Exploring the Paradoxical Light Clock Problem in Special Relativity Analysis

[altergnostic]

In a recent thread, a discussion developed on the subject of how we observe light and how this affects our understanding of SR. I called attention to the famous light clock diagrams:

http://home.comcast.net/~peter.m.brown/sr/image_gif/sr05-im-01.gif

In my view, the problem here is not with the resulting formulas, nor with time dilation, nor with the assumptions of SR, but that the light clock diagrams often lead to logical paradoxes, depending on how you interpret the problem, and I believe this happens because these diagrams are paradoxical to start with, and I would like to hear other opinions on this subject.

From where I stand, the problem with the diagrams is very simple:
You can't detect light at a distance. Detection of EM waves (or photons) in SR is strictly a local event. You can't be aware of light moving in any direction other than straight into your eyes (or detectors). So how can a non-local observer see those light rays? They are bouncing back and forth between the two mirrors, and anywhere else.

Suppose those are laser beams (so they don't radiate spherically). From the stationary frame, they go straight up and straight down, cross the distance between the mirrors at the speed of light and the time is proper time, so everything's fine. From the point of view of a distant observer, if you follow the logic in the diagrams, you could either deduce a change in the speed of light or time dilation. Since every experiment shows us that light always travels at c, we need time dilation to explain the angular light beams in the moving system.

First, if you assume that light was emitted at an angle, how does the emitter know the correct angle of emission?
Second, how can those light paths be part of anyone's data? If they reach the mirror, they don't reach the observer. We can assume that each mirror gives off a light signal every time it reflects the beam, and a local observer would measure the speed of the light as c, and no time dilation would be noticed. If a distant observer receives those light signals, than he must do the transforms with that light, and not the light that is being reflected inside the light clock. If you do that, than you will achieve the values for time dilation and length contraction before you diagram the light clock's beams, and when you finally get to diagram those beams, you will diagram them just like the local/stationary observer would.

Light would never be diagramed at those angles and nobody would even consider that light could go above c, or that it had to experience time dilation.

Isn't it weird to claim that the light has been time dilated? We use light to measure time dilation and length contraction on other things, so how can you time dilate the light itself?

Here is a very thorough and standard analysis:

http://physics.syr.edu/courses/modules/LIGHTCONE/LightClock/


I look for hearing different opinions on:

- How can we correctly diagram undetected light like that? How does the emitter adjust the angle of emission?
- Conversely, how can light be emitted at an angle if it's speed is not affected by the motion of the emitter?
- Shouldn't we apply SR transforms primarily with light that is actually observed and than use that information to diagram the interior of the light clock?
- Isn't diagraming the light clock in the moving frame like that illegal? Aren't those vectors purely imaginary?

 

[Mentz114]

In the rest frame of the emitter and detector the light travels vertically and hits the detector. This means that every observer will see that happening. It is an event where two worldlines intersect and is thus unchanged by any change of coordinates.

 

[nitsuj]

First, if you assume that light was emitted at an angle, how does the emitter know the correct angle of emission?

It's not a "for real" (proper?) angle. an angle is comparative, like speed, to your point of the diagrams being diagrams. Consider the light clock isn't necessarily moving at all. :uhh:

It's a longer distance/interval (chunk of spacetime) between events. ("ticking" of light clock, those events are what's measured)

For the light clock observer that longer distance is "contracted" (to the obvious straight up/down "path", more specifically the "shortest" length between events), if you measure the photon on an angle...I mean traveling a longer distance/interval between events.

From the other observers point of view, the longer distance/interval of the photon path is measured/observed as a purely length measurement.

Observation is an event, an event pin points length/time measurements to a specific location. The distance/interval between those ticking event locations of the light clock can be observed as any combination of length & time measurements.

apparently it all adds up to c, That is the continuum speeds along at c.
Is there such a thing as "proper angles". same idea as proper time/length ect? because of RoS, I'd guess there must be. It's time/length fusion:tongue2: Where the right(angle) place & the right(angle) time meet for an event. Well now I know what angles are. .

 

[Dale]

In my view, the problem here is not with the resulting formulas, nor with time dilation, nor with the assumptions of SR, but that the light clock diagrams often lead to logical paradoxes, depending on how you interpret the problem, and I believe this happens because these diagrams are paradoxical to start with

That would be truly amazing for a paradoxical start to yield correct formulas and correct values for time dilation. I think this is reason enough to be highly doubtful that the diagrams are paradoxical at all.

First, if you assume that light was emitted at an angle, how does the emitter know the correct angle of emission?

The emitter simply follows Maxwell's equations (or QED if quantum effects are important for the emitter). The laws of Maxwell and QED are such that if the light follows a straight path in one frame then it must follow a diagonal path in another frame.

Second, how can those light paths be part of anyone's data?

We have good experimental reasons to believe Maxwell's equations and QED. They predict those light paths. While a particular detector may not collect data from each point along the way, the data that any detector does collect is consistent with the light paths as drawn. So why not draw them? Should we just act surprised each time that some data is collected and pretend that we don't have a good understanding of the laws that governed that data?

 

[bahamagreen]

It may help to imagine that the light clock is "at rest" and that you as the remote observer are passing by it... what would the diagram look like then?
The deeper question might be why when setting out to measure time and space for things moving fast or at a distance that we would choose to use clocks and rulers when these are observed to change their times and lengths when employed in those situations... (as if we had a choice to use something else)?

 

[altergnostic]

That would be truly amazing for a paradoxical start to yield correct formulas and correct values for time dilation. I think this is reason enough to be highly doubtful that the diagrams are paradoxical at all.

Yes, it is truly amazing, but on the other hand, we already knew what we were trying to find, this is just one more way to visualize the problem and derive time dilation. A diagram full of logical holes can still yield the correct results, if you know the results from the start and build everything around that.

...if the light follows a straight path in one frame then it must follow a diagonal path in another frame.

This seems to contradict one of SR postulates, that the speed (and logically direction) of light is unaffected by the motion of the emitter.

The only workaround I see is if we always take the frame where light is detected as the stationary frame, as bahamagreen sugested. This is what actually happens in real life. A detector can't have motion relative to itself when it detects light. It is also what happens when we pick a rest frame in other SR thought problems, we take our frame as the rest frame and observe the relative positions and times of other things through light. We do the transforms with the light that reaches us from events, not with unseen light that never reaches our eyes.

[/QUOTE]While a particular detector may not collect data from each point along the way, the data that any detector does collect is consistent with the light paths as drawn.[/QUOTE]

Detections happen locally, just like in the diagram to the left. That's how a local observer would see the light paths, straight up and down. Are you saying that a local observer, in the stationary frame, would draw the light paths diagonally, as if he was on the moving frame, like the diagram to the right?

An observer in relative motion wrt the light clock would not even see those light beams, so how could he diagram them? Would't he need to receive some sort of signal that tells him when a beam has been reflected from a mirror? And if so, wouldn't you have to do the transforms with these light signals in the first place? Than we would find the coordinates as seen from the light clock's frame and the light beams would be diagramed just like in the stationary frame. The final results would be exactly the same, you would still find the same value for time dilation and everything, but isn't it way more consistent and logical? Doesn't it bother you that we are trying to diagram light as "seen" by an "observer" that isn't even aware of those light beams?

if you take the light clock system as the stationary frame, everything fits. This tells me that light is always a local event, and we should only be allowed to diagram or visualize light from a stationary frame. This would also explain why Einstein said that the speed of light was unaffected by the motion of the emitter in his 1905 paper, and only later introduced the receiver.

 

[Dale]

This seems to contradict one of SR postulates, that the speed (and logically direction) of light is unaffected by the motion of the emitter.

Nonsense. Speed is the magnitude of velocity, not its direction. There is no sense in which a postulate about the speed of light logically implies anything to do with its direction.

The only workaround I see is if we always take the frame where light is detected as the stationary frame, as bahamagreen sugested. This is what actually happens in real life. A detector can't have motion relative to itself when it detects light.

You (and others) seem stuck on this schizophrenic idea that events can happen in one frame and not in another. A reference frame is a mathematical construct, a simple way of labeling events and directions, a coordinate system. If a detector detects light in one frame then it must detect light in all frames. Simply changing your arbitrary mathematical labeling cannot change the physical fact that the light was detected.

Do you agree with that?

 

[harrylin]

In a recent thread, a discussion developed on the subject of how we observe light and how this affects our understanding of SR. I called attention to the famous light clock diagrams:
[..]
- How can we correctly diagram undetected light like that? How does the emitter adjust the angle of emission?
- [..]

On the first point ("undetected") I gave a counter example in that recent thread as follows:

Once more: a cloud chamber scatters light over the whole trajectory. It is technically feasible to observe the diagram on the right with an array of close-up lateral detectors that are in rest in the "stationary" frame, and the same can also be captured far away with a CMOS camera that is mounted "in rest" in the "stationary" frame. Both diagrams are equally observable, with real data.
As a matter of fact, SR was first of all concerned with comparing real data from real measurements, and that diagram illustrates what according to SR really can be measured.
- https://www.physicsforums.com/showthread.php?t=620279&page=2

And about the light angle, this was discussed for example here:
https://www.physicsforums.com/showthread.php?t=574757

 

[robinpike]

I look for hearing different opinions on:
...
How can we correctly diagram undetected light like that? How does the emitter adjust the angle of emission?
...

I too would like to see the explanation as to how matter adjusts the angle of the emitted light when the light clock is made to move?

For reference, the answer I have been given previously, is that: As far as the light clock is concerned, once it is at a steady speed, it is not moving - so the question of how the moving laser emits the light at a forward angle is irrelevant.

To me this doesn't explain how it works. (It's a bit like answering the question of: How does gravity work? And replying: Gravity works by pulling you towards the ground.)

 

[harrylin]

I too would like to see the explanation as to how matter adjusts the angle of the emitted light when the light clock is made to move?[..]

Explained in the second link of post #8

 

[nitsuj]

I too would like to see the explanation as to how matter adjusts the angle of the emitted light when the light clock is made to move?

For reference, the answer I have been given previously, is that: As far as the light clock is concerned, once it is at a steady speed, it is not moving - so the question of how the moving laser emits the light at a forward angle is irrelevant.

That's a a decent take on it.

But as far as the angle thing goes, you can't force your observation into another FoR.

For example just because your reality shows an angled path, doesn't mean that's the reality for the light clock observer. I believe they call this "relativity", it's symmetrical just like the light clock demonstrates (when including c postulate).

 

[Dale]

I too would like to see the explanation as to how matter adjusts the angle of the emitted light when the light clock is made to move?

The explanation is Maxwell's equations. Write down the equation in one frame, transform to another frame, and note two things:
1) the transformed wave moves at an angle
2) the transformed wave is a solution to Maxwell's equations

Regardless of the reference frame, the emitter simply follows Maxwell's equations.

 

[nitsuj]

To me this doesn't explain how it works. (It's a bit like answering the question of: How does gravity work? And replying: Gravity works by pulling you towards the ground.)

There is nothing there "working". It is a different perspective of the two events (photon hitting top/bottom of mirror).

This is a particularly big hurdle though; to accept two separate physical realities of the same events. Try to note the difference between the perspectives, at same time ignore the idea of light emitting at an angle as being the part of the physical events.

Get the picture in your head that the only difference between the frames is an increase in the path/distance/interval/chunk of spacetime the photon travels. You already know that path/distance/interval/chunk of spacetime can be observed either as time or length, depending on "perspective".

The angle is just a comparative result, while physical, is not a change in the physical light clock itself.

 

[harrylin]

On the first point ("undetected") I gave a counter example in that recent thread as follows:

Once more: a cloud chamber scatters light over the whole trajectory. It is technically feasible to observe the diagram on the right with an array of close-up lateral detectors that are in rest in the "stationary" frame, and the same can also be captured far away with a CMOS camera that is mounted "in rest" in the "stationary" frame. Both diagrams are equally observable, with real data.
As a matter of fact, SR was first of all concerned with comparing real data from real measurements, and that diagram illustrates what according to SR really can be measured.
- https://www.physicsforums.com/showthread.php?t=620279&page=2

And about the light angle, this was discussed for example here:
https://www.physicsforums.com/showthread.php?t=574757

It suddenly strikes me that of course this cloud chamber with glass walls can be used to better clarify that angle issue, as it was meant to bring home that this is exactly what necessarily must be measured.

Take a light ray going straight up from bottom to top, as depicted on the left, but in a cloud chamber with glass walls, and to which we attach the label S'; necessarily scattered light from halfway up (at Y=0.5L) is also at the same horizontal position in S'.*

However, what if this cloud chamber S' is moving at very high speed to the right as observed by a stationary system S, as depicted in the sketch on the right?

The scattering water molecules at the bottom in S' will be detected at for example x=0 in system S.
However, while the light moves up in S', S' moves to the right. Necessarily the scattering water molecules at 0.5L in S' are not at x=0 in S, but are slightly more to the right. And the scattered light at the top is even more to the right.

IOW, by geometric necessity this is what must be measured in S.

[EDIT:] I did not discuss here simultaneity; however, I think that for this picture that can only make a numerical difference, and not a qualitative one. It corresponds to physical reality (absolute, agreed by all) that S' moves like that relatively to S, with the light ray also progressing like that relative to the detectors of each. The sequence of local events as well as their respective locations is not an issue here. As a matter of fact, those are literally trajectories that can be traced simultaneously on photographic plates in S and in S'.

* Technically the reference system S' corresponds to that cloud chamber but with infinite extensions in all directions

 

[pervect]

In my view, the problem here is not with the resulting formulas, nor with time dilation, nor with the assumptions of SR, but that the light clock diagrams often lead to logical paradoxes, depending on how you interpret the problem, and I believe this happens because these diagrams are paradoxical to start with, and I would like to hear other opinions on this subject.

There's nothing particularly "paradoxical" about light clock diagrams. However, because they omit any discussion of the relativity of simultaneity, they aren't the whole story.

Since not understanding the relativity of simultaneity is responsible for (at a guess) at least 90% if not more of the problems people have with understanding relativity, and most likely the feature that leads you to believe that they are "paradoxical", it's an important omission.

 

[robphy]

There's nothing particularly "paradoxical" about light clock diagrams. However, because they omit any discussion of the relativity of simultaneity, they aren't the whole story.

Yes, that's one of the reasons why I made the animated spacetime diagrams
in the link provided by OP.

I think this animation might help with the discussion of photons associated with the moving clock:
http://physics.syr.edu/courses/modules/LIGHTCONE/LightClock/VisualizingProperTime-y-pair-A-with-photons.avi

 

[altergnostic]

Nonsense. Speed is the magnitude of velocity, not its direction.
If a detector detects light in one frame then it must detect light in all frames. Simply changing your arbitrary mathematical labeling cannot change the physical fact that the light was detected.

Do you agree with that?

Yes. If a detector detects light that detection is real. It is real because light has reached the detector, it is detected by contact. This detection is not an observation from a distant observer. A distant (moving or not) observer can't see light that is in contact with the mirrors, it can only see light that is in contact with himself.

I am not saying that the detection of light is not real from a distant (moving, in this particular case) observer, i am saying that this distant observer doesn't know that the light has been detected, he doesn't know where the light is, he is not receiving those light beams, he doesn't have any data from those light beams.

 

[altergnostic]

Take a light ray going straight up from bottom to top, as depicted on the left, but in a cloud chamber with glass walls, and to which we attach the label S'; necessarily scattered light from halfway up (at Y=0.5L) is also at the same horizontal position in S'.*

It is at the same horizontal y position at the moment of emission. But when it is detected some x distance away, the beam has moved in the y direction. So you will have the same y coordinate, but not the same t.

But more importantly, if the deterctors are stationary relative to the emitters, they will diagram the light as in the stationary frame. They will measure the distance between the mirrors in the stationary frame and diagram light in a straight path between them. The mirrors will not change position with time, and the time light takes to move from one mirror to the other will be in agreement with other onboard clocks, and the speed of light will remain constant.

Now, if your detectors are receiving scattered light from a moving frame, you must apply transforms to find the distance between mirrors and the time in the primed system using the the information brought by the scattered light that actually reaches you. Only after that you will have the coordinates of the primed system (the light clock) and can begin to correctly diagram those light paths, and as a matter of fact, if you do this, they will be diagramed just like the image on the left, the stationary system. The diagram to the right, with the diagonal beams, will not be diagramed that way using real data.

If light is reflected perpendicular to the mirror in one frame, it must do so in all frames, and if you apply SR transforms using real data, that is exactly what you will find.

This diagram yields the correct numbers because they are arrived at by applying transforms unrealistically, after the "fact". You are applying the same transforms you would in any SR problem. But the situation in the diagram in not a real, possible, situation. You can't observe light detected at a distance, because you have not detected that light, and you have no information of that light. So you can't diagram that light. The mathematical results are the same, but the diagram is falacious, it assumes you can see light that you can't, and it puts light moving at an angle, in contradiction with any real data. No detectors on the mirrors would detect light being reflected at an angle. If they did, the light would miss the other mirror.

 

[Mentz114]

The mathematical results are the same, but the diagram is falacious, it assumes you can see light that you can't, and it puts light moving at an angle, in contradiction with any real data. No detectors on the mirrors would detect light being reflected at an angle.

If a series of detectors is placed in a vertical line so a light on the detector is switched on when the beam passes, then the line of lights will not be vertical when observed ( photographed ?) by an observer in motion wrt the receiver/emitter.

Isn't this the same problem that Einstein treated directly, the passenger on the moving train drops a stone and sees it fall vertically, but the observer on the platform sees the stone move in a parabola.

 

[nitsuj]

1.) Principle of Relativity
2.) Invariance of c

As #1 says, yes the the physics are the same in both frames, this would have to include the light does not leave the emitter at an angle.

As #2 says, the light must leave the emitter at an angle, for c to be calculated as a constant.

While those two interpretations are at odds, it isn't a paradox. Those are perspectives from two different FoRs.

Considering them both as simultaneous realities isn't right, they are mutually exclusive perspectives (and physical realities). Would you ever try to imagine day & night at the same time and place as a reality?

 

[Austin0]

Quote by harrylin

Take a light ray going straight up from bottom to top, as depicted on the left, but in a cloud chamber with glass walls, and to which we attach the label S'; necessarily scattered light from halfway up (at Y=0.5L) is also at the same horizontal position in S'.*

It is at the same horizontal y position at the moment of emission. But when it is detected some x distance away, the beam has moved in the y direction. So you will have the same y coordinate, but not the same t.

But more importantly, if the deterctors are stationary relative to the emitters, they will diagram the light as in the stationary frame. They will measure the distance between the mirrors in the stationary frame and diagram light in a straight path between them. The mirrors will not change position with time, and the time light takes to move from one mirror to the other will be in agreement with other onboard clocks, and the speed of light will remain constant.

Now, if your detectors are receiving scattered light from a moving frame, you must apply transforms to find the distance between mirrors and the time in the primed system using the the information brought by the scattered light that actually reaches you. Only after that you will have the coordinates of the primed system (the light clock) and can begin to correctly diagram those light paths, and as a matter of fact, if you do this, they will be diagramed just like the image on the left, the stationary system. The diagram to the right, with the diagonal beams, will not be diagramed that way using real data.

If light is reflected perpendicular to the mirror in one frame, it must do so in all frames, and if you apply SR transforms using real data, that is exactly what you will find.

This diagram yields the correct numbers because they are arrived at by applying transforms unrealistically, after the "fact". You are applying the same transforms you would in any SR problem. But the situation in the diagram in not a real, possible, situation. You can't observe light detected at a distance, because you have not detected that light, and you have no information of that light. So you can't diagram that light. The mathematical results are the same, but the diagram is fallacious, it assumes you can see light that you can't, and it puts light moving at an angle, in contradiction with any real data. No detectors on the mirrors would detect light being reflected at an angle. If they did, the light would miss the other mirror.

HarryLin's view is perfectly applicable. It provides a means to directly track the path in another frame. No transformations required. A simple direct chart of the observed path provides the zig zag path. Transformation comes into play , specifically aberration, to derive the emission angle (90 deg.) in the clock frame from the observed angle in the lab frame (some angle less than 90 deg.).

To call such a chart fallacious is no more reasonable than to call any frame dependent measurements false. It is somewhat unclear on the concept... imo.

The basic physical principle involved is the intrinsic properties of light propagation.
That light conserves the forward momentum of the source. If this principle is itself false then the first postulate of SR fails as it is obvious that a simple laser would provide the means to determine absolute motion through deviation of a transverse beam from the point of aim.
is this what you think??

 

[Dale]

Yes. If a detector detects light that detection is real. It is real because light has reached the detector, it is detected by contact.

OK, since you agree that the detection is real in all frames, then every frame is justified in assigning coordinates to the event. Everything else follows from that and Maxwells laws (or even just the subset used for geometric optics).

 

[harrylin]

HarryLin's view is perfectly applicable. It provides a means to directly track the path in another frame. No transformations required.

Yes indeed. However, altergnostic seems to misunderstand what I meant:

It is at the same horizontal y position at the moment of emission. But when it is detected some x distance away, the beam has moved in the y direction. So you will have the same y coordinate, but not the same t.

That diagram isn't about "t"; it can be literally interpreted as being about light traces on S' and S, in the way that I explained.

But more importantly, if the deterctors are stationary relative to the emitters, they will diagram the light as in the stationary frame. [..]

That is the diagram of S' on the left; so you agree with that?

They will measure the distance between the mirrors in the stationary frame and diagram light in a straight path between them. Now, if your detectors are receiving scattered light from a moving frame, you must apply transforms to find the distance between mirrors [..]

:uhh: I'm afraid that you completely misunderstood what I meant with lateral detectors. They have nothing to do with the mirrors, in fact the mirror can be left out of my explanation. Forget about the mirror, I only discussed the light beam as it is going upwards. The detectors measure very close to the light beam the light that is scattered in the + and - z directions, which go in / come out of the picture.

And as Austin indicated, forget for the moment about transformations; I am instead talking about measurements, in this case very literal observations such as the traces on two photographic plates. On top of that, also the direction of scattering sequences (qualitative "time") can be determined without reference to a reference frame by exposing one additional plate in each system, each obscured with a shutter before the light has reached the top. As a result one can determine the time sequence (= the light propagation is from bottom to top, and not from top to bottom).

[..] The diagram to the right, with the diagonal beams, will not be diagramed that way using real data.

To the contrary, the two photographic plates will show traces exactly like on the diagrams on the left and on the right. For a useful discussion, you should comment step by step on my analysis of what will be observed on the two photographic plates. In fact, you stopped your analysis at the crucial point and instead started to talk about transformations. Please no transformations, just a single-system analysis of the scatter traces from the light beam on the two photographic plates!

 

[altergnostic]

OK, since you agree that the detection is real in all frames, then every frame is justified in assigning coordinates to the event. Everything else follows from that and Maxwells laws (or even just the subset used for geometric optics).

It is real in all frames, but an "observer" can't assign coordinats of events he is not aware of. Those light paths are not seen by a distant observer, so he can't diagram them. If you want him to know the times of the reflections, you have to send him a signal at each reflection event. As it is, the light clock diagram assumes he can actually see those light paths, which easily leads to paradoxes. Also notice that if he really sees those light paths at an angle he will actually measure light going faster than c, and only after the transforms will he have a dilated time and the speed of light normalized to c. But SR and experiment shows that no observer can measure light going faster (or slower) than c. That's why he should use light that reaches him directly, which he will directly measure going at c, to do the transforms in the first place.

 

[altergnostic]

That is the diagram of S' on the left; so you agree with that?

Yes.

The detectors measure very close to the light beam the light that is scattered in the + and - z directions, which go in / come out of the picture.

The point is that the detectors don't measure a moving frame, they would diagram no angles.

And as Austin indicated, forget for the moment about transformations; I am instead talking about measurements, in this case very literal observations such as the traces on two photographic plates. On top of that, also the direction of scattering sequences (qualitative "time") can be determined without reference to a reference frame by exposing one additional plate in each system, each obscured with a shutter before the light has reached the top. As a result one can determine the time sequence (= the light propagation is from bottom to top, and not from top to bottom).

To the contrary, the two photographic plates will show traces exactly like on the diagrams on the left and on the right. For a useful discussion, you should comment step by step on my analysis of what will be observed on the two photographic plates. In fact, you stopped your analysis at the crucial point and instead started to talk about transformations. Please no transformations, just a single-system analysis of the scatter traces from the light beam on the two photographic plates!

Ok. Where are these plates? If they are on the stationary system, they will measure the path straight up. If they are moving relative to the beam, they will draw a straight vertical line on each (considering a long exposure), but the scattered light will reach each plate at a right angle at each moment of detection. From the point of view of the moving plates, the path of the beam will seem to be receding the plate as time elapses, so it would directly measure a longer period of time for the beam to reach the top, since it will take longer for the scattered light to cross each subsequent distance between the beam and the plate, so this setup would conclude a longer distance and a longer time directly, which is what i mean when i say that you need to do the transforms with the light that reaches you (the plates) directly. That would be a correct diagram indeed, and it is because this setup is correct that the setup of the original diagram fails.

Do you see my point now?

 

[altergnostic]

The basic physical principle involved is the intrinsic properties of light propagation.
That light conserves the forward momentum of the source. If this principle is itself false then the first postulate of SR fails as it is obvious that a simple laser would provide the means to determine absolute motion through deviation of a transverse beam from the point of aim.
is this what you think??

No, i have no quarrel with the postulates if SR. What i dislike is the diagrams assuption that you can even see those beams. And on the fundamental issue, what I claim is that you always observe light from a stationary frame, relative to the light you receive directly. I think that's why the speed of light is independent of the motion of source / receiver. If not, you could measure light at a different speed than c. That's why assuming that you can see a distant light path moving transversally to your line of sight (not reaching your eyes) is bad diagraming. And that's why harrylin's proposition can be properly diagramed. (See my previous answer to harrylin.)

You see, my problem is not with theory, but with the assumptions of these light clock diagrams.

What do you think?

 

[altergnostic]

Just a curiosity, in his 1905 paper, Einstein said that the speed of light is the same regardless of the motion of the source. Only later he included the motion of receiver. The first claim is as complete as the second, and i actually think it is better: the speed of light is independent from the motion of the source, and the receiver is always stationary wrt the light it observes. Seeing light is strictly a local event, and you are never in motion relative to yourself. Seeing light is done by contact, and the only thing that matters is the relative velocity between the receiver and the light. Since you are stationary relative to yourself, and you can't assign a coordinate system to light itself, you must be stationary relative to detected light (detected by you).

I think that's what Einstein had in mind when he said that the speed of light is independent from the speed of the source and left the receiver out of that sentence. He had to include it later to prevent people from adding the speed of receiver to the speed of light, but he didn't have to. He had already proved that you can always choose to be in a rest frame in SR, and you can't assign any speed to light other than c.

 

[bahamagreen]

The OP's language in presenting the paradox is similar to that found here. I'm guessing the OP has read this... others might do so as well to get a sense of the proposed problem.

 

[altergnostic]

The OP's language in presenting the paradox is similar to that found here. I'm guessing the OP has read this... others might do so as well to get a sense of the proposed problem.

Indeed i have read it, but I started to take this seriously after a discussion on the twin paradox with an author of this site:
www.twinparadox.net

I disagreed with his interpretation and I started to realize some of the arguments I didn't agree with were related to the fact that sometimes he tried to solve SR problems visualizing light at a distance, so I started to look at this seriously.

I agree with harrylin's setup, and I think we should continue the discussion from there, I don't want to discuss that paper since it is someone else's reasoning, not mine. I don't know what logic took miles mathis to this problem, but I know how I arrived at my own conclusions and that's what I want to discuss.

You may read that paper to see another take on this problem, but I won't discuss the paper itself (I think that would actually be a distraction), at least not before we have exhausted the arguments already posted.

 

[altergnostic]

If a series of detectors is placed in a vertical line so a light on the detector is switched on when the beam passes, then the line of lights will not be vertical when observed ( photographed ?) by an observer in motion wrt the receiver/emitter.

Isn't this the same problem that Einstein treated directly, the passenger on the moving train drops a stone and sees it fall vertically, but the observer on the platform sees the stone move in a parabola.

I missed your post! Yes, i think it is pretty much the same problem as the train. Notice, though, that you must use the light directly reaching the observer from the detectors, and then he would measure a larger distance and a longer time span between reflections against measurements in the stationary system, and you must apply the transforms with those light signals. What I mean is that you would never consider light to move faster than c, even without experimental proof of it, the distance and the time increase proportionally and you don't even need to calculate time dilation on the reflecting beam to calculate light going at c.

Also notice you can't see the reflecting beams directly as the original problem assumes, which is the main point and the main source of my discomfort regarding the light clock diagram, since i have seen too many people trying to use distant light to incorrectly solve SR problems, not realising distant light can't be used becaused it hasn't been observed. It is imaginary data.

 

[marty1]

My difficulty with that drawing is that the moving emitter is giving sideways momentum to the photon. Is that possible? Shouldn't the photon leave the emitter and miss the mirror on the other side or miss the detector on the way back?

 

[bahamagreen]

OK... so I think what you are saying is that the light clock diagram must be viewed as an inferred and possibly flawed interpretation of what is remotely happening, not an existential representation.

If so, and the paradox stems from the current diagram, can the diagram be saved by altering it or is there no proper diagram possible for representing the remote behavior of the light path?

 

[marty1]

I think this diagram and the point it is trying to make can be saved by changing the direction of motion 90o

 

[Austin0]

It is real in all frames, but an "observer" can't assign coordinats of events he is not aware of. Those light paths are not seen by a distant observer, so he can't diagram them. If you want him to know the times of the reflections, you have to send him a signal at each reflection event. As it is, the light clock diagram assumes he can actually see those light paths, which easily leads to paradoxes. Also notice that if he really sees those light paths at an angle he will actually measure light going faster than c, and only after the transforms will he have a dilated time and the speed of light normalized to c. But SR and experiment shows that no observer can measure light going faster (or slower) than c. That's why he should use light that reaches him directly, which he will directly measure going at c, to do the transforms in the first place.

This is not correct. The observed diagonal paths in HarryLin's proposed set up would always produce c as a direct calculation of the distance and detection times (D/dt5) in the LAB frame.
No transformation whatsoever required in that frame.

Exactly as would be the case if the mirror was removed and the photon continued on to a detector at rest in the lab frame.

 

[marty1]

Please explain how photons are never given momentum from the source but a change in frequency due to the momentum of the source yet can be given the lateral momentum of the sideways motion of the source to cause them to continue across the moving lab and still hit the mirror and come back to the detector.