"Finally, we should note the basic metaphysical assumption of the classical laboratory — namely, that nature is neither capricious nor secretive. If nature were capricious, she would tell one observer one thing and another observer a quite different thing."