# Could SR not be built from only one postulate?

## Main Question or Discussion Point

Hello, I have a doubt regarding the postulates of SR.

The two postulates, according to Schutz, are:

1) No experiment can measure the absolute velocity of an observer; the results of any experiment performed by an observer do not depend on his speed relative to other observers who are not involved in the experiment.
2) The speed of light relative to any unaccelerated observer is c, regardless of the motion of the light's source relative to the observer.

Couldn't we combine these two postulates into one? I'm thinking of something along the lines of:

All the laws of physics are the same in every inertial frame of reference.

With this, I'm thinking you reach both of the original postulates- this new postulate implies that Maxwell's equations must hold in every inertial frame, so c must be the same in every inertial frame (which takes care of original postulate 2). Also, from this you can conclude that for an observer who is in an inertial frame of reference, the same laws of physics will hold as for another inertial observer moving at a different speed. Therefore, the first observer's experimental results will not be affected by their speed relative to the other observer (this takes care of original postulate 1).

What do you all think?

Related Special and General Relativity News on Phys.org
Dale
Mentor
All the laws of physics are the same in every inertial frame of reference.

With this, I'm thinking you reach both of the original postulates- this new postulate implies that Maxwell's equations must hold in every inertial frame
Only if you also postulate that Maxwell's equations is a law of physics.

Seriously? But that seems sort of superfluous to me; would it really be necessary?

Matterwave
Gold Member
Seriously? But that seems sort of superfluous to me; would it really be necessary?
Can you derive Maxwell's equations from your one postulate?

That's just the usual theoretical argument that establishes postulate 2. It's just that we'd rather not mention Maxwell's equations in some contexts because then we can just start with your 2) and you don't have to know E and M to understand. So, nothing really new here.

xox
Hello, I have a doubt regarding the postulates of SR.

The two postulates, according to Schutz, are:

1) No experiment can measure the absolute velocity of an observer; the results of any experiment performed by an observer do not depend on his speed relative to other observers who are not involved in the experiment.
2) The speed of light relative to any unaccelerated observer is c, regardless of the motion of the light's source relative to the observer.

Couldn't we combine these two postulates into one? I'm thinking of something along the lines of:
No, you cannot combine the two postulates into one but there are a lot of formulations of SR that drop the second postulate. You can do a google search for "single postulate formulation of SR". The most famous one dates from 1910(!), by Ignatowski.
A word of caution, SR is based on a lot more that the two postulates you listed, so , what you are really looking is for formulations that drop the principle of constancy of light speed.

Last edited:
bcrowell
Staff Emeritus
Gold Member
You can certainly combine the two postulates into one. We all know that Maxwell's equations are a law of physics, so there's no need to state explicitly that they're included. If Schutz's #2 were violated, then his #1 would also automatically be violated.

There is nothing special or sacred about Einstein's 1905 axiomatization of SR. From the modern point of view, it's awkward and archaic.

There's a more detailed description of this sort of thing in ch. 2 of my SR book: http://www.lightandmatter.com/sr/

Alright, thanks everyone for your responses! Homeomorphic, your argument made the most sense to me, thanks :) (Matterwave, I don't think you can derive Maxwell's equations from the two original postulates either, nor do I find it a necessary requirement of an SR postulate). And thank you xox, I looked a bit into it and it's all pretty interesting; and yeah, I did know about other assumptions made in SR (I'd seen in wiki before about things like the need to assume spatial homogeneity and isotropy).

Oh sorry Ben, just saw your post. Thanks :). I'll think of the postulates of SR in terms of only one principle, then- it seems simpler.

Ps. Your book looks great! I might read it alongside Schutz and Hartle.

Matterwave
Gold Member
Alright, thanks everyone for your responses! Homeomorphic, your argument made the most sense to me, thanks :) (Matterwave, I don't think you can derive Maxwell's equations from the two original postulates either, nor do I find it a necessary requirement of an SR postulate). And thank you xox, I looked a bit into it and it's all pretty interesting; and yeah, I did know about other assumptions made in SR (I'd seen in wiki before about things like the need to assume spatial homogeneity and isotropy).
You can't derive Maxwell's equations from only the original 2 postulates. But my point was not that it could be. My point was that if you are getting rid of one postulate, and you state that Maxwell's equations can be used to get rid of that postulate, then you have to postulate Maxwell's equations as a substitute unless you can derive Maxwell's equations from postulate 1.

In more formal language. Let's start with postulates A and B (the 1 and 2 originally we had), and you claim that A+C (where C is Maxwell's equations) implies B, then certainly you can use A and C as your fundamental postulates. However, unless you can also show A implies C, you cannot reduce to just A.

In other words, my argument was that the following argument is invalid: "A+C implies B, therefore to derive all the derivable facts given A and B, I only need A".

You can't derive Maxwell's equations from only the original 2 postulates. But my point was not that it could be. My point was that if you are getting rid of one postulate, and you state that Maxwell's equations can be used to get rid of that postulate, then you have to postulate Maxwell's equations as a substitute unless you can derive Maxwell's equations from postulate 1.

In more formal language. Let's start with postulates A and B (the 1 and 2 originally we had), and you claim that A+C (where C is Maxwell's equations) implies B, then certainly you can use A and C as your fundamental postulates. However, unless you can also show A implies C, you cannot reduce to just A.

In other words, my argument was that the following argument is invalid: "A+C implies B, therefore to derive all the derivable facts given A and B, I only need A".
But like Ben said, we already know Maxwell's equations and that they're laws of physics, so they're not required as an additional postulate. What I'm saying is we have postulates A and B, and an extra fact (Maxwell's equations) that isn't used in either postulate, but we know is true. So let's make a postulate C, which when considered along with that extra fact, will imply both A and B.

Dale
Mentor
Oh sorry Ben, just saw your post. Thanks :). I'll think of the postulates of SR in terms of only one principle, then- it seems simpler.
You should probably actually read the chapter. It involves much more than 1 principle. I think it is 5 or so.

Dale
Mentor
You can certainly combine the two postulates into one. We all know that Maxwell's equations are a law of physics, so there's no need to state explicitly that they're included.
The reason that I don't like this approach is because it is circular in motivation, if not in formulation.

If you just want to reduce the number of postulates you can always simply postulate the Lorentz transforms. But the point was to justify the Lorentz transforms on the basis of principles that physicists could be persuaded to accept.

The motivation for justifying the Lorentz transforms was that they were the symmetry group of Maxwell's equations. So including Maxwell's equations in the derivation (either directly or indirectly) makes the whole derivation silly. You may as well just state the fact that Maxwell's equations are invariant under the Lorentz transform and be done with it. That much was already recognized.

That said, your list is much cleaner and more thorough.

You should probably actually read the chapter. It involves much more than 1 principle. I think it is 5 or so.
I did. And by thinking of the SR postulates in terms of one principle, I was referring to the original post (which combined the original two into one).

Anyway, from Ben's book, I think the postulates P2, P4, and P5 are all implied by the definition of an inertial frame (which isn't a postulate itself). P3 regarding the isotropy and homogeneity of space I had mentioned previously and I'm not thinking of it as a postulate of SR, because (I could be completely wrong, but) I think it's a postulate for all physical theories. So for now I'll just think of SR as a physical theory built from one postulate.

Dale
Mentor
Sure, you can always build any theory from one postulate simply by postulating the theory with all of the underlying constructs. There is nothing wrong with that. Just think about your purpose in establishing a set of postulates (I understand Einstein's motivation, but I am not sure what yours is), and whether your choice of postulate accomplishes that purpose.

Last edited:
Matterwave
Gold Member
But like Ben said, we already know Maxwell's equations and that they're laws of physics, so they're not required as an additional postulate. What I'm saying is we have postulates A and B, and an extra fact (Maxwell's equations) that isn't used in either postulate, but we know is true. So let's make a postulate C, which when considered along with that extra fact, will imply both A and B.
But this is the same as postulating A and C (Maxwell's equations). I never said you could not do this. Just because you called it an "extra fact" and not a "postulate" does not mean you have removed it as a postulate...-.-

Ok, so the purpose of the postulates in a theory is to basically be the basis from which you can form arguments and make conclusions. My point is that having the rest of physics as a base (i.e. Maxwell's equations, the assumption that space is homogenous and isotropic, etc.), the only *additional* postulate required by SR is what I said in the original question, as opposed to the two that are usually cited.

xox
And thank you xox, I looked a bit into it and it's all pretty interesting; and yeah, I did know about other assumptions made in SR (I'd seen in wiki before about things like the need to assume spatial homogeneity and isotropy).
Correct. the interesting postulate to omit is the principle of constancy of light. As you can see, it has been done.

Matterwave
Gold Member
But how do you know which physics to base your theory on? After all, if instead of Maxwell's equations, we took Newton's laws to be the "rest of physics as a base" you mention, we get the wrong answer.

strangerep
Guitarphysics,

Let's go back to your original question...

The two postulates, according to Schutz, are:

1) No experiment can measure the absolute velocity of an observer; the results of any experiment performed by an observer do not depend on his speed relative to other observers who are not involved in the experiment.
2) The speed of light relative to any unaccelerated observer is c, regardless of the motion of the light's source relative to the observer.

Couldn't we combine these two postulates into one? I'm thinking of something along the lines of:

All the laws of physics are the same in every inertial frame of reference.
Yes, but some other things are required (which I'll explain below).

With this, I'm thinking you reach both of the original postulates- this new postulate implies that Maxwell's equations must hold in every inertial frame,
No, it doesn't. The most general transformation that preserves inertial motion for a given observer (located at the origin of his coordinate system) is fractional linear, i.e., of the form:
$$t' = \frac{At + Bx}{Ct + Dx} ~,~~~~~ x' = \frac{Et + Fx}{Ct + Dx} ~.$$The most general transformations that preserves the Maxwell wave eqns are conformal transformations -- which have a quadratic denominator in general. See special conformal transformation.

If one asks for a common subset of transformations that do both, one is reduced to ordinary linear transformations. If one assumes spatial isotropy, and a principle of "physical regularity", (i.e., that physical transformation must map finite values of observables to finite values), then the usual Lorentz transformations can be derived without further assumptions, and a universal constant limiting speed (called "c") is an additional output of the derivation. By examining the properties of material bodies whose relative speed is very close to "c", and taking a limit, one can deduce properties that coincide with those usually observed in light. I.e., one can use experiment to identify that "c" corresponds to lightspeed.

So let us drop the assumption that inertial-motion-preserving transformations ("IMTs") should also preserve Maxwell's eqns. Long ago, Bacry and Levy-Leblond figured out that the most general such algebras (larger than the Poincare algebra) are the deSitter algebras, and an additional universal constant with dimensions of length^2 is a further output of the derivation. This has lead to a modern exploration of ways to use this method to "derive" the cosmological constant $\Lambda$ -- since that's essentially what GR without matter boils down to: a deSitter universe.

Others have approached it in different ways. Kerner[2a,2b], and more recently Manida[3a,3b], explored different, more physically-motivated, generalizations -- by seeking the most general form of IMT that could reasonably be interpreted physically as a velocity boost. They arrived at deSitter geometries (surprise, surprise).

In these approaches, the local speed of light is still the usual "c", and Poincare-invariance is retained up to distance scales where cosmological effects become significant. Indeed, the apparent speed of light can vary over (large) times and distances -- but this is already familiar in cosmology, arising from expansion of space over time.

Buried within these approaches are different assumptions about time-reversal invariance. Bacry and Levy-Leblond assumed it explicitly. Manida initially didn't assume it, but later returned to it by embracing deSitter algebras. A slightly more general approach (relaxing the tacit demand for a co-moving transformed frame) might also be possible -- but that's not yet published (afaik) so I can't talk about it on PF.

References:

 H. Bacry, J.-M. L\'evy-Leblond, "Possible Kinematics",
J. Math. Phys., vol 9, no 10, 1605, (1968)

[2a] E. H. Kerner,
An extension of the concept of inertial frame and of Lorentz transformation,
Proc. Nat. Acad. Sci. USA, Vol. 73, No. 5, pp. 1418-1421, May 1976

[2b] E. H. Kerner,
Extended inertial frames and Lorentz transformations. II.
J. Math. Phys., Vol. 17, No. 10, (1976), p1797.

[3a] S. N. Manida,
Fock-Lorentz transformations and time-varying speed of light,
Available as: arXiv:gr-qc/9905046

[3b] S. N. Manida,
Generalized Relativistic Kinematics,
Theor. Math. Phys., vol 169, no 2, (2011), pp1643-1655.
Available as: arXiv:1111.3676 [gr-qc]

bcrowell
Staff Emeritus
Gold Member
But how do you know which physics to base your theory on? After all, if instead of Maxwell's equations, we took Newton's laws to be the "rest of physics as a base" you mention, we get the wrong answer.
This is a nice way of stating what's unsatisfactory about Einstein's 1905 axiomatization. It assumes the state of the art in 1905, which was that there were two main theories of physics: Newton's laws and Maxwell's equations. If you want an axiomatization that reads more like the modern view of how relativity works -- as a theory of the geometry of spacetime -- then you probably want something more like Ignatowsky's axiomatization.

Last edited:
strangerep, thanks very much for your detailed explanation, and the references! Unfortunately, I don't know much algebra so there's some of what you said that was beyond me, not to mention the papers you referenced (I could follow them pretty much through the introduction but nothing more :\ ). I had heard a bit about de Sitter and anti-de Sitter space, but didn't know what it was about. You made that a bit clearer for me, so thanks for that as well!

Matterwave, that's a good point- there's no guarantee that the current physics is correct either, so it would probably be better for *every* postulate of SR to be stated (like in Ben's book- again, Ben thanks for that, it looks like a very refreshing take on SR :D).

Thanks for the interesting responses everyone, you've given me a lot to think about.

Fredrik
Staff Emeritus
Gold Member
All the laws of physics are the same in every inertial frame of reference.

With this, I'm thinking you reach both of the original postulates- this new postulate implies that Maxwell's equations must hold in every inertial frame, so c must be the same in every inertial frame
I like Matterwave's reply (post #19) the best. You could also say that it implies that the velocity of a massive particle influenced by a constant force must satisfy the formula $v=(F/m)t+v_0$ in every inertial coordinate system. This implies that c is not the same in every inertial coordinate system.

I have a lot more to say about this subject, but unfortunately I don't have time right now.

atyy
Ok, so the purpose of the postulates in a theory is to basically be the basis from which you can form arguments and make conclusions. My point is that having the rest of physics as a base (i.e. Maxwell's equations, the assumption that space is homogenous and isotropic, etc.), the only *additional* postulate required by SR is what I said in the original question, as opposed to the two that are usually cited.
As Matterwave, DaleSpam and others have pointed out - you need to specify what the "laws of physics" are. If the laws of physics include Maxwell's equations then 2 is contained in 1. If Maxwell's equations are not in the "laws of physics" then 2 is not contained in 1.

The reason that Maxwell's equations are stated explicitly in 2 in most books is that for many years between Newton and Maxwell, Maxwell's equations were not in the "laws of physics". For example, if Newton's universal gravitation but not Maxwell's equations are in the "laws of physics", then 1 alone would produce Galilean relativity.

Whether you want to count the axioms as 1 or 2 is a matter of taste, depending on what you include in the "laws of physics".

xox
As Matterwave, DaleSpam and others have pointed out - you need to specify what the "laws of physics" are. If the laws of physics include Maxwell's equations then 2 is contained in 1. If Maxwell's equations are not in the "laws of physics" then 2 is not contained in 1.

The reason that Maxwell's equations are stated explicitly in 2 in most books is that for many years between Newton and Maxwell, Maxwell's equations were not in the "laws of physics". For example, if Newton's universal gravitation but not Maxwell's equations are in the "laws of physics", then 1 alone would produce Galilean relativity.

Whether you want to count the axioms as 1 or 2 is a matter of taste, depending on what you include in the "laws of physics".
The above is an excellent way of stating the answer.