I Light Acceleration: Does Light Really Accelerate?

[petrushkagoogol]

A beam of light bends when near a star. Since velocity is a vector, although the path still follows a null geodesic, a change in direction implies that acceleration is present. Light does then, in fact, accelerate. Is this correct ?

 

[A.T.]

a change in direction implies that acceleration is present

Coordinate acceleration.

Light does then, in fact, accelerate. Is this correct ?

It can have coordinate acceleration. Geodesic world-line corresponds to zero proper acceleration.

 

[phinds]

A beam of light bends when near a star. Since velocity is a vector, although the path still follows a null geodesic, a change in direction implies that acceleration is present. Light does then, in fact, accelerate. Is this correct ?

No it is not correct. A geodesic is followed BECAUSE there is no force being applied to the light. The geodesic is "bent" if looked at via Euclidean geometry, which would imply a force on the light, but space-time does not follow Euclidean geometry, It follows Riemann geometry and the geodesic is a straight line implying no force.

To restate: the light is not changing direction. It is going in a straight line.

EDIT: I see AT beat me to it and gave a more complete answer.

 

[Orodruin]

It follows Riemann geometry

Just to be specific:
Pseudo-Riemannian geometry. In Riemannian geometry the metric is positive definite.

 

[phinds]

Just to be specific:
Pseudo-Riemannian geometry. In Riemannian geometry the metric is positive definite.

OK. Thanks

 

[A.T.]

To restate: the light is not changing direction. It is going in a straight line.

The spatial path of light is curved and the propagation direction does change. Only its world-line is a geodesic (locally straight).