Section 5
Original German
§ 5.
Wir haben oben den mächtigen Unterschied der analytischen und synthetischen Urtheile gesehen. Die Möglichkeit analytischer Sätze konnte sehr leicht begriffen werden; denn sie gründet sich lediglich auf dem Satze des Widerspruchs. Die Möglichkeit synthetischer Sätze a posteriori, d. i. solcher, welche aus der Erfahrung geschöpft werden, bedarf auch keiner besondern Erklärung; denn Erfahrung ist selbst nichts anders, als eine continuirliche Zusammenfügung (Synthesis) der Wahrnehmungen. Es bleiben uns also nur synthetische Sätze a priori übrig, deren Möglichkeit gesucht oder untersucht werden muß, weil sie auf anderen Principien als dem Satze des Widerspruchs beruhen muß.
Wir dürfen aber die Möglichkeit solcher Sätze hier nicht zuerst suchen, d. i. fragen, ob sie möglich seien. Denn es sind deren gnug und zwar mit unstreitiger Gewißheit wirklich gegeben, und da die Methode, die wir jetzt befolgen, analytisch sein soll, so werden wir davon anfangen, daß dergleichen synthetische, aber reine Vernunfterkenntniß wirklich sei; aber alsdann müssen wir den Grund dieser Möglichkeit dennoch untersuchen und fragen, wie diese Erkenntniß möglich sei, damit wir aus den Principien ihrer Möglichkeit die Bedingungen ihres Gebrauchs, den Umfang und die Grenzen desselben zu bestimmen in Stand gesetzt werden. Die eigentliche, mit schulgerechter Präcision ausgedrückte Aufgabe, auf die alles ankommt, ist also:
Wie sind synthetische Sätze a priori möglich?
Ich habe sie oben der Popularität zu Gefallen etwas anders, nämlich als eine Frage nach dem Erkenntniß aus reiner Vernunft, ausgedrückt, welches ich dieses Mal ohne Nachtheil der gesuchten Einsicht wohl thun konnte: weil, da es hier doch lediglich um die Metaphysik und deren Quellen zu thun ist, man nach den vorher gemachten Erinnerungen sich, wie ich hoffe, jederzeit erinnern wird, daß, wenn wir hier von Erkenntniß aus reiner Vernunft reden, niemals von der analytischen, sondern lediglich der synthetischen die Rede sei.276.1
Auf die Auflösung dieser Aufgabe nun kommt das Stehen oder Fallen der Metaphysik und also ihre Existenz gänzlich an. Es mag jemand seine Behauptungen in derselben mit noch so großem Schein vortragen, Schlüsse auf Schlüsse bis zum Erdrücken aufhäufen; wenn er nicht vorher jene Frage hat gnugthuend beantworten können, so habe ich Recht zu sagen: es ist alles eitele, grundlose Philosophie und falsche Weisheit. Du sprichst durch reine Vernunft und maßest dir an, a priori Erkenntnisse gleichsam zu erschaffen, indem du nicht blos gegebene Begriffe zergliederst, sondern neue Verknüpfungen vorgiebst, die nicht auf dem Satze des Widerspruchs beruhen, und die du doch so ganz unabhängig von aller Erfahrung einzusehen vermeinst; wie kommst du nun hiezu, und wie willst du dich wegen solcher Anmaßungen rechtfertigen? Dich auf Beistimmung der allgemeinen Menschenvernunft zu berufen, kann dir nicht gestattet werden; denn das ist ein Zeuge, dessen Ansehen nur auf dem öffentlichen Gerüchte beruht. Quodcumque ostendis mihi sic, incredulus odi. Horat.
So unentbehrlich aber die Beantwortung dieser Frage ist, so schwer ist sie doch zugleich; und obzwar die vornehmste Ursache, weswegen man sie nicht schon längst zu beantworten gesucht hat, darin liegt, daß man sich nicht einmal hat einfallen lassen, daß so etwas gefragt werden könne, so ist doch eine zweite Ursache diese, daß eine gnugthuende Beantwortung dieser einen Frage ein weit anhaltenderes, tieferes und mühsameres Nachdenken erfordert, als jemals das weitläuftigste Werk der Metaphysik, das bei der ersten Erscheinung seinem Verfasser Unsterblichkeit versprach. Auch muß ein jeder einsehende Leser, wenn er diese Aufgabe nach ihrer Forderung sorgfältig überdenkt, anfangs durch ihre Schwierigkeit erschreckt, sie für unauflöslich, und gäbe es nicht wirklich dergleichen reine synthetische Erkenntnisse a priori, sie ganz und gar für unmöglich halten; welches dem David Hume wirklich begegnete, ob er sich zwar die Frage bei weitem nicht in solcher Allgemeinheit vorstellte, als es hier geschieht und geschehen muß, wenn die Beantwortung für die ganze Metaphysik entscheidend werden soll. Denn wie ist es möglich, sagte der scharfsinnige Mann, daß, wenn mir ein Begriff gegeben ist, ich über denselben hinausgehen und einen andern damit verknüpfen kann, der in jenem gar nicht enthalten ist, und zwar so, als wenn dieser nothwendig zu jenem gehöre? Nur Erfahrung kann uns solche Verknüpfungen an die Hand geben (so schloß er aus jener Schwierigkeit, die er für Unmöglichkeit hielt), und alle jene vermeintliche Nothwendigkeit oder, welches einerlei ist, dafür gehaltene Erkenntniß a priori ist nichts als eine lange Gewohnheit, etwas wahr zu finden und daher die subjective Nothwendigkeit für objectiv zu halten.
Wenn der Leser sich über Beschwerde und Mühe beklagt, die ich ihm durch die Auflösung dieser Aufgabe machen werde, so darf er nur den Versuch anstellen, sie auf leichtere Art selbst aufzulösen. Vielleicht wird er sich alsdann demjenigen verbunden halten, der eine Arbeit von so tiefer Nachforschung für ihn übernommen hat, und wohl eher über die Leichtigkeit, die nach Beschaffenheit der Sache der Auflösung noch hat gegeben werden können, einige Verwunderung merken lassen; auch hat es Jahre lang Bemühung gekostet, um diese Aufgabe in ihrer ganzen Allgemeinheit (in dem Verstande, wie die Mathematiker dieses Wort nehmen, nämlich hinreichend für alle Fälle) aufzulösen und sie auch endlich in analytischer Gestalt, wie der Leser sie hier antreffen wird, darstellen zu können.
Alle Metaphysiker sind demnach von ihren Geschäften feierlich und gesetzmäßig so lange suspendirt, bis sie die Frage: Wie sind synthetische Erkenntnisse a priori möglich? gnugthuend werden beantwortet haben. Denn in dieser Beantwortung allein besteht das Creditiv, welches sie vorzeigen müßten, wenn sie im Namen der reinen Vernunft etwas bei uns anzubringen haben; in Ermangelung desselben aber können sie nichts anders erwarten, als von Vernünftigen, die so oft schon hintergangen worden, ohne alle weitere Untersuchung ihres Anbringens abgewiesen zu werden.
Wollten sie dagegen ihr Geschäfte nicht als Wissenschaft, sondern als eine Kunst heilsamer und dem allgemeinen Menschenverstande anpassender Überredungen treiben, so kann ihnen dieses Gewerbe nach Billigkeit nicht verwehrt werden. Sie werden alsdann die bescheidene Sprache eines vernünftigen Glaubens führen, sie werden gestehen, daß es ihnen nicht erlaubt sei, über das, was jenseit der Grenzen aller möglichen Erfahrung hinausliegt, auch nur einmal zu muthmaßen, geschweige etwas zu wissen, sondern nur etwas (nicht zum speculativen Gebrauche, denn auf den müssen sie Verzicht thun, sondern lediglich zum praktischen) anzunehmen, was zur Leitung des Verstandes und Willens im Leben möglich und sogar unentbehrlich ist. So allein werden sie den Namen nützlicher und weiser Männer führen können, um desto mehr, je mehr sie auf den der Metaphysiker Verzicht thun: denn diese wollen speculative Philosophen sein, und da, wenn es um Urtheile a priori zu thun ist, man es auf schale Wahrscheinlichkeiten nicht aussetzen kann (denn was dem Vorgeben nach a priori erkannt wird, wird eben dadurch als nothwendig angekündigt), so kann es ihnen nicht erlaubt sein, mit Muthmaßungen zu spielen, sondern ihre Behauptung muß Wissenschaft sein, oder sie ist überall gar nichts.
Man kann sagen, daß die ganze Transscendentalphilosophie, die vor aller Metaphysik nothwendig vorhergeht, selbst nichts anders als blos die vollständige Auflösung der hier vorgelegten Frage sei, nur in systematischer Ordnung und Ausführlichkeit, und man habe also bis jetzt keine Transscendentalphilosophie. Denn was den Namen davon führt, ist eigentlich ein Theil der Metaphysik; jene Wissenschaft soll aber die Möglichkeit der letzteren zuerst ausmachen und muß also vor aller Metaphysik vorhergehen. Man darf sich also auch nicht wundern, da eine ganze und zwar aller Beihülfe aus andern beraubte, mithin an sich ganz neue Wissenschaft nöthig ist, um nur eine einzige Frage hinreichend zu beantworten, wenn die Auflösung derselben mit Mühe und Schwierigkeit, ja sogar mit einiger Dunkelheit verbunden ist.
Indem wir jetzt zu dieser Auflösung schreiten und zwar nach analytischer Methode, in welcher wir voraussetzen, daß solche Erkenntnisse aus reiner Vernunft wirklich sind: so können wir uns nur auf zwei Wissenschaften der theoretischen Erkenntniß (als von der allein hier die Rede ist) berufen, nämlich reine Mathematik und reine Naturwissenschaft; denn nur diese können uns die Gegenstände in der Anschauung darstellen, mithin, wenn etwa in ihnen ein Erkenntniß a priori vorkäme, die Wahrheit oder Übereinstimmung derselben mit dem Objecte in concreto, d. i. ihre Wirklichkeit, zeigen, von der alsdann zu dem Grunde ihrer Möglichkeit auf dem analytischen Wege fortgegangen werden könnte. Dies erleichtert das Geschäfte sehr, in welchem die allgemeine Betrachtungen nicht allein auf Facta angewandt werden, sondern sogar von ihnen ausgehen, anstatt daß sie in synthetischem Verfahren gänzlich in abstracto aus Begriffen abgeleitet werden müssen.
Um aber von diesen wirklichen und zugleich gegründeten reinen Erkenntnissen a priori zu einer möglichen, die wir suchen, nämlich einer Metaphysik als Wissenschaft, aufzusteigen, haben wir nöthig, das, was sie veranlaßt und als blos natürlich gegebene, obgleich wegen ihrer Wahrheit nicht unverdächtige Erkenntniß a priori jener zum Grunde liegt, deren Bearbeitung ohne alle kritische Untersuchung ihrer Möglichkeit gewöhnlichermaßen schon Metaphysik genannt wird, mit einem Worte die Naturanlage zu einer solchen Wissenschaft, unter unserer Hauptfrage mit zu begreifen; und so wird die transscendentale Hauptfrage, in vier andere Fragen zertheilt, nach und nach beantwortet werden.
1) Wie ist reine Mathematik möglich?
2) Wie ist reine Naturwissenschaft möglich?
3) Wie ist Metaphysik überhaupt möglich?
4) Wie ist Metaphysik als Wissenschaft möglich?
Man sieht, daß, wenn gleich die Auflösung dieser Aufgaben hauptsächlich den wesentlichen Inhalt der Kritik darstellen soll, sie dennoch auch etwas Eigenthümliches habe, welches auch für sich allein der Aufmerksamkeit würdig ist, nämlich zu gegebenen Wissenschaften die Quellen in der Vernunft selbst zu suchen, um dadurch dieser ihr Vermögen, etwas a priori zu erkennen, vermittelst der Tat selbst zu erforschen und auszumessen; wodurch denn diese Wissenschaften selbst, wenn gleich nicht in Ansehung ihres Inhalts, doch was ihren richtigen Gebrauch betrifft, gewinnen und, indem sie einer höheren Frage, wegen ihres gemeinschaftlichen Ursprungs, Licht zu verschaffen, zugleich Anlaß geben, ihre eigene Natur besser aufzuklären.
276.1. Es ist unmöglich zu verhüten, daß, wenn die Erkenntniß nach und nach weiter fortrückt, nicht gewisse schon classisch gewordne Ausdrücke, die noch von dem Kindheitsalter der Wissenschaft her sind, in der Folge sollten unzureichend und übel anpassend gefunden werden, und ein gewisser neuer und mehr angemessener Gebrauch mit dem alten in einige Gefahr der Verwechselung gerathen sollte. Analytische Methode, sofern sie der synthetischen entgegengesetzt ist, ist ganz was anderes als ein Inbegriff analytischer Sätze: sie bedeutet nur, daß man von dem, was gesucht wird, als ob es gegeben sei, ausgeht und zu den Bedingungen aufsteigt, unter denen es allein möglich. In dieser Lehrart bedient man sich öfters lauter synthetischer Sätze, wie die mathematische Analysis davon ein Beispiel giebt; und sie könnte besser die regressive Lehrart zum Unterschiede von der synthetischen oder progressiven heißen. Noch kommt der Name Analytik auch als ein Haupttheil der Logik vor, und da ist es die Logik der Wahrheit und wird der Dialektik entgegengesetzt, ohne eigentlich darauf zu sehen, ob die zu jener gehörige Erkenntnisse analytisch oder synthetisch seien.
Musean translation
Mouseia’s complete machine-assisted Musean translation, made directly from the complete German of the Prolegomena (AA IV 253–383, all §§1–60 and the Appendix), for fidelity, the steps of Kant’s argument, clarity and structure.
§ 5.
Above we saw the profound difference between analytic and synthetic judgments. The possibility of analytic propositions could be understood very easily, for it rests solely on the principle of contradiction. The possibility of synthetic propositions a posteriori—that is, those drawn from experience—likewise needs no special explanation, since experience itself is nothing but a continuous joining together, or synthesis, of perceptions. All that remains, then, are synthetic propositions a priori, whose possibility must be sought or investigated, since it must rest on principles other than the principle of contradiction.
But here we need not first seek the possibility of such propositions, that is, ask whether they are possible. Enough of them are actually given, and with indisputable certainty. Since the method we are now to follow is analytic, we shall begin from the actuality of such synthetic but pure rational cognition. Yet we must then investigate the ground of this possibility and ask how this cognition is possible, so that the principles of its possibility may enable us to determine the conditions of its use, its scope, and its boundaries. The proper problem, expressed with scholarly precision, on which everything depends, is therefore:
How Are Synthetic Propositions A Priori Possible?
For the sake of popular accessibility I expressed it somewhat differently above, namely as a question about cognition from pure reason. On this occasion I could do so without detriment to the insight sought: since our sole concern here is metaphysics and its sources, I hope the preceding reminders will always bring to mind that when we speak here of cognition from pure reason, we never mean analytic cognition, but only synthetic cognition.276.1
The entire standing or falling of metaphysics, and thus its existence, depends on the solution of this problem. Someone may present his assertions in it with the greatest show of plausibility, piling inference upon inference until they overwhelm us; if he has not first been able to answer that question satisfactorily, I have the right to say: it is all vain, groundless philosophy and false wisdom. You speak through pure reason and presume, as it were, to create cognitions a priori: you do not merely analyze given concepts but claim new connections, which do not rest on the principle of contradiction and which you nevertheless suppose you discern quite independently of all experience. How, then, do you arrive at this, and how will you justify such presumptions? You cannot be permitted to appeal to the assent of universal human reason; that is a witness whose standing rests only on public rumor. Quodcumque ostendis mihi sic, incredulus odi. Horat. Whatever you show me in this way, I disbelieve and loathe.
Yet indispensable as the answer to this question is, it is equally difficult. Although the chief reason no one has long since sought to answer it is that no one even thought such a question could be asked, a second reason is that a satisfactory answer to this one question requires far more sustained, profound, and laborious reflection than the most extensive work of metaphysics ever did, though at its first appearance that work promised its author immortality. Indeed, every discerning reader who carefully considers this problem in the light of its demands must at first be alarmed by its difficulty and regard it as insoluble; and if such pure synthetic cognitions a priori did not actually exist, he would regard it as altogether impossible. This is what really happened to David Hume, although he did not represent the question to himself in nearly the universality it has here and must have if the answer is to decide the fate of all metaphysics. For how is it possible, this acute thinker asked, that when a concept is given to me I can go beyond it and connect with it another that is not contained in it at all, and do so as though the latter necessarily belonged to the former? Only experience can supply us with such connections—so he concluded from that difficulty, which he took for impossibility—and all that supposed necessity, or, what amounts to the same thing, cognition held to be a priori, is nothing but a long habit of finding something true and therefore taking subjective necessity for objective.
If the reader complains of the trouble and effort I shall cause him by solving this problem, he need only try to solve it himself in an easier way. Perhaps he will then feel indebted to the person who has undertaken such deeply searching work for him, and be more inclined to express some surprise at the ease with which the solution could still be presented, given the nature of the matter. Indeed, it took years of effort to solve this problem in its full universality—in the sense in which mathematicians use the word, namely as sufficient for every case—and finally to be able to present it in the analytic form in which the reader will find it here.
All metaphysicians are accordingly solemnly and lawfully suspended from their occupations until they have satisfactorily answered the question: How are synthetic cognitions a priori possible? For this answer alone constitutes the credential they would have to produce if they have anything to submit to us in the name of pure reason. Without it, they can expect nothing but to be turned away by reasonable people, who have already been deceived so often, without any further examination of their submission.
If, on the other hand, they wish to pursue their occupation not as a science but as an art of salutary persuasion suited to common human understanding, they cannot fairly be denied this trade. They will then speak the modest language of rational faith. They will acknowledge that they are not permitted even to conjecture, let alone know, anything about what lies beyond the boundaries of all possible experience, but only to assume something—not for speculative use, which they must renounce, but solely for practical use—that is possible and even indispensable for guiding the understanding and the will in life. Only in this way can they bear the name of useful and wise men, all the more so the more they renounce that of metaphysicians. For the latter wish to be speculative philosophers; and since judgments a priori cannot be staked on insipid probabilities—what is claimed to be known a priori is thereby announced as necessary—they cannot be permitted to play with conjectures. Their assertion must be science, or it is nothing at all.
We may say that the whole of transcendental philosophy, which must necessarily precede all metaphysics, is itself nothing but the complete solution of the question presented here, only in systematic order and full detail; and that hitherto, therefore, there has been no transcendental philosophy. For what goes by that name is really a part of metaphysics, whereas this science is first to establish the possibility of metaphysics and must therefore precede all of it. Nor should we be surprised that, when an entire science—and one deprived of all help from other sciences, thus wholly new in itself—is needed to answer a single question adequately, its solution involves labor and difficulty, indeed even some obscurity.
As we now proceed to this solution, and do so by the analytic method, in which we presuppose that such cognitions from pure reason are actual, we can appeal to only two sciences of theoretical cognition—the only kind at issue here—namely pure mathematics and pure natural science. Only these can exhibit objects to us in intuition and thus, if a cognition a priori occurs in them, show its truth, or agreement with the object, in concreto, that is, its actuality. From this we could then proceed along the analytic path to the ground of its possibility. This greatly eases the task, in which general considerations are not merely applied to facts but actually proceed from them; whereas in the synthetic procedure they must be derived wholly in abstracto from concepts.
But to ascend from these actual and at the same time well-founded pure cognitions a priori to a possible cognition of the kind we seek, namely metaphysics as a science, the cognition that occasions it and underlies it as merely naturally given cognition a priori—though not beyond suspicion as regards its truth—and whose cultivation, without any critical investigation of its possibility, is ordinarily already called metaphysics—in a word, the natural predisposition toward such a science—must also, under our principal question, be included; and thus the principal transcendental question, divided into four further questions, will be answered step by step.
1) How is pure mathematics possible?
2) How is pure natural science possible?
3) How is metaphysics in general possible?
4) How is metaphysics as a science possible?
We see that, although the solution of these problems is chiefly to present the essential content of the Critique, it nevertheless has something distinctive of its own that deserves attention even in its own right: namely, to seek for given sciences their sources in reason itself, and thus, through reason’s actual achievement, to investigate and measure its capacity to cognize something a priori. In this way these sciences themselves gain, if not in their content, at least in their proper use; and as they illuminate a higher question concerning their common origin, they also provide an occasion to clarify their own nature better.
276.1. As cognition gradually advances, it is impossible to prevent certain expressions that have already become canonical, but still date from the science’s childhood, from eventually proving inadequate and ill-suited, or a new and more appropriate use from running some risk of confusion with the old. The analytic method, insofar as it is opposed to the synthetic, is something entirely different from a collection of analytic propositions. It means only that we begin from what is sought as though it were given and ascend to the conditions under which alone it is possible. This manner of exposition often employs nothing but synthetic propositions, as mathematical analysis shows; and it might better be called the regressive manner of exposition, as distinct from the synthetic or progressive. The name “analytic” also occurs as a principal division of logic, where it is the logic of truth and is opposed to dialectic, without regard, properly speaking, to whether the cognitions belonging to it are analytic or synthetic.
Plain English translation
Mouseia’s complete Plain English edition, made independently and directly from the same complete German text.
§ 5.
Above we saw the great difference between analytic and synthetic judgments. The possibility of analytic propositions was very easy to understand, because it rests solely on the principle of contradiction. The possibility of a posteriori synthetic propositions—that is, propositions drawn from experience—needs no special explanation either. Experience itself is nothing but a continuous joining together, or synthesis, of perceptions. So only synthetic a priori propositions remain whose possibility must be sought or investigated, because it must rest on principles other than the principle of contradiction.
But here we do not first have to seek the possibility of such propositions, that is, ask whether they are possible. Enough of them are actually given, with indisputable certainty. Since the method we are now following is to be analytic, we will start from the fact that such synthetic but pure cognition through reason actually exists. Even so, we must then investigate the ground of this possibility and ask how this cognition is possible. From the principles of its possibility we can then determine the conditions of its use, and the scope and limits of that use. The actual problem on which everything depends, expressed with scholarly precision, is therefore:
How Are Synthetic A Priori Propositions Possible?
Above, for the sake of accessibility, I expressed it a little differently, as a question about cognition from pure reason. On this occasion I could do so without harming the insight we are seeking. For our concern here is solely metaphysics and its sources. After the earlier reminders, I hope readers will always remember that when we speak here of cognition from pure reason, we never mean analytic cognition, but only synthetic cognition.276.1
On the solution to this problem depends the entire standing or falling of metaphysics, and therefore its existence. Someone may present claims in metaphysics with the greatest appearance of plausibility, piling up inferences on inferences until they overwhelm us. If that person has not first been able to answer this question adequately, I am entitled to say: it is all empty, groundless philosophy and false wisdom. You speak through pure reason and claim, as it were, to create a priori cognitions. You do not merely analyze given concepts, but assert new connections that do not rest on the principle of contradiction, yet you think you understand them entirely independently of all experience. How do you manage this, and how will you justify such claims? You cannot be allowed to appeal to the agreement of universal human reason. That is a witness whose reputation rests only on public rumor. Whatever you show me in this way, I disbelieve and detest. Horat.
But as indispensable as the answer to this question is, it is also difficult. The main reason why people did not try to answer it long ago is that it never even occurred to them that such a question could be asked. But there is a second reason: an adequate answer to this one question requires much more sustained, deep, and laborious reflection than the most extensive work of metaphysics ever did, even one that promised its author immortality on its first appearance. Every insightful reader who carefully considers what this problem requires must at first be frightened by its difficulty and think it insoluble. If such pure synthetic a priori cognitions did not actually exist, the reader would consider it completely impossible. That is what actually happened to David Hume, although he did not put the question in nearly so general a form as we do here, and must do if the answer is to be decisive for all metaphysics. For how is it possible, that perceptive man asked, that when a concept is given to me, I can go beyond it and connect with it another concept not contained in it at all, as though the latter necessarily belonged to the former? Only experience can supply such connections. That was his conclusion from this difficulty, which he took for an impossibility. All that supposed necessity—or, what amounts to the same thing, supposed a priori cognition—is nothing but a long habit of finding something true and therefore taking subjective necessity for objective necessity.
If readers complain about the trouble and effort my solution to this problem will cause them, they need only try to solve it more easily themselves. Perhaps they will then feel indebted to the person who has undertaken such deeply searching work for them. They may even express some surprise at how easy the solution could still be made, given the nature of the subject. It also took years of effort to solve this problem in its full generality—in the sense in which mathematicians use that word, namely, adequately for all cases—and finally to be able to present it in the analytic form readers will find here.
All metaphysicians are therefore formally and lawfully suspended from their business until they have adequately answered the question: How are synthetic a priori cognitions possible? Their answer alone constitutes the credentials they would have to show if they had something to present to us in the name of pure reason. Without those credentials, they can expect nothing but rejection, without further examination of their proposals, by reasonable people who have so often been deceived already.
If, instead, they wanted to pursue their business not as a science, but as an art of persuasion, beneficial and suited to ordinary human understanding, then fairness would not allow us to forbid them that trade. They would then speak in the modest language of rational belief. They would admit that they are not permitted even to conjecture about what lies beyond the limits of all possible experience, much less to know anything about it. They may only assume something—not for speculative use, which they must give up, but solely for practical use—that is possible and even indispensable for guiding the understanding and the will in life. Only in this way can they be called useful and wise, and the more they renounce the name of metaphysicians, the better their claim to those names. For metaphysicians want to be speculative philosophers. Where a priori judgments are at issue, we cannot risk relying on shallow probabilities: what is claimed to be known a priori is, by that very claim, presented as necessary. So they cannot be allowed to play with conjectures. Their claim must be science, or it is nothing at all.
We may say that the whole of transcendental philosophy, which must necessarily precede all metaphysics, is itself nothing but the complete solution of the question presented here, only in systematic order and full detail. We have therefore had no transcendental philosophy until now. What goes by that name is actually a part of metaphysics. But the former science must first establish the possibility of the latter, and must therefore precede all metaphysics. So we should not be surprised if solving the question involves effort and difficulty, and even some obscurity. To answer just this one question adequately requires an entire science, deprived of all help from other sciences and therefore entirely new in itself.
We now proceed to this solution by the analytic method, in which we presuppose that such cognitions from pure reason actually exist. We can appeal to only two sciences of theoretical cognition, which alone is at issue here: pure mathematics and pure natural science. Only these can present objects to us in intuition. So if any a priori cognition occurs in them, they can show concretely its truth or agreement with the object—that is, its actuality. From there we could then proceed analytically to the ground of its possibility. This greatly eases the work. General considerations are not only applied to facts but even start from them, whereas in the synthetic procedure they must be derived entirely in the abstract from concepts.
But to rise from these actual and at the same time well-grounded pure a priori cognitions to a possible cognition that we seek, namely metaphysics as a science, we must take what occasions metaphysics and underlies it as a merely naturally given a priori cognition, though one whose truth is suspect, and whose elaboration without any critical investigation of its possibility is usually already called metaphysics—in a word, the natural predisposition to such a science—into account under our main question; and so the main transcendental question, divided into four other questions, will be answered step by step.
1) How is pure mathematics possible?
2) How is pure natural science possible?
3) How is metaphysics in general possible?
4) How is metaphysics as a science possible?
We can see that, although the solution to these problems is mainly intended to present the essential content of the Critique, it also has something distinctive that deserves attention in its own right. For given sciences, it seeks their sources in reason itself, in order to investigate and measure, through reason's actual achievement, its ability to know something a priori. These sciences thereby gain, if not in content, at least in their correct use. In shedding light on a higher question about their common origin, they also provide an opportunity to clarify their own nature better.
276.1. As cognition gradually advances, it is impossible to prevent certain expressions that have already become established, but date from the science's childhood, from eventually proving inadequate and ill-fitting. A new and more suitable use then risks being confused with the old one. The analytic method, insofar as it is opposed to the synthetic method, is quite different from a collection of analytic propositions. It means only that we start from what is being sought, as though it were given, and ascend to the conditions under which alone it is possible. This method often uses nothing but synthetic propositions, as mathematical analysis shows. It might better be called the regressive method, as distinct from the synthetic or progressive method. The name Analytics also occurs as a main division of logic. There it is the logic of truth and is opposed to Dialectic, without really considering whether the cognitions belonging to it are analytic or synthetic.
About this section
Section 5 is part of Prolegomena to Any Future Metaphysics by Immanuel Kant. This page sets the German original beside 2 direct-source translations, so any line can be checked against the source. It is section 9 of 72 in the work, which is kept in the European tradition.
Open the work page for the full contents, every edition, and the download links, or go back to Explore to browse other complete works in the European tradition. The reader keeps your place when you are signed in and lets you compare the original with each translation, one edition at a time.