In Special Relativity, acceleration i.e. a changing velocity 4-vector results in time dilation, that is asymmetric aging of observers.
In General Relativity, the 4-vector does not change along a geodesic. Or rather, the co-ordinate components of the four-vector do change along a geodesic. But we have defined the covariant derivative such that the velocity-four vector differentiates to 0 along a geodesic.
Consequently, the acceleration along a geodesic is zero. And hence, it seems like the the GR time-dilation is of a different nature from the SR time dilation. The GR one being attributed to a varying metric while the SR one attributed to a non-zero acceleration.
However, what if we re-interpret the four-velocity in GR? What if we simply say that the four-velocity along a geodesic IS changing? We would do this by using partial derivatives instead of covariant derivatives. If we did this, wouldn't the GR time dilation get attributed to a non-zero acceleration just like in SR?
Does this mean that both time dilations are of the same underlying nature, and that attributing the GR time dilation to the metric is just a matter of semantics?
And how does all of this relate to "time dilation due to a graviton"? I've seen people argue that the reason graviton is special is that it somehow affects the metric of spacetime. If we simply re-interpret a changing metric as a non-zero acceleration, then the graviton is no longer special in this regard.