Much of mathematics isn’t empirically provable, except in a very fuzzy sort of way. You prove math... with math. Not by conjuring up physical experiments. That’s called, well, physics.