Why Haven't Two Clocks on a Table Been Used to Measure Light's One-Way Speed?

[Martin Miller]

Was Einstein really a genius?

Nereid noted:
"Einstein's theories ... have been tested in the
crucible of experiment and observation, and have
passed with flying colours."

Sorry to burst your bubble, "Mr. Nereid," but as
far as Einstein's special relativity goes, your
above is purely an urban legend.

There have been exactly zero tests of SR.

For example, the very basis of SR, Einstein's
light postulate (i.e., one-way, two-clock light
speed invariance) has not been tested.

(To explain: No one has ever used two clocks in
one frame to measure light's one-way speed.)
(In fact, no one has ever even shown on paper
how this could be done!)

For another example, actual time dilation effects
were not predicted by SR, so these effects do not
test or support SR.

(To explain: It is easy to prove that SR does not
pertain to actual or intrinsic time dilation {or to
an atomic clock's internal rhythm} -- all that needs
be done is to point out the very simple facts that [1]
any inertially-moving atomic clock always has only
_one_ atomic rhythm, and yet [2] Einstein's observers
in various frames find _many_ "rhythms" for one and the
same passing clock; these facts prove that SR can't
pertain to intrinsic clock rhythms.)

(And of course the same argument applies to the
"mass increase" and "length contraction" cases.)

Please check the historical record before posting
any more silly urban myths.

And as for Einstein's genius, he was indeed a very
brilliant person, a genius even, but he did not win
the Nobel for SR. Also, it is not genius-like to say
"I am merely stipulating one-way invariance purely
by definition" out of one side of one's mouth whilst
stating the exact opposite out of the other side (i.e.,
claiming that one-way invariance is a prediction or
a postulate or a law of physics per experiment).

 

[StarThrower]

Originally posted by Martin Miller

I would add that anyone who claims to be able to absolutely
synchronize clocks should also show what happens to light's
one-way speed when it is measured by said clocks.

Also, I should add that anyone who finds a way to absolutely
synchronize clocks will thereby disprove special relativity,
which of course has only relative simultaneity. (This includes
the scientists involved in the GPS system, but none of them has
publicly claimed to have found absolute synchronization.)

My prediction re the above is that no one in this forum will
ever show the required step-by-step proof of absolute clock
synchronization. (In fact, if I were wealthy, I would bet
big bucks on it!)

I must admit that I have come into this discussion a bit late, but I have read through the entire thread. The discussion seems to be focused on clock synchronization, so my question is on the meaning of clock synchronization. I just want to be sure we all agree on the meaning of "synchronization".

MEANING OF CLOCK SYNCHRONIZATION

We have two clocks, clock A and clock B. Let them be of identical construction, therefore if they are both in inertial reference frames, they tick at the same rate. Let us suppose that the two clocks are at the ends of an absolutely rigid rod, therefore if one clock is in an inertial reference frame, then so is the other. Let us suppose then, that the clocks are in an inertial reference frame (hence the rod isn't rotating), now focus on the meaning of synchronization. To say that the clocks are 'synchronized' means that when clock A reads zeta, clock B simultaneously reads zeta. So, for example, if one clock currently read 17, and the other clock simultaneously reads 9929, then the two clocks aren't synchronized. Do we all agree that this is the meaning of synchronization?

Thank you.

 

[Nereid]

Martin Miller wrote:[n] Nereid noted: "Einstein's theories ... have been tested in the crucible of experiment and observation, and have passed with flying colours."

Sorry to burst your bubble, "Mr. Nereid," but as far as Einstein's special relativity goes, your above is purely an urban legend.

There have been exactly zero tests of SR.

That's Ms Nereid to you Martin.

Do you have a copy of Y.Z.Zhang, Special Relativity and its Experimental Foundations, World Scientific (1997)? If so, please tell us which of the tests discussed in this book failed?

If you don't, please check this page, and tell us which tests failed.

Please note that I am interested FIRST in 'pass/fail' in the following sense:
1) was there a specific, objective prediction made from SR?
2) did the experiment or observation produce a clear, unambiguous result?
3) was the result the same as that predicted by SR (within the errors of the observation)?

For the avoidance of doubt, I'm not interested (at this stage) in whether you feel there may or may not be inconsistencies in SR.

 

[outandbeyond2004]

Eyesaw, you assumed that if source-independent light goes out at L+v and comes back at L-v, then the trip time would be the same as though we have average (L+v plus L-v)/2 = L. Nay, nay, not so fast.

Going out, the half-trip time is D/(L+v) (assuming of course that D is not contracted or expanded). Coming back, the halftrip time is D/(L-v).

The total trip time is then D (1/(L+v) + 1/(L-v)) = $$D\frac{2L}{(L^2 - v^2)} = (2D/L)\frac{1}{1 - (v/L)^2}$$

If v is a substantial fraction of L, then there would be a rather obvious difference. Anyway, the MMXI should be able to detect interference from very small fractions of light.

I will look at the rest of your post hoping to catch more errors.

 

[Eyesaw]

Originally posted by outandbeyond2004
Eyesaw, you assumed that if source-independent light goes out at L+v and comes back at L-v, then the trip time would be the same as though we have average (L+v plus L-v)/2 = L. Nay, nay, not so fast.

Going out, the half-trip time is D/(L+v) (assuming of course that D is not contracted or expanded). Coming back, the halftrip time is D/(L-v).

The total trip time is then D (1/(L+v) + 1/(L-v)) = $$D\frac{2L}{(L^2 - v^2)} = (2D/L)\frac{1}{1 - (v/L)^2}$$

If v is a substantial fraction of L, then there would be a rather obvious difference. Anyway, the MMXI should be able to detect interference from very small fractions of light.

I will look at the rest of your post hoping to catch more errors.

Actually, the v I used represents a distance- i.e. the displacement through space of the moving apparatus as a result of its velocity v. Thus, it's clear that in the direction of motion, the light travels an extra distance of +v while in the round trip, the lesser distance traveled is -v. Even so, there is an error in your formula for time of flight since the time of flight of the light cannot be D/c+v and D/c-v if the distance back and forth for the light is not the same. Your usage of c+v and c-v is also in error since we are assuming the speed of light as source independent. The actual time of flight should be D+Q/c + D-Q/c = 2D/c. That is, the roundtrip distance is 2D or 2L or 2M or 2 and any other alphabet.

 

[outandbeyond2004]

We need to do this problem more carefully, at least I do. I made at least two mistakes in my last post.

Okay, let's assume light independence.

Shall we assume that the interferometer is moving at velocity $$v$$ wrt the ether or the distant stars? To an observer moving with the interferometer, for light moving in the direction of the velocity, it appears to be moving at speed $$c - v$$? For light moving in the opposite direction, the observed speed is $$-c-v$$?

 

[russ_watters]

Originally posted by Martin Miller
FYI: There is only one way to prove absolute synchronization,
and that is by showing step-by-step how it can be done. (This
involves the necessary step of providing the verification
process.)

Except for the word "absolute" (it doesn't exist in science), you can prove quite easily that GPS clocks are synchronized to within nanoseconds: stand over a benchmark with a gps reciever and compare the positions.

MEANING OF CLOCK SYNCHRONIZATION

Good point: since MM hasn't defined what he means by it, we can only make assumptions and he can flippantly reject any attempt to guess his intent. To most people, "synchronized" isn't a binary situation: clocks are synchronized to within a certain arbitrary precision. For my example of GPS clocks, they are synchronized to within a few nanoseconds of each other.

So, MM, perhaps you could precisely define what you mean by "synchronized."

The postulate of physics being the same in all inertial frames makes any transformation equation redundant. Since the postulate is a statement about physics inside an inertial frame, of what relevance is it to transform an event into a different inertial frame? [emphasis added]

I'm sorry, but that statement by being self contradictory displays a fundamental misunderstanding of what Einstein's Relativity says. The key word there is "in," not "between." If the rules worked the same between different frames, there would be no need for transformations. But since they only work the same in different frames, you need to do transformations to go between them.

And different frames are still related. By knowing how they are related you can make the appropriate transformations.

 

[Eyesaw]

quote:
--------------------------------------------------------------------------------
The postulate of physics being the same in all inertial frames makes any transformation equation redundant. Since the postulate is a statement about physics inside an inertial frame, of what relevance is it to transform an event into a different inertial frame? [emphasis added]
--------------------------------------------------------------------------------

I'm sorry, but that statement by being self contradictory displays a fundamental misunderstanding of what Einstein's Relativity says. The key word there is "in," not "between." If the rules worked the same between different frames, there would be no need for transformations. But since they only work the same in different frames, you need to do transformations to go between them.

And different frames are still related. By knowing how they are related you can make the appropriate transformations.

If physics is the same in all inertial frames, why do you need to transform anything? You can just use any arbitrary set of coordinates as long as it is an inertial frame. The only reason to transform coordinates is if physics is not the same in different inertial frames. But if you are going to use one inertial frame to view the events in another inertial frame, it's a contradiction to say first there is velocity between them and then say there is no velocity between them. Transforming coordinates to make the laws of physics the same in all inertial frames can be done many ways- doing it by the Lorentz Transformation results in contradiction. The better way to do it is to decide on source dependency from the getgo (through experimentation) and then decide whether the laws of physics are the same or not inside inertial frames. If experiments show source independency then the laws of physics can't be the same in different inertial frames; if the show source dependency then they can. That's inertial frame physics in a nutshell.

 

[Martin Miller]

russ_watters wrote (in part):
"So, MM, perhaps you could precisely define what you mean
by "synchronized."

"Originally posted by Martin Miller
FYI: There is only one way to prove absolute synchronization,
and that is by showing step-by-step how it can be done. (This
involves the necessary step of providing the verification
process.)"

"Except for the word "absolute" (it doesn't exist in science),
you can prove quite easily that GPS clocks are synchronized to
within nanoseconds: stand over a benchmark with a gps reciever
and compare the positions."

MM replies:
Please pardon my protest, but the word "absolute" does exist in
science; indeed, it is simply the opposite of "relative," so if
"relative" exists, then so does "absolute." For example, since
Einstein's clocks are based on relative simultaneity, all I have
to do to define absolutely synchronous clocks is to say that they
are the opposite of Einstein's. And as I mentioned earlier, even
Einstein stated mathematically that the absolutely synchronous
clocks of classical physics yield a variable one-way light speed.
Here is the relevant Einsteinian quote (repeated):

"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." (_Relativity_ Chap. VII)

Furthermore, every physicist agrees that two touching clocks
can be absolutely synchronized.

And this last fact provides us with the following simple way of
defining absolute synchronization for separated clocks:

Begin with two unstarted atomic clocks which are inertially moving
along one's frame's x axis. As the clocks meet in passing, absolutely
synchronize them by letting one start the other on touching. Then
assume (as did Einstein) that these clocks have identical internal
atomic rhythms. Given this, we know that they will remain absolutely
synchronous forever, no matter how far apart in space they may be.

Here is yet another definition of absolute clock synchronization:
Begin with two unstarted clocks located on one's frame's x-axis
at some distance apart. Cut a rod to fit precisely between these
two clocks. Remove the rod, place it somewhere else on the x axis,
and then slide it (inertially) toward the clocks. If we assume
(as did Einstein) that this rod's physical (or intrinsic) length
does not really vary with rod speed, then the rod will absolutely
synchronize the clocks if the former's ends are used to start the
latter.

And here is a simple image-based definition of absolute clock
synchronization: Picture two clocks which are running and which
are 50 light-years apart. One of them is on your desk very close
to your eyes. You are to reach out and touch the face of this
nearby clock with your index finger. If, at this exact moment,
the distant clock happens to read the same time as your near
clock, then the two clocks are (very, very nearly) absolutely
synchronous. (Disclaimer: I am of course _not_ saying here
that we can somehow _know_ what the distant clock is reading
when you touch the near clock; all I am doing here is providing
a graphical image of absolute synchronization which anyone can
fully understand instantly.)

Here is a pretty good indication of absolute synchronization:
If there is no theoretical or empirical reason for the failure
of one's chosen method for absolute synchronization.

Here are two almost-certain indications of absolute synchronization:
[1] Light's one-way, two-clock speed would vary with frame velocity.
[2] Observers in all frames would find the same times for any given
set of events.

Finally, here is a mathematically-certain indication of absolute
synchronization: Clocks in all frames are absolutely synchronous
IFF any given frame's observer's self-measured speed is the same
as that as determined by any other frame's observer. (A self-
measured speed is found by using one's two supposedly synchronous
clocks to measure the one-way speed of a passing light ray; as
Einstein himself noted, this speed should vary with frame velocity.)

I have repeatedly asked for a step-by-step description of some
means of absolutely synchronizing two clocks; so far, none has
appeared. I am of course speaking of dead-on synchronization
for clocks located anywhere in space, and not just close (i.e.,
within nanoseconds) of closely-located clocks. I am also talking
about proper verification of the claimed absolute synchronization.

 

[Peterdevis]

Here are two almost-certain indications of absolute synchronization:
[1] Light's one-way, two-clock speed would vary with frame velocity.
[2] Observers in all frames would find the same times for any given
set of events.

So in other words this clocks don't exist in a world ruled be GR.
In GR it is impossible to say if two clocks are synchronizated when they are at a certain distance. The time dilitation ( = a measure of non synchronisity)depends of the path the information takes to come to us.

There is only one way to prove absolute synchronization,
and that is by showing step-by-step how it can be done. (This
involves the necessary step of providing the verification
process.)"

Since this absolute synchronization is a proof of the incorrectness of GR, it 's up to you (who claimed the whole academic world is wrong) to proof of an existing absolute synchronizated set of clocks.

Begin with two unstarted atomic clocks which are inertially moving
along one's frame's x axis. As the clocks meet in passing, absolutely
synchronize them by letting one start the other on touching. Then
assume (as did Einstein) that these clocks have identical internal
atomic rhythms. Given this, we know that they will remain absolutely
synchronous forever, no matter how far apart in space they may be.
moving

we know that= Martin Miller assume that
What reason do you have that two clocks must stay synchronous forever

If we assume
(as did Einstein) that this rod's physical (or intrinsic) length
does not really vary with rod speed, then the rod will absolutely
synchronize the clocks if the former's ends are used to start the
latter.

The poor man can't defend himself, so don't put words in hi mouth he didn't mend to say

 

[outandbeyond2004]

Eyeshaw:

I got lost at the L-M step. What is wrong with the following analysis?

upleg travel time $$t_u$$ can be found from this equation:

$$c = \frac{L + vt_u}{t_u}$$

Equation to find roundtrip travel time (downleg travel time $$t_d$$):

$$c = \frac{L + vt_u + L - vt_d}{t_u + t_d}$$

Eliminating $$t_u$$ and solving for $$t_d$$, I got

round trip time = $$\frac{L}{c-v} + \frac{L}{c+v}$$

That is not the same as $$\frac{2L}{c}$$, as we can see when we transform the equation just above:

roundtrip time = $$(\frac{2L}{c})\frac{1}{1-(\frac{v}{c})^2}$$

 

[wisp]

For example, the very basis of SR, Einstein's
light postulate (i.e., one-way, two-clock light
speed invariance) has not been tested.

(To explain: No one has ever used two clocks in
one frame to measure light's one-way speed.)
(In fact, no one has ever even shown on paper
how this could be done!)

For another example, actual time dilation effects
were not predicted by SR, so these effects do not
test or support SR.

Very true, but don't forget Roland DeWitte's experiment - two clocks and an electrical pulse (similar to the one-way light speed test). His results showed SR to be false.

And the physical cause of moving clocks running slow (time dilation)can be explained outside of SR using ether concepts.

 

[StarThrower]

Originally posted by Martin Miller

I have repeatedly asked for a step-by-step description of some
means of absolutely synchronizing two clocks; so far, none has
appeared. I am of course speaking of dead-on synchronization
for clocks located anywhere in space, and not just close (i.e.,
within nanoseconds) of closely-located clocks. I am also talking
about proper verification of the claimed absolute synchronization.

Martin, I do not understand why clock synchronization is so important to you, since actual clock readings fall out of the mathematical analysis. In SR all that matters is how the "amount of time of an event" in one inertial reference frame relates to the amount of time of the same event viewed in a different inertial frame. My point is fairly simple to understand:

1. Let us stipulate that we have two clocks of identical construction, and that the clocks are attached by a rigid rod.

2. Let it be stipulated that the clocks are in an inertial reference frame.

Therefore, the clocks will tick at the same rate.

Thus, if they are in sync then they remain in sync, and if they are out of sync, then the difference in their readings is constant.

For example, suppose that when one clock reads 4, the other clock simultaneously reads 7. Since the readings of the two clocks differ, the clocks aren't synchronized. The difference in readings is:

7-4=3

Thus, the clocks are not in sync by 3 units.

If the readings had been equivalent, then the two clocks would be truly synchronous.

So here is my point though. The difference in readings is constant, and falls out of the mathematical analysis. In other words, relativity talks about how an amount of time in one inertial system "transforms" into a different inertial system.

For example, suppose that some event lasts $$\Delta t$$ seconds measured by clock A, which is in an inertial reference frame. Now, suppose that clock B is moving relative to clock A, at a constant speed of v. According to SR, the time of the event will not be $$\Delta t$$ seconds, according to clock B. Instead, the time measured by clock B will be delta t times gamma.

My point is that it is the difference in readings that matters, not the actual readings themselves. So, I don't see why clock synchronization is an issue.

I think your real question isn't being addressed. The issue shouldn't be how do you make two clocks synchronous, I think the real issue is, how does one determine the absolute difference in clock readings.

For example, suppose that one clock reads X, and another clock simultaneously reads Y. The difference in clock readings is defined as follows:

X-Y

If X=Y then the clocks are synchronized.
if not (X=Y) then the clocks aren't synchronized.

I think the real question for you is, "can one empirically determine the value of X-Y?" You seem to be hung up on empirically determining that X-Y=0. Wouldn't it be just as good to empirically determine "X-Y"? For example, suppose you already determined that X-Y = 52. Thus, the reading of clock X always exceeds the reading of clock Y by 52 units. Thus, if you are stationed at clock Y, and your clock currently reads 733, then you know that clock X reads 733+52=785.

So I guess my question is, "why do you need us to show that two clocks cannot be brought into synchronization?"

 

[StarThrower]

Originally posted by Nereid
That's Ms Nereid to you Martin.

Do you have a copy of Y.Z.Zhang, Special Relativity and its Experimental Foundations, World Scientific (1997)? If so, please tell us which of the tests discussed in this book failed?

If you don't, please check this page, and tell us which tests failed.

Please note that I am interested FIRST in 'pass/fail' in the following sense:
1) was there a specific, objective prediction made from SR?
2) did the experiment or observation produce a clear, unambiguous result?
3) was the result the same as that predicted by SR (within the errors of the observation)?

For the avoidance of doubt, I'm not interested (at this stage) in whether you feel there may or may not be inconsistencies in SR.

Hello Ms. Nereid, I believe Martin said ALL of them.

:)

 

[outandbeyond2004]

It is not so much that clock A must show the same time as clock B, but simply that they have the same rate. Suppose A reads 10:00 and we can somehow simultaneously determine that B reads 10:30 PM. We wait about one hour, and check the clocks again. If we still can read them simultaneously and clock A reads 11:01 & B reads 11:31, then their rates are the same. Either we adjust one of the clocks so that they read the same, or simply take note of the initial times and take the time differences. For example we note Ta0 = 10:21, Tb0 = 10:31 and later Ta1 = 12:01 and Tb1 = 12:31. The time differences, deltaTa = Ta1 - Ta0 and deltaTb1 = Tb1 - Tb0 are still the same. Would this not be good enough?

Note that many physical processes can serve as relative clocks and cannot be adjusted to give absolute time. You can't look at me and exclaim, "Omy! 10:34 AM! I'm late!" Yet, some physicist might regard my beating heart as a sort of clock. How does anyone synchronize my heart with Greenwich mean time? My heart does beat at about 1 beat per second, sometimes, but I'll be darned if I am going to carry around a time read-out device for my heart.

How do you adjust the rate of all those binary pulsar systems out there? Can you attach a time-readout device to any of them? No, SR and GR really only work with time differences, not absolute time like Greenwich Mean Time or Universal Time.

 

[Hurkyl]

Since the interferometer went a distance of +v with respect to the vacuum, the light traveled L+M in the direction of motion and L-M back for a round trip distance of 2L.

Impossible. Did you even try to do the math on this one?

Suppose the interferometer lies on the x-axis and is traveling in the direction of the positive x-axis with a speed of v.

Suppose we emit a photon from the left edge of the interferometer at time 0 when the left edge of the interferometer is at x-coordinate 0. (So the right edge has coordinate L).

IOW, at time 0:
The left edge has x-coordinate 0
The right edge has x-coordinate L
The photon has x-coordinate 0

Now, let's find at what time the photon strikes the right edge:
The formula for the photons position is

x = 0 + ct

The formula for the right edge's position is

x = L + vt

Setting the x-coordinates equal gives:

L + vt = ct

or

t = L / (c - v)


So, at time t = L / (c - v):
The left edge is at x-coordinate L v / (c - v)
The right edge is at x-coordinate L c / (c - v)
The photon is at x-coordinate L c / (c - v)

Now, let's find out when the photon strikse the left edge again.

The formula for the position of the left edge is

x = v t

The formula for the position of the photon is

x - L c / (c - v) = -c (t - L / (c-v))
simplifying:
x - L c / (c - v) = -c t + L c / (c - v)
x = 2 L c / (c - v) - ct

Setting the x-coordinates equal and solving gives:

vt = 2 L c / (c - v) - ct
t = 2 L c / ( (c - v) (c + v) ) = 2 L c / (c^2 - v^2)


So the photon travels for time 2 L c / (c^2 - v^2). In that time, it travels a distance of 2 L c^2 / (c^2 - v^2). Funny, but that seems to be inequal to 2 L.


And you get an even different distance if the inferometer is moving in a different direction!

 

[Hurkyl]

If physics is the same in all inertial frames, why do you need to transform anything?

Practically, one very useful reason to transform anything is because computation may be simpler in another frame.


Conceptually, careful application of the transform laws help one work through errors in their understanding.

 

[russ_watters]

Originally posted by Eyesaw
If physics is the same in all inertial frames, why do you need to transform anything?

I'm not sure where you're getting this idea from, but the whole point of "relativity," a concept which predates Einstein's version, is that the laws are the same in all reference frames. Under classical (also called Galilean) relativity, the fundamental postulate is the same as in Einstein's: The laws of the universe are the same for all observers regardless of inertial frame of reference.

This does not and never did imply that two observers in different frames would see the same thing - transformations are needed even under classical relativity. This should be self-evident from the commonly cited thought experiment with a man walking on a train:

A man is walking forward on a train at 1m/s relative to the train. The train is moving forward at 10m/s relative to the station. To a man standing on the platform, how fast is the man on the train moving? To a man sitting on the train, how fast is the walking man moving.

Clearly, the man sitting on the train and the man standing on the platform will not agree on the speed of the man walking unless they have a transformation equation by which to relate the two different frames of reference.

 

[outandbeyond2004]

One can indeed assert that physics is the same from frame to frame, in the sense that the laws of physics are the same in all frames. Yet, someone in free fall is not the same as someone lying on the ground. Nevertheless, we ought to be able to apply the _same_ laws to both persons' experiences. Transforming the laws from frame to frame is a necessity, because physics do differ from frame to frame, at least in evidence or observations.

 

[russ_watters]

Originally posted by wisp
Very true, but don't forget Roland DeWitte's experiment - two clocks and an electrical pulse (similar to the one-way light speed test). His results showed SR to be false.

If true, it would be relatively (no pun intended) trivial to duplicate his results. From what I have heard, his results aren't accepted as credible by the scientific community.

And the physical cause of moving clocks running slow (time dilation)can be explained outside of SR using ether concepts.

That's true, but it is never preferable in science to add unsupported assumptions to a theory in order to make it fit some preconcieved notion of how the universe "should" work. So until someone finds real, scientific, positive evidence to show that there is an ether, we cannot assume there is one. At the very least, ether theory fails on those grounds.

Hello Ms. Nereid, I believe Martin said ALL of them.

That's nice, but I say all of them are valid. Do we have a Mexican standoff now, or should we practice some science by explaining why?

 

[Nereid]

Originally posted by StarThrower
Hello Ms. Nereid, I believe Martin said ALL of them.

:)

What Martin actually wrote was: "There have been exactly zero tests of SR." IOW, no one has tested SR. What he didn't say was whether any of the tests in the list I provided were a) predictions of SR, b) done, and c) consistent.

The strange thing is, I've asked this of all SR and GR naysayers here at PF, and (apart from wisp) they are all mute (except re MMx).

 

[Martin Miller]

StarThrower wrote:
"Martin, I do not understand why clock synchronization is so
important to you, since actual clock readings fall out of the
mathematical analysis. In SR all that matters is how the
'amount of time of an event' in one inertial reference frame
relates to the amount of time of the same event viewed in a
different inertial frame."

I would love to see the reaction if you posted the above in the
relativity newsgroup. Anyway, I can now see why my messages are
not getting through here. In no way do clock readings "fall out
of the math analysis." For example, it is only because Einstein's
clocks are asynchronous that light's one-way, two-clock speed is
found to be invariant and isotropic, so it is only because of
the asynchronousness of Einstein's clocks that SR exists.

Given absolutely synchronous clocks (or their equivalent, namely,
clocks with a known difference), light's one-way speed will *NOT*
be invariant, and it will *NOT* be isotropic, and we would say
"Goodbye" to all of SR! [see diagrammed examples below]

Does this sound as if clock readings "fall out"?
I don't think so!

StarThrower also wrote:
"I think the real question for you is, 'can one empirically
determine the value of X-Y?'"

Well, of course all we need to know is the absolute difference
between the clocks; that goes without saying, so why the heck
am I having to say it!?

StarThrower finally wrote:
"So I guess my question is, 'why do you need us to show that two
clocks cannot be brought into synchronization?'"

Because clock synchronization controls all one-way speed values,
including light's. (And there are one-way speed values in both
Einstein's transformation equations and his composition of
velocities theorem, so these are controlled primarily by clock
synchronization.)

You have to understand that the only difference between Galilean/
Newtonian physics and Einstein's is the two-clock value of light's
one-way speed, and this value is controlled by **synchronization**.
Galileo assumed truly or absolutely synchronous clocks, and this
gave him c ± v as light's one-way, two-clock speed. Einstein
mandated absolutely ****asynchronous****, which gave him the
_incorrect_ result of c for all.

Special relativity stands or falls on the single question of
clock synchronization. This is why clock synchronization is
of the utmost importance re flat space-time physics.

If I produce truly synchronous clocks (or their equivalent,
clocks with a known difference), then I will overthrow SR.

outandbeyond2004 noted:
"It is not so much that clock A must show the same time as
clock B, but simply that they have the same rate."

Apparently you never simply used a couple of clocks to measure
light's one-way speed on paper; I very strongly suggest that
you do this ASAP.

In fact, I will help you by doing it _now_, as follows:

First, I will use Einstein's asynchronous clocks
in a single inertial frame:
(frame speed is 0.6c wrt the light source S, about which
an emitted light sphere is assumed to remain symmetrical)
(distance is per an in-frame ruler which is contracted
by 20% due to its motion wrt the light source S)
(time T is per a clock with is at rest wrt S; time t
is per the clock which is moving wrt S at speed 0.6c)
([t] represents a clock which now reads time t)

[0]----1 LY----[-0.6]->0.6c
S~>light ray

...[1.6]----1 LY----[1]->0.6c
S~~~~~~~~~~~~~>light ray

Due to length contraction, the rod which is measured by the
in-frame rule to be 1 LY long is only 0.8 LY long. Due to the
frame's motion wrt S and its emitted light, the ray's effective
speed on paper is c - v. Thus, the time T is 0.8/(c - v) =
0.8/(1 - 0.6) = 2 yrs. Thus, the time t = gamma * 2 = 0.8 * 2
= 1.6 yrs. Of course, the light ray's speed per the frame's
observers is their ruler-measured frame distance of 1 LY divided
by their two-clock-measured time span of [1.6 + (-0.6)] - 0 = 1 yr.
So the light ray's measured speed is simply c, as it simply _must_
be in all Einsteinian frames.

It should be clear from the above that it is impossible to obtain
Einstein's necessary one-way light speed invariance for any frame
which moves wrt the light source _unless_ the frame's clocks are
absolutely _asynchronous_, but just to nail home this critical point,
I will now show what happens when truly synchronous clocks are used
to measure light's one-way speed:

[0]----1 LY----[0]->0.6c
S~>light ray

...[1.6]----1 LY----[1.6]->0.6c
S~~~~~~~~~~~~~~>light ray

As shown, the initial speed result is 1 LY/1.6 yrs = 0.625c, which
seems to tell the observers that their frame speed is 0.375c, but
they need to correct for clock slowing and rod shrinkage as follows:
The actual on-paper rod length is 0.8 LY, and the on-paper clock
time T is 2 yrs, so the corrected value of the light ray's one-way
speed is 0.8 LY/2 yrs = 0.4c, which yields the correct frame speed
of 0.6c.

 

[russ_watters]

Originally posted by Martin Miller
Given absolutely synchronous clocks (or their equivalent, namely,
clocks with a known difference), light's one-way speed will *NOT*
be invariant, and it will *NOT* be isotropic, and we would say
"Goodbye" to all of SR!

You are operating under the incorrect and unfounded assumption that SR is incorrect. You'll need to do better than using your own assumption as a proof.

It also appears you are using a definition of "synchonized" which has clocks operating at precisely the same rates without transformations in all frames. I'm not sure though, because as much as you are harping on synchronization, you have yet to actually define what you mean by it.

In any case, you've purposely structured the your assumptions in way that logically excludes the possibility of SR being correct. While that may be compelling to you, you can't simply assume anything you want and base a logical proof on it. Even the assumptions have to be grounded in reality. Yours are not.

You (and others) keep asserting your position to be correct without explanation. You can't simply assert you are correct and demand to be proven wrong. Science doesn't work that way. You have to actively prove your position correct with scientifically valid data and calculations.

In addition, a number of people have shown a number of flaws in your reasoning and you have declined to address the problems we found. Particularly, Nereid, several pages ago (in response to an assertion that SR hasn't been demonstrated in any experiments), linked a list of experiments proving SR and asked for specific problems you (and others) have with these experiments. No response. Here's that post again:

Eyesaw, in another thread in this very same sub-forum, I post two links to lists (with references) of tests of SR and GR. Your reply to my post was (excerpts): "Yes, I have looked at that webpage before. But before we go over these experiments, I'd still like an answer to how any test can be claimed to have confirmed SR ..."

In a nutshell, the answer to your question is 'you can make quite specific predictions from SR; you can do the experiments and make the observations; when you do, you find that the predictions are correct, to within the experimental/observational errors'. IMHO, that's all you can ask of a theory.

So, let's go look at the experiments on the lists:
1) was a specific prediction from SR made?
2) was that prediction made correctly (e.g. no screw-up in the math)?
3) did the researchers do the experiment/make the observation?
4) were the results consistent with the prediction?

Please tell us which of the experiments, in your mind, have "NO" as the answer to any question.

Until you guys can begin to address what the existing theories actually say and mean (not what you assert they say and mean, but what actual scientists say they say and mean) and tell us precisely where the flaws lie in the mountains of papers/experiments supporting them (pick one or two for a start), you haven't proven your point in our eyes.

 

[outandbeyond2004]

Martin Miller:

I think now what you refer to as absolute synchronization is in fact what others refer to as the absolute time of Newton and Galileo -- a single standard of time for all observers.

Back in Galileo and Newton's time, experimental equipment and technique were not as advanced as ours are. The best Galileo could say regarding the speed of light was that it must be very great if not infinite. However, by the time Einstein was beginning to ponder the matters that led to his 'miracle year' (1905), experiments began to show that Galilean-Newton relativity didn't seem to account for the facts of Nature very well. In particular the Michelson-Morley experiment shook physics quite a bit.

Eyesaw is not doing a good job of explaining the null results of experiments like the MMx. Can you do a good job? It seems as though light speed was not really infinite or practically so ('must be very great'), but was finite. Maxwell's electrodynamic equations predicted a single, fixed speed for EM radiation, but until Einstein, nobody really understood what it implied for physics.

Galilean relativity works fine as long as the speed of light is 'very great.' But, Einstein showed contradictions in Galilean Rel. when he analyzed in a single frame of reference what two observers moving wrt each other should observe.

SR (and GenRel) does permit each observer to use a single standard of time, but one standard for all observers turned out to be unworkable. Instead, at least conceptually, each observer was allowed to set up and use his own standard of time.

Your "not absolutely synchronized" objections misses the point that one can always analyze the physics of two observers moving wrt each other in a different frame, with a single standard of time.

One thing you must do if you continue to insist on Galilean relativity is to show that observers can get information faster than the speed of light. I know of no such experiment showing that this is possible.

 

[jdavel]

Martin Miller posted (again!):

"For example, the very basis of SR, Einstein's light postulate (i.e., one-way, two-clock light speed invariance) has not been tested."


But that doesn't matter! To validate a theory in physical science, it's not necessary to prove (or even to test) the POSTULATES of the theory. The theory is validated or refuted by testing its PREDICTIONS. It doesn't even matter whether the postulates are testable. IT DOESN'T EVEN MATTER WHETHER THE POSTULATES ARE TRUE! If the theory is a good predictor, it's a good theory. If it's the best predictor, it's the best theory.

SR predicts (among other things) that measurements of time made with a moving clock will show it to be running slow compared to measurements made with the same clock when it's stationary, that measurements of the length and mass of a moving object will show it to be shorter and more massive than measurements on the same object when it's standing still. Experiments, with ever increasing precision, have tested these predictions (ad nauseum) for nearly 100 years. At the present time, it is by far the best predictor for the results of these experiments; no other theory even comes close.

 

[Martin Miller]

Was Einstein really a genius?

'Ms Nereid' (not 'Mr.') wrote:
"... please check this page, and tell us which tests failed."

Hello, Ms Nereid, I know where you are coming from because
I have seen that page many times. It contains only round-trip
cases, rotating clock cases, and intrinsic time dilation, not
one of which applies to, supports, or tests SR.
For example, see the following site which proves that all
rotating clock cases are expected to have null results: http://www.geocities.com/antirelativity/Rotating_Clock_Analysis.html

You have to know SR to know what could test it.

SR did not predict round-trip invariance or isotropy.
This was proved - at least the latter was proved - prior to SR.

SR does not pertain to intrinsic time dilation, but only to a
trivial and irrelevant point-of-view (apparent) "clock slowing"
caused by the asynchronousness of Einstein's clocks. (That SR's
"time dilation" does not pertain to the actual atomic rhythms
of clocks is extremely easy to prove, and here is a proof:
Any single inertially-moving atomic clock will always have only
one intrinsic atomic rhythm, and yet Einstein's observers in
various frames always will find _different_ "rhythms" for one
and the same passing atomic clock; therefore, SR's "time dilation"
does _not_ pertain to the physical or intrinsic atomic rhythm of
an atomic clock.) (Similarly, SR does not pertain to either
intrinsic rod lengths or to intrinsic mass.)

What does SR predict?
To what does SR pertain?
What does SR say?
Not surprisingly, since SR is wrong, the website you
referenced made no mention of the real SR, but, as you
requested, I will not go into the "SR is wrong stuff"
yet, but will merely explain what SR predicts.

SR began when Einstein saw round-trip isotropy (given by the
Michelson-Morley experiment of course). Einstein then assumed
that round-trip isotropy implied both round-trip invariance and
one-way isotropy/invariance. Seeing that clock synchronization
controls all two-clock times, he saw that light's one-way,
two-clock speed depends on clock synchronization, so he then
related his clocks to obtain one-way invariance/isotropy.

At this critical point, we must decide whether Einstein had
a postulate or merely a clock synchronization definition.
But either way, Einstein loses.

Case I - Assuming that Einstein had a light postulate -
If Einstein had a light postulate, and if this postulate says
one-way, two-clock light speed invariance/isotropy, then _how_
can this postulate be tested?

The answer of course is as follows:
Only by using correctly synchronized clocks.

Case II - Assuming that Einstein had a mere definition -
If all Einstein had was a mere convention or definition for
relating clocks, then how can it be proved that his clocks
are correctly related?

One answer is as follows:
By proving that light's one-way, two-clock speed is indeed
invariant and isotropic - independently of his _definition_
(which _forces_ invariance and isotropy).

Do you see the logical circle which is wound tightly around
the neck of relativity?

Adding to the problems for SR is the fact that Einstein himself
explicitly (mathematically) admitted that the absolutely
synchronous clocks of classical physics would NOT find his
precious invariance and isotropy in the one-way case. And
he was unable to prove that such clocks cannot exist.
[REF: "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."
(From _Relativity_ Chap. VII)]

SR has never been tested because no one has found a way to
correctly synchronize clocks, not even on paper.

 

[Martin Miller]

russ_watters noted:
"You (and others) keep asserting your position to be correct without
explanation. You can't simply assert you are correct and demand to be
proven wrong. Science doesn't work that way. You have to actively prove
your position correct with scientifically valid data and calculations."

Wrong. I need not prove anything. It is Einstein who claimed that
light's one-way speed is invariant and isotropic. Thus it is he
(or his followers) who must prove this claim. (OTOH, I have proved
that the claim is dead wrong.)

russ_watters noted:
"In addition, a number of people have shown a number of flaws in your
reasoning and you have declined to address the problems we found.
Particularly, Nereid, several pages ago (in response to an assertion
that SR hasn't been demonstrated in any experiments), linked a list of experiments proving SR and asked for specific problems you (and others) have with these experiments. No response."

No one has found a single flaw. The Nereid-cited site does not
have a single experiment which proves Einstein's sole claim of
the invariance/isotropy of light's one-way, two-clock speed
(where the clocks are in the same frame, as they must be).

Can you prove Einstein's claim?
Can you even show - even on paper - _how_ it could be proved?

No, you cannot.

Therefore, SR's sole basis (Einstein's claim) is unproved, and
cannot be proved even on paper.

Does this sound as if SR is a valid scientific theory?

 

[Martin Miller]

outandbeyond2004 noted:
"One thing you must do if you continue to insist on Galilean
relativity is to show that observers can get information faster
than the speed of light. I know of no such experiment showing
that this is possible."

One thing you must do if you continue to insist on Einsteinian
relativity is to show that observers can get invariance and
isotropy for the one-way speed of light, but I know of no way
to show this experimentally, and neither do you.

(BTW, getting information faster than light has nothing to
do with the facts that Einstein's clocks are incorrectly
related and that truly synchronous clocks will yield a
variable one-way light speed.)

 

[Martin Miller]

outandbeyond2004 wrote:
"Eyesaw is not doing a good job of explaining the null results
of experiments like the MMx. Can you do a good job?"

I don't need to explain it, but many have claimed that SR
explains it. Can you explain how SR is supposed to have
explained the MMx null result? (Or do you believe that SR
cannot explain it?)

 

[outandbeyond2004]

Martin Miller, is your preferred alternative to Einsteinian relativity Galilean Relativity? If not, what is it? Why continue to ride something that seems to be realistic no more?

Do you know about the binary pulsar systems? They involve one-way EM signals.

Look, nothing can be proven in science. You have to start with assumptions. This is a fact of life that I need not prove: YOU show us a fact that need not be assumed apart from facts of your own existence or facts inferred from them. You are going the wrong way by insisting on proof of something that may need to be assumed. Ultimately all that can be asked of a theory is that it predicts things correctly. Do you really believe Galiliean R. still does so?

 

[jdavel]

Martin Miller wrote:

"SR has never been tested because no one has found a way to correctly synchronize clocks, not even on paper."


As I posted previously, the test of a scientific theory is a test of its predictions not its postulates. However, since you're so adamant about the a measurement of the one-way speed of light being the only true test of SR, here's a test you could do, at least "on paper"?

Put a clock into the same orbit around the sun that the Earth travels in but trailing behind the Earth by, say, 90 degrees (i.e., three months). Now watch the clock through a telescope. If light speed depends on direction, you'll observe the clock speeding up and slowing down over the course of a year. If it doesn't, you won't.

How's that?

 

[DrChinese]

Originally posted by jdavel
Martin Miller posted (again!):

"For example, the very basis of SR, Einstein's light postulate (i.e., one-way, two-clock light speed invariance) has not been tested."


But that doesn't matter! To validate a theory in physical science, it's not necessary to prove (or even to test) the POSTULATES of the theory. The theory is validated or refuted by testing its PREDICTIONS. It doesn't even matter whether the postulates are testable. IT DOESN'T EVEN MATTER WHETHER THE POSTULATES ARE TRUE! If the theory is a good predictor, it's a good theory. If it's the best predictor, it's the best theory.

Yes, quite so! Utility is the easiest way to determine the value of a theory. It makes predictions. Einstein made plenty of predictions via his theories, and they have plenty of utility. Recall E=Mc^2? What was that, a lucky guess? SR, GR, numerous contributions to early quantum theory.

No one is going to prove his theories wrong, but even more accurate ones may come along in the future. The same thing happened with Newton's laws of gravitational attraction, which are still quite useful today. Einstein did not invalidate Newton's work when he advanced General Relativity. So Newton's contributions are not diminished.

 

[russ_watters]

Originally posted by Martin Miller
Wrong. I need not prove anything. It is Einstein who claimed that
light's one-way speed is invariant and isotropic. Thus it is he
(or his followers) who must prove this claim. (OTOH, I have proved
that the claim is dead wrong.)...

The Nereid-cited site does not have a single experiment which proves Einstein's sole claim of the invariance/isotropy of light's one-way, two-clock speed (where the clocks are in the same frame, as they must be).

So basically then your argument boils down to 'my assertions are correct and the entire scientific community is wrong.' Still not a very compelling argument.

 

[Nereid]

Ha, outandbeyond2004 beat me to it!

Consider the following thought experiment:

A very reliable clock is located at a very great distance, in a stable environment (= nothing of any significance changes over a period of a century). The clock emits signals across the EM spectrum, at a rate (according to itself) of 1 'tick' per second.

The clock is located, from our perspective, on the ecliptic* (= the plane of the orbit of the Earth around the solar system's baricentre). Martin has a radio, X-ray (some caveats), and optical observatory fixed on the surface of the Earth, and his position, relative to the solar system barycentre, is known to him - and all other PF members - to an accuracy of 1cm**. Eyesaw, wisp, and many others visit his observatory regularly, and have full access to all equipment. Indeed, they are free to install their own equipment at his observatory, to be sure that their own concerns about SR and GR are properly addressed.

What one-way test does Martin perform?

*To ensure that gravitational effects of the Sun are properly accounted for, there is, in fact, a set of accurate, distant clocks, at angular distances of 10^o, 20^o, 30^o, and 40^o above and below the ecliptic.

**Of course there's several years' effort to explain and convince all that the +/- 1cm is acceptable; for the purpose of this thought experiment, please assume this huge effort has resulted in consensus.

 

[Nereid]

Martin Miller wrote: *SNIP
No one has found a single flaw. The Nereid-cited site does not have a single experiment which proves Einstein's sole claim of the invariance/isotropy of light's one-way, two-clock speed (where the clocks are in the same frame, as they must be).

Can you prove Einstein's claim?
Can you even show - even on paper - _how_ it could be proved?

No, you cannot.

Therefore, SR's sole basis (Einstein's claim) is unproved, and cannot be proved even on paper. Does this sound as if SR is a valid scientific theory?
*SNAP

At the risk of boring readers to death, the list contains:
- predictions from SR
- experiments to test those predictions
- results that are, within the experimental error limits, consistent with the predictions.

Are the predictions correctly derived (e.g. no mistakes in the math)?

Did the experimenters do what they said they did?

Are the results consistent with the predictions?

How hard is it to answer these questions?

Now to 'Einstein's sole claim of the invariance/isotropy of light's one-way, two-clock speed (where the clocks are in the same frame, as they must be)'. We could say "cool, let's go test it!", or we could say "assume this to be true for now; what predictions follow? how well do those predictions match experiments?"

Martin would like to do the former - OK, let's help Martin set up an experiment to do the test he wants.

Others are OK with the latter - and *all* such tests of SR pass with flying colours. Further, putting SR into QM produced QED (OK, this is a gross over-simplification), and QED, when tested in the lab, turns out to be the most accurate physics we've got, so far.

I hope your nose isn't too put out of joint Martin, but I personally am quite comfortable with proceeding to devise cosmological experiments, based on the assumption that SR (actually GR) is a valid description for the whole universe, at least from the time of last (neutrino) scattering.