[Physics] Clocks in relativity

[Ilja Schmelzer]

2016-11-16 12:00 GMT+01:00, physics-request at tuks.nl <physics-request at tuks.nl>:
> From: "Thomas Goodey" <thomas at flyingkettle.com>
> It depends what you mean by "physically real". Each one of
> two moving observers, when he looks at (considered in his
> own reference frame, of course) the clock of the other
> observer, sees that it is running slower than his clock.
> That is the brute fact. I don't know whether you would
> consider that fact to be "physically real" or not; but
> anyway, it's what is always actually observed in that
> situation.

It is part of a well-known thing often emphasized by Popper,
namely that observation is always interpretation in the light
of existing theories.

In fact, if we compare two inertial observers, then there is no
such animal as a comparison of the speed of their clocks.
They can meet only once, and this is not enough.

What can they do?  They can introduce some synchronization
for clocks far away of each other.  But this is already nothing
they observe, but a consequence of the particular synchronization
procedure they choose.

And, in particular, if they accept the same synchronization procedure,
they will obtain the same results too.  Above usind the "Einstein
synchronization procedure" does not mean they use the same, because
this procedure depends on the observer which is assumed to be at rest.
For observers not at rest the procedure is a quite nonsensical one.

>> Einstein  himself said (and I am not going to look it up
>> to reference it) "It does not matter what type of clock
>> is used."
>
> But obviously he meant, an accurate clock. If the clock is
> bad, and its running speed fluctuates widely, nothing
> whatever can be said about it. Of course. For example, if
> you use your heartbeat as a clock, you will not reach any
> physically meaningful results.

If the result is physically meaningful depends on what are
your requirements of accuracy are. So for some questions
it could be accurate enough.

Then, Einstein meant all clocks.  Because all clocks will
be time-dilated in the same way.  Of course, if this gives
a time dilation factor 1.00000000000001 and your clock
has only a 1% accuracy, this gives no measurable time
dilation. But that the effect is not measurable with that
particular clock does not mean that the effect itself
does not exist.