The rationalist, then, has his bias like another. Though it takes in part a critical or negative form, it is fundamentally as positive as another. He has come to crave for coherence and consistency in narratives, statements, explanations, arguments, propositions, and systems of thought; even as his 'contrary' or competitor has come to crave for evidence that something 'supernatural' wields a purposive and 'intelligent' control, mediate or immediate, over all things, using among others 'supernormal' means. This 'contrary' thinker may or may not believe in 'spirits' in the ordinary sense, may or may not believe in the immortality of human minds; but if he is really to be an opponent of the rationalist bias he is to be classed as having a bias to traditional or authoritative views of the cosmos, to religious as against naturalistic explanations of history, to a conception of the human as of the extra-human processes in terms of a controlling will and purpose. He too, it is true, must have some craving for coherence and consistency--else he could not debate and reason at all; but the other craving in him has primed that. It is a fallacy, we may note in passing, to suppose that the 'agnostic' attitude, so-called, is something between the two main forms of bias here posited. Agnosticism, logically carried out, can differentiate from other forms of rationalism only in local limitation of belief; and in practice it is not often found to do even that. The agnostic inevitably begins in terms of the rationalist bias, in craving for coherence and consistency of statement; and his most circumspect negations stand for precaution against inconsistent credulity. But precisely in virtue of that bias, he is the opponent of the supernaturalist bias. He does not in effect merely say, 'I do not know': he implicitly says 'You do not know' to the professor of non-natural knowledge.