jartsa said:
Here's a little story:
There's a bakery in a valley making 100 buns a day. Valley people buy 100 buns a day.
Then the bakery is moved to the top of a mountain. Now bakery makes 101 buns a day. After million days bakery has million unsold buns.
The point of the story is that the extra buns are real buns. The reason for extra real buns is real extra swiftness of the staff.
I'm a little uncomfortable with this description, because it makes it seem that there is an objective, coordinate-independent notion of clock rates.
Being a little more precise, what's true is this:
In a spacetime diagram, we can identify 4 events:
e_1: The start of one day at the bottom of the mountain.
e_2: The start of that day at the top of the mount.
e_3: The end of the day at the bottom of the mountain.
e_4: The end of the day at the top of the mountain.
It's an objective fact that a clock at the top of the mountain advances 1% farther between events e_2 and e_4 than a clock at the bottom of the mountain advances between e_1 and e_3. What's relative, rather than objective, is the claim, necessary to make a case for the mountaintop clock running faster, that e_1 and e_2 are simultaneous, and e_3 and e_4 are simultaneous. They are simultaneous in the noninertial frame in which the clocks are at rest and the gravitational field is time-independent. But not in all frames.
Here's a Euclidean analogy: You have a flat region on Earth. There are roads going in all different directions, some straight and some winding. Each road has associated roadside markers on the side of road with consecutive integers on them: marker #1, marker #2, etc. It's not clear whether the markers have a standard separation (say, every 10 meters) or whether different roads use different separations. Let me use the phrase "marker rate" to mean the number of markers per kilometer.
So let's suppose that there are two roads that are parallel--they stay the same distance apart. What you notice is that where Road A has marker number 1, you can look to the closest marker on Road B, and it's marker number 2. Where Road A has marker number 2, you can look to the closest marker on Road B, and it's marker #4. Etc. So clearly, the "marker rate" of Road B is twice that of Road A. Right?
No. What if in reality, both roads are circular, with the same center. Road A is a circle of radius 1 kilometer, and Road B is a circle of radius 2 kilometers. They actually have exactly the same "marker rate" of one marker every 10 meters. It's just that traveling through the same angle on Road B gets you twice as far as traveling through that same angle on Road A.
Time dilation in a weak gravitational field is analogous to traveling on circular roads. The analogy of "number of markers between two points" is "elapsed time between two events". The analogy of "angle traveled" is coordinate time. The analogy of "radius of the circle" is "height of the clock". Different clocks travel different amounts of elapsed time for the same amount of coordinate time in the same way that different roads pass different numbers of markers for the same amount of angular distance traveled.