Immanuel Kant's Critique of Pure Reason — trans. F. Max Müller

ical knowledge cannot claim this advantage, for here we must always consider the general in the abstract (by concepts), while in mathematics we may consider the general in the concrete, in each single intuition, and yet through pure representation a priori, where every mistake becomes at once manifest.

I should prefer, [p. 735] therefore, to call the former acromnatic, or audible (discursive) proofs, because they can be carried out by words only (the object in thought), rather than demonstrations, which, as the very term implies, depend on the intuition of the object.

It follows from all this that it is not in accordance with the very nature of philosophy to boast of its dogmatical character, particularly in the field of pure reason, and to deck itself with the titles and ribands of mathematics, an order to which it can never belong, though it may well hope for co-operation with that science. All those attempts are vain pretensions which can never be successful, nay, which can only prove an obstacle in the discovery of the illusions of reason, when ignoring its own limits, and which must mar our success in calling back, by means of a sufficient explanation of our concepts, the conceit of speculation to the more modest and thorough work of selfknowledge. Reason ought not, therefore, in its transcendental endeavours, to look forward with such confidence, as if the path which it has traversed must lead straight to its goal, nor depend with such assurance on its premisses as to consider it unnecessary to look back from time to time, to find out whether, in the progress of its conclusions, errors may come to light, which were overlooked in the principles, and which render it nee- [p. 736] essary either to determine those principles more accurately or to change them altogether.

Discipline of Piwe Reason 591

I divide all apodictic propositions, whether demonstrable or immediately certain, into Dogmata and Mathemata. A directly synthetical proposition, based on concepts, is a Dogma; a proposition of the same kind, arrived at by the construction of concepts, is a Mathema, Analytical judgments teach us really no more of an object than w^hat the concept which we have of it contains in itself. They cannot enlarge our knowledge beyond the concept, but only clear it. They cannot, therefore, be properly called dogmas (a word which might perhaps best be translated by precepts^ LeJirspriiche). According to our ordinary mode of speech, we could apply that name to that class only of the two above-mentioned classes of synthetical propositions a priori which refers to philosophical knowledge, and no one would feel inclined to give the name of Dogma to the propositions of arithmetic or geometry. In this way the usage of language confirms our explanation that those judgments only which are based on conceptions, and" not those which are arrived at by the construction of concepts, can be called dogmatic.

Now in the whole domain of pure reason, in its purely speculative use, there does not exist a single directly synthetical judgment based on concepts. We have shown that reason, by means of ideas, is incapable of any synthetical judgments which could claim objective validity, while by means of the concepts of our understanding it establishes no doubt some perfectly certain prin- [p. 737] ciples, but not directly from concepts, but indirectly only, by referring such concepts to something purely contingent, namely, possible experience. When such experience (anything as an object of possible experience) is presupposed, these principles are, no doubt, apodictically certain, but