> The most eye-opening thing I've learned about relativity in the past few years is that the notion that space has no preferential direction is an axiom in the

> The most eye-opening thing I've learned about relativity in the past few years is that the notion that space has no preferential direction is an axiom in the theory. No, it's not. It's a geometric property of particular solutions in the theory. Those solutions include the ones we use to describe the universe as a whole. But there are plenty of other solutions that don't have this property. (For example, the family of solutions that describe black holes.) > There's nothing about the way we measure the speed of light that would disambiguate if light traveled instantaneously in one direction and at half the measured speed of light in the other. You're confusing two different concepts here. The concept of spatial isotropy, which is what "space has no preferential direction" refers to, is different from the concept that the one-way speed of light could vary by direction. The first concept, spatial isotropy, is an invariant concept: it's a geometric property that either is or is not possessed by particular solutions of the Einstein Field Equation. The second concept, anisotropy of the one-way speed of light, is not an invariant concept or a geometric property: it's an artifact of your choice of coordinates. You can take a spacetime that is spatially isotropic, and choose coordinates on it that make it seem like the one-way speed of light varies by direction. Or you can take a spacetime that is not spatially isotropic, and choose coordinates on it that make it seem like the one-way speed of light does not vary by direction. So the one-way speed of light is simply the wrong thing to think about.