# Why is light speed constant in all reference frames?

So, given that I know nothing about physics, this is probably going to sound like a stupid question. But I've always wondered how it is possible that a person A moving at speed observes light emitted from their frame of reference at c, and then a person B sitting still (relative to person A of course) sees that same light travelling also at c. It seems like that just shouldn't happen.

So I think I'm right when I say it has something to do with space and time being perceived differently depending on how fast you are going.
"The Michelson-Morley experiment (MMX) was intended to measure the velocity of the Earth relative to the “lumeniferous aether” which was at the time presumed to carry electromagnetic phenomena. The failure of it and the other early experiments to actually observe the Earth's motion through the aether became significant in promoting the acceptance of Einstein's theory of Special Relativity, as it was appreciated from early on that Einstein's approach (via symmetry) was more elegant and parsimonious of assumptions than were other approaches (e.g. those of Maxwell, Hertz, Stokes, Fresnel, Lorentz, Ritz, and Abraham)."

The idea was that as earth is seen to move with the solarsystem as well as around the sun you would find a different speed of light relative what motion you measured that speed in. That as the aether was thought of as some absolute 'frame of reference' which all bodies moved relative. But the light was measured the same speed everywhere (approximately, as good as could be done by that time.). And that was a headache because even if you assumed no aether, you still had to explain why that light speed didn't change with Earths motion, against or with it.

Michelson–Morley experiment.

To make sense of that, and later even more refined experiments, came two assumptions, that light was a 'constant' and that the reason why it constantly gave the same speed ignoring the motion of the object it rested on, like earth, must be that something happened with the measuring equipment, a Lorentz contraction. Learning about this Poincaré suggested that this local time, as indicated by clocks 'moving in the aether', could be synchronized under the assumption of a constant light speed.

It is said that Einstein based his assumptions not so much on this, but instead on Maxwell's equations, that also gave radiation a absolute speed. But he must have known of those results and ideas too. And as far I know all experiments done so far has validated his definitions. The thing with it is that you start from those experimental results, and then you keep on building. That was what Einstein did when he later presented GR, in where 'gravity' became equivalent to a uniform constant acceleration.

As all 'clocks' differ with their motion, as defined relative some observer, he now could bind how clocks differed in a motion to gravity. And all of this came from one postulate, as I see it. That lights speed in a vacuum was a constant 'c', no matter from where it was sent, if in motion relative the observer or not. It's a whole 'frame work', and I all to often put it all together :) but it all works out experimentally, and what it comes down to for me is 'c'.

To assume differently you need to give a definition to those experiments already done, where you prove how light can be a variable at the same time as it always give you a constant 'c' when measured locally as I see it. And that one should be tricky, especially as gravity and 'clocks' works out beautifully in Einsteins definitions, and also fits very well with Special relativity following his definitions.

experimental basis of Special Relativity.

Last edited:
When I try to convey the block universe concept to someone, I try to give it my best shot and not include personal bias. But, it remains a deep mystery for me personally. Many physicists who discuss it feel that this model is directly manifest by the Lorentz frames. I have tried and tried for many years to find material that effectively counters it, without success. I've tried to think up scenarios without success. I can't counter the argument for the block universe, but at the same time I just don't see how we can reconcile it when you consider the bizarre implications. Take your choice between zombies or a consciousness that is present over the entire neuronal world line. Einstein actually seemed to favor the block universe with consciousness(es) present over the entire world line (his letter to Besso's wife upon the death of his close friend and colleague, Besso).

harrylin, I've been impressed with the quality of your posts and consider you one of the reliable physicists here. So, I'm intrigued by your comment about the "Lorentzian picture." I have always felt that the block universe model is the most direct picture of the Lorentz frames that could be envisioned. Thanks for your comments.

You may have noticed that I've posted comments based on the block universe from time to time. I always hope for someone who can post counters to the model, but I have decided that the subject is not too interesting to forum members (or perhaps they feel it is too close to the edge of philosophy--or maybe way over the line and far too deeply into philosophy).
Yes, it's very close to philosophy - but then, "why" is very close to philosophy, and physics started out as a branch of philosophy ("PhD" isn't an error). Anyway, since you presented the 4D interpretation, it's only fair to also mention the Lorentzian 1+3D interpretation.

That interpretation treats Minkowski space-time as a mathematical model or method for calculating physical events (that was at least around 1920 also Einstein's position; only he was rather ambiguous about his interpretation). There is no reason to assume that a calculation method should itself have a 1-to-1 correspondence to physical reality; in general that is also not the case in physics.

If you are familiar with Newton's absolutist interpretation of classical relativity, then SR is easy to understand as a refinement, or correction, of Newtonian mechanics. Time and space are still regarded as concepts that refer to physically very different things. The relativistic effects are explained as effects from absolute motion, just as Newton explained the effects from rotation as due to absolute motion.* Absolute motion has absolute effects, but those effects don't allow us to detect an "absolute frame", even if there is one. That this should not be regarded as a "conspiracy" can be made plausible by for example arguing that everything, even matter, is made up of fields and radiation, or that the conservation laws demand this.

So, it's a trade-off: the block universe model suggests a 1-to-1 correspondence to what physically "really" happens, but with rather bizarre implications; while the absolute motion model suggests that we may retain some of our intuition about "space" and "time", but the downside is that the beautiful symmetry is only in the phenomena.

Harald

PS: The answer to why c has the value that it has, is from that point of view that it is the propagation constant of space.

* This was elaborated in Langevin's rather long-winding expose on The evolution of space and time (1911):
http://en.wikisource.org/wiki/The_Evolution_of_Space_and_Time

Last edited:
Yes, it's very close to philosophy - but then, "why" is very close to philosophy, and physics started out as a branch of philosophy ("PhD" isn't an error). Anyway, since you presented the 4D interpretation, it's only fair to also mention the Lorentzian 1+3D interpretation.

That interpretation treats Minkowski space-time as a mathematical model or method for calculating physical events (that was at least around 1920 also Einstein's position; only he was rather ambiguous about his interpretation). There is no reason to assume that a calculation method should itself have a 1-to-1 correspondence to physical reality; in general that is also not the case in physics.

If you are familiar with Newton's absolutist interpretation of classical relativity, then SR is easy to understand as a refinement, or correction, of Newtonian mechanics. Time and space are still regarded as concepts that refer to physically very different things. The relativistic effects are explained as effects from absolute motion, just as Newton explained the effects from rotation as due to absolute motion.* Absolute motion has absolute effects, but those effects don't allow us to detect an "absolute frame", even if there is one. That this should not be regarded as a "conspiracy" can be made plausible by for example arguing that everything, even matter, is made up of fields and radiation, or that the conservation laws demand this.

So, it's a trade-off: the block universe model suggests a 1-to-1 correspondence to what physically "really" happens, but with rather bizarre implications; while the absolute motion model suggests that we may retain some of our intuition about "space" and "time", but the downside is that the beautiful symmetry is only in the phenomena.

Harald

PS: The answer to why c has the value that it has, is from that point of view that it is the propagation constant of space.

* This was elaborated in Langevin's rather long-winding expose on The evolution of space and time (1911):
http://en.wikisource.org/wiki/The_Evolution_of_Space_and_Time
Great response. Thanks a lot.

rbj
The answer to why c has the value that it has, is from that point of view that it is the propagation constant of space.
and that value can be whatever you want, as long as it's real, positive, and finite. otherwise it doesn't matter and the question is sorta meaningless.

and that value can be whatever you want, as long as it's real, positive, and finite. otherwise it doesn't matter and the question is sorta meaningless.
I think I know what you mean, rbj. But, could you clarify whether you are referring mainly to

a) The particular numerical value and units associated with the speed (186,000 mi/sec, ... etc.) or,

b) The speed of light is the same for all observers.

I think it is based on the basic assumption of relativity, which is the equivalence of natural laws of physics are the same in all inertial reference frames. Then light must have the same speed in all frames, since speed of light is derived from Maxwell's equations which is a valid law of nature. And this principle is quite tenable, as there is really no distinction between different inertial reference frames.

How do you put on your shoes? Doesn't the light between your hands and your shoes change wavelength and frequency. v= lambda*frequency. HMMMMM It even says in some physics books blue light travels faster than red light. But don't be sure about that either. DERRRRRRR AT ALL LIGHT IS MASSLESS AND MOVES AT V=C.

Great response. Thanks a lot.
You're welcome.

I forgot to mention that Langevin's "absolutist" argument starts at p.47, with the "twin" scenario as illustration.

And perhaps an elaboration is at its place:

- The answer to why c has the value that it has, is from that point of view that it is the propagation constant of space;
- An intuitive answer to why light speed is constant in all reference frames, is from that point of view that apparently everything, even matter, is made up of fields and radiation, so that everything is affected the same;
- The invariance of the speed of light between reference frames is imposed by the conservation laws (indeed, the modern PoR has even been derived based on those laws). That argument is independent of the physical model that one uses.

Harald

You're welcome.

I forgot to mention that Langevin's "absolutist" argument starts at p.47, with the "twin" scenario as illustration.

And perhaps an elaboration is at its place:

- The answer to why c has the value that it has, is from that point of view that it is the propagation constant of space;
- An intuitive answer to why light speed is constant in all reference frames, is from that point of view that apparently everything, even matter, is made up of fields and radiation, so that everything is affected the same;
- The invariance of the speed of light between reference frames is imposed by the conservation laws (indeed, the modern PoR has even been derived based on those laws). That argument is independent of the physical model that one uses.

Harald
Thanks again for the comments and reference. The reference you gave may not have been the one you were remembering. There are only 16 pages in this reference. Although he did use an illustration analogous to the twin scenario, he did not actually present the twin paradox in this paper.

Also, there was no discussion of the interpretation of special relativity. From what you say, I assume he takes the position of logical positivists, wherein physics should no longer be concerned with external objective reality--or models of an assumed external material world. This is in contrast to Einstein's comment,

"The belief in an external world independent of the perceiving subject is the basis of all natural science." (from "Clerk Maxwell's Influence on the Evolution of the Idea of Physical Reality", from Einstein's "The World As I See It").

Also, Einstein seems to imply from some of his writings and lectures that there is a three-fold distinction between an "external world", the observer's "perception of that external world", and "our notions of it."

In another comment in an Address at Columbia University, "Behind the tireless efforts of the investigator there lurks a stronger, more mysterious drive: it is existence and reality that one wishes to comprehend."

The main point I would make in this context (in conjunction with our discussion of the "Block Universe") is that physicists such as Langevin and others, who reject the reality of a physical 4-dimensional universe, do it without proposing any other concept of an external physical world. And the reason is simple: Special relativity has no description of an external physical world other than a 4-dimensional space in which observers have different cross-section views of that 4-dimensional world. And in any case they do not embrace any external world (that is, an external world as implied in Einstein's comments above).

Einstein himself was careful to not force such a model on his audience, as can be seen with his comment that followed immediately after the above comment in his Columbia University address: "But one shrinks from the use of such words, for one soon gets into difficulties when one has to explain what is really meant by 'reality' and by 'comprehend' in such a general statement." Of course Einstein was well read and schooled in philosophical writings and understood quite well the problem of reality.

But my primary point here is, again, that a physicist chooses one of three stances: 1) physics pursues the comprehension of an external physical world (which from special relativity directly implies a a 4-dimensional world), 2) the logical positivist or operational view, in which physics should only be interested in predicting the outcome of experiments, making measurements and advancing mathematical models that agree with experiments (the models are mathematically symbolic only and have no implications about an actual external physical reality), or finally 3) One may simply take the stance of no committment to either 1) or 2).

it's not just the speed of light (which is the speed of the EM interaction), it's the speed of all of the "instantaneous" interactions. it's a property of space and time, not of any particular interaction; EM, gravity, strong.
i think you got the cause and effect sorta backwards. the speed of communication is limited to c because the fundamental interactions are not really instantaneous and have their effect speed limited.
It seems TGLAD was evaluating the universe from a designer's perspective - which is intriguing and probably educational but potentially deceptive (if such a design isn't constrained by human imagination).

it has to be real, positive, and finite. other than that, it doesn't really matter. it really becomes just a matter of units and Nature doesn't give a damn about what units humans (or the aliens on the planet Zog) choose to use.
Curious. What is the motivation for that restriction? The mathematics behind Quantum Mechanics pretty much tells us that when dealing with more than one dimension, energy translation must be considered with regard to complex values, even hyper-complex values when more than two dimensions are involved. Though I can understand that SR is in some sense limited in application to a single spatial dimension (a linear path between emitter and absorber).

Last edited:
It is interesting to note, that one can derive the "equations of special relativity" (i.e. the Lorentz transformations) without requiring that the speed of light is constant. The principle of causality (i.e. that an event cannot be caused by a future event), is enough to impose a maximum speed of transmission of information. It turns out that this coincides with the speed of light.
I don't see how this is possible. From my understanding, the derivation itself comes from the fact that the speed of light is constant. If d = c t and c is constant then only d and t can change when compared to another distance c t. Then d and t have to be assigned as d' and t' from the frame of reference of another observer where the constant c wouldn't allow for the same d and t value. (d' and t' can not equal d and t) If you where to say there was instead a c and a c' to make the equations valid with each other you would get completely different equations. It would be like solving for the speed of an object seen from two different point of views.

Why, 300,000 km/s? Why is the sky blue? It just is. I think it is amazing with spacetime dialation that we can even observe an object traveling at a limited constant speed that is so fast that spacetime itself approuches zero for an object at that speed.

Thanks again for the comments and reference. The reference you gave may not have been the one you were remembering. There are only 16 pages in this reference.
Scientia 10 p.31-54, http://en.wikisource.org/wiki/The_Evolution_of_Space_and_Time - the page numbers are indicated on the left.
Although he did use an illustration analogous to the twin scenario, he did not actually present the twin paradox in this paper.
Also, there was no discussion of the interpretation of special relativity.
As far as I know that was the first full presentation of the twin scenario (viewed from both perspectives) - but indeed he did not call them "twins". He did not at all present it as something paradoxical but as another illustration of his physical interpretation of the theory (p.47):

"We therefore have hold on the ether through accelerations, and acceleration has an absolute sense as determining the production of waves from matter that has undergone a change in velocity, and the aether manifests its reality as the vehicle, as the carrier of energy transported by these waves."

Evidently he would have fully agreed with Einstein's citation by you here under:
From what you say, I assume he takes the position of logical positivists, wherein physics should no longer be concerned with external objective reality--or models of an assumed external material world. This is in contrast to Einstein's comment,

"The belief in an external world independent of the perceiving subject is the basis of all natural science." (from "Clerk Maxwell's Influence on the Evolution of the Idea of Physical Reality", from Einstein's "The World As I See It").
Einstein himself was careful to not force such a model on his audience [..]
As SR does not directly depend on a physical model, it united people with such different interpretations as Lorentz and Minkowski. However, in his Leyden inauguration speech of 1920 Einstein did indicate that SR corresponds to the Lorentzian ether.
But my primary point here is, again, that a physicist chooses one of three stances: 1) physics pursues the comprehension of an external physical world (which from special relativity directly implies a a 4-dimensional world), 2) the logical positivist or operational view, in which physics should only be interested in predicting the outcome of experiments, making measurements and advancing mathematical models that agree with experiments (the models are mathematically symbolic only and have no implications about an actual external physical reality), or finally 3) One may simply take the stance of no committment to either 1) or 2).
Nearly so. I cannot follow that exact separation of options, and I'd say that it already doesn't match such individuals as Einstein and Langevin. My variant on your statement:

A physicist chooses one of three stances (or flip-flops between them!):
1) physics pursues the comprehension of an external (or "real") physical world. Special relativity seems to imply either a Lorentzian ether (3D ether) or a physical Minkowski Spacetime (a 4D block universe, not just the "world of events");
2) the logical positivist or operational view, in which physics should only be interested in predicting the outcome of experiments, making measurements and advancing mathematical models that agree with experiments. The formulation of SR by Einstein in 1905 reflects that operational view.
3) One may simply take the stance of no commitment to either 1) or 2).

Mathematical models do certainly not imply the actual physical reality; however linking that argument to the above numbering requires an additional sub division.

I don't see how this is possible. From my understanding, the derivation itself comes from the fact that the speed of light is constant. [..].
It is possible to derive the same based on the assumption that there is a limit speed, and that was also understood in 1905. It just happened that the speed of light was an easy boundary condition for the derivation; it was known to correspond (at least to very good approximation) to that limit speed.

Scientia 10 p.31-54, http://en.wikisource.org/wiki/The_Evolution_of_Space_and_Time - the page numbers are indicated on the left.

As far as I know that was the first full presentation of the twin scenario (viewed from both perspectives) - but indeed he did not call them "twins". He did not at all present it as something paradoxical but as another illustration of his physical interpretation of the theory (p.47):

"We therefore have hold on the ether through accelerations, and acceleration has an absolute sense as determining the production of waves from matter that has undergone a change in velocity, and the aether manifests its reality as the vehicle, as the carrier of energy transported by these waves."

Evidently he would have fully agreed with Einstein's citation by you here under:

As SR does not directly depend on a physical model, it united people with such different interpretations as Lorentz and Minkowski. However, in his Leyden inauguration speech of 1920 Einstein did indicate that SR corresponds to the Lorentzian ether.

Nearly so. I cannot follow that exact separation of options, and I'd say that it already doesn't match such individuals as Einstein and Langevin. My variant on your statement:

A physicist chooses one of three stances (or flip-flops between them!):
1) physics pursues the comprehension of an external (or "real") physical world. Special relativity seems to imply either a Lorentzian ether (3D ether) or a physical Minkowski Spacetime (a 4D block universe, not just the "world of events");
2) the logical positivist or operational view, in which physics should only be interested in predicting the outcome of experiments, making measurements and advancing mathematical models that agree with experiments. The formulation of SR by Einstein in 1905 reflects that operational view.
3) One may simply take the stance of no commitment to either 1) or 2).

Mathematical models do certainly not imply the actual physical reality; however linking that argument to the above numbering requires an additional sub division.
Good job, as usual, Harrylin.

I'm afraid I'm the hopeless pursuer of the external reality (although I'm still repulsed by some of the implications of the "block universe"). I would embrace Einstein's comment, "The belief in an external world independent of the perceiving subject is the basis of all natural science." And I would embrace it without Einstein's follow-up apologetics.

I wish I could rediscover the reference in which Einstein once expressed the sentiment (and I could be wrong) that to pursue something other than the external physical continuum world is to leave one vulnerable to solipsism. I do remember vividly his statement that included the phrase, "...there is no escape from solipsism."

I'm not sure of the logic, but it may have been something to the effect that if your reality is not that of an external world, then you evidently have an internal reality in mind ("inside the mind"). But then all of these other observers in your world are in your mind and that's getting dangerously close to solipsism.

Some may claim their reality is external--it's just of an ethereal sort--not physical or material. But, now we leave physics. The external 4-dimensional space is physics (it is directly described by special and general relativity); the other stuff is philosophy and metaphysics--unless considered strictly from the standpoint of mathematical modeling (with no implications about reality--as you clarified in your post).

 Post-Script: If I cannot have an external physical world, then I would rather take the fall-back position of just not pursuing reality at all. Just make predictions about the outcome of experiments and then do the measurements--and you can do the math modeling if don't take the models literally. As Dalespam put it, "...time is the t in the equations of physics." (or something to that effect).

Last edited:
[..]
Some may claim their reality is external--it's just of an ethereal sort--not physical or material. But, now we leave physics. The external 4-dimensional space is physics (it is directly described by special and general relativity); the other stuff is philosophy and metaphysics--unless considered strictly from the standpoint of mathematical modeling (with no implications about reality--as you clarified in your post).

 Post-Script: If I cannot have an external physical world, then I would rather take the fall-back position of just not pursuing reality at all. Just make predictions about the outcome of experiments and then do the measurements--and you can do the math modeling if don't take the models literally. As Dalespam put it, "...time is the t in the equations of physics." (or something to that effect).
I'm afraid that I didn't express myself very well (by mistake I omitted the word "directly"). I did not mean that mathematical models have no implications at all about reality, but I meant that there doesn't need to be a 1-to-1 correspondence to what physically "really" happens. Mathematical models help us to imagine possible physical models of nature, but often there are several proposed explanations that perfectly match the same mathematical predictions.

Perhaps it's better to clarify this with an example - here is a rather silly one:
When we bring two equal volumes of gas of different temperatures together, the standard mathematical model for predicting the resulting temperature is T=(T1+T2)/2. But the equation does not reflect what really happens: nature doesn't add the two temperatures and then cuts that temperature in half. :tongue2: In reality the temperature evolves from the extremes to the average and that's not at all what the (simplest) standard mathematical model appears to suggest.

Last edited: