So far, then, although we are compelled, whatever may be our pretensions to empiricism, to employ some sort of a standard of theological probability of our own whenever we assume to estimate the fruits of other men’s religion, yet this very standard has been begotten out of the drift of common life. It is the voice of human experience within us, judging and condemning all gods that stand athwart the pathway along which it feels itself to be advancing. Experience, if we take it in the largest sense, is thus the parent of those disbeliefs which, it was charged, were inconsistent with the experiential method. The inconsistency, you see, is immaterial, and the charge may be neglected.
If we pass from disbeliefs to positive beliefs, it seems to me that there is not even a formal inconsistency to be laid against our method. The gods we stand by are the gods we need and can use, the gods whose demands on us are reinforcements of our demands on ourselves and on one another. What I then propose to do is, briefly stated, to test saintliness by common sense, to use human standards to help us decide how far the religious life commends itself as an ideal kind of human activity. If it commends itself, then any theological beliefs that may inspire it, in so far forth will stand accredited. If not, then they will be discredited, and all without reference to anything but human working principles. It is but the elimination of the humanly unfit, and the survival of the humanly fittest, applied to religious beliefs; and if we look at history candidly and without prejudice, we have to admit that no religion has ever in the long run established or proved itself in any other way. Religions have approved themselves; they have ministered to sundry vital needs which they found reigning. When they violated other needs too strongly, or when other faiths came which served the same needs better, the first religions were supplanted.
The needs were always many, and the tests were never sharp. So the reproach of vagueness and subjectivity and “on the whole”-ness, which can with perfect legitimacy be addressed to the empirical method as we