Dismiss Notice
Join Physics Forums Today!
The friendliest, high quality science and math community on the planet! Everyone who loves science is here!

4D spacetime Light cone Twins paradox

  1. Oct 26, 2005 #1
    I'm an "on-my-own-free-time" arm-chair student of physics. Lol.
    So if this question is way off the mark my apologies.
    Feel free to let me know where I’m off base.


    For me, a great visual example of the twin paradox was found at this site:
    http://www.phy.syr.edu/courses/modules/LIGHTCONE/LightClock/default.html [Broken]
    (Thanks Janus it was your link posted on another thread I used to get there)

    So, the second to last AVI or RAM at the bottom of the page shows the twins or (A) and (B)s full journey within their own world lines travelling through 4d-spacetime.

    The twin (A) stays in his inertial frame and the twin (B) travels away from the first twin and then turns around.
    It was said that at the point of the turn it is revealed who has the preferred reference frame.

    So the paradox is resolved when the twin (B) turns around and returns to meet twin (A)

    Twin (B)s change of direction shows that he can no longer be considered to be in the preferred reference frame. He will age more slowly also the faster he goes.

    But my question is: When the twin (B) turns to return to twin (A) why can't it be said instead that twin (A) was the one doing the turning WRT to twin (B).
    Or a third possibility could be that they both share half the turn, if one can put it that way.

    So for the first half the trip twin (B)s clock runs slowly compared to twin(A) and for the second half of the trip twin (A)s clock runs slowly compared to twin (B)s. If they are doing a circular orbit around each other then perhaps the time dilation of both twins would just cancel out in real time.

    Sorry for the convoluted example but there you go. That’s my question. Is this possible? At the end of the trip both twins aged at the same rate?

    One can’t prove who was turning around whom and no one was discovered to be in the preferred inertial reference frame for the entire duration of the trip.

    Both reference frames during all turns or orbits equally cancel out each others time dilation. But time dilation still occurs as per usual.


    Hope that made some sense.


    Last edited by a moderator: May 2, 2017
  2. jcsd
  3. Oct 26, 2005 #2
    The answer to this is in your own post a couple of lines up.
    Twin (B)s change of direction shows that he can no longer be considered to be in the preferred reference frame.
    In other words, it is twin (B) and twin (B) only that gets mushed up into the nose cone of his rocket when it slows down to turn around. That's how he knows he's the one that will be younger when the two meet up.

    In that case, they would both get mushed up and indeed be the same age when they meet. There is still no paradox because as in the first case, the problem goes away when you account for all the frames being used. At the base of the twins seeming-paradox is the pretense that you are using two frames of reference when you are using three. In this case you are using five. All of the paradox goes away when you account for what happens as you jump from one frame to another.
    Last edited: Oct 26, 2005
  4. Oct 26, 2005 #3
    Thanks for the reply

    Thanks for the reply.
    What you say makes sense.
    But I'm still not sure on some of these points.

    Lets say twin B is unaware of his flight path. So too twin A.

    So the knowledge of being mushed up makes him age?
    Let's say the acceleration or change in direction were slow enough or the orbit very large (computer controlled) as to make the change in G's unnoticable then why would anyone need to get mushed up or be aware that they were turning. It can be set up so that the twins are unaware of the flight path and unable to tell.

    The data for both twins trips could be compiled at the end of the journey.Do they suddenly age as soon as the data is compiled?

    Five? I believe you but I'm only counting 2.

    Thanks again,


    PS... Please don't hit me with advanced maths of the lorentz tranformation as I'm sure I can plug in the numbers with a little help, but I'm trying to think through this conceptually. BTW I say that with humor..LOL...it's all good.
    Last edited: Oct 26, 2005
  5. Oct 26, 2005 #4
    Ok. I get this part. :smile:

    Many Thanks

  6. Oct 26, 2005 #5
    Again, that cramped feeling up there in the nose cone will tell him that something is going on. Twin A never experiences that feeling.

    Well, that's sort of what I said, but it's not true. Here's the real skinny. When the rocket accelerates, the frame changes. It is the frame change that causes the difference in ages. Being all mushed and stuff tells you who is experiencing the frame change and who is not.

    Yes it can. But again, it is not the awareness that is the issue, it is the frame change. Traditionally, the twins seeming-paradox is presented with no acceleration for most of the journey so that special relativity can be invoked. If you accelerate slowly, you lose this angle. You are accelerating a great deal of the time and now you need general relativity.

    No, but there is a lot of insight in this question. If we assume that the travelling twin is not accelerating on the outward leg, accelerates instantaneously to turn 180 degrees and then no more acceleration for the return journey, then the travelling twin will 'see' the stationary twin age suddenly at the moment of turning. Actually see is a poor word for it. As you put it, they will need to wait until the data is compiled. But although they must wait, when they do finally see the data, it will show that the aging took place at the turn in the journey.

    1. A's outward leg
    2. A's inward leg.
    3. B's outward leg.
    4. B's inward leg.
    5. The frame of the meeting point.

    It is common to think of this last frame as not moving in space, but in time only, although that is a frame dependent concept. Interestingly, travelling with this 'non-moving' frame takes the longest proper time to traverse the two meeting events. That's how it is that the stationary twin ages more than the travelling one.
  7. Oct 26, 2005 #6
    Roger that.
    Cool. Well said.
    Well said. Frame change and G.R. involved here.
    Got it. Cool.
    Again. Really understandable and no maths! thanks.
    The insight was yours. I just had a hunch.
    The rapid aging is strange but interesting.
    Sure enough that's IS 5. Very Clear.
    Appreciate your expertise and your time.

    all the best,

  8. Oct 26, 2005 #7


    User Avatar

    Staff: Mentor

    No, you do not. This is a common misconception. Special relativity can handle acceleration by integrating over a sequence of "instantaneously co-moving inertial reference frames," one for each velocity the accelerating observer "passes through."
    What special relativity can't handle is situations where the effects of gravity are significant, that is, where you have to deal with curved spacetime.
  9. Oct 26, 2005 #8
    eon_rider, jtbell may be right about this. An easier way to see that there is a problem with slowly accelerating is that in the end you still need to turn the frame around 180 degrees. Although it does matter whether you do that quickly or slowly, the mere fact that you did it is enough to void your twin paradox warranty.

    I'm not 100% sure that jtbell is right though. After all, acceleration is equivalent to gravity. How much acceleration do you need to curve space? I would have guessed any at all. A while back I was told that the concept of CMRF was a stop-gap measure to approximate GR in the years between 1905 and 1915 when there was no GR. I was told that it is accurate enough and convenient enough for some calculations, but theoretically eclipsed after GR was introduced. Am I misinformed?
  10. Oct 26, 2005 #9


    User Avatar
    Science Advisor

    You'd have to be careful about what you integrate, though--you couldn't just use the same quantities you'd use if you were sticking to a single inertial frame. For example, if you find the function for the velocity of another clock v(t) as a function of its velocity in your instantaneous co-moving inertial frame at proper time t on your own clock, you cannot then just integrate [tex]\int_{t_0}^{t_1} \sqrt{1 - v(t)^2/c^2} \, dt[/tex] to find the total time elapsed on that clock between two proper times [tex]t_0[/tex] and [tex]t_1[/tex] on your clock (say, between the time you departed from the other clock and the time you reunited with it), because that would fail to take into account the way your definition of simultaneity is changing as your instantantaneous co-moving inertial frame is changing.
  11. Oct 26, 2005 #10


    User Avatar
    Science Advisor

    The effects experienced by an observer in a small region undergoing acceleration are equivalent to the effect the observer would experience in a gravitational field, but that doesn't mean that acceleration itself causes spacetime to be curved, you can talk about accelerating objects in flat spacetime (in practice I guess the energy used to accelerate an object would have to contribute very slightly to the curvature of spacetime, but I think that for ordinary human-scale objects undergoing human-scale accelerations the effect would be tiny, too small to measure).

    I think the equivalence principle does mean that if you have an object experiencing constant acceleration in flat spacetime, you can come up with a new coordinate system where the object is at rest and the G-forces experienced by the object are now explained in terms of a uniform gravitational field--that's what this page on the GR analysis of the twin paradox seems to be saying. A question for the GR experts here--I think I remember from a previous discussion someone telling me that in GR the curvature of spacetime does not depend on your coordinate system, so can a "uniform gravitational field" exist in flat spacetime? If the curvature of space alone can vary depending on your choice of coordinate system, does a uniform gravitational field in flat spacetime mean that space is curved even though spacetime is not?
    Last edited: Oct 26, 2005
  12. Oct 26, 2005 #11
    It sounds fishy to me. You seem to be saying that acceleration is equivalent to a special kind of gravity that doesn't curve spacetime.

    Here is my take on things. I don't think that use of the MCRF is strictly allowable in SR, you just get away with it because the calculations work. In that respect it is like using Newton's equations to calculate rocket trajectories. It is adequate for calculations, but theoretically incorrect. In other words, there are three ways of handling the twins:

    1. SR
    2. SR + MCRF
    3. GR

    In this scheme, SR by itself is not adequate to explain the 'slowly accelerated' version of the twins seeming-paradox. However, I agree with jtbell that for computational purposes, GR is not necessary either. Instead SR + MCRF can be used for the resolution of the seeming-paradox. But only because it gets the right numbers. A full, theoretically correct analysis requires GR. These are the opinions of a non-physicist.
  13. Oct 26, 2005 #12


    User Avatar
    Science Advisor

    Well, the equivalence principle only deals with local regions, and locally spacetime is always near-flat...if you're in free fall, you experience all the same phenomena in your local region as you would if you were moving inertially in flat spacetime. The feeling of being pulled by gravity is because some non-gravitational force (like the electromagnetic force from the ground pushing up on your feet) is preventing you from taking the geodesic that you would in free-fall, and you feel the same thing as you would if a non-gravitational force was accelerating you in empty space (which also prevents you from following a geodesic).
    Sorry, what does MCRF stand for?
    It's theoretically incorrect because the predictions of Newtonian mechanics deviate slightly from those of relativity, but not much if the velocities are small compared to light. On the other hand, GR should give exactly the same predictions as SR in flat spacetime. And again, it's certainly possible to have acceleration in flat spacetime...although as I said it would take some energy to accelerate a massive object and that energy would curve spacetime slightly, you can always get around this by talking about a "test particle" of arbitrarily small mass. Acceleration itself does not curve spacetime, only matter/energy does.
  14. Oct 27, 2005 #13
    Just a thought.

    Is there a tension between the non-locality of Q.M and the private nature of inertial frames and world lines in S.R?
    In other words. Let’s say the twins make their flight and return.
    But the trip is done in a double blind way where the twins don’t know their flight path nor does the scientist. (at the NASA of the future let's pretend)

    Then the data is compiled. At the moment of compiling the data (doing the math) could some kind of non-binary wave function collapse and suddenly one twin ages the correct amount according to S.R.

    LOL. I realise this is a complete fishing expedition and I’m just throwing out 3
    ideas or more as a non-expert but its interesting to me.
    Those concepts being:

    1) non-locality in Q.M. – Copenhagen interpretation or some other well respected one.

    2) collapse of a wave function (state of existence?) (is the cat alive or not) and can this be applied to does one twin age or not only at the moment of observing the data and compiling it.

    3) The rules of S.R stating that each world line is inertial and private?

    I’m so very much reaching here so I don’t mind if you all laugh.

    But can Q.M issues run into S.R. issues when dealing with the twin paradox?


    Last edited: Oct 27, 2005
  15. Oct 27, 2005 #14
    No. The nature of QM plays no role in relativity whatsoever. The intervals that determine relative aging rates is completely defined and measurable at all time in relativity. There are no collapsing wavefunctions and such. Changes in relative aging rates are completely determined by who accelerates, a little over a long time or a lot over a short time doesn't matter. Who accelerates can be agreed upon by all observers in all frames of reference and has nothing to do with whether anyone observed any part of it or not. Once you understand what's involved it's not even a real paradox. The apparent contradiction comes from the fact that you both appear to aging faster than the other until one of you accelerates to other. This is because you both disagree on when now is.
  16. Oct 27, 2005 #15
    I have some of the same issues in dealing with GR so I posted a question https://www.physicsforums.com/showthread.php?t=96927" in hopes of fully articulating the situation.
    Last edited by a moderator: Apr 21, 2017
  17. Oct 27, 2005 #16
    Well said my_wan. Cool.

    I should have googled it first.

    When I did articles and papers from universities and arXiv.org
    came up on the subject of Q.M and S.R and the peaceful co-existence between the two for the most part, but not always.
    But as you say for the most part there is no issue. I believe.

    Thanks for the reply.


    Last edited: Oct 27, 2005
  18. Oct 27, 2005 #17
    Momentarily Comoving Reference Frame. The postulates of SR constrain the theory to inertial reference frames. As I understand it, the use of MCRF grew up in the years between the introduction of SR in 1905 and that of GR in 1915 as a way of dealing with non-inertial frames. It is still used now because it gets the right answer in restricted cases. It takes advantage of the fact that slightly non-inertial frames are almost inertial. You integrate across inertial reference frames that are momentarily comoving with the non-inertial frame. It is computationally convenient and accurate (definition of slightly non-inertial: if MCRF gets the wrong answer then the non wasn't slight) but has no theoretical underpinings. SR doesn't allow you to work with non-inertial frames, and MCRF doesn't correctly account for all of the GR effects of spacetime curvature.

    Edited to removed references to acceleration.
    Last edited: Oct 27, 2005
  19. Oct 27, 2005 #18
    We don't seem to disagree. Near-flat is far enough away from flat for SR to fail. The postulates of SR state that they are limited to inertial frames. Thus MCRF.

    I don't fully understand your distinction between acceleration curving spacetime and energy curving spacetime. Can there be acceleration without energy? Using an admitedly Newtonian point of view: Acceleration is proportional to force and an accelerating particle cannot be stationary for any period of time. Multiply the force times the distance moved and you get energy.

    Below is a famous picture of acceleration causing curvature. I have been told that there is a flaw in this picture, but I don't know what the flaw is, so I present it with the hopes that someone can explain it to me.

    Consider a disk spinning about its center like a roulette wheel. Consider a portion of the circumference small enough to be nearly linear. As the length of this small piece is parallel to its motion, a stationary observer at a point on the axis of rotation will observe a foreshortening. Now consider the radius connecting the center of the disk to the center of the small piece of the circumerence. This distance is perpendicular to the motion and so is not foreshortened. As the piece of circumference was arbitrary, the entire circumference is foreshortened. So the observer notes that one of the properties of flat space [itex]c = 2\pi{r}[/itex] does not hold. The cause is explained entirely by the acceleration of the disk without any mention of forces or energy.
    Last edited: Oct 27, 2005
  20. Oct 27, 2005 #19
    SR is about looking at the world from an inertial reference frame. That's easy enough. So I can sit still in an inertial reference frame, and watch the famous twins, both going off in opposite directions, and then one turning around and going back to the other. I sit in my inertial armchair and watch all this occur. There has been no need to do any physics in a non-inertial reference frame.

    The proper time elapsed by a particle in a path is given by:
    [tex]\Delta\tau = \int^{\lambda=b}_{\lambda=a} \sqrt{-\eta_{\mu\nu}\frac{dx^\mu}{d\lambda}\frac{dx^\nu}{d\lambda}} d\lambda[/tex]
    where [itex]x^\mu(\lambda)[/itex] is a parametrization of the worldline of that particle (I'm quoting more or less straight from Carroll). Furthermore, [itex]\eta_{\mu\nu} = \mbox{diag}(-1, 1, 1, 1)[/itex].

    No GR involved whatsoever.
    Last edited: Oct 27, 2005
  21. Oct 27, 2005 #20


    User Avatar
    Science Advisor
    Gold Member

    When does inetgrating over the MCIFs get a wrong answer? The answers certainly aren't any different from GR sans curvature.

    The postulates of SR apply to inertial frmaes, but there's ntohing to stop us constructing non-inertial frames. Plus once you accept that SR naturally leads to the description of spacetime in terms of a Minkowski manifold, the only difference between inertial and non-inertial farmes is that inertial frames have constant basis fields.
    Last edited: Oct 27, 2005
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook