When one admits that nothing is certain one must, I think, also add that some things are more nearly certain than others