On reductionism, science and the reversal of the burden of proof