83
$\begingroup$

I often hear the story of how Einstein came up to the conclusion that time would slow down the faster you move, because the speed of light has to remain the same.

My question is, how did Einstein know that measuring the speed of light wouldn't be affected by the speed at which you are moving. Was this common knowledge already before Einstein published his paper on special relativity? If not, what led him to that conclusion?

$\endgroup$
  • 6
    $\begingroup$ See also this question: Did Einstein know about the Michelson-Morley experiment?. $\endgroup$ – rghome Jul 2 at 8:41
  • 4
    $\begingroup$ This would be a better fit at hsm.SE. AFAICT, Einstein's thought processes are not really known, and his recollections are sometimes unreliable when compared with historical circumstances. $\endgroup$ – Ben Crowell Jul 2 at 12:04
  • 2
    $\begingroup$ @BenCrowell on the other hand, the papers he wrote at the time are exceptionally clear $\endgroup$ – Nathaniel Jul 2 at 18:58
  • $\begingroup$ Consider pitt.edu/~jdnorton/Goodies/Chasing_the_light $\endgroup$ – Roman Odaisky Jul 2 at 22:24
  • 2
    $\begingroup$ I will just throw in my two pennies worth: strictly speaking, he didn't 'know' it was a constant. He postulated that it had to be correct in order for classical mechanics to continue functioning in the setting he was discussing. There was logically no other alternative to postulating that the speed of light is a constant. $\endgroup$ – Tom Jul 3 at 12:04
107
$\begingroup$

Besides Michelson and Morley experimental results, Einstein also considered the theoretical aspects. It can be derived from Maxwell's equations that the speed at which electromagnetic waves travel is: $c=\left(\epsilon_{0}\mu_{0}\right)^{-1/2}$. Since light is an electromagnetic wave, that means that the speed of light is equal to the speed of the electromagnetic waves. $\epsilon_{0}$ and $\mu_{0}$ are properties of the vacuum and are constants, so $c$ will also be a constant. Thus from Maxwell's theory of electromagnetism alone we can already see that the speed of light in vacuum should be constant.

On the other hand, Galilean invariance tells us that the laws of motion have the same form in all inertial frames. There is no special inertial frame (as far as Newton's laws are concerned).

Another key element here is Galilean transformation, which was the tool used for transforming from one inertial frame to another. It can be easily seen that considering the first two elements to be valid:

  • Maxwell's theory of electromagnetism - speed of light is constant
  • Galilean invariance - the laws of motion have the same form in all inertial frames

means that we can no longer apply the Galilean transformation, because otherwise we will get a contradiction. Thus at least one of these three "key elements" must be wrong.

  • Maxwell's theory of electromagnetism - speed of light is constant
  • Galilean invariance - the laws of motion have the same form in all inertial frames
  • Galilean transformation

It turned out that the last one (Galilean transformation) was wrong. Einstein considered the first two correct and built the special theory of relativity. The correct transformation from one inertial frame to another, in the assumption of the validity of the Maxwell's theory and Galilean invariance, turns out to be Lorentz transformation . It is nice to check that the Lorentz transformation does indeed reduce to the Galilean transformation in the $v\ll c$ limit. That's why, in a sense, Galilean transformation is not wrong, but rather incomplete or a particular case. We can say that Galilean transformation needed to be generalized, and this was acomplished by introducing the invariance of the speed of light and maintaining the Galilean invariance.

How do Maxwell's equations predict that the speed of light is constant

Maxwell's equations in differential form: $$\tag{1}\nabla\cdot \mathbf{E}=\frac{\rho}{\epsilon_{0}}\label{1}$$ $$\tag{2}\nabla\cdot \mathbf{B}=0\label{2}$$ $$\tag{3}\nabla\times\mathbf{E}=-\frac{\partial \mathbf{B}}{\partial t}\label{3}$$ $$\tag{4}\nabla\times \mathbf{B}=\mu_{0}\mathbf{J}+\mu_{0}\epsilon_{0}\frac{\partial \mathbf{E}}{\partial t}\label{4}$$

We can try to derive a wave equation in vacuum. Since we are considering the vacuum, we do not have charge densities, so equation ($\ref{1}$) becomes: $$\tag{5}\nabla\cdot \mathbf{E}=0\label{5}$$ In vacuum we do not have current densities either, so equation ($\ref{4}$) becomes: $$\tag{6}\nabla\times \mathbf{B}=\mu_{0}\epsilon_{0}\frac{\partial \mathbf{E}}{\partial t}\label{6}$$

Now if we apply the curl to equation ($\ref{3}$), we get: $$\tag{7}\nabla\times\left(\nabla\times\mathbf{E}\right)=-\frac{\partial}{\partial t}\left(\nabla\times\mathbf{B}\right)\label{7}$$

We can use vector identity to evaluate the LHS of equation ($\ref{7}$): $$\tag{8}\nabla\times\left(\nabla\times\mathbf{E}\right)=\nabla\left(\underbrace{\nabla\cdot\mathbf{E}}_{=0}\right)-\nabla^2\mathbf{E}\label{8}$$ $$\tag{9}\nabla\times\left(\nabla\times\mathbf{E}\right)=-\nabla^2\mathbf{E}\label{9}$$

For the RHS of equation ($\ref{7}$), we can replace $\nabla\times\mathbf{B}$ with the expression we have from equation ($\ref{6}$): $$\tag{10}-\frac{\partial}{\partial t}\left(\nabla\times\mathbf{B}\right)=-\frac{\partial}{\partial t}\left(\mu_{0}\epsilon_{0}\frac{\partial \mathbf{E}}{\partial t}\right)=-\mu_{0}\epsilon_{0}\frac{\partial^2 \mathbf{E}}{\partial t^2}\label{10}$$ Putting all together: $$\tag{11}-\nabla^2\mathbf{E}=-\mu_{0}\epsilon_{0}\frac{\partial^2 \mathbf{E}}{\partial t^2}\label{11}$$ $$\tag{12}\nabla^2\mathbf{E}-\mu_{0}\epsilon_{0}\frac{\partial^2 \mathbf{E}}{\partial t^2}\label{12}=0$$ The general form of a wave equation is: $$\tag{13}\nabla^2\mathbf{\Psi}-\frac{1}{v^2}\frac{\partial^2 \mathbf{\Psi}}{\partial t^2}\label{13}=0$$ where $v$ is the velocity of the wave. Equation ($\ref{12}$) decribes an electromagnetic wave moving with velocity $v=\frac{1}{\sqrt{\epsilon_{0}\mu_{0}}}$. Since light is an electromagnetic wave, that means that light is also propagating at this speed in vacuum. And since both $\epsilon_{0}$ and $\mu_{0}$ are constant, that means that $\frac{1}{\sqrt{\epsilon_{0}\mu_{0}}}$ is also a constant. Hence light moves at a constant speed in vacuum.

$\endgroup$
  • 4
    $\begingroup$ How does maxwell equation tell you speed of light has to be constant? You would have to know the transformatiln properties in different frames of E and B. Its definitely not that simple. $\endgroup$ – lalala Jul 2 at 16:17
  • 13
    $\begingroup$ Your three points are right, but the derivation from Maxwell eqns sort of misses the point. The maths is correct, but the whole issue is the nature of "vacuum". Saying "and there you are" at the end assumes the hindsight of knowing how to interpret the role of what we now call vacuum but was thought to be aether. $\endgroup$ – Andrew Steane Jul 2 at 17:36
  • 30
    $\begingroup$ This doesn't really address the question because classical physics is full of waves whose wave equation you can derive to find a speed that depends only on constants (albeit constant properties of matter). And the speed of those waves did depend on the motion of the observer because they were only correct relative the medium. This is the reason that many 19th century physicist pre-supposed the existence of a medium for light. $\endgroup$ – dmckee Jul 2 at 20:31
  • 12
    $\begingroup$ You could have all three of your second set of bullets if you had light moving in a medium just as you can have all three of those ideas and still have sound moving in a medium. $\endgroup$ – dmckee Jul 2 at 20:33
  • 4
    $\begingroup$ So in the end Einstein didn't know that the speed of light should be the same in all inertial frames. He assumed that, based on the experimental evidence that was around back then. He found a way of combining all these experimental results and all the theoretical predictions of the electromagnetic theory into a consistent framework. But, of course, in the end the invariance of the speed of light in all inertial frames is a postulate. $\endgroup$ – AWanderingMind Jul 3 at 9:59
17
$\begingroup$

In 1887 the Michelson Morley experiment provided evidence that the speed of light was independent of the direction of travel of the observer (in their case, they used the movement of the Earth around the sun).

Morley conducted additional experiments with Dayton Miller from 1902 to 1904 which confirmed the results.

Einstein was aware of this (at least later), but it is not clear (and he himself seems to have been unsure) what influence it had, if any, on his 1905 paper. On this matter, there is more detail given in the answers to the question linked in the comment above, which supports the idea that Einstein was more influenced by theoretical arguments.

The fact that the experiment yielded a negative result (they were looking for evidence of the "Ether") may have made it less noteworthy at the time. But its influence was indirect in that it did not invalidate the theories that relied on a constant speed of light.

$\endgroup$
  • 8
    $\begingroup$ The Michelson Morley experiment was not a definitive statement about the speed of light; there is more than one way of interpreting its implications. The question here is about what influenced Einstein and the record shows that this experiment was not crucial as far as he was concerned. The important information was the electromagnetic theory and the experiments by Fizeau. It follows that an answer to the question, as posed, should not go first to Michelson and Morley. $\endgroup$ – Andrew Steane Jul 2 at 9:37
  • $\begingroup$ @Andrew Steane thank you for your comment. I have qualified even further my response in the last paragraph to make it clearer that this was not the primary argument influencing Einstein. $\endgroup$ – rghome Jul 2 at 11:22
  • $\begingroup$ Albert Einstein was not reportedly aware of MMX. $\endgroup$ – Poutnik Jul 4 at 7:03
  • 2
    $\begingroup$ @Poutnik There is a whole question on that: Did Einstein know about the Michelson-Morley experiment?. The best I can make out, he was aware, but after 1905. $\endgroup$ – rghome Jul 4 at 7:08
10
$\begingroup$

Einstein was influenced primarily by the electromagnetic theory developed by several people and culminating in the Maxwell equations, and by the experiments on the speed of light in moving water carried out by Fizeau.

At the time it was not common knowledge what light would do in general; the interpretation of the electromagnetic theory was much puzzled over. The Michelson Morley experiment was a further piece of evidence, but not a crucial one for Einstein.

He was certainly struck by the fact that Maxwell's equations suggest a physical interpretation in which light will recede from you no matter how fast you go to try and catch it up. This was not self-evident and not the only way to interpret the equations, but Einstein seems to have realised that it would be a valuable exercise to abandon the type of aether ideas being tried out at the time, and just stick to the notion that light speed is independent of the source, and then push this way of thinking through to its logical conclusions. Upon doing this, he discovered that one can still get a self-consistent set of ideas about space and time, but they are different from the more familiar (Galilean) ones.

$\endgroup$
  • 2
    $\begingroup$ "Maxwell's equations predict light will recede from you no matter how fast you go" -- isn't that overselling Maxwell somewhat? The equations themselves make no claim of being valid in all inertial frames; that is an external context for them. (Of course this doesn't contradict the point that Einstein seems to have been motivated primarily by a conviction that the relativity of mechanics ought to be reflected in a corresponding relativity of other laws of nature, more than by the failures to empirically pinpoint a privileged "Maxwell frame".) $\endgroup$ – Henning Makholm Jul 2 at 22:23
  • 2
    $\begingroup$ @HenningMakholm Thanks for this; I modified the text a little to put it somewhat closer to what I think may have been the sequence of ideas. $\endgroup$ – Andrew Steane Jul 3 at 7:18
4
$\begingroup$

I'm sorry but @AWanderingMind answer is incorrect. The definition used for Galilean invariance was wrong (which the whole answer was based on), and it does not resemble what happened. The speed of light from Maxwell's equations was not considered constant in all inertial frames, that's what this whole dilema was about. You can read about this in Griffith's electrodynamics book.

https://en.wikipedia.org/wiki/Special_relativity

Galilean invariance applied to equations of motion, not all laws of physics. Einstein's theory of relativity said that all laws of physics are the same in all inertial frames. He came to this conclusion with the idea of a loop on a train moving through a magnetic field. For the people not on the train, they consider a motional emf is created, with the flux rule,

$ε=- d \Phi /dt$

Whereas the people on the train would use Faraday's law and get the same results, even though their physical explanation for the process is completely different. Through transformations Einstein was able to connect the two explanations to each other. Like he was able to show that a how a moving charge with no magnetic field in its rest frame translate to a magnetic field in the frame it is moving through. This is the new stuff he figured out, not EM wave speed from Maxwell's equation, that was already known. He thus came up with his relativity conclusions, since the speed of light would be generated at the same speed for all (as the wave generated from Maxwell's equations show) when all laws of physics (including electromagnetics) are the same in all inertial frames, not just the laws of motion (which did not apply to EM wave generation). There is no absolute rest system.

The Michaelson-Morley experiment he was supposedly vaguely aware of, and many had conflicting interpretations of it (like one was that the earth dragged the ether with it, thus keeping the ether hypothesis alive). For him theoretical reasons were good enough. No ether was required, thus there was no ether. He looked at the Lorentz transformation equations and said they did reflect physical reality, rather than just being an interesting mathematical outcome.

$\endgroup$
  • 1
    $\begingroup$ Pre-special relativity may seem counter intuitive now, but that was the thought at the time, though Einstein's conjecture that the two loop/magnet processes represent the same thing may seem obvious nowadays. $\endgroup$ – user47014 Jul 4 at 2:01
  • $\begingroup$ I did not figure this out or am speculating about what happend, I'm just repeating what is in a textbook and what Einstein said. $\endgroup$ – user47014 Jul 11 at 2:08
2
$\begingroup$

When using the term 'the speed of light' it is sometimes necessary to make the distinction between its one-way speed and its two-way speed.

The "one-way" speed of light, from a source to a detector, cannot be measured independently of a convention as to how to synchronize the clocks at the source and the detector. What can however be experimentally measured is the round-trip speed (or "two-way" speed of light) from the source to the detector and back again.

Albert Einstein chose a synchronization convention see Einstein synchronization that made the one-way speed equal to the two-way speed. The constancy of the one-way speed in any given inertial frame is the basis of his special theory of relativity.

Experiments that attempted to directly probe the one-way speed of light independent of synchronization have been proposed, but none has succeeded in doing so.

The paper A Comparison between Lorentz's Ether Theory and Special Relativity in the Light of the Experiments of Trouton and Noble by Prof. Michel Janssen provides comprehensive historical and theoretical analysis of development of the theory.

Some thought and references on conventionality of distant simultaneity can be found here

$\endgroup$
  • 1
    $\begingroup$ I believe this is the correct answer. The maths supports a different one way and two way speed but Occam's razor seems to apply. $\endgroup$ – CramerTV Jul 2 at 16:49
  • $\begingroup$ @CramerTV Does the maths actually "support" different speeds (which I, perhaps mistakenly, read somewhat as "positively supports"), or would "the maths allows ..." be more accurate? $\endgroup$ – TripeHound Jul 2 at 19:59
  • 1
    $\begingroup$ @TripeHound, Yes, I think allowed would be quite proper to supports. $\endgroup$ – CramerTV Jul 3 at 1:29

protected by AccidentalFourierTransform Jul 13 at 14:32

Thank you for your interest in this question. Because it has attracted low-quality or spam answers that had to be removed, posting an answer now requires 10 reputation on this site (the association bonus does not count).

Would you like to answer one of these unanswered questions instead?

Not the answer you're looking for? Browse other questions tagged or ask your own question.