Abductive reasoning, which is compatible with determinism. Even if we had to resort to blind trial and error after enumerating all possible theories, science would still work, just less efficiently. Like I said, I don't see the problem.
I think you might be joking. I assert that we cannot in fact enumerate all possible theories, even if you could write down a representation, they must be enumerated in meatspace to conduct an actual comparison.
Of course we can enumerate all possible theories, in principle. This is a trivial corollary of the fact that we can enumerate all Turing machines.
It's not practical but that's not the point. The point is that if we are free enough to enumerate all possible theories in a deterministic universe, then science can be conducted. Clearly we can enumerate all possible theories, therefore even if our universe is deterministic, we can do science.
