Harry wrote:
> "Tom Roberts" <tjroberts@lucent.com> wrote in message
> news:cicnjv$qi2@netnews.proxy.lucent.com...
>>
>> Well, it's really that COMPARISONS of clocks at different gravitational
>> potentials indicate that a signal between them is redshifted or
>> blueshifted. In GR this is most naturally described as being a result of
>> spacetime curvature, and neither clock is truly "affected"; nor is the
>> signal "affected" either, at any point along its trajectory -- this is a
>> global effect induced by the geometry of the manifold.
> 
> Tom, are you sure that corresponds to Einstein's GRT? IMO that would imply
> that he would have lost the insight he had in 1911:
> "If we measure time in S1 with the clock U1, then we must measure time in S2
> with a clock which goes 1+phi/c^2 times more slowly than the clock U when
> compared with U at one and the same place".

Einstein did not "lose" this insight, but it evolved into a more complex 
and subtle relationship.


> GRT as expressed in the Am.J.Phys. of 2000 corresponds well to that way of
> formulating it:
> "the phenomenon is [correctly] explained through the behavior of clocks
> which run faster the higher they are located in the potential"
> 
> Thus, if you disagree with that, then whose GRT do you represent?

GR is a diverse theory because of its coordinate independence. Yet we 
humans are used to using coordinates to describe physical phenomena -- 
so much so that we often do it without realizing it, as here.

The physical observation is redshift in a light signal sent from a lower 
clock+source to a higher clock+detector.

A: One can choose one set of coordinates and describe this as
changes in clock rates with altitude (as in your quote above).

B: One can choose a different set of coordinates and describe this as a 
change in the light signal between the clocks (light "loses energy" as 
it climbs upward).

   [One could choose still different coordinates that give
    various mixtures of A and B. I'll ignore this...]

C: One can choose to use a coordinate-independent approach and describe 
this as no change in either clocks or signal, but as an instance of 
geometrical perspective.


The problem with A is that when one checks each clock by comparing to a 
standard clock, both clocks check out identically; so how can this be 
due to "change" in the clocks?

The problem with B is that one can check the light signal everywhere
along its path, and find that over no small region does it change either
frequency, wavelength, or speed; so how can this be due to "change" in the
signal?

There are no such problems with C, which is why it is the preferred
description of this, IMHO. C also has the virtue that it is right in
line with "length contraction" and "time dilation" in SR.


Tom Roberts tjroberts@lucent.com