Defending Empiricism About Causation