This fact is construed by the pragmatist as a significant indication of the way out of the epistemological _impasse_. The 'relative' truths, which Intellectualism passed by with contempt, may differ in _practical value_ and lead to the conceptions of _practical truth_ and certainty which may be better adapted to the requirements of human life than the elusive and discredited ideals of absolute truth and certainty, and may enable us to justify the distinctions we make between the 'true' and the 'false. At any rate, this suggestion seems worth following up. To begin with, we must radically disabuse our minds of the idea that thinking _starts from certainty._ Even the self-evident and self-confident 'intuitions' that impress the uncritical so much with their claim to infallibility are really the results of antecedent doubts and ponderings, and would never be enunciated unless there were thought to be a dispute about them. In real life thought starts from perplexities, from situations in which, as Professor Dewey says, beliefs have to be 'reconstructed,' and it aims at setting doubts at rest. It is psychologically impossible for a rational mind to assert what it knows to be true, and supposes everyone else to admit the truth of. This is why even a philosopher's conversation does not consist of a rehearsal of all the unchallenged truisms that he can remember. Being thus conditioned by a doubt, every judgment is a challenge. It claims truth, and backs its claims by the authority of its maker; but it would be folly to imagine that it thereby becomes _ipso facto_ true, or is meant to be universally accepted without testing. Its maker must know this as well as anyone, unless his dogmatism has quite blotted out his common sense. Indeed, he may himself have given preference to the judgment he made over the alternatives that occurred to him only after much debate and hesitation, and may propound it only as a basis for further discussion and testing. Initially, then, every judgment is a _truth-claim_, and this claim is merely _formal_. It does not _mean_ that the claim is absolutely true, and that it is impious to question it. On the contrary, it has still to be validated by others, and may work in such a way that its own maker withdraws it, and corrects it by a better. The intellectualist accounts of truth have all failed to make this vital distinction between 'truth-claim' and validated truth. They rest on a _confusion of formal with absolute truth_, and it is on this account that they cannot distinguish between 'truth' and error. For false judgments also formally claim 'truth,' No judgment alleges that it is false.[C]