In 1971, Joseph Hafele and Richard Keating used atomic clocks to test the prediction of time dilation resulting from motion (special relativity) and gravity (general relativity). In 1972,they reported in Science (SCIENCE, 14 Jul 1972, Vol 177, Issue 4044, pp. 168-170) the detection of relativistic time loss as a result of motion. Laying aside for the moment the issue of gravity on the effect of time, was the detection of time loss as a result of motion in accordance with what is predicted by special relativity?
In the experiment, an atomic clock A was at the naval observatory. An atomic clock B was on a commercial plane that flew eastward around the world. At the end of the flight, Atomic clock B lost time with respect to atomic clock A.
From the reference frame of clock A, this would appear to be in accordance with prediction of special relativity. That is, from the reference frame of clock A, the relative motion of clock B was expected to slow the passage of time of clock B, so clock B would be expected to lose time to with respect to clock A. The data confirmed this prediction.
How about from the reference frame of clock B? In special relativity no reference frame is privileged over another, and the same result is to be expected regardless of the reference frame from which measurements are made.
From the reference frame of clock B, clock B was stationary and clock A was moving at a high speed with respect to clock B. From the reference frame of clock B, the relative motion of clock A would be expected to slow the passage of time of clock A, so clock A would be expected to lose time with respect to clock B. Yet in the experiment, clock A gained time with respect to clock B. The data contradicted the prediction.
How could special relativity be confirmed in one reference frame and simultaneously contradicted in another reference frame. Is this a fatal flaw for the prediction of time dilation in special relativity or is there a way to resolve this conundrum?
Dear Professor Weller,
Thank you for your well described question.
The same query has been considered by Professor McCausland, cf. below:
https://youtu.be/_jQ49W6sqiU?t=592
It's a variation of the animadversion first propounded (I think) by Professor Dingle.
You may or may not be aware, but there is also a problem with the "Relativity of Simultaneity".
Einstein is definitely in error (in my opinion) when he says the light-rays meet the man on the moving carriage at M' one after the other. They have to meet the man on the moving carriage and the man on the embankment simultaneously, for logical reasons.
Cf. in particular the observation of Professor Kılıç, here:
"Thought this was a brilliant observation by Professor Kılıç...."
https://www.researchgate.net/post/Is_there_a_misspeak_in_Einsteins_train_and_embankment_thought_experiment_as_described_by_Einstein_in_the_1952_edition_of_his_book/90
Gary: Thanks for the heads up. I see now that Herbert Dingle raised the issue in 1950. And that there is a long discussion on this at https://www.researchgate.net/post/Is_there_a_solid_counter-argument_against_Dingles_old_objection_to_Relativity_Theory/730).
I understand how gravity can affect time, particularly if time is measure by frequency of light. But never have been able to understand how unaccelerated motion as in special relativity could account for the disparate passage of time purportedly detected by the data gathered from experiments like that of Hafele and Keating.
Because time isn't invariant under Lorentz transformations. What matters are the quantities that are invariant under these transformations. Individual coordinates aren't invariant under Lorentz transformations, spacetime distances (computed using the Minkowski metric) are. It's remarkable that this isn't now common knowledge.
All the ``confusion'' and ``paradoxes" about special relativity are due to focusing on quantities that aren't invariant under Lorentz transformations and not focusing on the quantities that are invariant.
Dear Professor Nicolis,
Well like you, I'm keen for Special Relativity not to be a load of baloney. A mushrooming cloud of nebulous hot-air, that disappears at the slightest scrutiny.
And I do think there is something in it. In particular the Minkowski scheme is something real (I think)-- but it is more to do with the extraordinary properties of light, rather than anything else. You could say it is a restatement that the reality has to be consistent among the (relatively moving) observers.
However, you do appear to be saying we shouldn't focus on quantities like where the clock-hands are, or the length of a rod, or the mass of an object, in order to understand Special Relativity? That's all fine and dandy, but it means quantities that can be measured are to be brushed under the carpet as "unspeakables" leaving a verifiable-free theory?
In effect you are saying Special Relativity has no verifiable results. No clocks that run slow? no rods that are contracted? no mass that is augmented?
If so, then what's left?
Also I'd say there is a part of Special Relativity-- the "Relativity of Simultaneity" which is just flat wrong.
Cf.
"Thought this was a brilliant observation by Professor Kılıç...."
here
https://www.researchgate.net/post/Is_there_a_misspeak_in_Einsteins_train_and_embankment_thought_experiment_as_described_by_Einstein_in_the_1952_edition_of_his_book/90
The mathematical description of natural phenomena relies on identifying the symmetries, those transformations that leave the equations of motion invariant. It, also, implies that only quantities that are invariant under these transformations have any meaning, since quantities that aren't depend on the frame and invariants don't. So the mass of an object can be shown to be invariant under Lorentz transformations; the length of a rod, however isn't invariant under Lorentz transformations.
It is possible to test special relativity, by checking for the appearance of terms that aren't invariant under Lorentz transformations in the equations of motion and this is always done for background. Cf. for instance https://lorentz.sitehost.iu.edu/kostelecky/faq.html
Dear Professor Nicolis,
"It is possible to test special relativity, by checking for the appearance of terms that aren't invariant under Lorentz transformations in the equations of motion and this is always done for background".
These checks are for mathematical properties of the theory, it's not "testing" as is commonly understood, experimentally, because you have provided nothing to test, unless I have misunderstood here.
What physical things-- clock hand positions, lengths of rods, the mass of an object, temperature, induction E, voltage, &c. are to be measured experimentally?
Or are you saying Special Relativity is akin to String Theory? and is "unfalsifiable".
Dear Professor Nicolis,
"So the mass of an object can be shown to be invariant under Lorentz transformations; the length of a rod, however isn't invariant under Lorentz transformations. "
I think you mean the energy-momentum is invariant. The "mass of an object" varies as indicated by the equation in the above text.
if everything moves in straight lines at constant velocities, then SR's time dilation and length-contraction effects are interpretative, not physical.
Suppose that an object recedes from you at v=0.6c
- If you believe that the speed of light is globally fixed in its frame, you'll expect to see a redshift of E'/E = (c-v)/c = 0.4 .
- If you believe that the speed of light is globally fixed in your frame, you'll expect to see a weaker redshift of E'/E = c/(c+v) = 1/1.6 = 0.625.
Special relativity removes this discrepancy by averaging together the two conflicting predictions ... it multiplies the two earlier predictions together and takes the square root.
- This gives the SR prediction, of E'/E = sqrt[ (c-v) / (c+v) ] that you''ll find in section 7 of the translated 1905 electrodynamics paper. This gives sqrt[0.4*0.625] = sqrt[0.25] = 0.5
---
- If you divide 0.5 by 0.4, you'll see that the disagreement when we move to SR from "observer-centric c" is 5/4 = 1.25
- If you divide 0.5 by 0.625, you'll see that the disagreement when we move to SR from "emitter-centric c" is 0.5/0.625 = 1/1.25
... so for v=0.6c, you can get to the SR prediction, either by assuming that the s.o.l. if fixed in your frame and that the other guy is ageing 1.25 times slower, or by assuming that the s.o.l. is fixed in their frame, and that you are moving wrt the magical frame, and are time dilated, and are therefore seeing the signals coming in 1.25 times faster.
It makes no difference. Whether they are "really" ageing faster than you, or you are "really" aging faster than them has no immediate physical consequences, it's a made-up explanation. As long as nobody accelerates, and there are no gravitational fields, there is no physical way to test whose interpretation is right. It's metaphysics.
Where does the 1.25 come from? It's the Lorentz factor.
E'/E = sqrt[ 1 - v^2/c^2 ] = sqrt[ 1- 0.6^2] = sqrt[ 1 - 0.36] = sqrt[ 0.64 ] = 0.8 ... and 1/0.8 = 1.25
So, think of SR as an averaging approach, and think of the Lorentz factor as a correction factor used as a shortcut method to get the same result, and you should end up with a pretty good intuition of what SR ought to predict for any simple situation.
Preprint Special relativity considered as an average of earlier theories
Ignoring the acceleration differences and gravitational (height) differences still leaves 2 issues for me:
1) we don't know the physical mechanism of decay in the atomic clocks. For example, suppose a decay event is triggered by a neutrino impact on the nucleus ...
2) I think there is an interpretation issue: What B (or A) observes is their measure, in their system, of what the other's clock is doing - NOT the real action in the comoving frame.
Preprint Animadversions upon the wiki figure purporting to demonstrat...
- Badges
- Science topic