Einstein's general theory of relativity was accepted once astronomers observed the bending of light from a star during a solar eclipse. So gravity bends light.

 

How do we know that the stars we see at night are really in the locations (or directions) that we see them in? That is, if the light from these stars have travelled hundred, thousands, or even millions of light years, wouldn't their paths have been bent and warped several times over by nearby gravitational sources? In which case, they would most likely not be located where it looks like they are.

 

So how do we know where the stars really are - and if we have some way of knowing, how much does the light from a typical star get diverged from a straight path?

One could measure the deflection over the course of the year, so that our location changes a little. By parallax, one would expect a slight change in deflection angle, though that may be smaller than what could be measured. But if the deflections were large to begin with, that would increase the chances that our motion would introduce a measurable change in deflection.

 

In most cases there would be other indications that there was a large gravity field in the path.