In general, attempts to establish the impossibility of metaphysics founder on the difficulty that we cannot show that metaphysics is from the nature of the case impossible without already presupposing some metaphysical beliefs.2 To show that the nature of reality is such that we cannot know anything of it, we must already make some presuppositions about reality and about the ultimate nature of the human mind. How much metaphysics we are to admit is another question. The persistent disagreement of metaphysicians and the frequent discoveries of fallacies in their arguments certainly do not give much encouragement to hope that we shall be able to arrive at a system of metaphysics which is at once elaborate and securely founded. I think, however, that the significance of the perennial disagreement in question may easily be exaggerated. It may well be that, when two metaphysicians are disputing, what often happens is this. One philosopher asserts the truth of p and another the truth of q, p and q being apparently quite incompatible. Now presumably the facts are likely to be such that some of the criteria required if we are to assert p are present and also some of the criteria required if we are to assert q, otherwise there would be no dispute. It is in consequence of this that there will be both a tendency to assert p and a tendency to assert q. In that case we cannot say without reservations that either is right, because all the criteria required to justify the assertion are not present, but it does not matter whether we say p or q provided we