harrylin said:
SR asserts in essence the contrary of what you here assert, just as phrased by Einstein in the citation in my last post: it is assumed that effectively the same "patch" of light reaches both observers. That according to SR this is in practice not exactly true (and according to QM not at all), is totally irrelevant. Einstein assumes that both observers see the same light flash, and this is not a problem because according to SR the fact of observation cannot affect the speed of light before the observation. It's a mix-up between QM and SR to think that in SR the fact of observation changes what is observed in an essential way.
SR states the opposite? All data in a real SR problem has to be either given information or data brought by light, reaching one observer. You can't directly use data that another observer is receiving at the same time he is receiving it. That's why you need transforms in the first place, because two observers have different perspectives, they are at different positions, times and velocities relative to the point of emission/reflection (i.e.: the back of the train). Hence, they may describe the coordinates in space and time of the same event in different ways.
We may assume they see the same light only for simplicity, as Einstein often tries to do. It is obvious that this assumption has no grounds on any real situation, or are you saying that if light is emitted at the midpoint between us, when we see it we see with the same patch of light/photons? This is exactly the same situation as if two laser beams are sent in opposite directions, one towards you, another towards me. I see one, you see the other. I don't see the one who reaches your eyes, and vice-versa. If you saw the same beam as I do, we are either at the same place or the beam has been reflected from me back to you, and then we wouldn't see the beam at the same time.
And where did I say that the act of observing changes what is observed? Nowhere i had this in mind. That's not SR.
Strictly speaking one cannot observe a beam through vacuum; in such a case one can only observe the detections, for example with sensors at the mirror and the light source. But one can also have stationary detectors on both sides with which one registers the "ticks" immediately when they arrive, either by partly transparent mirrors or by means of relaying the detection. Such things are merely technological issues; for SR those are valid detections of the bouncing light in the "stationary" system.
Moreover, one can do the same in a cloud chamber, and then detect the scattered light simultaneously in both systems, by means of very nearby detectors. For SR this isn't an issue, and to think that it could matter reveals a misunderstanding of SR.
All these examples show no contradiction to what I have been saying. In a cloud chamber, you need to scatter light so that two detectors can detect anything, meaning, you need light to move towards them. All detections are valid, of course, but they occur locally, and that's why they are described from a stationary system, as you yourself pointed out, just like in the left diagram above, but never like the one on the right, that's supposedly a diagram made by a distant observer in relative motion, but how can he possibly have seen those beams? And if he uses the detections, which are done in a "stationary" system, why on Earth would he try to diagram them like "seen" from a moving system? There's no possible way that you can diagram that triangle with real data. The detections are made locally, in a stationary system, they can only be diagrammed from a stationary system. There's no light detection at a distance.
If you don't see the relevance of this to SR, just look at the diagram I posted. That diagram is a mistake. It comes from wrong assumptions and poor visualization, as most SR paradoxes and misunderstandings come from. People have a hard time visualizing light, and they try to do it in ways they shouldn't. Strictly speaking, you don't observe light, you observe with light. Einstein was very careful to tell us that light is a special case. Diagraming unseen light moving at an angle due to relative motion is forbidden. First because light would have to go faster than c, since it has to cover a larger distance in the same time than in the stationary system. Second, because it is unrealistic, you have no way whatsoever to see that light, you can only guess that it traveled at those angles, and it is a wrong guess, because you didn't do the SR transforms on any light that was reaching you. Again, if you were sent light signals at each reflection, you'd need to apply the transforms on those light signals, and you would end up with a diagram just like if you diagrammed it locally, in the stationary system.
Note also that in the train example, both reference systems are "close-up" to both detections.
It doesn't matter how close they are, they still see different patches / beams / photons. This is irrelevant.
I was afraid of something like that.
- The "fundamental issue" that you may be thinking of, is that distant simultaneity cannot be detected except as a function of definition. However:
- There is nothing that binds "stationary system" to "nearby detection", or "moving system" to "far away detection".
Such arguments are therefore only helping a
misunderstanding of how the speed of light can be constant for everyone.
Let me clarify what i mean by distant observer. I am not claiming that distance alone causes relativistic effects (although sometimes it does) and I am not relating these distant observers to any specific type of frame, stationary or otherwise. I use distant in a non-relativistic sense. A distant observer is just another observer, distant is just to emphasize that it doesn't receive the same light as you do. Very fundamental. It is the fact that you can't directly see the same light as someone else, even if they are "close-up", that allows for SR operations in the first place. You have to transform your observed coordinates into the other system's (locally) observed coordinates.
I am not talking about simultaneity, although it is directly related to what I'm saying. I am talking about a more correct understanding of light and how to diagram it, or visualize it, properly. You don't even need equations to understand thought problems like the embankment/train. You need them to get the numbers, but the equations don't explain anything. That's why Einstein was so lengthy in logical explanations, he tried to explain SR in as many different ways as he could, without math, so that anyone could understand why you needed the Lorentz transforms in the first place. I am presenting no contradiction to Einstein, all I am saying is that you have to understand exactly how we interact with light before you go deep into SR, otherwise you end up with bad diagrams and visualizations like above.
No, that is not correct. Instead, I cited how he used the terminology of relative motion.
