE-Quoting Einstein:] "w is the required velocity of light with respect to the carriage, and we have

w = c - v.

The velocity of propagation of a ray of light relative to the carriage thus comes out smaller than c."

Einstein did not disprove his w = c - v. He merely discarded it because he disliked it.

Even though SR must have one-way invariance, it cannot get it without forcing it to happen (by rigging the one-way time on the clocks), so SR's one-way invariance cannot exist experimentally. On the other hand, one-way variance can exist experimentally, if absolutely synchronous clocks are used to measure light's one-way speed.

So where does this leave the 2nd postulate?

It leaves it saying the following:

Since I, Einstein, firmly (but, to be honest, baselessly) believe in "all null results," I believe that clocks will not be properly related unless they are forced by definition to obtain "c" for light's one-way speed. Furthermore, I must assume that absolute synchronization is not possible because it can quickly overturn my theory.

Translation (again speaking as Einstein): Although my clocks are admittedly not correctly related (i.e., they are not truly (or absolutely) synchronous), we must nevertheless accept their results (e.g., the transformation equations, the composition of velocities theorem, one-way light speed invariance) as meaningful parts of physics. But since I cannot prove a negative, I have to admit that it may be possible to absolutely synchronize clocks, thereby totally disproving special relativity, despite round-trip nullness or node nullness or whatever. Cadwgan Gedrych 14:25, 20 April 2006 (UTC)

continued in 2nd Postulate Update II

Expansion of explanation?

I really like that this article has both the basics and the details/math. However, this section left me wondering:

Because of the freedom one has to select how one defines units of length and time in physics, it is possible to make one of the two postulates of relativity a tautological consequence of the definitions, but one cannot do this for both postulates simultaneously, as when combined they have consequences which are independent of one's choice of definition of length and time.

Could someone please expand it a bit? Or if it is immediately apparent to everyone (and I'm just a dumbass), can someone please explain this in more mathematical detail for me?

Thanks OhSoCurious 18:44, 14 April 2006 (UTC)

Clean Up

Firstly, this article if pretty difficult to understand. I don't have the know-how to tune it up, but it needs simpler wording. Also, near the bottom their is alot of bolded, bright red HTML coding, which shouldn't be there. Thanks, Theonlyedge 22:47, 19 April 2006 (UTC)

Agreed. Can any expert in this field attempt to make this article a bit more simple to read? --Siva1979Talk to me 21:10, 25 April 2006 (UTC)
The special relativity article is not understandable because critical facts were omitted, such as the simple fact that Einstein's clocks are not synchronous. Cadwgan Gedrych 13:50, 26 April 2006 (UTC)

2nd Postulate Update II

continued from 2nd postulate update

Cadwgan Gedrych, here's why I am not happy with the definition of absolute time: "Simultaneity" is not defined (presumably you don't accept the light signal definition, because then the absolute time just defined does not exist; that has been shown e. g. by GPS satellites). You have yet to explain how any (simpler) alternative to special relativity explains Alfred Centauri's standing wave example; this example is really a better way to measure one-way light speed, voiding the "time-rigging" argument. It is a way to test the second postulate. You did not point out any logical flaws in my arguments, you only urged me to show you that the experiments have already been done. However, I found a possible flaw myself: Given my previously stated hidden assumption, we don't know which frame is the "rest frame". However, Alfred Centauri's approach does not have this problem. Icek 17:50, 20 April 2006 (UTC) Cadwgan, without speaking to the merits (or lack thereof) of any of your arguments, I remind you that Wikipedia talk pages are not intended to be forums for advancing a particular view over any other view. "Wikipedians generally oppose the use of talk pages just for the purpose of partisan talk about the main subject. Wikipedia is not a soapbox; it's an encyclopedia. In other words, talk about the article, not about the subject." Talk_page#Basic_rules_for_all_talk_pages So, please go ahead and make whatever edits you feel are appropriate to the main article. Of couse, as with all contributions, your edits are subject to modification and, if they are sufficiently controversial, being reverted. So, perhaps you should create a new article entitled "Cadwgans' intepretation of what Einstein REALLY said" Alfred Centauri 18:05, 20 April 2006 (UTC)

I qoute Cadwgan: "Surely you admit that if this is not the case, then SR falls, regardless of round-trip nullness or node nullness or whatever." It is straightforward to show that, assuming only the homogeniety and isotropy of space and time and the principle of relativity, the most general space and time coordinate transformation mixes space and time coordinates. Further, this transformation has a scaling that depends on an invariant speed and the relative velocity between the coordinate systems. If we allow this invariant speed to go to infinity, the space and time components decouple to yield the Galilean transformation where the notion of absolute time and absolutely synchronized clocks actually means something. In this context, the 2nd postulate is simply the assertion that the invariant speed of this general coordinate transformation is c. Now, it has been claimed to be shown that, by experiment, the invariant speed of the universe can be determined. It occurs to me that although this experiment may not give the exact result (due to experimental error), it certainly can determine if the invariant speed is finite. If this is true, it follows that if this experiment gives a finite result, the notion of absolute time and thus, absolutely synchronized clocks, is without meaning and, rather than breaking SR, would simply change the invariant speed of the transformation equations. Alfred Centauri 21:05, 20 April 2006 (UTC)

Cadwgan Gedrych, here's why I am not happy with the definition of absolute time: "Simultaneity" is not defined (presumably you don't accept the light signal definition, because then the absolute time just defined does not exist; that has been shown e. g. by GPS satellites). You have yet to explain how any (simpler) alternative to special relativity explains Alfred Centauri's standing wave example; this example is really a better way to measure one-way light speed, voiding the "time-rigging" argument. It is a way to test the second postulate. You did not point out any logical flaws in my arguments, you only urged me to show you that the experiments have already been done. However, I found a possible flaw myself: Given my previously stated hidden assumption, we don't know which frame is the "rest frame". However, Alfred Centauri's approach does not have this problem. Icek 17:50, 20 April 2006 (UTC) Cadwgan, without speaking to the merits (or lack thereof) of any of your arguments, I remind you that Wikipedia talk pages are not intended to be forums for advancing a particular view over any other view.

"Wikipedians generally oppose the use of talk pages just for the purpose of partisan talk about the main subject. Wikipedia is not a soapbox; it's an encyclopedia. In other words, talk about the article, not about the subject." Talk_page#Basic_rules_for_all_talk_pages

So, please go ahead and make whatever edits you feel are appropriate to the main article. Of couse, as with all contributions, your edits are subject to modification and, if they are sufficiently controversial, being reverted. So, perhaps you should create a new article entitled "Cadwgans' intepretation of what Einstein REALLY said" Alfred Centauri 18:05, 20 April 2006 (UTC)

I qoute Cadwgan: "Surely you admit that if this is not the case, then SR falls, regardless of round-trip nullness or node nullness or whatever."

It is straightforward to show that, assuming only the homogeniety and isotropy of space and time and the principle of relativity, the most general space and time coordinate transformation mixes space and time coordinates. Further, this transformation has a scaling that depends on an invariant speed and the relative velocity between the coordinate systems. If we allow this invariant speed to go to infinity, the space and time components decouple to yield the Galilean transformation where the notion of absolute time and absolutely synchronized clocks actually means something. In this context, the 2nd postulate is simply the assertion that the invariant speed of this general coordinate transformation is c. Now, it has been claimed to be shown that, by experiment, the invariant speed of the universe can be determined. It occurs to me that although this experiment may not give the exact result (due to experimental error), it certainly can determine if the invariant speed is finite. If this is true, it follows that if this experiment gives a finite result, the notion of absolute time and thus, absolutely synchronized clocks, is without meaning and, rather than breaking SR, would simply change the invariant speed of the transformation equations. Alfred Centauri 21:05, 20 April 2006 (UTC)

It has been said that I am pushing some sort of purely personal view; hopefully, the following will refute this:
Here is why only the two-clock version of the 2nd postulate is the only version that's relevant:
Only the 2-clock, one-way version (as opposed to all other versions, e.g., the round-trip, one-clock version, the nodes version, etc.) contains Einstein's clock-setting procedure, and this unique procedure controls all of the two-clock results of special relativity (SR).
For example, it controls all one-way speed values in SR, including the most critical one, the one-way speed of light.
For another example, it controls the relativistic transformation equations.
For yet another example, it controls the composition of velocities in SR.
It causes Einstein's relativity of simultaneity.
It causes Einstein's "time dilation."
It causes Einstein's "length contraction."
It causes Einstein's "momentum variance."
However, as critical as the one-way, 2-clock case is, this specific experiment has never been performed, not even on paper.
That is, c invariance has never been shown, nor can it be shown!
From this it follows that Galileo's one-way light speed variance has not been disproved.
Here are two important things to be asked of any article that purports to pertain to the 2nd postulate:
1. Does the article tell what the 2nd postulate says about the one-way, 2-clock light speed case?
2. Does the article dare to mention that the one-way, 2-clock light speed case remains open?
3. Does the article say that the one-way, 2-clock light speed case controls SR's results such as the Einsteinian transformation equations?
4. Does the article state clearly that no one has ever measured light's speed from Clock A to Clock B?
5. Does the article admit that Einstein's clocks are incorrectly related, so all of SR's 2-clock results are incorrect?
6. Does the article admit that no one has ever disproved the Galilean transformation equations (taking into account clock slowing and length contraction, which were added by Lorentz)?
7 Does said article explicitly state that light's one-way, 2-clock speed will vary with frame velocity if absolutely synchronous clocks were used instead of Einstein's absolutely asynchronous clocks?
I do not want to create some silly new article called "Cadwgan's Intepretation of what Einstein REALLY Said"; I DO want some sort of consensus here in TALK:SR about the current article's dismal failures.

Cadwgan Gedrych 19:54, 24 April 2006 (UTC)

Referring to the collaborative work of others as a 'dismal failure' is an inspired approach to achieving the consensus you desire. Please Don't be a dick Alfred Centauri 23:53, 24 April 2006 (UTC)

About the above points,
- Point 1: It was set by definition - in fact that convention was already commonplace, as Poincare indicated in an earlier article. A procedural definition isn't a postulate about physics, eventhough it appears that Einstein suggested so in his 1905 paper.
- Point 2: Regretfully the sources are either disagreeing or unclear. As this is an encyclopdia, we can't (not allowed!) do better. Still, according to http://arxiv.org/abs/gr-qc/0409105, Einstein reiterated in 1907 that "clocks can be adjusted in such a way that [OWLS] becomes [..] c". That's obviously not a postulate but a procedure, and it would be helpful if the article cites that.
- Point 3: If you mean the Lorentz transformations: they are indeed dependent on that convention, but SRT's results are not (read for example the above arxiv article to understand why).
- Point 4: covered by point 2.
- Point 5: "incorrect" is a misnomer for a definition.
- Point 6. That Lorentz developed the LT based on the GT can be found in the corresponding links.
- Point 7: It may be enlightening to mention the fact that in GPS the speed of radio waves relative to the (moving) satellite is not c, and that that in no way contradicts SRT.
Regards, Harald88 11:04, 25 April 2006 (UTC)
To Alfred Centauri:
If the facts show (and they do) that the article is a dismal failure, then I have every right (indeed, a duty) to state just that, so it is you who is being the dick, not to mention the snide remark previously made about me making my own little article.
Here is a clear-cut example of a failure:
Even though the 2nd postulate fully controls how Einstein's clocks are related temporally, and even though this given temporal relationship must control all two-clock measurements in SR (including light's one-way, two-clock speed), and even though the only difference between Lorentz and Einstein is how their clocks are related, the article does not tell us the difference between Lorentz's (and Galileo's) clocks and Einstein's. Can you, Alfred Centuri, do this? Or will you cop out by changing the subject?
To Harald88:
Thanks for trying, but no cigar.
Do you mean to flatly contradict Einstein's own mathematical statement (given above by me) that the absolutely synchronous clocks of classical physics yield a variable one-way speed for light?
RE your "Point 5: "incorrect" is a misnomer for a definition.":
Look, it doesn't matter if Einstein's clocks are set by definition or by a herd of howler monkeys; if his clocks are asynchronous, then they are asynchronous, and this renders all of their results incorrect, including light's one-way, two-clock speed, the special relativity transformation equations, the Einsteinian composition of velocities theorem, the relativity of simultaneity, etc., as I have mentioned over and over.
Can you prove that Einstein's clocks are absolutely synchronous? If not, then they are absolutely asynchronous, and all of special relativity's two-clock results are incorrect.
You can't get correct results from incorrectly related clocks.
Regards Cadwgan Gedrych 13:33, 26 April 2006 (UTC)
Please read the article on Relativity of simultaneity for an introduction to this subject. Next, in 1905 Poincare demonstrated that according to the new mechanics no "absolute speed" can be measured, from which follows that his light speed convention is the simplest solution. Harald88 20:01, 27 April 2006 (UTC)

Re "the difference between Lorentz's (and Galileo's) clocks and Einstein's": As I mentioned above, neither theory says anything about how the clocks themselves work locally.

Re "RE your "Point 5: "incorrect" is a misnomer for a definition."": You do not seem to understand that it's not determined a priori what simultaneity is for spacially displaced clocks.

Re "Can you prove that Einstein's clocks are absolutely synchronous? If not, then they are absolutely asynchronous": see above. "absolutely synchronous" does not have an inherent meaning.

I'm getting tired of this discussion (but I'm still interested in your explanation for the standing wave example). Icek 14:36, 26 April 2006 (UTC)

Icek, all will be clear only after you show how Einstein derived his simple equation w = c - v.
(This equation was the cause of special relativity, and it has zero to do with a standing wave.) Good luck! (And please note that I did not say that Einstein said that the equation was correct, just that he derived it.)
Cadwgan Gedrych 18:39, 27 April 2006 (UTC)
See above; and if you here refer to the closing speed formula, that is a straightforward and fundamental measurement rule (vector addition of velocities). IOW, there is nothing to "derive", relative velocities are defined as such in physics. Harald88 20:01, 27 April 2006 (UTC)

To repeat myself again:

[Quoting Einstein:] "w is the required velocity of light with respect to the carriage, and we have

w = c - v.

The velocity of propagation of a ray of light relative to the carriage thus comes out smaller than c.

But this result comes into conflict with the principle of relativity...." http://www.bartleby.com/173/7.html

Surely, Harald88, you have seen this Einsteinian statement, and just as surely, Harald88, you know that it is not a closing velocity, but is simply a direct, two-clock measurement of light's one-way speed, given in the context of a single inertial frame in classical (Galilean/Newtonian) physics.

Please, Harald88, do us all a favor, and show how Einstein derived that simple equation "w = c - v" if you can.

The payoff for such a small act would be huge: You will then fully understand Einstein's 2nd postulate!

Cadwgan Gedrych 19:01, 28 April 2006 (UTC)

I have already tried to explain to you here above that that relative velocity is a two clock measurement in which the speed of light has been "set" to be uniformly c relative to the frame of choice; and I do understand Einstein's 1905 derivation (there is a minor glitch of no relevance in it, which was discussed about one year ago (or two?!)) on sci.physiocs.relativity. And that is the place for such opinion discussions, not here. Harald88 22:09, 28 April 2006 (UTC)
Einstein, in VI of "Relativity" applies the (then standard) addition of velocities theorem to a man walking (with speed w) inside a train car (moving with speed v) to calculate the speed of the man relative to the embankment: W=v+w. Then, in the next chapter entitled "The apparent incompatibility of the law of propagation of light with the principle of relativity", Einstein employs the same theorem to light propagating at c relative to the embankment: w = c - v. (hmmmmmm - I'm still waiting for my huge payoff..................... nope - no payoff here). Look's like Einstein straightforwardly applied the Galilean addition of velocities theorem.
He then says that this is in conflict the the (restricted) principle of relativity. Why? Because, at the beginning of the same chapter, he refers to the 'simple' law of physics that light propagates at c in the vacuum. He also says this: "The assumption that the velocity of propagation is dependent on the direction 'in space' is in itself improbable." So, in 1916 when Einstein wrote this "popular exposition" as "A clear explanation theat anyone can understand", he non-rigorously explains that (a) the Galilean addition of velocities theorem yields a speed of light that can be essentially any value, (b) EM theory says the speed of light is c, (c) physical observations indicate this speed does not depend on frequency (color) or the motion of the emitting body, and (d) that the notion this speed is direction dependent is improbable. Do you believe that this 'laymans' relativity explanation should reveal to us that the 2nd postulate is whatever it is you think it is? Alfred Centauri 21:30, 28 April 2006 (UTC)

If I might throw my two cents into this debate, I might point out that it is possible to "measure" the speed of light in two different frames without needing synchronized clocks or such. According to the first postulate, "...the same laws of electrodynamics and optics will be valid for all frames of reference for which the equations of mechanics hold good", or more coloqiually all the laws of physics are valid in all inertial frames, including as insinuated in the above quote, the laws of electromagnetism, usch as Maxwell's Equations. By testing the strength of the electric and magnetic interactions, one can get (or confirm or whatever) the values of   and  , the permeability and permittivity of free space, respectively. The speed of light in vacuum, however, is  . This measurement does not require any two-way trips (or much of any trip whatsoever). Thus the speed of light is constant for two observers in inertial frames of reference, no matter how measured. Indeed, this is the reason why some people consider the second postulate redundant - it's included in some sense in the first one. DAG 23:10, 29 April 2006 (UTC)

None of you found the payoff because none of you told how the clocks were set during Einstein's derivation of his w = c - v. I, too, am getting tired of this discussion - it is not worth the effort - let Wiki remain clueless re the truth of the 2nd postulate.
Cadwgan Gedrych 01:07, 30 April 2006 (UTC)
Isn't it well known that the Galilean addition of velocities theorem is based on absolute, universal, no kidding, don't even think about changin, time? Is this the big payoff you've been alluding to? Bummer! I was hoping you actually had something of merit to bring to the discussion. Alfred Centauri 03:46, 30 April 2006 (UTC)

Pjacobi, Centauri: Neither of you hit the target, so the payoff eluded you both.

There is only one way to hit the target, and that is by simply telling us how Einstein's clocks are temporally related. A bonus can be won by then going on to tell us how Galileo's clocks' temporal relationship compares with Einstein's. Cadwgan Gedrych 19:14, 1 May 2006 (UTC)

You can read that below. If something is unclear, you can comment on that Talk page; and I'll add a link to it in this article. Harald88 20:13, 1 May 2006 (UTC)

Where below? Which Talk page?

Cadwgan Gedrych 13:19, 2 May 2006 (UTC)

Einstein synchronisation
Talk:Einstein synchronisation
Icek 21:24, 2 May 2006 (UTC)

Wouldn't it be much, much, much, much simpler if you simply copied and pasted the thing you claim is there to here? I read the article, and I did not see anything at all about how Einstein's and Galileo's clocks differ temporally in any given frame. And if you do find it, then it should be put into the 2nd postulate section because the 2nd postulate is precisely and only about how Einstein's time for two clocks differs from the classical 2-clock time. Cadwgan Gedrych 15:22, 3 May 2006 (UTC)

What are "Galileo's clocks"? Classical physics does not care about clock synchronization. Icek 02:06, 4 May 2006 (UTC)

To Icek: How do you think the Galilean transformation was derived unless Galileo's clocks were used with some sort of clock synchronization? This is elementary. Cadwgan Gedrych 18:48, 4 May 2006 (UTC)

As far as I know, Galileo did not really think of this to be a problem and implicitely assumed that information transmission speed is unlimited (though he thought about measuring the speed of light). Icek 20:11, 4 May 2006 (UTC)

To Icek: Did not think of what as a problem? Cadwgan Gedrych 18:08, 5 May 2006 (UTC)

Synchronization of locally displaced clocks. Icek 07:00, 6 May 2006 (UTC)

To Icek: Please tell us how clocks are temporally related when they are used to derive the Galilean transformation equations. Or, alternatively, tell us how Einstein's clocks are related when the SR transformations are derived. Then, and only then, will you see what is meant by "Galileo's clocks." Cadwgan Gedrych 19:36, 8 May 2006 (UTC)

Why should I tell you the obvious again? Because one cannot answer my simple question in a way that would be advantageous to your point of view? The only thing I see from this is that you are not able to, you don't want to, or you pretend to not understand the problem. As others pointed out, this is not the right place to talk about this. End of discussion. Icek 22:08, 8 May 2006 (UTC)