FactChecker said:
During any inertial-flight part of the traveler's trip, he is correct that the stationary twin is aging slower.
No, he isn't. All he can say is that
in the inertial frame in which he is currently at rest, the stationary twin is aging slower. But this is a frame-dependent statement. Frame-dependent statements are not about actual real things; they're about calculated abstractions.
FactChecker said:
Each observer sees that time is passing slower for the other.
This is false. Each observer
calculates that, in his current inertial rest frame, time is passing slower for the other. But that is not what each observer actually
sees. What each observer actually
sees is what I have been describing all along about Doppler shifts. In other words:
The traveling twin
sees the stay-at-home twin's clock running slower than his until he turns around; then he
sees the stay-at-home twin's clock running faster than his. The speed-up in the second part outweighs the slow-down in the first part, so when the twins meet up again the traveling twin has
seen the stay-at-home twin's clock have more total elapsed time than his.
The stay-at-home twin
sees the traveling twin's clock running slower than his for most of the time they are apart; then, not long before the twins meet again, the stay-at-home twin
sees the traveling twin's clock speed up so it is running faster than his. But the slow-down in the first part outweighs the speed-up in the second part, so when the twins meet up again the stay-at-home twin has
seen the traveling twin's clock have less total elapsed time than his.
The Doppler Shift Explanation page in the Usenet FAQ article that I linked to earlier describes this in somewhat more detail.
The crucial point here is that you have to distinguish the frame-dependent concept of "time dilation", which by itself
cannot be used to make accurate predictions about differential aging, from the invariant concept of "directly
seen Doppler shift/clock rate", which
can be used to make accurate predictions about differential aging. But the latter is
not the same as the former; the directly
seen Doppler shift/clock rate factor is
not the same as the calculated "time dilation" factor.
This is actually one of the limitations of the most common way of teaching SR, that it focuses on inertial frames and calculated frame-dependent quantities, and invites confusion of frame-dependent quantities with invariants.
FactChecker said:
I should have said that any physical process in one IRF appears slower to an observer in a different IRF that is moving relative to the process IRF.
No. What
appears is what I described as
seen above. There is
no direct observable that corresponds to the frame-dependent
calculated time dilation. (There can't be, because frame-dependent quantities can never be direct observables.)