Thus, the question is, whence do we derive our reasons for _believing_ in the truth of such propositions? No, whence does our belief get its cause? But the _origin of a belief,_ of a strong conviction, is a psychological problem: and very limited and narrow experience frequently brings about such a belief! _It already presupposes_ that there are not only "data _a posteriori_" but also "data _a priori_"-- that is to say, "previous to experience." Necessary and universal truth cannot be given by experience: it is therefore quite clear that it has come to us without experience at all? There is no such thing as an isolated judgment! An isolated judgment is never "true," it is never knowledge; only in _connection with,_ and when _related to,_ many other judgments, is a guarantee of its truth forthcoming. What is the difference between true and false belief? What is knowledge? He "knows" it, that is heavenly! Necessary and universal truth cannot be given by experience! It is therefore independent of experience, _of_ all experience! The view which comes quite _a priori,_ and therefore independent of all experience, _merely out of reason,_ is "pure knowledge"! "The principles of logic, the principle of identity and of contradiction, are examples of pure knowledge, because they precede all experience."--But these principles are not cognitions, but _regulative articles of faith._ In order to establish the _a priori_ character (the pure rationality) of mathematical axioms, space _must be conceived as a form of pure reason._ Hume had declared that there were no _a priori_ synthetic judgments. Kant says there are--the mathematical ones! And if there are such judgments, there may also be such things as metaphysics and a knowledge of things by means of pure reason! Mathematics is possible under conditions which are _not_ allowed to metaphysics. All human knowledge is either experience or mathematics. A judgment is synthetic--that is to say, it co-ordinates various ideas. It is _a priori_--that is to say, this co-ordination is universally true and necessary, and is arrived at, not by sensual experience, but by pure reason. If there are such things as _a priori_ judgments, then reason must be able to co-ordinate: co-ordination is a form. Reason must _possess a formative faculty._ 531. _Judging_ is our oldest faith; it is our habit of believing this to be true or false, of asserting or denying, our certainty that something is thus and not otherwise, our belief that we really "know"--_what_ is believed to be true in all judgments?