Whereas reasons may, and usually do, figure among the proximate causes of belief, and thus play a part in both kinds of series (cognitive and causal), it is always possible to trace back the causal series to a point where every trace of rationality vanishes ; where we are left face to face with conditions of beliefs social, physiological, and physical— which, considered in themselves, are quite a-logical in their character. /.../ on any merely naturalistic hypothesis, the rational elements in the causal series lie always on the surface. Penetrate but a short way down, and they are found no more.
Arthur Balfour