That doesn't completely defeat empiricism though. Controlled experiments are best, then uncontrolled experiments (checking that the theories are backed up by re

That doesn't completely defeat empiricism though. Controlled experiments are best, then uncontrolled experiments (checking that the theories are backed up by real data), then pure reason (non-reductionism) is a cop-out.