Aristotle hit on two principles of physics that are just as accepted today as when he found them:
1.) Physics is about motion
2.) All motions reduce to/ are based on change of place.
The disagreement occurs at the next step in trying to give an account of place. The history of physics can be told as the attempt to move from the idea of (a) place as a container to (b) place being a relation between bodies.
Both see place as arising from a body’s existence in a larger complex, and there is much to draw out of this, but for the moment I’ll follow the shift from (a) type to (b) type places.
1.) Aristotle and the Ancient-Medieval Cosmos. Since place is a container and all physics is about it, the physical world is a container. This container locates all, and all that moves is moving with respect to it. What is natural is a tendency for some location in the container. Motion is absolutely different from rest, and all suggestions that nature might treat them as equivalent under certain conditions is impossible.
2.) The Classical shift. The idea that the Cosmos was a place became hard to believe when its Crystalline walls were seen as impossible. Some motion had to now be perpetual without being pushed by a perpetual motor on a sphere, and so had to be perpetual from within. The simplest solution suggested itself by the feeling of stasis one can experience on a moving ship: uniform motions needed no more of an extrinsic account of its motion than rest needs an account of its rest. The question “how does that uniformly moving body manage such a uniform motion?”is seen as being just as ridiculous as asking, when you see a dead cat on the road, “how does it manage to keep so still?”
But a container still remained: absolute space. This was a mathematical (i.e. abstract) entity and so fell victim to an interaction problem. The postulate of aether was a way to solve the problem.
We ought not to pretend that it is an easy thing to do away with aether. Besides the (still outstanding) problem of Newton’s Bucket, aether was necessary to keep physics from falling into pure formalism and therefore into idealism. We can’t have something waving (the action) with nothing waving (like a hand). Absent aether, however, this is exactly what you are saying “light” is. Light becomes a sort of nothing that does stuff. Given aether, place is thus still ultimately a container, or a way in which all bodies can be located by touching something that is immobile.
There were problems with aether from the beginning. Wavelengths are inverse to the density of the medium, and so the tiny light wave required a medium 10,000 times denser than steel which, somehow, everything flowed through with no effort.
3.) The Relativist shift. Aether was dropped. There is now a body-space system. If this space is taken as mathematical it gives the same problems as absolute space. What “space” seems to be a shorthand for is a way to maintain the relationships between bodies, i.e. to have a purely (b) type place as mentioned above.
Something of a container sense of place still remains in the claims of a block universe, but it might be possible to account for these as setting up a set of equivalences between any even in space-time and any other postulated event.
QM, however, is background dependent and so still preserves an (a) type notion of space. Smolin is right to see this as the most suspect element in the theory.