So much for his working out of the thought. But on the other hand, Kant never transcended the absoluteness of the distinction between the principles of synthesis and analysis, of sufficient reason and contradiction. The result was that he regarded the synthetic principle as the principle only of our knowledge, while perfect knowledge he still considered to follow the law of identity, of mere analysis. He worked out the factor of negation, of differentiation, contained in the notion of synthesis, but limited it to synthesis upon material of sense, presupposing that there is another kind of knowledge, not limited to sense, not depending upon the synthetic principle, but resting upon the principle of contradiction, or analysis, and that this kind is the type, the norm, of the only perfect knowledge. In other words, while admitting the synthetic principle of differentiation as a necessary element within _our_ knowledge, he held that on account of this element our knowledge is limited to the phenomenal realm. Leibniz's error was in supposing that the pure principles of the logical understanding, resting on contradiction, could give _us_ knowledge of the noumenal world; his truth was in supposing that only by such principles _could_ they be known. Thus, in substance, Kant. Like Leibniz, in short, he failed to transcend the absoluteness of the value of the scholastic method; but he so worked out another and synthetic method,--the _development_ of the idea of sufficient reason,--that he made it necessary for his successors to transcend it. The second point concerns the relations of the sensuous and the super-sensuous. Here, besides setting right the ordinary misconception of Leibniz, Kant did nothing but render him consistent with himself. Leibniz attempted to prove the existence of God, as we have seen, by the principles both of sufficient reason and contradiction. Kant denies the validity of the proof by either method. God is the sufficient cause, or reason, of the contingent sense world. But since Leibniz admits that this contingent world may, after all, be but a dream, how shall we rise from it to the notion of God? It is not our dreams that demonstrate to us the existence of reality. Or, again, sense-knowledge is confused knowledge. How shall this knowledge, by hypothesis imperfect, guarantee to us the existence of a perfect being? On the other hand, since the synthetic principle, or that of sufficient reason, _is_ necessary to give us knowledge of matters of fact, the principle of contradiction, while it may give us a consistent and even necessary notion of a supreme being, cannot give this notion reality. Leibniz, while admitting, with regard to all other matters of fact, that the principles of formal logic can give no unconditional knowledge, yet supposes that, with regard to the one unconditional reality, they are amply sufficient. Kant but renders him self-consistent on this point.