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

Help wanted with an explanation

  1. Sep 27, 2010 #1
    I have a BS in physics from some time ago. In another forum on another board I have become embroiled in a discussion with another poster about SR and the need to adjust the clocks on GPS satellites to account for SR effects. I need some help trying to explain things. As expressed by the other poster, our positions are

    "And you keep missing the point about HOW they correct for the supposed time dilation. Time dilation, according to theory, DOES affect time itself. You could not, therefore, correct for it by adjusting the clock. IF it were really being observed, you would have to adjust for the variance via a different method of correction, but not through physically adjusting the clock to run slower.

    "And no, the difference between what you are saying and what I am saying is this: you are saying that what is being observed is SRT/time dilation (effectively saying that time is what affected by the velocity of the satellites). I am saying the CLOCK is affected by the motion, NOT TIME. It may seem to give the appearance of affecting time, but it is the clock's mechanism that is affected."

    Any suggestions on how to approach this? Nothing I've tried has worked to get through so far.
  2. jcsd
  3. Sep 27, 2010 #2


    User Avatar
    Science Advisor

    Well, what does it mean to say that time dilation affects clocks but not "time itself"? Does this imply that there is some objective truth about the factor by which a given clock is running slow relative to "time itself"? As a purely philosophical matter you can choose to adopt an "interpretation" of SR where there is a single inertial frame whose views on time/simultaneity/length are the objectively correct ones, so a clock moving at a speed of v relative to this preferred frame is "objectively" running slow by a factor of [tex]\sqrt{1 - v^2/c^2}[/tex] (this type of interpretation is sometimes called a Lorentz ether theory). But as long as SR is correct on an experimental level, the laws of physics should work the same in all inertial frames, meaning there would be no possible experiment you could do that would tell you which of the infinite possible inertial frames is the one whose opinions on clock ticking rates are "objectively correct". Assuming you are the person who takes the position that time really slows down, I guess you could explain to the other poster that different inertial frames can disagree about which of a given pair of clocks is ticking slower, and ask if the poster thinks there is any experimental way to determine the "truth" about this matter.
  4. Sep 27, 2010 #3


    User Avatar
    Science Advisor

    Since a GPS essentially is nothing but a clock, it's of course enough to adjust the clock. How could it not be enough? Time goes faster, the clock is going slower, and everything is fine again. All other effects of time dilation are uninteresting for GPS.
    Tell him what "time" means in this context: time is what a clock shows.
    Operationally, this means: A satellite is a zero g environment. You go and build clocks that work in a zero g environment. Build all varieties you can conceive of, decaying muons, cesium, quartz, biological, whatever. All these clocks will have different accuracies, as can be seen by comparnig clocks with identical technology.
    Now you adjust them once. Now observe the clocks at different velocites, with exactly this adjustment.
    a) If they all show the same time within their accuracies, but different times than similar clocks with different veocities, it is said that time is affected, not the clocks. By definition.
    b) If they all show different times according to their mechanisms, you'd say that the clocks are affected by something. But that's not what we observe.

    If another poster believes that all conceivable clocks are affected exactly the same way by velocity, that's also case a) by definition.
  5. Sep 27, 2010 #4
    Well every physical process aboard that satellite is affected the same way, not just clock mechanisms. So let me ask you this: what do you think the problem is in saying that time is affected?
  6. Sep 27, 2010 #5
    Thanks for all of the responses. They are appreciated.

    I did, by the way, try to get across the point that the "motion" of which he was speaking is simply the motion of the inertial frame, which is accounted for in SR and appears as time dilation - and got nowhere. It is almost as though he is on a hidden variable kick.
    Last edited: Sep 27, 2010
  7. Sep 27, 2010 #6


    User Avatar
    Science Advisor

    You might get somewhere by pointing out it is permissable to adopt an interpretation like the Lorentz ether theory I mentioned where there is a preferred frame and things like time dilation and motion are defined relative to it, but then mention that it would be absolutely impossible to decide which frame it is experimentally--that way, if he disagrees with the second part you can press him on what experiment he would propose to do it, and if he doesn't disagree with the second part you can just "agree to disagree" on the philosophical issue, pointing out that the complete "hiddenness" of the preferred frame is seen by many as a good reason to abandon such a notion.
  8. Sep 27, 2010 #7
    Just a general comment about GPS clocks. They run faster than the clocks on Earth. This you cannot explain by applying SR, it can only be explained by applying GR.
Share this great discussion with others via Reddit, Google+, Twitter, or Facebook