Let us now examine some attempts to solve this problem, and first of all let us take the pragmatist solution. Pragmatism claims that it has no difficulty in explaining error, because, as we have already seen, it acknowledges no other test or criterion of truth except a pragmatic one. Every proposition or judgment that we make must, in order to have any meaning whatever, be relevant to some human purpose; every such proposition is a truth-claim; and every truth-claim is tested by its workability. Consequently, error is simply the failure of a proposition to establish its claim by the practical test of working. Propositions marked by such failure are errors. As there is no truth independent of time, place, and circumstance, no irrelevant truth, no truth independent of the conditions under which its claim is put forward, there is no truth that may not become error. No judgment, according to pragmatism, is an error pure and simple--that is to say, it cannot come into existence as error, for it comes claiming truth, and maintaining that claim until challenged; it becomes an error in retrospect only, and always in relation to another judgment which corrects it. Error does not characterise a class of judgments; it is something that happens to a judgment, it is a judgment whose truth-claim is rejected in reference to another judgment {80} which succeeds. The essential thing in the pragmatist doctrine of error is that in claiming to be true a judgment is not challenging comparison with some independent reality, nor is it claiming to belong to a timeless order of existence--to be eternal; it is claiming to fulfil the particular purpose for which it has been called forth, whether that purpose be practical or theoretical. Let us now consider the explanation of error offered by the idealist philosophy. In this view only the whole truth is wholly true; the Absolute, as a perfect, concrete, individual system, is the ideal, and all that falls short of it can only possess a degree of truth--a degree which is greater or less according as it approximates to the ideal. The degrees of truth are not quantitative, not a mixture of truth and error, but a nearer or more distant approach to the ideal. There can be no absolute error, because if truth is the whole, error, if it exists at all, must in some way be included in truth. Clearly error cannot as such be truth, and therefore it must follow that, in the whole, error loses its character of error, and finds reconciliation of its contradiction to truth. Error, then, if it is something, and not a pure negation, is partial or incomplete truth; the perplexity and contradiction that it gives rise to are incidental to our partial view. Knowledge, it must seem to us, can exist only for omniscience. Unless we know everything, we know nothing.