--- type: article status: Published slug: "kant-prolegomena" title: "Prolegomena to Any Future Metaphysics" author: "Immanuel Kant" tradition: "european" language: "de" genre: "philosophy" period: "1783" source_url: "https://archive.org/details/kantsgesammeltes04kant" certainty: observed source: "https://archive.org/details/kantsgesammeltes04kant" --- # Prolegomena to Any Future Metaphysics ## Original ### Title Prolegomena zu einer jeden künftigen Metaphysik, die *als Wissenschaft* wird auftreten können, von Immanuel Kant. ### Preface Diese Prolegomena sind nicht zum Gebrauch für Lehrlinge, sondern für künftige Lehrer und sollen auch diesen nicht etwa dienen, um den Vortrag einer schon vorhandnen Wissenschaft anzuordnen, sondern um diese Wissenschaft selbst allererst zu erfinden. Es giebt Gelehrte, denen die Geschichte der Philosophie (der alten sowohl, als neuen) selbst ihre Philosophie ist; für diese sind gegenwärtige Prolegomena nicht geschrieben. Sie müssen warten, bis diejenigen, die aus den Quellen der Vernunft selbst zu schöpfen bemüht sind, ihre Sache werden ausgemacht haben, und alsdann wird an ihnen die Reihe sein, von dem Geschehenen der Welt Nachricht zu geben. Widrigenfalls kann nichts gesagt werden, was ihrer Meinung nach nicht schon sonst gesagt worden ist, und in der That mag dieses auch als eine untrügliche Vorhersagung für alles Künftige gelten; denn da der menschliche Verstand über unzählige Gegenstände viele Jahrhunderte hindurch auf mancherlei Weise geschwärmt hat, so kann es nicht leicht fehlen, daß nicht zu jedem Neuen etwas Altes gefunden werden sollte, was damit einige Ähnlichkeit hätte. Meine Absicht ist, alle diejenigen, so es werth finden, sich mit Metaphysik zu beschäftigen, zu überzeugen, daß es unumgänglich nothwendig sei, ihre Arbeit vor der Hand auszusetzen, alles bisher Geschehene als ungeschehen anzusehen und vor allen Dingen zuerst die Frage aufzuwerfen: ob auch so etwas als Metaphysik überall nur möglich sei. Ist sie Wissenschaft, wie kommt es, daß sie sich nicht wie andre Wissenschaften in allgemeinen und daurenden Beifall setzen kann? Ist sie keine, wie geht es zu, daß sie doch unter dem Scheine einer Wissenschaft unaufhörlich groß thut und den menschlichen Verstand mit niemals erlöschenden, aber nie erfüllten Hoffnungen hinhält? Man mag also entweder sein Wissen oder Nichtwissen demonstriren, so muß doch einmal über die Natur dieser angemaßten Wissenschaft etwas Sicheres ausgemacht werden; denn auf demselben Fuße kann es mit ihr unmöglich länger bleiben. Es scheint beinahe belachenswerth, indessen daß jede andre Wissenschaft unaufhörlich fortrückt, sich in dieser, die doch die Weisheit selbst sein will, deren Orakel jeder Mensch befrägt, beständig auf derselben Stelle herumzudrehen, ohne einen Schritt weiter zu kommen. Auch haben sich ihre Anhänger gar sehr verloren, und man sieht nicht, daß diejenigen, die sich stark genug fühlen, in andern Wissenschaften zu glänzen, ihren Ruhm in dieser wagen wollen, wo jedermann, der sonst in allen übrigen Dingen unwissend ist, sich ein entscheidendes Urtheil anmaßt, weil in diesem Lande in der That noch kein sicheres Maß und Gewicht vorhanden ist, um Gründlichkeit von seichtem Geschwätze zu unterscheiden. Es ist aber eben nicht so etwas Unerhörtes, daß nach langer Bearbeitung einer Wissenschaft, wenn man Wunder denkt, wie weit man schon darin gekommen sei, endlich sich jemand die Frage einfallen läßt: ob und wie überhaupt eine solche Wissenschaft möglich sei. Denn die menschliche Vernunft ist so baulustig, daß sie mehrmals schon den Thurm aufgeführt, hernach aber wieder abgetragen hat, um zu sehen, wie das Fundament desselben wohl beschaffen sein möchte. Es ist niemals zu spät, vernünftig und weise zu werden; es ist aber jederzeit schwerer, wenn die Einsicht spät kommt, sie in Gang zu bringen. Zu fragen, ob eine Wissenschaft auch wohl möglich sei, setzt voraus, daß man an der Wirklichkeit derselben zweifle. Ein solcher Zweifel aber beleidigt jedermann, dessen ganze Habseligkeit vielleicht in diesem vermeinten Kleinode bestehen möchte; und daher mag sich der, so sich diesen Zweifel entfallen läßt, nur immer auf Widerstand von allen Seiten gefaßt machen. Einige werden in stolzem Bewußtsein ihres alten und eben daher für rechtmäßig gehaltenen Besitzes mit ihren metaphysischen Compendien in der Hand auf ihn mit Verachtung herabsehen; andere, die nirgend etwas sehen, als was mit dem einerlei ist, was sie schon sonst irgendwo gesehen haben, werden ihn nicht verstehen, und alles wird einige Zeit hindurch so bleiben, als ob gar nichts vorgefallen wäre, was eine nahe Veränderung besorgen oder hoffen ließe. Gleichwohl getraue ich mir vorauszusagen, daß der selbstdenkende Leser dieser Prolegomenen nicht blos an seiner bisherigen Wissenschaft zweifeln, sondern in der Folge gänzlich überzeugt sein werde, daß es dergleichen gar nicht geben könne, ohne daß die hier geäußerte Forderungen geleistet werden, auf welchen ihre Möglichkeit beruht, und da dieses noch niemals geschehen, daß es überall noch keine Metaphysik gebe. Da sich indessen die Nachfrage nach ihr doch auch niemals verlieren kann[^257-1], weil das Interesse der allgemeinen Menschenvernunft mit ihr gar zu innigst verflochten ist, so wird er gestehen, daß eine völlige Reform, oder vielmehr eine neue Geburt derselben nach einem bisher ganz unbekannten Plane unausbleiblich bevorstehe, man mag sich nun eine Zeitlang dagegen sträuben, wie man wolle. Seit *Lockes* und *Leibnizens* Versuchen, oder vielmehr seit dem Entstehen der Metaphysik, so weit die Geschichte derselben reicht, hat sich keine Begebenheit zugetragen, die in Ansehung des Schicksals dieser Wissenschaft hätte entscheidender werden können, als der Angriff, den *David Hume* auf dieselbe machte. Er brachte kein Licht in diese Art von Erkenntniß, aber er schlug doch einen Funken, bei welchem man wohl ein Licht hätte anzünden können, wenn er einen empfänglichen Zunder getroffen hätte, dessen Glimmen sorgfältig wäre unterhalten und vergrößert worden. *Hume* ging hauptsächlich von einem einzigen, aber wichtigen Begriffe der Metaphysik, nämlich dem der *Verknüpfung der Ursache und Wirkung*, (mithin auch dessen Folgebegriffe der Kraft und Handlung etc.) aus und forderte die Vernunft, die da vorgiebt, ihn in ihrem Schooße erzeugt zu haben, auf, ihm Rede und Antwort zu geben, mit welchem Rechte sie sich denkt: daß etwas so beschaffen sein könne, daß, wenn es gesetzt ist, dadurch auch etwas anderes nothwendig gesetzt werden müsse; denn das sagt der Begriff der Ursache. Er bewies unwidersprechlich: daß es der Vernunft gänzlich unmöglich sei, a priori und aus Begriffen eine solche Verbindung zu denken, denn diese enthält Nothwendigkeit; es ist aber gar nicht abzusehen, wie darum, weil Etwas ist, etwas anderes nothwendiger Weise auch sein müsse, und wie sich also der Begriff von einer solchen Verknüpfung a priori einführen lasse. Hieraus schloß er, daß die Vernunft sich mit diesem Begriffe ganz und gar betrüge, daß sie ihn fälschlich für ihr eigen Kind halte, da er doch nichts anders als ein Bastard der Einbildungskraft sei, die, durch Erfahrung beschwängert, gewisse Vorstellungen unter das Gesetz der Association gebracht hat und eine daraus entspringende subjective Nothwendigkeit, d. i. Gewohnheit, für eine objective, aus Einsicht, unterschiebt. Hieraus schloß er: die Vernunft habe gar kein Vermögen, solche Verknüpfungen auch selbst nur im Allgemeinen zu denken, weil ihre Begriffe alsdann bloße Erdichtungen sein würden; und alle ihre vorgeblich a priori bestehende Erkenntnisse wären nichts als falsch gestempelte gemeine Erfahrungen, welches eben so viel sagt, als: es gebe überall keine Metaphysik und könne auch keine geben.[^258-1] So übereilt und unrichtig auch seine Folgerung war, so war sie doch wenigstens auf Untersuchung gegründet; und diese Untersuchung war es wohl werth, daß sich die guten Köpfe seiner Zeit vereinigt hätten, die Aufgabe in dem Sinne, wie er sie vortrug, wo möglich glücklicher aufzulösen, woraus denn bald eine gänzliche Reform der Wissenschaft hätte entspringen müssen. Allein das der Metaphysik von je her ungünstige Schicksal wollte, daß er von keinem verstanden wurde. Man kann es, ohne eine gewisse Pein zu empfinden, nicht ansehen, wie so ganz und gar seine Gegner *Reid*, *Oswald*, *Beattie* und zuletzt noch *Priestley* den Punkt seiner Aufgabe verfehlten und, indem sie immer das als zugestanden annahmen, was er eben bezweifelte, dagegen aber mit Heftigkeit und mehrentheils mit großer Unbescheidenheit dasjenige bewiesen, was ihm niemals zu bezweifeln in den Sinn gekommen war, seinen Wink zur Verbesserung so verkannten, daß alles in dem alten Zustande blieb, als ob nichts geschehen wäre. Es war nicht die Frage, ob der Begriff der Ursache richtig, brauchbar und in Ansehung der ganzen Naturerkenntniß unentbehrlich sei, denn dieses hatte *Hume* niemals in Zweifel gezogen; sondern ob er durch die Vernunft a priori, gedacht werde und auf solche Weise eine von aller Erfahrung unabhängige innre Wahrheit und daher auch wohl weiter ausgedehnte Brauchbarkeit habe, die nicht blos auf Gegenstände der Erfahrung eingeschränkt sei: hierüber erwartete *Hume* Eröffnung. Es war ja nur die Rede von dem Ursprunge dieses Begriffs, nicht von der Unentbehrlichkeit desselben im Gebrauche: wäre jener nur ausgemittelt, so würde es sich wegen der Bedingungen seines Gebrauches und des Umfangs, in welchem er gültig sein kann, schon von selbst gegeben haben. Die Gegner des berühmten Mannes hätten aber, um der Aufgabe ein Gnüge zu thun, sehr tief in die Natur der Vernunft, so fern sie blos mit reinem Denken beschäftigt ist, hineindringen müssen, welches ihnen ungelegen war. Sie erfanden daher ein bequemeres Mittel, ohne alle Einsicht trotzig zu thun, nämlich die Berufung auf den *gemeinen Menschenverstand*. In der That ists eine große Gabe des Himmels, einen geraden (oder, wie man es neuerlich benannt hat, schlichten) Menschenverstand zu besitzen. Aber man muß ihn durch Thaten beweisen, durch das Überlegte und Vernünftige, was man denkt und sagt, nicht aber dadurch, daß, wenn man nichts Kluges zu seiner Rechtfertigung vorzubringen weiß, man sich auf ihn als ein Orakel beruft. Wenn Einsicht und Wissenschaft auf die Neige gehen, alsdann und nicht eher sich auf den gemeinen Menschenverstand zu berufen, das ist eine von den subtilen Erfindungen neuerer Zeiten, dabei es der schalste Schwätzer mit dem gründlichsten Kopfe getrost aufnehmen und es mit ihm aushalten kann. So lange aber noch ein kleiner Rest von Einsicht da ist, wird man sich wohl hüten, diese Nothhülfe zu ergreifen. Und beim Lichte besehen, ist diese Appellation nichts anders als eine Berufung auf das Urtheil der Menge; ein Zuklatschen, über das der Philosoph erröthet, der populäre Witzling aber triumphirt und trotzig thut Ich sollte aber doch denken, *Hume* habe auf einen gesunden Verstand eben so wohl Anspruch machen können, als *Beattie* und noch überdem auf das, was dieser gewiß nicht besaß, nämlich eine kritische Vernunft, die den gemeinen Verstand in Schranken hält, damit er sich nicht in Speculationen versteige, oder, wenn blos von diesen die Rede ist, nichts zu entscheiden begehre, weil er sich über seine Grundsätze nicht zu rechtfertigen versteht; denn nur so allein wird er ein gesunder Verstand bleiben. Meißel und Schlägel können ganz wohl dazu dienen, ein Stück Zimmerholz zu bearbeiten, aber zum Kupferstechen muß man die Radirnadel brauchen. So sind gesunder Verstand sowohl als speculativer beide, aber jeder in seiner Art brauchbar: jener, wenn es auf Urtheile ankommt, die in der Erfahrung ihre unmittelbare Anwendung finden, dieser aber, wo im Allgemeinen, aus bloßen Begriffen, geurtheilt werden soll, z. B. in der Metaphysik, wo der sich selbst, aber oft per antiphrasin so nennende gesunde Verstand ganz und gar kein Urtheil hat. Ich gestehe frei: die Erinnerung des *David Hume* war eben dasjenige, was mir vor vielen Jahren zuerst den dogmatischen Schlummer unterbrach und meinen Untersuchungen im Felde der speculativen Philosophie eine ganz andre Richtung gab. Ich war weit entfernt, ihm in Ansehung seiner Folgerungen Gehör zu geben, die blos daher rührten, weil er sich seine Aufgabe nicht im Ganzen vorstellte, sondern nur auf einen Theil derselben fiel, der, ohne das Ganze in Betracht zu ziehen, keine Auskunft geben kann. Wenn man von einem gegründeten, obzwar nicht ausgeführten Gedanken anfängt, den uns ein anderer hinterlassen, so kann man wohl hoffen, es bei fortgesetztem Nachdenken weiter zu bringen, als der scharfsinnige Mann kam, dem man den ersten Funken dieses Lichts zu verdanken hatte. Ich versuchte also zuerst, ob sich nicht *Hume’s* Einwurf allgemein vorstellen ließe, und fand bald: daß der Begriff der Verknüpfung von Ursache und Wirkung bei weitem nicht der einzige sei, durch den der Verstand a priori sich Verknüpfungen der Dinge denkt, vielmehr daß Metaphysik ganz und gar daraus bestehe. Ich suchte mich ihrer Zahl zu versichern, und da dieses mir nach Wunsch, nämlich aus einem einzigen Princip, gelungen war, so ging ich an die Deduction dieser Begriffe, von denen ich nunmehr versichert war, daß sie nicht, wie *Hume* besorgt hatte, von der Erfahrung abgeleitet, sondern aus dem reinem Verstande entsprungen seien. Diese Deduction, die meinem scharfsinnigen Vorgänger unmöglich schien, die niemand außer ihm sich auch nur hatte einfallen lassen, obgleich jedermann sich der Begriffe getrost bediente, ohne zu fragen, worauf sich denn ihre objective Gültigkeit gründe, diese, sage ich, war das Schwerste, das jemals zum Behuf der Metaphysik unternommen werden konnte; und was noch das Schlimmste dabei ist, so konnte mir Metaphysik, so viel davon nur irgendwo vorhanden ist, hiebei auch nicht die mindeste Hülfe leisten, weil jene Deduction zuerst die Möglichkeit einer Metaphysik ausmachen soll. Da es mir nun mit der Auflösung des Humischen Problems nicht blos in einem besondern Falle, sondern in Absicht auf das ganze Vermögen der reinen Vernunft gelungen war: so konnte ich sichere, obgleich immer nur langsame Schritte thun, um endlich den ganzen Umfang der reinen Vernunft in seinen Grenzen sowohl, als seinem Inhalt vollständig und nach allgemeinen Principien zu bestimmen, welches denn dasjenige war, was Metaphysik bedarf, um ihr System nach einem sicheren Plan aufzuführen. Ich besorge aber, daß es der *Ausführung* des Humischen Problems in seiner möglich größten Erweiterung (nämlich der Kritik der reinen Vernunft) eben so gehen dürfte, als es dem *Problem* selbst erging, da es zuerst vorgestellt wurde. Man wird sie unrichtig beurtheilen, weil man sie nicht versteht; man wird sie nicht verstehen, weil man das Buch zwar durchzublättern, aber nicht durchzudenken Lust hat; und man wird diese Bemühung darauf nicht verwenden wollen, weil das Werk trocken, weil es dunkel, weil es allen gewohnten Begriffen widerstreitend und überdem weitläuftig ist. Nun gestehe ich, daß es mir unerwartet sei, von einem Philosophen Klagen wegen Mangels an Popularität, Unterhaltung und Gemächlichkeit zu hören, wenn es um die Existenz einer gepriesenen und der Menschheit unentbehrlichen Erkenntniß selbst zu thun ist, die nicht anders als nach den strengsten Regeln einer schulgerechten Pünktlichkeit ausgemacht werden kann, auf welche zwar mit der Zeit auch Popularität folgen, aber niemals den Anfang machen darf. Allein was eine gewisse Dunkelheit betrifft, die zum Theil von der Weitläuftigkeit des Plans herrührt, bei welcher man die Hauptpunkte, auf die es bei der Untersuchung ankommt, nicht wohl übersehen kann: so ist die Beschwerde deshalb gerecht, und dieser werde ich durch gegenwärtige *Prolegomena* abhelfen. Jenes Werk, welches das reine Vernunftvermögen in seinem ganzen Umfange und Grenzen darstellt, bleibt dabei immer die Grundlage, worauf sich die Prolegomena nur als Vorübungen beziehen; denn jene Kritik muß als Wissenschaft systematisch und bis zu ihren kleinsten Theilen vollständig dastehen, ehe noch daran zu denken ist, Metaphysik auftreten zu lassen, oder sich auch nur eine entfernte Hoffnung zu derselben zu machen. Man ist es schon lange gewohnt, alte, abgenutzte Erkenntnisse dadurch neu aufgestutzt zu sehen, daß man sie aus ihren vormaligen Verbindungen herausnimmt, ihnen ein systematisches Kleid nach eigenem beliebigen Schnitte, aber unter neuen Titeln anpaßt; und nichts anders wird der größte Theil der Leser auch von jener Kritik zum voraus erwarten. Allein diese Prolegomena werden ihn dahin bringen, einzusehen, daß es eine ganz neue Wissenschaft sei, von welcher niemand auch nur den Gedanken vorher gefaßt hatte, wovon selbst die bloße Idee unbekannt war, und wozu von allem bisher Gegebenen nichts genutzt werden konnte, als allein der Wink, den *Humes* Zweifel geben konnten, der gleichfalls nichts von einer dergleichen möglichen förmlichen Wissenschaft ahndete, sondern sein Schiff, um es in Sicherheit zu bringen, auf den Strand (den Scepticism) setzte, da es denn liegen und verfaulen mag, statt dessen es bei mir darauf ankommt, ihm einen Piloten zu geben, der nach sicheren Principien der Steuermannskunst, die aus der Kenntniß des Globus gezogen sind, mit einer vollständigen Seekarte und einem Compaß versehen, das Schiff sicher führen könne, wohin es ihm gut dünkt. Zu einer neuen Wissenschaft, die gänzlich isolirt und die einzige ihrer Art ist, mit dem Vorurtheil gehen, als könne man sie vermittelst seiner schon sonst erworbenen vermeinten Kenntnisse beurtheilen, obgleich die es eben sind, an deren Realität zuvor gänzlich gezweifelt werden muß, bringt nichts anders zuwege, als daß man allenthalben das zu sehen glaubt, was einem schon sonst bekannt war, weil etwa die Ausdrücke jenem ähnlich lauten: nur daß einem alles äußerst verunstaltet, widersinnisch und kauderwelsch vorkommen muß, weil man nicht die Gedanken des Verfassers, sondern immer nur seine eigene, durch lange Gewohnheit zur Natur gewordene Denkungsart dabei zum Grunde legt. Aber die Weitläuftigkeit des Werks, so fern sie in der Wissenschaft selbst und nicht dem Vortrage gegründet ist, die dabei unvermeidliche Trockenheit und scholastische Pünktlichkeit sind Eigenschaften, die zwar der Sache selbst überaus vortheilhaft sein mögen, dem Buche selbst aber allerdings nachtheilig werden müssen. Es ist zwar nicht jedermann gegeben, so subtil und doch zugleich so anlockend zu schreiben als *David Hume*, oder so gründlich und dabei so elegant als *Moses Mendelssohn*; allein Popularität hätte ich meinem Vortrage (wie ich mir schmeichele) wohl geben können, wenn es mir nur darum zu thun gewesen wäre, einen Plan zu entwerfen und dessen Vollziehung andern anzupreisen, und mir nicht das Wohl der Wissenschaft, die mich so lange beschäftigt hielt, am Herzen gelegen hätte; denn übrigens gehörte viel Beharrlichkeit und auch selbst nicht wenig Selbstverläugnung dazu, die Anlockung einer früheren, günstigen Aufnahme der Aussicht auf einen zwar späten, aber dauerhaften Beifall nachzusetzen. *Plane machen* ist mehrmals eine üppige prahlerische Geistesbeschäftigung, dadurch man sich ein Ansehen von schöpferischem Genie giebt, indem man fordert, was man selbst nicht leisten, tadelt, was man doch nicht besser machen kann, und vorschlägt, wovon man selbst nicht weiß, wo es zu finden ist; wiewohl auch nur zum tüchtigen Plane einer allgemeinen Kritik der Vernunft schon etwas mehr gehört hätte, als man wohl vermuthen mag, wenn er nicht blos wie gewöhnlich eine Declamation frommer Wünsche hätte werden sollen. Allein reine Vernunft ist eine so abgesonderte, in ihr selbst so durchgängig verknüpfte Sphäre, daß man keinen Theil derselben antasten kann, ohne alle übrige zu berühren, und nichts ausrichten kann, ohne vorher jedem seine Stelle und seinen Einfluß auf den andern bestimmt zu haben: weil, da nichts außer derselben ist, was unser Urtheil innerhalb berichtigen könnte, jedes Theiles Gültigkeit und Gebrauch von dem Verhältnisse abhängt, darin er gegen die übrige in der Vernunft selbst steht, und wie bei dem Gliederbau eines organisirten Körpers der Zweck jedes Gliedes nur aus dem vollständigen Begriff des Ganzen abgeleitet werden kann. Daher kann man von einer solchen Kritik sagen, daß sie niemals zuverlässig sei, wenn sie nicht *ganz* und bis auf die mindesten Elemente der reinen Vernunft *vollendet* ist, und daß man von der Sphäre dieses Vermögens entweder *alles*, oder *nichts* bestimmen und ausmachen müsse. Ob aber gleich ein bloßer Plan, der vor der *Kritik* der reinen Vernunft vorhergehen möchte, unverständlich, unzuverlässig und unnütz sein würde, so ist er dagegen um desto nützlicher, wenn er darauf folgt. Denn dadurch wird man in den Stand gesetzt, das Ganze zu übersehen, die Hauptpunkte, worauf es bei dieser Wissenschaft ankommt, stückweise zu prüfen und manches dem Vortrage nach besser einzurichten, als es in der ersten Ausfertigung des Werks geschehen konnte. Hier ist nun ein solcher *Plan* nach vollendetem Werke, der nunmehr nach *analytischer Methode* angelegt sein darf, da das Werk selbst durchaus nach *synthetischer Lehrart* abgefaßt sein mußte, damit die Wissenschaft alle ihre Articulationen, als den Gliederbau eines ganz besondern Erkenntnißvermögens, in seiner natürlichen Verbindung vor Augen stelle. Wer diesen Plan, den ich als Prolegomena vor aller künftigen Metaphysik voranschicke, selbst wiederum dunkel findet, der mag bedenken: daß es eben nicht nöthig sei, daß jedermann Metaphysik studire, daß es manches Talent gebe, welches in gründlichen und selbst tiefen Wissenschaften, die sich mehr der Anschauung nähern, ganz wohl fortkommt, dem es aber mit Nachforschungen durch lauter abgezogene Begriffe nicht gelingen will, und daß man seine Geistesgaben in solchem Fall auf einen andern Gegenstand verwenden müsse; daß aber derjenige, der Metaphysik zu beurtheilen, ja selbst eine abzufassen unternimmt, den Forderungen, die hier gemacht werden, durchaus ein Gnüge thun müsse, es mag nun auf die Art geschehen, daß er meine Auflösung annimmt, oder sie auch gründlich widerlegt und eine andere an deren Stelle setzt — denn abweisen kann er sie nicht —, und daß endlich die so beschrieene Dunkelheit (eine gewohnte Bemäntelung seiner eigenen Gemächlichkeit oder Blödsichtigkeit) auch ihren Nutzen habe: da alle, die in Ansehung aller andern Wissenschaften ein behutsames Stillschweigen beobachten, in Fragen der Metaphysik meisterhaft sprechen und dreust entscheiden, weil ihre Unwissenheit hier freilich nicht gegen anderer Wissenschaft deutlich absticht, wohl aber gegen ächte kritische Grundsätze, von denen man also rühmen kann: ignavum, fucos, pecus a praesepibus arcent. *Virg.* [^257-1]: Rusticus exspectat, dum defluat amnis; at ille Labitur et labetur in omne volubilis aevum. *Horat.* [^258-1]: Gleichwohl nannte *Hume* eben diese zerstörende Philosophie selbst Metaphysik und legte ihr einen hohen Werth bei. „Metaphysik und Moral“, sagt er (Versuche 4ter Theil, Seite 214, deutsche Übers.), „sind die wichtigsten Zweige der Wissenschaft; Mathematik und Naturwissenschaft sind nicht halb so viel werth.“ Der scharfsinnige Mann sah aber hier blos auf den negativen Nutzen, den die Mäßigung der übertriebenen Ansprüche der speculativen Vernunft haben würde, um so viel endlose und verfolgende Streitigkeiten, die das Menschengeschlecht verwirren, gänzlich aufzuheben; aber er verlor darüber den positiven Schaden aus den Augen, der daraus entspringt, wenn der Vernunft die wichtigsten Aussichten genommen werden, nach denen allein sie dem Willen das höchste Ziel aller seiner Bestrebungen ausstecken kann. ### Preliminary reminder Prolegomena. Vorerinnerung von dem *Eigenthümlichen aller metaphysischen Erkenntniß.* ### Section 1 § 1. *Von den Quellen der Metaphysik.* Wenn man eine Erkenntniß als *Wissenschaft* darstellen will, so muß man zuvor das Unterscheidende, was sie mit keiner andern gemein hat, und was ihr also *eigenthümlich* ist, genau bestimmen können; widrigenfalls die Grenzen aller Wissenschaften in einander laufen, und keine derselben ihrer Natur nach gründlich abgehandelt werden kann. Dieses Eigenthümliche mag nun in dem Unterschiede des *Objects* oder der *Erkenntnißquellen* oder auch der *Erkenntnißart* oder einiger, wo nicht aller dieser Stücke zusammen bestehen, so beruht darauf zuerst die Idee der möglichen Wissenschaft und ihres Territorium. Zuerst, was die *Quellen* einer metaphysischen Erkenntniß betrifft, so liegt es schon in ihrem Begriffe, daß sie nicht empirisch sein können. Die Principien derselben (wozu nicht blos ihre Grundsätze, sondern auch Grundbegriffe gehören) müssen also niemals aus der Erfahrung genommen sein: denn sie soll nicht physische, sondern metaphysische, d. i. jenseit der Erfahrung liegende, Erkenntniß sein. Also wird weder äußere Erfahrung, welche die Quelle der eigentlichen Physik, noch innere, welche die Grundlage der empirischen Psychologie ausmacht, bei ihr zum Grunde liegen. Sie ist also Erkenntniß a priori, oder aus reinem Verstande und reiner Vernunft. Hierin würde sie aber nichts Unterscheidendes von der reinen Mathematik haben; sie wird also *reine philosophische* Erkenntniß heißen müssen; wegen der Bedeutung dieses Ausdrucks aber beziehe ich mich auf Kritik d. r. V. Seite 712 u. f., wo der Unterschied dieser zwei Arten des Vernunftgebrauchs einleuchtend und gnugthuend ist dargestellt worden. — So viel von den Quellen der metaphysischen Erkenntniß. ### Section 2 § 2. *Von der Erkenntnißart, die allein metaphysisch heißen kann.* a) *Von dem Unterschiede synthetischer und analytischer Urtheile überhaupt.* Metaphysische Erkenntniß muß lauter Urtheile a priori enthalten, das erfordert das Eigenthümliche ihrer Quellen. Allein Urtheile mögen nun einen Ursprung haben, welchen sie wollen, oder auch ihrer logischen Form nach beschaffen sein, wie sie wollen, so giebt es doch einen Unterschied derselben dem Inhalte nach, vermöge dessen sie entweder blos *erläuternd* sind und zum Inhalte der Erkenntniß nichts hinzuthun, oder *erweiternd* und die gegebene Erkenntniß vergrößern; die erstern werden *analytische*, die zweiten *synthetische* Urtheile genannt werden können. Analytische Urtheile sagen im Prädicate nichts als das, was im Begriffe des Subjects schon wirklich, obgleich nicht so klar und mit gleichem Bewußtsein gedacht war. Wenn ich sage: alle Körper sind ausgedehnt, so habe ich meinen Begriff vom Körper nicht im mindesten erweitert, sondern ihn nur aufgelöset, indem die Ausdehnung von jenem Begriffe schon vor dem Urtheile, obgleich nicht ausdrücklich gesagt, dennoch wirklich gedacht war; das Urtheil ist also analytisch. Dagegen enthält der Satz: einige Körper sind schwer, etwas im Prädicate, was in dem allgemeinen Begriffe vom Körper nicht wirklich gedacht wird; er vergrößert also meine Erkenntniß, indem er zu meinem Begriffe etwas hinzuthut, und muß daher ein synthetisches Urtheil heißen. b) *Das gemeinschaftliche Prinzip aller analytischen Urtheile ist der Satz des Widerspruchs.* Alle analytische Urtheile beruhen gänzlich auf dem Satze des Widerspruchs und sind ihrer Natur nach Erkenntnisse a priori, die Begriffe, die ihnen zur Materie dienen, mögen empirisch sein, oder nicht. Denn weil das Prädicat eines bejahenden analytischen Urtheils schon vorher im Begriffe des Subjects gedacht wird, so kann es von ihm ohne Widerspruch nicht verneint werden; ebenso wird sein Gegentheil in einem analytischen, aber verneinenden Urtheile nothwendig von dem Subject verneint und zwar auch zufolge dem Satze des Widerspruchs. So ist es mit den Sätzen: Jeder Körper ist ausgedehnt, und: Kein Körper ist unausgedehnt (einfach), beschaffen. Eben darum sind auch alle analytische Sätze Urtheile a priori, wenn gleich ihre Begriffe empirisch sind, z. B. Gold ist ein gelbes Metall; denn um dieses zu wissen, brauche ich keiner weitern Erfahrung außer meinem Begriffe vom Golde, der enthielte, daß dieser Körper gelb und Metall sei: denn dieses machte eben meinen Begriff aus, und ich durfte nichts thun als diesen zergliedern, ohne mich außer demselben wornach anders umzusehen. c) *Synthetische Urtheile bedürfen ein anderes Prinzip als den Satz des Widerspruchs.* Es giebt synthetische Urtheile a posteriori, deren Ursprung empirisch ist; aber es giebt auch deren, die a priori gewiß sind, und die aus reinem Verstande und Vernunft entspringen. Beide kommen aber darin überein, daß sie nach dem Grundsatze der Analysis, nämlich dem Satze des Widerspruchs, allein nimmermehr entspringen können; sie erfordern noch ein ganz anderes Prinzip, ob sie zwar aus jedem Grundsatze, welcher er auch sei, jederzeit dem Satze des Widerspruchs *gemäß* abgeleitet werden müssen, denn nichts darf diesem Grundsatze zuwider sein, obgleich eben nicht alles daraus abgeleitet werden kann. Ich will die synthetischen Urtheile zuvor unter Classen bringen. 1) *Erfahrungsurtheile* sind jederzeit synthetisch. Denn es wäre ungereimt, ein analytisches Urtheil auf Erfahrung zu gründen, da ich doch aus meinem Begriffe gar nicht hinausgehen darf, um das Urtheil abzufassen, und also kein Zeugniß der Erfahrung dazu nötig habe. Daß ein Körper ausgedehnt sei, ist ein Satz, der a priori feststeht, und kein Erfahrungsurtheil. Denn ehe ich zur Erfahrung gehe, habe ich alle Bedingungen zu meinem Urtheile schon in dem Begriffe, aus welchem ich das Prädicat nach dem Satze des Widerspruchs nur herausziehen und dadurch zugleich der *Nothwendigkeit* des Urtheils bewußt werden kann, welche mir Erfahrung nicht einmal lehren würde. 2) *Mathematische Urtheile* sind insgesammt synthetisch. Dieser Satz scheint den Bemerkungen der Zergliederer der menschlichen Vernunft bisher ganz entgangen, ja allen ihren Vermutungen gerade entgegengesetzt zu sein, ob er gleich unwidersprechlich gewiß und in der Folge sehr wichtig ist. Denn weil man fand, daß die Schlüsse der Mathematiker alle nach dem Satze des Widerspruches fortgehen (welches die Natur einer jeden apodiktischen Gewißheit erfordert), so überredete man sich, daß auch die Grundsätze aus dem Satze des Widerspruchs erkannt würden, worin sie sich sehr irrten; denn ein synthetischer Satz kann allerdings nach dem Satze des Widerspruchs eingesehen werden, aber nur so, daß ein anderer synthetischer Satz vorausgesetzt wird, aus dem er gefolgert werden kann, niemals aber an sich selbst. Zuvörderst muß bemerkt werden: daß eigentliche mathematische Sätze jederzeit Urtheile a priori und nicht empirisch sind, weil sie Nothwendigkeit bei sich führen, welche aus Erfahrung nicht abgenommen werden kann. Will man mir aber dieses nicht einräumen, wohlan so schränke ich meinen Satz auf die *reine Mathematik* ein, deren Begriff es schon mit sich bringt, daß sie nicht empirische, sondern blos reine Erkenntniß a priori enthalte. Man sollte anfänglich wohl denken, daß der Satz 7 + 5 = 12 ein blos analytischer Satz sei, der aus dem Begriffe einer Summe von Sieben und Fünf nach dem Satze des Widerspruches erfolge. Allein wenn man es näher betrachtet, so findet man, daß der Begriff der Summe von 7 und 5 nichts weiter enthalte, als die Vereinigung beider Zahlen in eine einzige, wodurch ganz und gar nicht gedacht wird, welches diese einzige Zahl sei, die beide zusammenfaßt. Der Begriff von Zwölf ist keinesweges dadurch schon gedacht, daß ich mir blos jene Vereinigung von Sieben und Fünf denke; und ich mag meinen Begriff von einer solchen möglichen Summe noch so lange zergliedern, so werde ich doch darin die Zwölf nicht antreffen. Man muß über diese Begriffe hinausgehen, indem man die Anschauung zu Hülfe nimmt, die einem von beiden correspondirt, etwa seine fünf Finger oder (wie *Segner* in seiner Arithmetik) fünf Punkte, und so nach und nach die Einheiten der in der Anschauung gegebenen Fünf zu dem Begriffe der Sieben hinzuthut. Man erweitert also wirklich seinen Begriff durch diesen Satz 7 + 5 = 12 und thut zu dem ersteren Begriff einen neuen hinzu, der in jenem gar nicht gedacht war, d. i. der arithmetische Satz ist jederzeit synthetisch, welches man desto deutlicher inne wird, wenn man etwas größere Zahlen nimmt; da es denn klar einleuchtet, daß, wir möchten unsern Begriff drehen und wenden, wie wir wollen, wir, ohne die Anschauung zu Hülfe zu nehmen, vermittelst der bloßen Zergliederung unserer Begriffe, die Summe niemals finden könnten. Eben so wenig ist irgend ein Grundsatz der reinen Geometrie analytisch. Daß die gerade Linie zwischen zwei Punkten die kürzeste sei, ist ein synthetischer Satz. Denn mein Begriff vom Geraden enthält nichts von Größe, sondern nur eine Qualität. Der Begriff des Kürzesten kommt also gänzlich hinzu und kann durch keine Zergliederung aus dem Begriffe der geraden Linie gezogen werden. Anschauung muß also hier zu Hülfe genommen werden, vermittelst deren allein die Synthesis möglich ist. Einige andere Grundsätze, welche die Geometer voraussetzen, sind zwar wirklich analytisch und beruhen auf dem Satze des Widerspruchs; sie dienen aber nur, wie identische Sätze, zur Kette der Methode und nicht als Principien, z. B. a = a, das Ganze ist sich selber gleich, oder (a + b) > a, d. i. das Ganze ist größer als sein Theil. Und doch auch diese selbst, ob sie gleich nach bloßen Begriffen gelten, werden in der Mathematik nur darum zugelassen, weil sie in der Anschauung können dargestellt werden. Was uns hier gemeiniglich glauben macht, als läge das Prädicat solcher apodiktischen Urtheile schon in unserm Begriffe, und das Urtheil sei also analytisch, ist blos die Zweideutigkeit des Ausdrucks. Wir *sollen* nämlich zu einem gegebenen Begriffe ein gewisses Prädicat hinzudenken, und diese Nothwendigkeit haftet schon an den Begriffen. Aber die Frage ist nicht, was wir zu dem gegebenen Begriffe hinzu *denken sollen*, sondern was wir *wirklich in ihm*, obzwar nur dunkel, *denken*; und da zeigt sich, daß das Prädicat jenem Begriffe zwar nothwendig, aber nicht unmittelbar, sondern vermittelst einer Anschauung, die hinzukommen muß, anhänge. ### Section 3 § 3. Anmerkung *zur allgemeinen Eintheilung der Urtheile in analytische und synthetische.* Diese Eintheilung ist in Ansehung der Kritik des menschlichen Verstandes unentbehrlich und verdient daher in ihr *classisch* zu sein; sonst wüßte ich nicht, daß sie irgend anderwärts einen beträchtlichen Nutzen hätte. Und hierin finde ich auch die Ursache, weswegen dogmatische Philosophen, die die Quellen metaphysischer Urtheile immer nur in der Metaphysik selbst, nicht aber außer ihr, in den reinen Vernunftgesetzen überhaupt, suchten, diese Eintheilung, die sich von selbst darzubieten scheint, vernachlässigten und wie der berühmte *Wolff*, oder der seinen Fußstapfen folgende scharfsinnige *Baumgarten* den Beweis von dem Satze des zureichenden Grundes, der offenbar synthetisch ist, im Satze des Widerspruchs suchen konnten. Dagegen treffe ich schon in *Lockes* Versuchen über den menschlichen Verstand einen Wink zu dieser Eintheilung an. Denn im vierten Buch, dem dritten Hauptstück § 9 u. f., nachdem er schon vorher von der verschiedenen Verknüpfung der Vorstellungen in Urtheilen und deren Quellen geredet hatte, wovon er die eine in der Identität oder Widerspruch setzt (analytische Urtheile), die andere aber in der Existenz der Vorstellungen in einem Subject (synthetische Urtheile), so gesteht er § 10, daß unsere Erkenntniß (a priori) von der letztern sehr enge und beinahe gar nichts sei. Allein es herrscht in dem, was er von dieser Art der Erkenntniß sagt, so wenig Bestimmtes und auf Regeln Gebrachtes, daß man sich nicht wundern darf, wenn niemand, sonderlich nicht einmal Hume Anlaß daher genommen hat, über Sätze dieser Art Betrachtungen anzustellen. Denn dergleichen allgemeine und dennoch bestimmte Principien lernt man nicht leicht von andern, denen sie nur dunkel obgeschwebt haben. Man muß durch eigenes Nachdenken zuvor selbst darauf gekommen sein, hernach findet man sie auch anderwärts, wo man sie gewiß nicht zuerst würde angetroffen haben, weil die Verfasser selbst nicht einmal wußten, daß ihren eigenen Bemerkungen eine solche Idee zum Grunde liege. Die, so niemals selbst denken, besitzen dennoch die Scharfsichtigkeit, alles, nachdem es ihnen gezeigt worden, in demjenigen, was sonst schon gesagt worden, aufzuspähen, wo es doch vorher niemand sehen konnte. ### General questions Der Prolegomenen Allgemeine Frage: Ist überall Metaphysik möglich? ### Section 4 § 4. Wäre Metaphysik, die sich als Wissenschaft behaupten könnte, wirklich; könnte man sagen: hier ist Metaphysik, die dürft ihr nur lernen, und sie wird euch unwiderstehlich und unveränderlich von ihrer Wahrheit überzeugen: so wäre diese Frage unnöthig, und es bliebe nur diejenige übrig, die mehr eine Prüfung unserer Scharfsinnigkeit, als den Beweis von der Existenz der Sache selbst beträfe, nämlich *wie sie möglich sei*, und wie Vernunft es anfange, dazu zu gelangen. Nun ist es der menschlichen Vernunft in diesem Falle so gut nicht geworden. Man kann kein einziges Buch aufzeigen, so wie man etwa einen *Euklid* vorzeigt, und sagen: das ist Metaphysik, hier findet ihr den vornehmsten Zweck dieser Wissenschaft, das Erkenntniß eines höchsten Wesens und einer künftigen Welt, bewiesen aus Principien der reinen Vernunft. Denn man kann uns zwar viele Sätze aufzeigen, die apodiktisch gewiß sind und niemals gestritten worden; aber diese sind insgesammt analytisch und betreffen mehr die Materialien und den Bauzeug zur Metaphysik, als die Erweiterung der Erkenntniß, die doch unsere eigentliche Absicht mit ihr sein soll (§ 2. litt. c). Ob ihr aber gleich auch synthetische Sätze (z. B. den Satz des zureichenden Grundes) vorzeigt, die ihr niemals aus bloßer Vernunft, mithin, wie doch eure Pflicht war, a priori bewiesen habt, die man euch aber doch gerne einräumt: so gerathet ihr doch, wenn ihr euch derselben zu eurem Hauptzwecke bedienen wollt, in so unstatthafte und unsichere Behauptungen, daß zu aller Zeit eine Metaphysik der anderen entweder in Ansehung der Behauptungen selbst, oder ihrer Beweise widersprochen und dadurch ihren Anspruch auf daurenden Beifall selbst vernichtet hat. Sogar sind die Versuche, eine solche Wissenschaft zu Stande zu bringen, ohne Zweifel die erste Ursache des so früh entstandenen Scepticismus gewesen, einer Denkungsart, darin die Vernunft so gewaltthätig gegen sich selbst verfährt, daß diese niemals, als in völliger Verzweiflung an Befriedigung in Ansehung ihrer wichtigsten Absichten hätte entstehen können. Denn lange vorher, ehe man die Natur methodisch zu befragen anfing, befrug man blos seine abgesonderte Vernunft, die durch gemeine Erfahrung in gewisser Maße schon geübt war, weil Vernunft uns doch immer gegenwärtig ist, Naturgesetze aber gemeiniglich mühsam aufgesucht werden müssen: und so schwamm Metaphysik oben auf wie Schaum, doch so, daß so wie der, den man geschöpft hatte, zerging, sich sogleich ein anderer auf der Oberfläche zeigte, den immer einige begierig aufsammleten, wobei andere, anstatt in der Tiefe die Ursache dieser Erscheinung zu suchen, sich damit weise dünkten, daß sie die vergebliche Mühe der erstern belachten. Das Wesentliche und Unterscheidende der reinen *mathematischen* Erkenntniß von aller andern Erkenntniß a priori ist, daß sie durchaus *nicht aus Begriffen*, sondern jederzeit nur durch die Construction der Begriffe (Kritik S. 713) vor sich gehen muß. Da sie also in ihren Sätzen über den Begriff zu demjenigen, was die ihm correspondirende Anschauung enthält, hinausgehen muß: so können und sollen ihre Sätze auch niemals durch Zergliederung der Begriffe, d. i. analytisch, entspringen und sind daher insgesammt synthetisch. Ich kann aber nicht umhin, den Nachteil zu bemerken, den die Vernachlässigung dieser sonst leichten und unbedeutend scheinenden Beobachtung der Philosophie zugezogen hat. *Hume*, als er den eines Philosophen würdigen Beruf fühlte, seine Blicke auf das ganze Feld der reinen Erkenntniß a priori zu werfen, in welchem sich der menschliche Verstand so große Besitzungen anmaßt, schnitt unbedachtsamer Weise eine ganze und zwar die erheblichste Provinz derselben, nämlich reine Mathematik, davon ab in der Einbildung, ihre Natur und so zu reden ihre Staatsverfassung beruhe auf ganz andern Principien, nämlich lediglich auf dem Satze des Widerspruchs; und ob er zwar die Eintheilung der Sätze nicht so förmlich und allgemein oder unter der Benennung gemacht hatte, als es von mir hier geschieht, so war es doch gerade so viel, als ob er gesagt hätte: reine Mathematik enthält blos *analytische* Sätze, Metaphysik aber synthetische a priori. Nun irrte er hierin gar sehr, und dieser Irrthum hatte auf seinen ganzen Begriff entscheidend nachtheilige Folgen. Denn wäre das von ihm nicht geschehen, so hätte er seine Frage wegen des Ursprungs unserer synthetischen Urtheile weit über seinen metaphysischen Begriff der Causalität erweitert und sie auch auf die Möglichkeit der Mathematik a priori ausgedehnt; denn diese mußte er eben sowohl für synthetisch annehmen. Alsdann aber hätte er seine metaphysische Sätze keinesweges auf bloße Erfahrung gründen können, weil er sonst die Axiomen der reinen Mathematik ebenfalls der Erfahrung unterworfen haben würde, welches zu thun er viel zu einsehend war. Die gute Gesellschaft, worin Metaphysik alsdann zu stehen gekommen wäre, hätte sie wider die Gefahr einer schnöden Mißhandlung gesichert; denn die Streiche, welche der letzern zugedacht waren, hätten die erstere auch treffen müssen, welches aber seine Meinung nicht war, auch nicht sein konnte: und so wäre der scharfsinnige Mann in Betrachtungen gezogen worden, die denjenigen hätten ähnlich werden müssen, womit wir uns jetzt beschäftigen, die aber durch seinen unnachahmlich schönen Vortrag unendlich würden gewonnen haben. *Eigentlich metaphysische* Urtheile sind insgesammt synthetisch. Man muß *zur Metaphysik gehörige* von eigentlich *metaphysischen* Urtheilen unterscheiden. Unter jenen sind sehr viele analytisch, aber sie machen nur die Mittel zu metaphysischen Urtheilen aus, auf die der Zweck der Wissenschaft ganz und gar gerichtet ist, und die allemal synthetisch sind. Denn wenn Begriffe zur Metaphysik gehören, z. B. der von Substanz, so gehören die Urtheile, die aus der bloßen Zergliederung derselben entspringen, auch nothwendig zur Metaphysik, z. B. Substanz ist dasjenige, was nur als Subject existirt etc., und vermittelst mehrerer dergleichen analytischen Urtheile suchen wir der Definition der Begriffe nahe zu kommen. Da aber die Analysis eines reinen Verstandesbegriffs (dergleichen die Metaphysik enthält) nicht auf andere Art vor sich geht, als die Zergliederung jedes andern, auch empirischen Begriffs, der nicht in die Metaphysik gehört (z. B. Luft ist eine elastische Flüssigkeit, deren Elasticität durch keinen bekannten Grad der Kälte aufgehoben wird), so ist zwar der Begriff, aber nicht das analytische Urtheil eigenthümlich metaphysisch: denn diese Wissenschaft hat etwas Besonderes und ihr Eigenthümliches in der Erzeugung ihrer Erkenntnisse a priori, die also von dem, was sie mit allen andern Verstandeserkenntnissen gemein hat, muß unterschieden werden; so ist z. B. der Satz: alles, was in den Dingen Substanz ist, ist beharrlich, ein synthetischer und eigenthümlich metaphysischer Satz. Wenn man die Begriffe a priori, welche die Materie der Metaphysik und ihr Bauzeug ausmachen, zuvor nach gewissen Principien gesammlet hat, so ist die Zergliederung dieser Begriffe von großem Werthe; auch kann dieselbe als ein besonderer Theil (gleichsam als philosophia definitiva), der lauter analytische, zur Metaphysik gehörige Sätze enthält, von allen synthetischen Sätzen, die die Metaphysik selbst ausmachen, abgesondert vorgetragen werden. Denn in der That haben jene Zergliederungen nirgend anders einen beträchtlichen Nutzen, als in der Metaphysik, d. i. in Absicht auf die synthetischen Sätze, die aus jenen zuerst zergliederten Begriffen sollen erzeugt werden. Der Schluß dieses Paragraphs ist also: daß Metaphysik es eigentlich mit synthetischen Sätzen a priori zu thun habe, und diese allein ihren Zweck ausmachen, zu welchem sie zwar allerdings mancher Zergliederungen ihrer Begriffe, mithin analytischer Urtheile bedarf, wobei aber das Verfahren nicht anders ist, als in jeder andern Erkenntnißart, wo man seine Begriffe durch Zergliederung blos deutlich zu machen sucht. Allein die Erzeugung der Erkenntniß a priori sowohl der Anschauung als Begriffen nach, endlich auch synthetischer Sätze a priori und zwar im philosophischen Erkenntnisse machen den wesentlichen Inhalt der Metaphysik aus. Überdrüssig also des Dogmatismus, der uns nichts lehrt, und zugleich des Scepticismus, der uns gar überall nichts verspricht, auch nicht einmal den Ruhestand einer erlaubten Unwissenheit, aufgefordert durch die Wichtigkeit der Erkenntniß, deren wir bedürfen, und mißtrauisch durch lange Erfahrung in Ansehung jeder, die wir zu besitzen glauben, oder die sich uns unter dem Titel der reinen Vernunft anbietet, bleibt uns nur noch eine kritische Frage übrig, nach deren Beantwortung wir unser künftiges Betragen einrichten können: *Ist überall Metaphysik möglich?* Aber diese Frage muß nicht durch sceptische Einwürfe gegen gewisse Behauptungen einer wirklichen Metaphysik (denn wir lassen jetzt noch keine gelten), sondern aus dem nur noch *problematischen* Begriffe einer solchen Wissenschaft beantwortet werden. In der *Kritik der reinen Vernunft* bin ich in Absicht auf diese Frage synthetisch zu Werke gegangen, nämlich so, daß ich in der reinen Vernunft selbst forschte und in dieser Quelle selbst die Elemente sowohl, als auch die Gesetze ihres reinen Gebrauchs nach Prinzipien zu bestimmen suchte. Diese Arbeit ist schwer und erfordert einen entschlossenen Leser, sich nach und nach in ein System hinein zu denken, was noch nichts als gegeben zum Grunde legt außer die Vernunft selbst und also, ohne sich auf irgend ein Factum zu stützen, die Erkenntniß aus ihren ursprünglichen Keimen zu entwickeln sucht. *Prolegomena* sollen dagegen Vorübungen sein; sie sollen mehr anzeigen, was man zu thun habe, um eine Wissenschaft wo möglich zur Wirklichkeit zu bringen, als sie selbst vortragen. Sie müssen sich also auf etwas stützen, was man schon als zuverlässig kennt, von da man mit Zutrauen ausgehen und zu den Quellen aufsteigen kann, die man noch nicht kennt, und deren Entdeckung uns nicht allein das, was man wußte, erklären, sondern zugleich einen Umfang vieler Erkenntnisse, die insgesammt aus den nämlichen Quellen entspringen, darstellen wird. Das methodische Verfahren der Prolegomenen, vornehmlich derer, die zu einer künftigen Metaphysik vorbereiten sollen, wird also *analytisch* sein. Es trifft sich aber glücklicher Weise, daß, ob wir gleich nicht annehmen können, daß Metaphysik als Wissenschaft *wirklich* sei, wir doch mit Zuversicht sagen können, daß gewisse reine synthetische Erkenntniß a priori wirklich und gegeben sei, nämlich *reine Mathematik* und *reine Naturwissenschaft*; denn beide enthalten Sätze, die theils apodiktisch gewiß durch bloße Vernunft, theils durch die allgemeine Einstimmung aus der Erfahrung und dennoch als von Erfahrung unabhängig durchgängig anerkannt werden. Wir haben also einige wenigstens *unbestrittene* synthetische Erkenntniß a priori und dürfen nicht fragen, ob sie möglich sei (denn sie ist wirklich), sondern nur *wie sie möglich sei*, um aus dem Princip der Möglichkeit der gegebenen auch die Möglichkeit aller übrigen ableiten zu können. Prolegomena. Allgemeine Frage: Wie ist Erkenntniß aus reiner Vernunft möglich? ### Section 5 § 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. ### First part: How is pure mathematics possible? Der transscendentalen Hauptfrage Erster Theil. Wie ist reine Mathematik möglich? ### Section 6 § 6. Hier ist nun eine große und bewährte Erkenntniß, die schon jetzt von bewundernswürdigem Umfange ist und unbegrenzte Ausbreitung auf die Zukunft verspricht, die durch und durch apodiktische Gewißheit, d. i. absolute Nothwendigkeit, bei sich führt, also auf keinen Erfahrungsgründen beruht, mithin ein reines Product der Vernunft, überdem aber durch und durch synthetisch ist. „Wie ist es nun der menschlichen Vernunft möglich, eine solche Erkenntniß gänzlich a priori zu Stande zu bringen?“ Setzt dieses Vermögen, da es sich nicht auf Erfahrungen fußt, noch fußen kann, nicht irgend einen Erkenntnißgrund a priori voraus, der tief verborgen liegt, der sich aber durch diese seine Wirkungen offenbaren dürfte, wenn man den ersten Anfängen derselben nur fleißig nachspürte? ### Section 7 § 7. Wir finden aber, daß alle mathematische Erkenntniß dieses Eigenthümliche habe, daß sie ihren Begriff vorher *in der Anschauung* und zwar a priori, mithin einer solchen, die nicht empirisch, sondern reine Anschauung ist, darstellen müsse, ohne welches Mittel sie nicht einen einzigen Schritt thun kann; daher ihre Urtheile jederzeit *intuitiv* sind, an statt daß Philosophie sich mit *discursiven* Urtheilen, *aus bloßen Begriffen*, begnügen muß und ihre apodiktische Lehren wohl durch Anschauung erläutern, niemals aber daher ableiten kann. Diese Beobachtung in Ansehung der Natur der Mathematik giebt uns nun schon eine Leitung auf die erste und oberste Bedingung ihrer Möglichkeit: nämlich es muß ihr irgend eine *reine Anschauung* zum Grunde liegen, in welcher sie alle ihre Begriffe in concreto und dennoch a priori darstellen oder, wie man es nennt, sie *construiren* kann.[^281-1] Können wir diese reine Anschauung und die Möglichkeit einer solchen ausfinden, so erklärt sich daraus leicht, wie synthetische Sätze a priori in der reinen Mathematik, und mithin auch, wie diese Wissenschaft selbst möglich sei; denn so wie die empirische Anschauung es ohne Schwierigkeit möglich macht, daß wir unseren Begriff, den wir uns von einem Object der Anschauung machen, durch neue Prädicate, die die Anschauung selbst darbietet, in der Erfahrung synthetisch erweitern, so wird es auch die reine Anschauung thun; nur mit dem Unterschiede: daß im letztern Falle das synthetische Urtheil a priori gewiß und apodiktisch, im ersteren aber nur a posteriori und empirisch gewiß sein wird, weil diese nur das enthält, was in der zufälligen empirischen Anschauung angetroffen wird, jene aber, was in der reinen nothwendig angetroffen werden muß, indem sie als Anschauung a priori mit dem Begriffe *vor aller Erfahrung* oder einzelnen Wahrnehmung unzertrennlich verbunden ist. [^281-1]: Siehe Kritik S. 713. ### Section 8 § 8. Allein die Schwierigkeit scheint bei diesem Schritte eher zu wachsen, als abzunehmen. Denn nunmehr lautet die Frage: *wie ist es möglich, etwas a priori anzuschauen?* Anschauung ist eine Vorstellung, so wie sie unmittelbar von der Gegenwart des Gegenstandes abhängen würde. Daher scheint es unmöglich, a priori *ursprünglich* anzuschauen, weil die Anschauung alsdann ohne einen weder vorher, noch jetzt gegenwärtigen Gegenstand, worauf sie sich bezöge, stattfinden müßte und also nicht Anschauung sein könnte. Begriffe sind zwar von der Art, daß wir uns einige derselben, nämlich die, so nur das Denken eines Gegenstandes überhaupt enthalten, ganz wohl a priori machen können, ohne daß wir uns in einem unmittelbaren Verhältnisse zum Gegenstande befänden, z. B. den Begriff von Größe, von Ursache etc.; aber selbst diese bedürfen doch, um ihnen Bedeutung und Sinn zu verschaffen, einen gewissen Gebrauch in concreto, d. i. Anwendung auf irgend eine Anschauung, dadurch uns ein Gegenstand derselben gegeben wird. Allein wie kann Anschauung des Gegenstandes vor dem Gegenstande selbst vorhergehen? ### Section 9 § 9. Müßte unsre Anschauung von der Art sein, daß sie Dinge vorstellte, *so wie sie an sich selbst sind*, so würde gar keine Anschauung a priori stattfinden, sondern sie wäre allemal empirisch. Denn was in dem Gegenstande an sich selbst enthalten sei, kann ich nur wissen, wenn er mir gegenwärtig und gegeben ist. Freilich ist es auch alsdann unbegreiflich, wie die Anschauung einer gegenwärtigen Sache mir diese sollte zu erkennen geben, wie sie an sich ist, da ihre Eigenschaften nicht in meine Vorstellungskraft hinüber wandern können; allein die Möglichkeit davon eingeräumt, so würde doch dergleichen Anschauung nicht a priori stattfinden, d. i. ehe mir noch der Gegenstand vorgestellt würde: denn ohne das kann kein Grund der Beziehung meiner Vorstellung auf ihn erdacht werden, sie müßte denn auf Eingebung beruhen. Es ist also nur auf eine einzige Art möglich, daß meine Anschauung vor der Wirklichkeit des Gegenstandes vorhergehe und als Erkenntniß a priori stattfinde, *wenn sie nämlich nichts anders enthält, als die Form der Sinnlichkeit, die in meinem Subject vor allen wirklichen Eindrücken vorhergeht, dadurch ich von Gegenständen afficirt werde*. Denn daß Gegenstände der Sinne dieser Form der Sinnlichkeit gemäß allein angeschaut werden können, kann ich a priori wissen. Hieraus folgt: daß Sätze, die blos diese Form der sinnlichen Anschauung betreffen, von Gegenständen der Sinne möglich und gültig sein werden, imgleichen umgekehrt, daß Anschauungen, die a priori möglich sind, niemals andere Dinge als Gegenstände unsrer Sinne betreffen können. ### Section 10 § 10. Also ist es nur die Form der sinnlichen Anschauung, dadurch wir a priori Dinge anschauen können, wodurch wir aber auch die Objecte nur erkennen, wie sie uns (unsern Sinnen) *erscheinen* können, nicht wie sie an sich sein mögen; und diese Voraussetzung ist schlechterdings nothwendig, wenn synthetische Sätze a priori als möglich eingeräumt, oder, im Falle sie wirklich angetroffen werden, ihre Möglichkeit begriffen und zum voraus bestimmt werden soll. Nun sind Raum und Zeit diejenigen Anschauungen, welche die reine Mathematik allen ihren Erkenntnissen und Urtheilen, die zugleich als apodiktisch und nothwendig auftreten, zum Grunde legt; denn Mathematik muß alle ihre Begriffe zuerst in der Anschauung und reine Mathematik in der reinen Anschauung darstellen, d. i. sie construiren, ohne welche (weil sie nicht analytisch, nämlich durch Zergliederung der Begriffe, sondern synthetisch verfahren kann) es ihr unmöglich ist, einen Schritt zu thun, so lange ihr nämlich reine Anschauung fehlt, in der allein der Stoff zu synthetischen Urtheilen a priori gegeben werden kann. Geometrie legt die reine Anschauung des Raums zum Grunde. Arithmetik bringt selbst ihre Zahlbegriffe durch successive Hinzusetzung der Einheiten in der Zeit zu Stande, vornehmlich aber reine Mechanik kann ihre Begriffe von Bewegung nur vermittelst der Vorstellung der Zeit zu Stande bringen. Beide Vorstellungen aber sind blos Anschauungen; denn wenn man von den empirischen Anschauungen der Körper und ihrer Veränderungen (Bewegung) alles Empirische, nämlich was zur Empfindung gehört, wegläßt, so bleiben noch Raum und Zeit übrig, welche also reine Anschauungen sind, die jenen a priori zum Grunde liegen und daher selbst niemals weggelassen werden können, aber eben dadurch, daß sie reine Anschauungen a priori sind, beweisen, daß sie bloße Formen unserer Sinnlichkeit sind, die vor aller empirischen Anschauung, d. i. der Wahrnehmung wirklicher Gegenstände, vorhergehen müssen, und denen gemäß Gegenstände a priori erkannt werden können, aber freilich nur, wie sie uns erscheinen. ### Section 11 § 11. Die Aufgabe des gegenwärtigen Abschnitts ist also aufgelöset. Reine Mathematik ist als synthetische Erkenntniß a priori nur dadurch möglich, daß sie auf keine andere als bloße Gegenstände der Sinne geht, deren empirischer Anschauung eine reine Anschauung (des Raums und der Zeit) und zwar a priori zum Grunde liegt und darum zum Grunde liegen kann, weil diese nichts anders als die bloße Form der Sinnlichkeit ist, welche vor der wirklichen Erscheinung der Gegenstände vorhergeht, indem sie dieselbe in der That allererst möglich macht. Doch betrifft dieses Vermögen, a priori anzuschauen, nicht die Materie der Erscheinung, d. i. das, was in ihr Empfindung ist, denn diese macht das Empirische aus, sondern nur die Form derselben, Raum und Zeit. Wollte man im mindesten daran zweifeln, daß beide gar keine den Dingen an sich selbst, sondern nur bloße ihrem Verhältnisse zur Sinnlichkeit anhängende Bestimmungen sind, so möchte ich gerne wissen, wie man es möglich finden kann, a priori und also vor aller Bekanntschaft mit den Dingen, ehe sie nämlich uns gegeben sind, zu wissen, wie ihre Anschauung beschaffen sein müsse, welches doch hier der Fall mit Raum und Zeit ist. Dieses ist aber ganz begreiflich, so bald beide für nichts weiter als formale Bedingungen unserer Sinnlichkeit, die Gegenstände aber blos für Erscheinungen gelten; denn alsdann kann die Form der Erscheinung, d. i. die reine Anschauung, allerdings aus uns selbst, d. i. a priori, vorgestellt werden. ### Section 12 § 12. Um etwas zur Erläuterung und Bestätigung beizufügen, darf man nur das gewöhnliche und unumgänglich nothwendige Verfahren der Geometern ansehen. Alle Beweise von durchgängiger Gleichheit zweier gegebenen Figuren (da eine in allen Stücken an die Stelle der andern gesetzt werden kann) laufen zuletzt darauf hinaus, daß sie einander decken, welches offenbar nichts anders als ein auf der unmittelbaren Anschauung beruhender synthetischer Satz ist; und diese Anschauung muß rein und a priori gegeben werden, denn sonst könnte jener Satz nicht für apodiktisch gewiß gelten, sondern hätte nur empirische Gewißheit. Es würde nur heißen: man bemerkt es jederzeit so, und er gilt nur so weit, als unsre Wahrnehmung bis dahin sich erstreckt hat. Daß der vollständige Raum (der selbst keine Grenze eines anderen Raumes mehr ist) drei Abmessungen habe, und Raum überhaupt auch nicht mehr derselben haben könne, wird auf den Satz gebaut, daß sich in einem Punkte nicht mehr als drei Linien rechtwinklicht schneiden können; dieser Satz aber kann gar nicht aus Begriffen dargethan werden, sondern beruht unmittelbar auf Anschauung und zwar reiner a priori, weil er apodiktisch gewiß ist; daß man verlangen kann, eine Linie solle ins Unendliche gezogen (in indefinitum), oder eine Reihe Veränderungen (z. B. durch Bewegung zurückgelegte Räume) solle ins Unendliche fortgesetzt werden, setzt doch eine Vorstellung des Raumes und der Zeit voraus, die blos an der Anschauung hängen kann, nämlich so fern sie an sich durch nichts begrenzt ist; denn aus Begriffen könnte sie nie geschlossen werden. Also liegen doch wirklich der Mathematik reine Anschauungen a priori zum Grunde, welche ihre synthetische und apodiktisch geltende Sätze möglich machen; und daher erklärt unsere transscendentale Deduction der Begriffe im Raum und Zeit zugleich die Möglichkeit einer reinen Mathematik, die ohne eine solche Deduction, und ohne daß wir annehmen, „alles, was unsern Sinnen gegeben werden mag (den äußeren im Raume, dem inneren in der Zeit), werde von uns nur angeschauet, wie es uns erscheint, nicht wie es an sich selbst ist,“ zwar eingeräumt, aber keineswegs eingesehen werden könnte. ### Section 13 § 13. Diejenigen, welche noch nicht von dem Begriffe loskommen können, als ob Raum und Zeit wirkliche Beschaffenheiten wären, die den Dingen an sich selbst anhingen, können ihre Scharfsinnigkeit an folgendem Paradoxon üben und, wenn sie dessen Auflösung vergebens versucht haben, wenigstens auf einige Augenblicke von Vorurtheilen frei, vermuthen, daß doch vielleicht die Abwürdigung des Raumes und der Zeit zu bloßen Formen unsrer sinnlichen Anschauung Grund haben möge. Wenn zwei Dinge in allen Stücken, die an jedem für sich nur immer können erkannt werden, (in allen zur Größe und Qualität gehörigen Bestimmungen) völlig einerlei sind, so muß doch folgen, daß eins in allen Fällen und Beziehungen an die Stelle des andern könne gesetzt werden, ohne daß diese Vertauschung den mindesten kenntlichen Unterschied verursachen würde. In der That verhält sich dies auch so mit ebenen Figuren in der Geometrie; allein verschiedene sphärische zeigen unerachtet jener völligen innern Übereinstimmung doch eine solche Verschiedenheit im äußeren Verhältniß, daß sich eine an die Stelle der andern gar nicht setzen läßt; z. B. zwei sphärische Triangel von beiden Hemisphären, die einen Bogen des Äquators zur gemeinschaftlichen Basis haben, können völlig gleich sein in Ansehung der Seiten sowohl als Winkel, so daß an keinem, wenn er allein und zugleich vollständig beschrieben wird, nichts angetroffen wird, was nicht zugleich in der Beschreibung des andern läge, und dennoch kann einer nicht an die Stelle des andern (nämlich auf dem entgegengesetzten Hemisphär) gesetzt werden; und hier ist denn doch eine innere Verschiedenheit beider Triangel, die kein Verstand als innerlich angeben kann, und die sich nur durch das äußere Verhältniß im Raume offenbart. Allein ich will gewöhnlichere Fälle anführen, die aus dem gemeinen Leben genommen werden können. Was kann wohl meiner Hand oder meinem Ohr ähnlicher und in allen Stücken gleicher sein, als ihr Bild im Spiegel? Und dennoch kann ich eine solche Hand, als im Spiegel gesehen wird, nicht an die Stelle ihres Urbildes setzen; denn wenn dieses eine rechte Hand war, so ist jene im Spiegel eine linke, und das Bild des rechten Ohres ist ein linkes, das nimmermehr die Stelle des ersteren vertreten kann. Nun sind hier keine innre Unterschiede, die irgend ein Verstand nur denken könnte; und dennoch sind die Unterschiede innerlich, so weit die Sinne lehren, denn die linke Hand kann mit der rechten unerachtet aller beiderseitigen Gleichheit und Ähnlichkeit doch nicht zwischen denselben Grenzen eingeschlossen sein (sie können nicht congruiren); der Handschuh der einen Hand kann nicht auf der andern gebraucht werden. Was ist nun die Auflösung? Diese Gegenstände sind nicht etwa Vorstellungen der Dinge, wie sie an sich selbst sind, und wie sie der pure Verstand erkennen würde, sondern es sind sinnliche Anschauungen, d. i. Erscheinungen, deren Möglichkeit auf dem Verhältnisse gewisser an sich unbekannten Dinge zu etwas anderem, nämlich unserer Sinnlichkeit, beruht. Von dieser ist nun der Raum die Form der äußern Anschauung, und die innere Bestimmung eines jeden Raumes ist nur durch die Bestimmung des äußeren Verhältnisses zu dem ganzen Raume, davon jener ein Theil ist (dem Verhältnisse zum äußeren Sinne), d. i. der Theil ist nur durchs Ganze möglich, welches bei Dingen an sich selbst als Gegenständen des bloßen Verstandes niemals, wohl aber bei bloßen Erscheinungen stattfindet. Wir können daher auch den Unterschied ähnlicher und gleicher, aber doch incongruenter Dinge (z. B. widersinnig gewundener Schnecken) durch keinen einzigen Begriff verständlich machen, sondern nur durch das Verhältniß zur rechten und linken Hand, welches unmittelbar auf Anschauung geht. *Anmerkung I.* Die reine Mathematik und namentlich die reine Geometrie kann nur unter der Bedingung allein objective Realität haben, daß sie blos auf Gegenstände der Sinne geht, in Ansehung deren aber der Grundsatz feststeht: daß unsre sinnliche Vorstellung keinesweges eine Vorstellung der Dinge an sich selbst, sondern nur der Art sei, wie sie uns erscheinen. Daraus folgt, daß die Sätze der Geometrie nicht etwa Bestimmungen eines bloßen Geschöpfs unserer dichtenden Phantasie sind und also nicht mit Zuverlässigkeit auf wirkliche Gegenstände könnten bezogen werden, sondern daß sie nothwendiger Weise vom Raume und darum auch von allem, was im Raume angetroffen werden mag, gelten, weil der Raum nichts anders ist, als die Form aller äußeren Erscheinungen, unter der uns allein Gegenstände der Sinne gegeben werden können. Die Sinnlichkeit, deren Form die Geometrie zum Grunde legt, ist das, worauf die Möglichkeit äußerer Erscheinungen beruht; diese also können niemals etwas anderes enthalten, als was die Geometrie ihnen vorschreibt. Ganz anders würde es sein, wenn die Sinne die Objecte vorstellen müßten, wie sie an sich selbst sind. Denn da würde aus der Vorstellung vom Raume, die der Geometer a priori mit allerlei Eigenschaften desselben zum Grunde legt, noch gar nicht folgen, daß alles dieses sammt dem, was daraus gefolgert wird, sich gerade so in der Natur verhalten müsse. Man würde den Raum des Geometers für bloße Erdichtung halten und ihm keine objective Gültigkeit zutrauen, weil man gar nicht einsieht, wie Dinge nothwendig mit dem Bilde, das wir uns von selbst und zum voraus von ihnen machen, übereinstimmen müßten. Wenn aber dieses Bild oder vielmehr diese formale Anschauung die wesentliche Eigenschaft unserer Sinnlichkeit ist, vermittelst deren uns allein Gegenstände gegeben werden, diese Sinnlichkeit aber nicht Dinge an sich selbst, sondern nur ihre Erscheinungen vorstellt, so ist ganz leicht zu begreifen und zugleich unwidersprechlich bewiesen: daß alle äußere Gegenstände unsrer Sinnenwelt nothwendig mit den Sätzen der Geometrie nach aller Pünktlichkeit übereinstimmen müssen, weil die Sinnlichkeit durch ihre Form äußerer Anschauung (den Raum), womit sich der Geometer beschäftigt, jene Gegenstände als bloße Erscheinungen selbst allererst möglich macht. Es wird allemal ein bemerkungswürdiges Phänomen in der Geschichte der Philosophie bleiben, daß es eine Zeit gegeben hat, da selbst Mathematiker, die zugleich Philosophen waren, zwar nicht an der Richtigkeit ihrer geometrischen Sätze, sofern sie blos den Raum beträfen, aber an der objectiven Gültigkeit und Anwendung dieses Begriffs selbst und aller geometrischen Bestimmungen desselben auf Natur zu zweifeln anfingen; da sie besorgten, eine Linie in der Natur möchte doch wohl aus physischen Punkten, mithin der wahre Raum im Objecte aus einfachen Theilen bestehen, obgleich der Raum, den der Geometer in Gedanken hat, daraus keinesweges bestehen kann. Sie erkannten nicht: daß dieser Raum in Gedanken den physischen, d. i. die Ausdehnung der Materie selbst, möglich mache; daß dieser gar keine Beschaffenheit der Dinge an sich selbst, sondern nur eine Form unserer sinnlichen Vorstellungskraft sei; daß alle Gegenstände im Raume bloße Erscheinungen, d. i. nicht Dinge an sich selbst, sondern Vorstellungen unsrer sinnlichen Anschauung seien, und, da der Raum, wie ihn sich der Geometer denkt, ganz genau die Form der sinnlichen Anschauung ist, die wir a priori in uns finden, und die den Grund der Möglichkeit aller äußern Erscheinungen (ihrer Form nach) enthält, diese nothwendig und auf das präciseste mit den Sätzen des Geometers, die er aus keinem erdichteten Begriff, sondern aus der subjectiven Grundlage aller äußern Erscheinungen, nämlich der Sinnlichkeit selbst, zieht, zusammen stimmen müssen. Auf solche und keine andere Art kann der Geometer wider aller Chicanen einer seichten Metaphysik wegen der ungezweifelten objectiven Realität seiner Sätze gesichert werden, so befremdend sie auch dieser, weil sie nicht bis zu den Quellen ihrer Begriffe zurückgeht, scheinen müssen. *Anmerkung II.* Alles, was uns als Gegenstand gegeben werden soll, muß uns in der Anschauung gegeben werden. Alle unsere Anschauung geschieht aber nur vermittelst der Sinne; der Verstand schauet nichts an, sondern reflectirt nur. Da nun die Sinne nach dem jetzt Erwiesenen uns niemals und in keinem einzigen Stück die Dinge an sich selbst, sondern nur ihre Erscheinungen zu erkennen geben, diese aber bloße Vorstellungen der Sinnlichkeit sind, „so müssen auch alle Körper mit sammt dem Raume, darin sie sich befinden, für nichts als bloße Vorstellungen in uns gehalten werden und existiren nirgend anders, als blos in unsern Gedanken.“ Ist dieses nun nicht der offenbare Idealismus? Der Idealismus besteht in der Behauptung, daß es keine andere als denkende Wesen gebe, die übrige Dinge, die wir in der Anschauung wahrzunehmen glauben, wären nur Vorstellungen in den denkenden Wesen, denen in der That kein außerhalb diesen befindlicher Gegenstand correspondirte. Ich dagegen sage: es sind uns Dinge als außer uns befindliche Gegenstände unserer Sinne gegeben, allein von dem, was sie an sich selbst sein mögen, wissen wir nichts, sondern kennen nur ihre Erscheinungen, d. i. die Vorstellungen, die sie in uns wirken, indem sie unsere Sinne afficiren. Demnach gestehe ich allerdings, daß es außer uns Körper gebe, d. i. Dinge, die, obzwar nach dem, was sie an sich selbst sein mögen, uns gänzlich unbekannt, wir durch die Vorstellungen kennen, welche ihr Einfluß auf unsre Sinnlichkeit uns verschafft, und denen wir die Benennung eines Körpers geben; welches Wort also blos die Erscheinung jenes uns unbekannten, aber nichts desto weniger wirklichen Gegenstandes bedeutet. Kann man dieses wohl Idealismus nennen? Es ist ja gerade das Gegentheil davon. Daß man unbeschadet der wirklichen Existenz äußerer Dinge von einer Menge ihrer Prädicate sagen könne: sie gehörten nicht zu diesen Dingen an sich selbst, sondern nur zu ihren Erscheinungen und hätten außer unserer Vorstellung keine eigene Existenz, ist etwas, was schon lange vor *Lockes* Zeiten, am meisten aber nach diesen allgemein angenommen und zugestanden ist. Dahin gehören die Wärme, die Farbe, der Geschmack ꝛc. Daß ich aber noch über diese aus wichtigen Ursachen die übrigen Qualitäten der Körper, die man primarias nennt, die Ausdehnung, den Ort und überhaupt den Raum mit allem, was ihm anhängig ist (Undurchdringlichkeit oder Materialität, Gestalt ꝛc.), auch mit zu bloßen Erscheinungen zähle, dawider kann man nicht den mindesten Grund der Unzulässigkeit anführen; und so wenig wie der, so die Farben nicht als Eigenschaften, die dem Object an sich selbst, sondern nur den Sinn des Sehens als Modificationen anhängen, will gelten lassen, darum ein Idealist heißen kann: so wenig kann mein Lehrbegriff idealistisch heißen, blos deshalb weil ich finde, daß noch mehr, ja *alle Eigenschaften, die die Anschauung eines Körpers ausmachen*, blos zu seiner Erscheinung gehören; denn die Existenz des Dinges, was erscheint, wird dadurch nicht wie beim wirklichen Idealism aufgehoben, sondern nur gezeigt, daß wir es, wie es an sich selbst sei, durch Sinne gar nicht erkennen können. Ich möchte gerne wissen, wie denn meine Behauptungen beschaffen sein müßten, damit sie nicht einen Idealism enthielten. Ohne Zweifel müßte ich sagen: daß die Vorstellung vom Raume nicht blos dem Verhältnisse, was unsre Sinnlichkeit zu den Objecten hat, vollkommen gemäß sei, denn das habe ich gesagt, sondern daß sie sogar dem Object völlig ähnlich sei; eine Behauptung, mit der ich keinen Sinn verbinden kann, so wenig als daß die Empfindung des Rothen mit der Eigenschaft des Zinnobers, der diese Empfindung in mir erregt, eine Ähnlichkeit habe. *Anmerkung III.* Hieraus läßt sich nun ein leicht vorherzusehender, aber nichtiger Einwurf gar leicht abweisen: „daß nämlich durch die Idealität des Raums und der Zeit die ganze Sinnenwelt in lauter Schein verwandelt werden würde.“ Nachdem man nämlich zuvörderst alle philosophische Einsicht von der Natur der sinnlichen Erkenntniß dadurch verdorben hatte, daß man die Sinnlichkeit blos in einer verworrenen Vorstellungsart setzte, nach der wir die Dinge immer noch erkennten, wie sie sind, nur ohne das Vermögen zu haben, alles in dieser unserer Vorstellung zum klaren Bewußtsein zu bringen; dagegen von uns bewiesen worden, daß Sinnlichkeit nicht in diesem logischen Unterschiede der Klarheit oder Dunkelheit, sondern in dem genetischen des Ursprungs der Erkenntniß selbst bestehe, da sinnliche Erkenntniß die Dinge gar nicht vorstellt, wie sie sind, sondern nur die Art, wie sie unsere Sinnen afficiren, und also, daß durch sie blos Erscheinungen, nicht die Sachen selbst dem Verstande zur Reflexion gegeben werden: nach dieser nothwendigen Berichtigung regt sich ein aus unverzeihlicher und beinahe vorsetzlicher Mißdeutung entspringender Einwurf, als wenn mein Lehrbegriff alle Dinge der Sinnenwelt in lauter Schein verwandelte. Wenn uns Erscheinung gegeben ist, so sind wir noch ganz frei, wie wir die Sache daraus beurtheilen wollen. Jene, nämlich Erscheinung, beruhte auf den Sinnen, diese Beurtheilung aber auf dem Verstande, und es frägt sich nur, ob in der Bestimmung des Gegenstandes Wahrheit sei oder nicht. Der Unterschied aber zwischen Wahrheit und Traum wird nicht durch die Beschaffenheit der Vorstellungen, die auf Gegenstände bezogen werden, ausgemacht, denn die sind in beiden einerlei, sondern durch die Verknüpfung derselben nach den Regeln, welche den Zusammenhang der Vorstellungen in dem Begriffe eines Objects bestimmen, und wie fern sie in einer Erfahrung beisammen stehen können oder nicht. Und da liegt es gar nicht an den Erscheinungen, wenn unsere Erkenntniß den Schein für Wahrheit nimmt, d. i. wenn Anschauung, wodurch uns ein Object gegeben wird, für Begriff vom Gegenstande oder auch der Existenz desselben, die der Verstand nur denken kann, gehalten wird. Den Gang der Planeten stellen uns die Sinne bald rechtläufig, bald rückläufig vor, und hierin ist weder Falschheit noch Wahrheit, weil, so lange man sich bescheidet, daß dieses vorerst nur Erscheinung ist, man über die objective Beschaffenheit ihrer Bewegung noch gar nicht urtheilt. Weil aber, wenn der Verstand nicht wohl darauf Acht hat, zu verhüten, daß diese subjective Vorstellungsart nicht für objectiv gehalten werde, leichtlich ein falsches Urtheil entspringen kann, so sagt man: sie scheinen zurückzugehen; allein der Schein kommt nicht auf Rechnung der Sinne, sondern des Verstandes, dem es allein zukommt, aus der Erscheinung ein objectives Urtheil zu fällen. Auf solche Weise, wenn wir auch gar nicht über den Ursprung unserer Vorstellungen nachdächten und unsre Anschauungen der Sinne, sie mögen enthalten, was sie wollen, im Raume und Zeit nach Regeln des Zusammenhanges aller Erkenntniß in einer Erfahrung verknüpfen: so kann, nachdem wir unbehutsam oder vorsichtig sind, trüglicher Schein oder Wahrheit entspringen; das geht lediglich den Gebrauch sinnlicher Vorstellungen im Verstande und nicht ihren Ursprung an. Eben so wenn ich alle Vorstellungen der Sinne sammt ihrer Form, nämlich Raum und Zeit, für nichts als Erscheinungen und die letztern für eine bloße Form der Sinnlichkeit halte, die außer ihr an den Objecten gar nicht angetroffen wird, und ich bediene mich derselben Vorstellungen nur in Beziehung auf mögliche Erfahrung: so ist darin nicht die mindeste Verleitung zum Irrthum, oder ein Schein enthalten, daß ich sie für bloße Erscheinungen halte; denn sie können dessen ungeachtet nach Regeln der Wahrheit in der Erfahrung richtig zusammenhängen. Auf solche Weise gelten alle Sätze der Geometrie vom Raume eben sowohl von allen Gegenständen der Sinne, mithin in Ansehung aller möglichen Erfahrung, ob ich den Raum als eine bloße Form der Sinnlichkeit, oder als etwas an den Dingen selbst Haftendes ansehe; wiewohl ich im ersteren Falle allein begreifen kann, wie es möglich sei, jene Sätze von allen Gegenständen der äußeren Anschauung a priori zu wissen; sonst bleibt in Ansehung aller nur möglichen Erfahrung alles eben so, wie wenn ich diesen Abfall von der gemeinen Meinung gar nicht unternommen haͤtte. Wage ich es aber mit meinen Begriffen von Raum und Zeit über alle mögliche Erfahrung hinauszugehen, welches unvermeidlich ist, wenn ich sie für Beschaffenheiten ausgebe, die den Dingen an sich selbst anhingen (denn was sollte mich da hindern, sie auch von eben denselben Dingen, meine Sinnen möchten nun auch anders eingerichtet sein und für sie passen oder nicht, dennoch gelten zu lassen?), alsdann kann ein wichtiger Irrthum entspringen, der auf einem Scheine beruht, da ich das, was eine blos meinem Subject anhängende Bedingung der Anschauung der Dinge war und sicher für alle Gegenstände der Sinne, mithin alle nur mögliche Erfahrung galt, für allgemein gültig ausgab, weil ich sie auf die Dinge an sich selbst bezog und nicht auf Bedingungen der Erfahrung einschränkte. Also ist es so weit gefehlt, daß meine Lehre von der Idealität des Raumes und der Zeit die ganze Sinnenwelt zum bloßen Scheine mache, daß sie vielmehr das einzige Mittel ist, die Anwendung einer der allerwichtigsten Erkenntnisse, nämlich derjenigen, welche Mathematik a priori vorträgt, auf wirkliche Gegenstände zu sicheren und zu verhüten, daß sie nicht für bloßen Schein gehalten werde, weil ohne diese Bemerkung es ganz unmöglich wäre auszumachen, ob nicht die Anschauungen von Raum und Zeit, die wir von keiner Erfahrung entlehnen, und die dennoch in unserer Vorstellung a priori liegen, bloße selbstgemachte Hirngespinste wären, denen gar kein Gegenstand, wenigstens nicht adäquat, correspondirte, und also Geometrie selbst ein bloßer Schein sei; dagegen ihre unstreitige Gültigkeit in Ansehung aller Gegenstände der Sinnenwelt eben darum, weil diese bloße Erscheinungen sind, von uns hat dargethan werden können. Es ist zweitens so weit gefehlt, daß diese meine Principien darum, weil sie aus den Vorstellungen der Sinne Erscheinungen machen, statt der Wahrheit der Erfahrung sie in bloßen Schein verwandeln sollten, daß sie vielmehr das einzige Mittel sind, den transscendentalen Schein zu verhüten, wodurch Metaphysik von je her getäuscht und eben dadurch zu den kindischen Bestrebungen verleitet worden, nach Seifenblasen zu haschen, weil man Erscheinungen, die doch bloße Vorstellungen sind, für Sachen an sich selbst nahm; woraus alle jene merkwürdige Auftritte der Antinomie der Vernunft erfolgt sind, davon ich weiter hin Erwähnung thun werde, und die durch jene einzige Bemerkung gehoben wird: daß Erscheinung, so lange als sie in der Erfahrung gebraucht wird, Wahrheit, sobald sie aber über die Grenze derselben hinausgeht und transscendent wird, nichts als lauter Schein hervorbringt. Da ich also den Sachen, die wir uns durch Sinne vorstellen, ihre Wirklichkeit lasse und nur unsre sinnliche Anschauung von diesen Sachen dahin einschränke, daß sie in gar keinem Stücke, selbst nicht in den reinen Anschauungen von Raum und Zeit etwas mehr als blos Erscheinung jener Sachen, niemals aber die Beschaffenheit derselben an ihnen selbst vorstelle, so ist dies kein der Natur von mir angedichteter durchgängiger Schein; und meine Protestation wider alle Zumuthung eines Idealisms ist so bündig und einleuchtend, daß sie sogar überflüssig scheinen würde, wenn es nicht unbefugte Richter gäbe, die, indem sie für jede Abweichung von ihrer verkehrten, obgleich gemeinen Meinung gerne einen alten Namen haben möchten und niemals über den Geist der philosophischen Benennungen urtheilen, sondern blos am Buchstaben hingen, bereit ständen, ihren eigenen Wahn an die Stelle wohl bestimmter Begriffe zu setzen und diese dadurch zu verdrehen und zu verunstalten. Denn daß ich selbst dieser meiner Theorie den Namen eines transscendentalen Idealisms gegeben habe, kann keinen berechtigen, ihn mit dem empirischen Idealism des *Cartes* (wiewohl dieser nur eine Aufgabe war, wegen deren Unauflöslichkeit es nach *Cartesens* Meinung jedermann frei stand, die Existenz der körperlichen Welt zu verneinen, weil sie niemals gnugthuend beantwortet werden könnte) oder mit dem mystischen und schwärmerischen des *Berkeley* (wowider und andre ähnliche Hirngespinste unsre Kritik vielmehr das eigentliche Gegenmittel enthält) zu verwechseln. Denn dieser von mir sogenannte Idealism betraf nicht die Existenz der Sachen (die Bezweifelung derselben aber macht eigentlich den Idealism in recipirter Bedeutung aus), denn die zu bezweifeln, ist mir niemals in den Sinn gekommen, sondern blos die sinnliche Vorstellung der Sachen, dazu Raum und Zeit zu oberst gehören, und von diesen, mithin überhaupt von allen Erscheinungen habe ich nur gezeigt: daß sie nicht Sachen (sondern bloße Vorstellungsarten), auch nicht den Sachen an sich selbst angehörige Bestimmungen sind. Das Wort transscendental aber, welches bei mir niemals eine Beziehung unserer Erkenntniß auf Dinge, sondern nur aufs *Erkenntnißvermögen* bedeutet, sollte diese Mißdeutung verhüten. Ehe sie aber dieselbe doch noch fernerhin veranlasse, nehme ich diese Benennung lieber zurück und will ihn den *kritischen* genannt wissen. Wenn es aber ein in der That verwerflicher Idealism ist, wirkliche Sachen (nicht Erscheinungen) in bloße Vorstellungen zu verwandeln: mit welchem Namen will man denjenigen benennen, der umgekehrt bloße Vorstellungen zu Sachen macht? Ich denke, man könne ihn den *träumenden* Idealism nennen zum Unterschiede von dem vorigen, der der *schwärmende* heißen mag, welche beide durch meinen sonst sogenannten transscendentalen, besser *kritischen* Idealism haben abgehalten werden sollen. ### Second part: How is pure natural science possible? *Der transscendentalen Hauptfrage* **Zweiter Theil.** *Wie ist reine Naturwissenschaft möglich?* ### Section 14 § 14. *Natur* ist das *Dasein* der Dinge, so fern es nach allgemeinen Gesetzen bestimmt ist. Sollte Natur das Dasein der Dinge an sich selbst bedeuten, so würden wir sie niemals weder a priori noch a posteriori erkennen können. Nicht a priori; denn wie wollen wir wissen, was den Dingen an sich selbst zukomme, da dieses niemals durch Zergliederung unserer Begriffe (analytische Sätze) geschehen kann, weil ich nicht wissen will, was in meinem Begriffe von einem Dinge enthalten sei (denn das gehört zu seinem logischen Wesen), sondern was in der Wirklichkeit des Dinges zu diesem Begriff hinzukomme, und wodurch das Ding selbst in seinem Dasein außer meinem Begriffe bestimmt sei. Mein Verstand und die Bedingungen, unter denen er allein die Bestimmungen der Dinge in ihrem Dasein verknüpfen kann, schreibt den Dingen selbst keine Regel vor; diese richten sich nicht nach meinem Verstande, sondern mein Verstand müßte sich nach ihnen richten; sie müßten also mir vorher gegeben sein, um diese Bestimmungen von ihnen abzunehmen; alsdann aber wären sie nicht a priori erkannt. Auch a posteriori wäre eine solche Erkenntniß der Natur der Dinge an sich selbst unmöglich. Denn wenn mich Erfahrung *Gesetze*, unter denen das Dasein der Dinge steht, lehren soll, so müßten diese, so fern sie Dinge an sich selbst betreffen, auch außer meiner Erfahrung ihnen *nothwendig* zukommen. Nun lehrt mich die Erfahrung zwar, was dasei, und wie es sei, niemals aber, daß es nothwendiger Weise so und nicht anders sein müsse. Also kann sie die Natur der Dinge an sich selbst niemals lehren. ### Section 15 § 15. Nun sind wir gleichwohl wirklich im Besitze einer reinen Naturwissenschaft, die a priori und mit aller derjenigen Nothwendigkeit, welche zu apodiktischen Sätzen erforderlich ist, Gesetze vorträgt, unter denen die Natur steht. Ich darf hier nur diejenige Propädeutik der Naturlehre, die unter dem Titel der allgemeinen Naturwissenschaft vor aller Physik (die auf empirische Principien gegründet ist) vorhergeht, zum Zeugen rufen. Darin findet man Mathematik, angewandt auf Erscheinungen, auch blos discursive Grundsätze (aus Begriffen), welche den philosophischen Theil der reinen Naturerkenntniß ausmachen. Allein es ist doch auch manches in ihr, was nicht ganz rein und von Erfahrungsquellen unabhängig ist: als der Begriff der *Bewegung*, der *Undurchdringlichkeit* (worauf der empirische Begriff der Materie beruht), der *Trägheit* u. a. m., welche es verhindern, daß sie nicht ganz reine Naturwissenschaft heißen kann; zudem geht sie nur auf die Gegenstände äußerer Sinne, also giebt sie kein Beispiel von einer allgemeinen Naturwissenschaft in strenger Bedeutung, denn die muß die Natur überhaupt, sie mag den Gegenstand äußerer Sinne oder den des innern Sinnes (den Gegenstand der Physik sowohl als Psychologie) betreffen, unter allgemeine Gesetze bringen. Es finden sich aber unter den Grundsätzen jener allgemeinen Physik etliche, die wirklich die Allgemeinheit haben, die wir verlangen, als der Satz: *daß die Substanz bleibt* und beharrt, *daß alles, was geschieht*, jederzeit *durch eine Ursache* nach beständigen Gesetzen vorher *bestimmt sei*, u. s. w. Diese sind wirklich allgemeine Naturgesetze, die völlig a priori bestehen. Es giebt also in der That eine reine Naturwissenschaft, und nun ist die Frage: *wie ist sie möglich?* ### Section 16 § 16. Noch nimmt das Wort Natur eine andre Bedeutung an, die nämlich das *Object* bestimmt, indessen daß in der obigen Bedeutung sie nur die *Gesetzmäßigkeit* der Bestimmungen des Daseins der Dinge überhaupt andeutete. Natur also, materialiter betrachtet, ist der *Inbegriff aller Gegenstände der Erfahrung*. Mit dieser haben wir es hier nur zu thun, da ohnedem Dinge, die niemals Gegenstände einer Erfahrung werden können, wenn sie nach ihrer Natur erkannt werden sollten, uns zu Begriffen nöthigen würden, deren Bedeutung niemals in concreto (in irgend einem Beispiele einer möglichen Erfahrung) gegeben werden könnte, und von deren Natur wir uns also lauter Begriffe machen müßten, deren Realität, d. i. ob sie wirklich sich auf Gegenstände beziehen, oder bloße Gedankendinge sind, gar nicht entschieden werden könnte. Was nicht ein Gegenstand der Erfahrung sein kann, dessen Erkenntniß wäre hyperphysisch, und mit dergleichen haben wir hier gar nicht zu thun, sondern mit der Naturerkenntniß, deren Realität durch Erfahrung bestätigt werden kann, ob sie gleich a priori möglich ist und vor aller Erfahrung hervorgeht. ### Section 17 § 17. Das *Formale* der Natur in dieser engern Bedeutung ist also die Gesetzmäßigkeit aller Gegenstände der Erfahrung und, sofern sie a priori erkannt wird, die *nothwendige* Gesetzmäßigkeit derselben. Es ist aber eben dargethan: daß die Gesetze der Natur an Gegenständen, so fern sie nicht in Beziehung auf mögliche Erfahrung, sondern als Dinge an sich selbst betrachtet werden, niemals a priori können erkannt werden. Wir haben es aber hier auch nicht mit Dingen an sich selbst (dieser ihre Eigenschaften lassen wir dahin gestellt sein), sondern blos mit Dingen als Gegenständen einer möglichen Erfahrung zu thun, und der Inbegriff derselben ist es eigentlich, was wir hier Natur nennen. Und nun frage ich, ob, wenn von der Möglichkeit einer Naturerkenntniß a priori die Rede ist, es besser sei, die Aufgabe so einzurichten: wie ist die nothwendige Gesetzmäßigkeit *der Dinge* als Gegenstände der Erfahrung, oder: wie ist die nothwendige Gesetzmäßigkeit *der Erfahrung* selbst in Ansehung aller ihrer Gegenstände überhaupt a priori zu erkennen möglich? Beim Lichte besehen, wird die Auflösung der Frage, sie mag auf die eine oder die andre Art vorgestellt sein, in Ansehung der reinen Naturerkenntniß (die eigentlich den Punkt der Quästion ausmacht) ganz und gar auf einerlei hinauslaufen. Denn die subjectiven Gesetze, unter denen allein eine Erfahrungserkenntniß von Dingen möglich ist, gelten auch von diesen Dingen als Gegenständen einer möglichen Erfahrung (freilich aber nicht von ihnen als Dingen an sich selbst, dergleichen aber hier auch in keine Betrachtung kommen). Es ist gänzlich einerlei, ob ich sage: ohne das Gesetz, daß, wenn eine Begebenheit wahrgenommen wird, sie jederzeit auf etwas, was vorhergeht, bezogen werde, worauf sie nach einer allgemeinen Regel folgt, kann niemals ein Wahrnehmungsurtheil für Erfahrung gelten; oder ob ich mich so ausdrücke: alles, wovon die Erfahrung lehrt, daß es geschieht, muß eine Ursache haben. Es ist indessen doch schicklicher, die erstere Formel zu wählen. Denn da wir wohl a priori und vor allen gegebenen Gegenständen eine Erkenntniß derjenigen Bedingungen haben können, unter denen allein eine Erfahrung in Ansehung ihrer möglich ist, niemals aber, welchen Gesetzen sie ohne Beziehung auf mögliche Erfahrung an sich selbst unterworfen sein mögen, so werden wir die Natur der Dinge a priori nicht anders studiren können, als daß wir die Bedingungen und allgemeine (obgleich subjective) Gesetze erforschen, unter denen allein ein solches Erkenntniß als Erfahrung (der bloßen Form nach) möglich ist, und darnach die Möglichkeit der Dinge als Gegenstände der Erfahrung bestimmen; denn würde ich die zweite Art des Ausdrucks wählen und die Bedingungen a priori suchen, unter denen Natur als *Gegenstand* der Erfahrung möglich ist, so würde ich leichtlich in Mißverstand gerathen können und mir einbilden, ich hätte von der Natur als einem Dinge an sich selbst zu reden, und da würde ich fruchtlos in endlosen Bemühungen herumgetrieben werden, für Dinge, von denen mir nichts gegeben ist, Gesetze zu suchen. Wir werden es also hier blos mit der Erfahrung und den allgemeinen und a priori gegebenen Bedingungen ihrer Möglichkeit zu thun haben und daraus die Natur als den ganzen Gegenstand aller möglichen Erfahrung bestimmen. Ich denke, man werde mich verstehen, daß ich hier nicht die Regeln der *Beobachtung* einer Natur, die schon gegeben ist, verstehe, die setzen schon Erfahrung voraus; also nicht, wie wir (durch Erfahrung) der Natur die Gesetze ablernen können, denn diese wären alsdann nicht Gesetze a priori und gäben keine reine Naturwissenschaft; sondern wie die Bedingungen a priori von der Möglichkeit der Erfahrung zugleich die Quellen sind, aus denen alle allgemeine Naturgesetze hergeleitet werden müssen. ### Section 18 § 18. Wir müssen denn also zuerst bemerken: daß, obgleich alle Erfahrungsurtheile empirisch sind, d. i. ihren Grund in der unmittelbaren Wahrnehmung der Sinne haben, dennoch nicht umgekehrt alle empirische Urtheile darum Erfahrungsurtheile sind, sondern daß über das Empirische und überhaupt über das der sinnlichen Anschauung Gegebene noch besondere Begriffe hinzukommen müssen, die ihren Ursprung gänzlich a priori im reinen Verstande haben, unter die jede Wahrnehmung allererst subsumirt und dann vermittelst derselben in Erfahrung kann verwandelt werden. *Empirische Urtheile, sofern sie objective Gültigkeit haben*, sind **Erfahrungsurtheile**; die aber, *so nur subjectiv gültig* sind, nenne ich bloße **Wahrnehmungsurtheile**. Die letztern bedürfen keines reinen Verstandesbegriffs, sondern nur der logischen Verknüpfung der Wahrnehmungen in einem denkenden Subject. Die erstern aber erfordern jederzeit über die Vorstellungen der sinnlichen Anschauung noch besondere, im *Verstande ursprünglich erzeugte Begriffe*, welche es eben machen, daß das Erfahrungsurtheil *objectiv gültig* ist. Alle unsere Urtheile sind zuerst bloße Wahrnehmungsurtheile: sie gelten blos für uns, d. i. für unser Subject, und nur hinten nach geben wir ihnen eine neue Beziehung, nämlich auf ein Object, und wollen, daß es auch für uns jederzeit und eben so für jedermann gültig sein solle; denn wenn ein Urtheil mit einem Gegenstande übereinstimmt, so müssen alle Urtheile über denselben Gegenstand auch unter einander übereinstimmen, und so bedeutet die objective Gültigkeit des Erfahrungsurtheils nichts anders, als die nothwendige Allgemeingültigkeit desselben. Aber auch umgekehrt, wenn wir Ursache finden, ein Urtheil für nothwendig allgemeingültig zu halten (welches niemals auf der Wahrnehmung, sondern dem reinen Verstandesbegriffe beruht, unter dem die Wahrnehmung subsumirt ist), so müssen wir es auch für objectiv halten, d. i. daß es nicht blos eine Beziehung der Wahrnehmung auf ein Subject, sondern eine Beschaffenheit des Gegenstandes ausdrücke; denn es wäre kein Grund, warum anderer Urtheile nothwendig mit dem meinigen übereinstimmen müßten, wenn es nicht die Einheit des Gegenstandes wäre, auf den sie sich alle beziehen, mit dem sie übereinstimmen und daher auch alle unter einander zusammenstimmen müssen. ### Section 19 § 19. Es sind daher objective Gültigkeit und nothwendige Allgemeingültigkeit (für jedermann) Wechselbegriffe, und ob wir gleich das Object an sich nicht kennen, so ist doch, wenn wir ein Urtheil als gemeingültig und mithin nothwendig ansehen, eben darunter die objective Gültigkeit verstanden. Wir erkennen durch dieses Urtheil das Object (wenn es auch sonst, wie es an sich selbst sein möchte, unbekannt bliebe) durch die allgemeingültige und nothwendige Verknüpfung der gegebenen Wahrnehmungen; und da dieses der Fall von allen Gegenständen der Sinne ist, so werden Erfahrungsurtheile ihre objective Gültigkeit nicht von der unmittelbaren Erkenntniß des Gegenstandes (denn diese ist unmöglich), sondern blos von der Bedingung der Allgemeingültigkeit der empirischen Urtheile entlehnen, die, wie gesagt, niemals auf den empirischen, ja überhaupt sinnlichen Bedingungen, sondern auf einem reinen Verstandesbegriffe beruht. Das Object bleibt an sich selbst immer unbekannt; wenn aber durch den Verstandesbegriff die Verknüpfung der Vorstellungen, die unsrer Sinnlichkeit von ihm gegeben sind, als allgemeingültig bestimmt wird, so wird der Gegenstand durch dieses Verhältniß bestimmt, und das Urtheil ist objectiv. Wir wollen dieses erläutern: daß das Zimmer warm, der Zucker süß, der Wermuth widrig sei[^299-1], sind blos subjectiv gültige Urtheile. Ich verlange gar nicht, daß ich es jederzeit oder jeder andre es eben so wie ich finden soll; sie drücken nur eine Beziehung zweier Empfindungen auf dasselbe Subject, nämlich mich selbst und auch nur in meinem diesmaligen Zustande der Wahrnehmung, aus und sollen daher auch nicht vom Objecte gelten; dergleichen nenne ich Wahrnehmungsurtheile. Eine ganz andere Bewandtniß hat es mit dem Erfahrungsurtheile. Was die Erfahrung unter gewissen Umständen mich lehrt, muß sie mich jederzeit und auch jedermann lehren, und die Gültigkeit derselben schränkt sich nicht auf das Subject oder seinen damaligen Zustand ein. Daher spreche ich alle dergleichen Urtheile als objectiv gültige aus; als z. B. wenn ich sage, die Luft ist elastisch, so ist dieses Urtheil zunächst nur ein Wahrnehmungsurtheil, ich beziehe zwei Empfindungen in meinen Sinnen nur auf einander. Will ich, es soll Erfahrungsurtheil heißen, so verlange ich, daß diese Verknüpfung unter einer Bedingung stehe, welche sie allgemein gültig macht. Ich will also, daß ich jederzeit und auch jedermann dieselbe Wahrnehmung unter denselben Umständen nothwendig verbinden müsse. [^299-1]: Ich gestehe gern, daß diese Beispiele nicht solche Wahrnehmungsurtheile vorstellen, die jemals Erfahrungsurtheile werden könnten, wenn man auch einen Verstandesbegriff hinzu thäte, weil sie sich blos aufs Gefühl, welches jederman als blos subjectiv erkennt und welches also niemals dem Object beigelegt werden darf, beziehen und also auch niemals objectiv werden können; ich wollte nur vor der Hand ein Beispiel von dem Urtheile geben, was blos subjectiv gültig ist und in sich keinen Grund zur nothwendigen Allgemeingültigkeit und dadurch zu einer Beziehung aufs Object enthält. Ein Beispiel der Wahrnehmungsurtheile, die durch hinzugesetzten Verstandesbegriff Erfahrungsurtheile werden, folgt in der nächsten Anmerkung. ### Section 20 § 20. Wir werden daher Erfahrung überhaupt zergliedern müssen, um zu sehen, was in diesem Product der Sinne und des Verstandes enthalten und wie das Erfahrungsurtheil selbst möglich sei. Zum Grunde liegt die Anschauung, deren ich mir bewußt bin, d. i. Wahrnehmung (perceptio), die blos den Sinnen angehört. Aber zweitens gehört auch dazu das Urtheilen (das blos dem Verstande zukommt). Dieses Urtheilen kann nun zwiefach sein: erstlich, indem ich blos die Wahrnehmungen vergleiche und in einem Bewußtsein meines Zustandes, oder zweitens, da ich sie in einem Bewußtsein überhaupt verbinde. Das erstere Urtheil ist blos ein Wahrnehmungsurtheil und hat so fern nur subjective Gültigkeit, es ist blos Verknüpfung der Wahrnehmungen in meinem Gemüthszustande ohne Beziehung auf den Gegenstand. Daher ist es nicht, wie man gemeiniglich sich einbildet, zur Erfahrung gnug, Wahrnehmungen zu vergleichen und in einem Bewußtsein vermittelst des Urtheilens zu verknüpfen; dadurch entspringt keine Allgemeingültigkeit und Nothwendigkeit des Urtheils, um deren willen es allein objectiv gültig und Erfahrung sein kann. Es geht also noch ein ganz anderes Urtheil voraus, ehe aus Wahrnehmung Erfahrung werden kann. Die gegebene Anschauung muß unter einem Begriff subsumirt werden, der die Form des Urtheilens überhaupt in Ansehung der Anschauung bestimmt, das empirische Bewußtsein der letzteren in einem Bewußtsein überhaupt verknüpft und dadurch den empirischen Urtheilen Allgemeingültigkeit verschafft; dergleichen Begriff ist ein reiner Verstandesbegriff a priori, welcher nichts thut, als blos einer Anschauung die Art überhaupt zu bestimmen, wie sie zu Urtheilen dienen kann. Es sei ein solcher Begriff der Begriff der Ursache, so bestimmt er die Anschauung, die unter ihm subsumirt ist, z. B. die der Luft in Ansehung des Urtheilens überhaupt, nämlich daß der Begriff der Luft in Ansehung der Ausspannung in dem Verhältniß der Antecedens zum Consequens in einem hypothetischen Urtheile diene. Der Begriff der Ursache ist also ein reiner Verstandesbegriff, der von aller möglichen Wahrnehmung gänzlich unterschieden ist und nur dazu dient, diejenige Vorstellung, die unter ihm enthalten ist, in Ansehung des Urtheilens überhaupt zu bestimmen, mithin ein allgemeingültiges Urtheil möglich zu machen. Nun wird, ehe aus einem Wahrnehmungsurtheil ein Urtheil der Erfahrung werden kann, zuerst erfordert: daß die Wahrnehmung unter einem dergleichen Verstandesbegriffe subsumirt werde; z. B. die Luft gehört unter den Begriff der Ursache, welche das Urtheil über dieselbe in Ansehung der Ausdehnung als hypothetisch bestimmt.[^301-1] Dadurch wird nun nicht diese Ausdehnung als blos zu meiner Wahrnehmung der Luft in meinem Zustande, oder in mehrern meiner Zustände, oder in dem Zustande der Wahrnehmung anderer gehörig, sondern als dazu *nothwendig* gehörig vorgestellt; und das Urtheil: die Luft ist elastisch, wird allgemeingültig und dadurch allererst Erfahrungsurtheil, daß gewisse Urtheile vorhergehen, die die Anschauung der Luft unter den Begriff der Ursache und Wirkung subsumiren und dadurch die Wahrnehmungen nicht blos respective auf einander in meinem Subjecte, sondern in Ansehung der Form des Urtheilens überhaupt (hier der hypothetischen) bestimmen und auf solche Art das empirische Urtheil allgemeingültig machen. Zergliedert man alle seine synthetische Urtheile, so fern sie objectiv gelten, so findet man, daß sie niemals aus bloßen Anschauungen bestehen, die blos, wie man gemeiniglich dafür hält, durch Vergleichung in einem Urtheil verknüpft worden, sondern daß sie unmöglich sein würden, wäre nicht über die von der Anschauung abgezogene Begriffe noch ein reiner Verstandesbegriff hinzugekommen, unter dem jene Begriffe subsumirt und so allererst in einem objectiv gültigen Urtheile verknüpft worden. Selbst die Urtheile der reinen Mathematik in ihren einfachsten Axiomen sind von dieser Bedingung nicht ausgenommen. Der Grundsatz: die gerade Linie ist die kürzeste zwischen zwei Punkten, setzt voraus, daß die Linie unter den Begriff der Größe subsumirt werde, welcher gewiß keine bloße Anschauung ist, sondern lediglich im Verstande seinen Sitz hat und dazu dient, die Anschauung (der Linie) in Absicht auf die Urtheile, die von ihr gefällt werden mögen, in Ansehung der Quantität derselben, nämlich der Vielheit, (als iudicia plurativa)[^302-1] zu bestimmen, indem unter ihnen verstanden wird, daß in einer gegebenen Anschauung vieles Gleichartige enthalten sei. [^301-1]: Um ein leichter einzusehendes Beispiel zu haben, nehme man folgendes: wenn die Sonne den Stein bescheint, so wird er warm. Dieses Urtheil ist ein bloßes Wahrnehmungsurtheil und enthält keine Nothwendigkeit, ich mag dieses noch so oft und andere auch noch so oft wahrgenommen haben; die Wahrnehmungen finden sich nur gewöhnlich so verbunden. Sage ich aber: die Sonne *erwärmt* den Stein, so kommt über die Wahrnehmung noch der Verstandesbegriff der Ursache hinzu, der mit dem Begriff des Sonnenscheins den der Wärme *nothwendig* verknüpft, und das synthetische Urtheil wird nothwendig allgemeingültig, folglich objectiv, und aus einer Wahrnehmung in Erfahrung verwandelt. [^302-1]: So wollte ich lieber die Urtheile genannt wissen, die man in der Logik particularia nennt. Denn der letztere Ausdruck enthält schon den Gedanken, daß sie nicht allgemein sind. Wenn ich aber von der Einheit (in einzelnen Urtheilen) anhebe und so zur Allheit fortgehe, so kann ich noch keine Beziehung auf die Allheit beimischen: ich denke nur die Vielheit ohne Allheit, nicht die Ausnahme von derselben. Dieses ist nöthig, wenn die logische Momente den reinen Verstandesbegriffen untergelegt werden sollen; im logischen Gebrauche kann man es beim Alten lassen. ### Section 21 § 21. Um nun also die Möglichkeit der Erfahrung, so fern sie auf reinen Verstandesbegriffen a priori beruht, darzulegen, müssen wir zuvor das, was zum Urtheilen überhaupt gehört, und die verschiedene Momente des Verstandes in denselben in einer vollständigen Tafel vorstellen; denn die reinen Verstandesbegriffe, die nichts weiter sind als Begriffe von Anschauungen überhaupt, so fern diese in Ansehung eines oder des andern dieser Momente zu Urtheilen an sich selbst, mithin nothwendig und allgemeingültig bestimmt sind, werden ihnen ganz genau parallel ausfallen. Hiedurch werden auch die Grundsätze a priori der Möglichkeit aller Erfahrung als einer objectiv gültigen empirischen Erkenntniß ganz genau bestimmt werden. Denn sie sind nichts anders als Sätze, welche alle Wahrnehmung (gemäß gewissen allgemeinen Bedingungen der Anschauung) unter jene reine Verstandesbegriffe subsumiren. **Logische Tafel** *der Urtheile.* | | 1. | | |---|---|---| | | *Der Quantität nach* | | | | Allgemeine | | | | Besondere | | | | Einzelne | | | 2. | | 3. | | *Der Qualität nach* | | *Der Relation nach* | | Bejahende | | Kategorische | | Verneinende | | Hypothetische | | Unendliche | | Disjunctive | | | 4. | | | | *Der Modalität nach* | | | | Problematische | | | | Assertorische | | | | Apodiktische. | | **Transscendentale Tafel** *der Verstandesbegriffe.* | | 1. | | |---|---|---| | | *Der Quantität nach* | | | | Einheit (das Maß) | | | | Vielheit (die Größe) | | | | Allheit (das Ganze) | | | 2. | | 3. | | *Der Qualität* | | *Der Relation* | | Realität | | Substanz | | Negation | | Ursache | | Einschränkung | | Gemeinschaft | | | 4. | | | | *Der Modalität* | | | | Möglichkeit | | | | Dasein | | | | Nothwendigkeit. | | **Reine physiologische Tafel** allgemeiner Grundsätze der Naturwissenschaft. | | 1. | | |---|---|---| | | *Axiomen* | | | | der Anschauung | | | 2. | | 3. | | *Anticipationen* | | *Analogien* | | der Wahrnehmung | | der Erfahrung | | | 4. | | | | *Postulate* | | | | des empirischen Denkens | | | | überhaupt. | | ### Section 21a § 21. [a]. Um alles Bisherige in einen Begriff zusammenzufassen, ist zuvörderst nöthig, die Leser zu erinnern, daß hier nicht von dem Entstehen der Erfahrung die Rede sei, sondern von dem, was in ihr liegt. Das erstere gehört zur empirischen Psychologie und würde selbst auch da ohne das zweite, welches zur Kritik der Erkenntniß und besonders des Verstandes gehört, niemals gehörig entwickelt werden können. Erfahrung besteht aus Anschauungen, die der Sinnlichkeit angehören, und aus Urtheilen, die lediglich ein Geschäfte des Verstandes sind. Diejenige Urtheile aber, die der Verstand lediglich aus sinnlichen Anschauungen macht, sind noch bei weitem nicht Erfahrungsurtheile. Denn in einem Fall würde das Urtheil nur die Wahrnehmungen verknüpfen, so wie sie in der sinnlichen Anschauung gegeben sind; in dem letztern Falle aber sollen die Urtheile sagen, was Erfahrung überhaupt, mithin nicht, was die bloße Wahrnehmung, deren Gültigkeit blos subjectiv ist, enthält. Das Erfahrungsurtheil muß also noch über die sinnliche Anschauung und die logische Verknüpfung derselben (nachdem sie durch Vergleichung allgemein gemacht worden) in einem Urtheile etwas hinzufügen, was das synthetische Urtheil als nothwendig und hiedurch als allgemeingültig bestimmt; und dieses kann nichts anders sein als derjenige Begriff, der die Anschauung in Ansehung einer Form des Urtheils vielmehr als der anderen als an sich bestimmt vorstellt, d. i. ein Begriff von derjenigen synthetischen Einheit der Anschauungen, die nur durch eine gegebne logische Function der Urtheile vorgestellt werden kann. ### Section 22 § 22. Die Summe hievon ist diese: die Sache der Sinne ist, anzuschauen; die des Verstandes, zu denken. Denken aber ist Vorstellungen in einem Bewußtsein vereinigen. Diese Vereinigung entsteht entweder blos relativ aufs Subject und ist zufällig und subjectiv, oder sie findet schlechthin statt und ist nothwendig oder objectiv. Die Vereinigung der Vorstellungen in einem Bewußtsein ist das Urtheil. Also ist Denken so viel als Urtheilen, oder Vorstellungen auf Urtheile überhaupt beziehen. Daher sind Urtheile entweder blos subjectiv, wenn Vorstellungen auf ein Bewußtsein in einem Subject allein bezogen und in ihm vereinigt werden; oder sie sind objectiv, wenn sie in einem Bewußtsein überhaupt, d. i. darin nothwendig, vereinigt werden. Die logische Momente aller Urtheile sind so viel mögliche Arten, Vorstellungen in einem Bewußtsein zu vereinigen. Dienen aber eben dieselben als Begriffe, so sind sie Begriffe von der *nothwendigen* Vereinigung derselben in einem Bewußtsein, mithin Principien objectiv gültiger Urtheile. Diese Vereinigung in einem Bewußtsein ist entweder analytisch, durch die Identität, oder synthetisch, durch die Zusammensetzung und Hinzukunft verschiedener Vorstellungen zu einander. Erfahrung besteht in der synthetischen Verknüpfung der Erscheinungen (Wahrnehmungen) in einem Bewußtsein, so fern dieselbe nothwendig ist. Daher sind reine Verstandesbegriffe diejenige, unter denen alle Wahrnehmungen zuvor müssen subsumirt werden, ehe sie zu Erfahrungsurtheilen dienen können, in welchen die synthetische Einheit der Wahrnehmungen als nothwendig und allgemeingültig vorgestellt wird.[^305-1] [^305-1]: Wie stimmt aber dieser Satz, daß Erfahrungsurtheile Nothwendigkeit in der Synthesis der Wahrnehmungen enthalten sollen, mit meinem oben vielfältig eingeschärften Satze, daß Erfahrung als Erkenntniß a posteriori blos zufällige Urtheile geben könne? Wenn ich sage: Erfahrung lehrt mir etwas, so meine ich jederzeit nur die Wahrnehmung, die in ihr liegt, z. B. daß auf die Beleuchtung des Steins durch die Sonne jederzeit Wärme folge, und also ist der Erfahrungssatz so fern allemal zufällig. Daß diese Erwärmung nothwendig aus der Beleuchtung durch die Sonne erfolge, ist zwar in dem Erfahrungsurtheile (vermöge des Begriffs der Ursache) enthalten, aber das lerne ich nicht durch Erfahrung, sondern umgekehrt, Erfahrung wird allererst durch diesen Zusatz des Verstandesbegriffs (der Ursache) zur Wahrnehmung erzeugt. Wie die Wahrnehmung zu diesem Zusatze komme, darüber muß die Kritik im Abschnitte von der transscendentalen Urtheilskraft Seite 137 u. f. nachgesehen werden. ### Section 23 § 23. Urtheile, so fern sie blos als die Bedingung der Vereinigung gegebener Vorstellungen in einem Bewußtsein betrachtet werden, sind Regeln. Diese Regeln, so fern sie die Vereinigung als nothwendig vorstellen, sind Regeln a priori, und so fern keine über sie sind, von denen sie abgeleitet werden, Grundsätze. Da nun in Ansehung der Möglichkeit aller Erfahrung, wenn man an ihr blos die Form des Denkens betrachtet, keine Bedingungen der Erfahrungsurtheile über diejenige sind, welche die Erscheinungen nach der verschiedenen Form ihrer Anschauung unter reine Verstandesbegriffe bringen, die das empirische Urtheil objectiv-gültig machen, so sind diese die Grundsätze a priori möglicher Erfahrung. Die Grundsätze möglicher Erfahrung sind nun zugleich allgemeine Gesetze der Natur, welche a priori erkannt werden können. Und so ist die Aufgabe, die in unsrer vorliegenden zweiten Frage liegt: *Wie ist reine Naturwissenschaft möglich?*, aufgelöset. Denn das Systematische, was zur Form einer Wissenschaft erfordert wird, ist hier vollkommen anzutreffen, weil über die genannte formale Bedingungen aller Urtheile überhaupt, mithin aller Regeln überhaupt, die die Logik darbietet, keine mehr möglich sind, und diese ein logisches System, die darauf gegründeten Begriffe aber, welche die Bedingungen a priori zu allen synthetischen und nothwendigen Urtheilen enthalten, eben darum ein transscendentales, endlich die Grundsätze, vermittelst deren alle Erscheinungen unter diese Begriffe subsumirt werden, ein physiologisches, d. i. ein Natursystem ausmachen, welches vor aller empirischen Naturerkenntniß vorhergeht, diese zuerst möglich macht und daher die eigentliche allgemeine und reine Naturwissenschaft genannt werden kann. ### Section 24 § 24. Der erste[^306-1] jener physiologischen Grundsätze subsumirt alle Erscheinungen, als Anschauungen im Raum und Zeit, unter den Begriff der *Größe* und ist so fern ein Princip der Anwendung der Mathematik auf Erfahrung. Der zweite subsumirt das eigentlich Empirische, nämlich die Empfindung, die das Reale der Anschauungen bezeichnet, nicht geradezu unter den Begriff der *Größe*, weil Empfindung keine Anschauung ist, die Raum oder Zeit *enthielte*, ob sie gleich den ihr correspondirenden Gegenstand in beide setzt; allein es ist zwischen Realität (Empfindungsvorstellung) und der Null, d. i. dem gänzlich Leeren der Anschauung, in der Zeit doch ein Unterschied, der eine Größe hat, da nämlich zwischen einem jeden gegebenen Grade Licht und der Finsterniß, zwischen einem jeden Grade Wärme und der gänzlichen Kälte, jedem Grad der Schwere und der absoluten Leichtigkeit, jedem Grade der Erfüllung des Raumes und dem völlig leeren Raume immer noch kleinere Grade gedacht werden können, so wie selbst zwischen einem Bewußtsein und dem völligen Unbewußtsein (psychologischer Dunkelheit) immer noch kleinere stattfinden; daher keine Wahrnehmung möglich ist, welche einen absoluten Mangel bewiese, z. B. keine psychologische Dunkelheit, die nicht als ein Bewußtsein betrachtet werden könnte, welches nur von anderem, stärkerem überwogen wird, und so in allen Fällen der Empfindung; weswegen der Verstand sogar Empfindungen, welche die eigentliche Qualität der empirischen Vorstellungen (Erscheinungen) ausmachen, anticipiren kann vermittelst des Grundsatzes, daß sie alle insgesammt, mithin das Reale aller Erscheinung Grade habe, welches die zweite Anwendung der Mathematik (mathesis intensorum) auf Naturwissenschaft ist. [^306-1]: Diese drei auf einander folgende Paragraphen werden schwerlich gehörig verstanden werden können, wenn man nicht das, was die Kritik über die Grundsätze sagt, dabei zur Hand nimmt; sie können aber den Nutzen haben, das Allgemeine derselben leichter zu übersehen und auf die Hauptmomente Acht zu haben. ### Section 25 § 25. In Ansehung des Verhältnisses der Erscheinungen und zwar lediglich in Absicht auf ihr Dasein ist die Bestimmung dieses Verhältnisses nicht mathematisch, sondern dynamisch und kann niemals objectiv gültig, mithin zu einer Erfahrung tauglich sein, wenn sie nicht unter Grundsätzen a priori steht, welche die Erfahrungserkenntniß in Ansehung derselben allererst möglich machen. Daher müssen Erscheinungen unter den Begriff der Substanz, welcher aller Bestimmung des Daseins als ein Begriff vom Dinge selbst zum Grunde liegt, oder zweitens, so fern eine Zeitfolge unter den Erscheinungen, d. i. eine Begebenheit, angetroffen wird, unter den Begriff einer Wirkung in Beziehung auf Ursache, oder, so fern das Zugleichsein objectiv, d. i. durch ein Erfahrungsurtheil, erkannt werden soll, unter den Begriff der Gemeinschaft (Wechselwirkung) subsumirt werden, und so liegen Grundsätze a priori objectiv gültigen, obgleich empirischen Urtheilen, d. i. der Möglichkeit der Erfahrung, so fern sie Gegenstände dem Dasein nach in der Natur verknüpfen soll, zum Grunde. Diese Grundsätze sind die eigentlichen Naturgesetze, welche dynamisch heißen können. Zuletzt gehört auch zu den Erfahrungsurtheilen die Erkenntniß der Übereinstimmung und Verknüpfung: nicht sowohl der Erscheinungen unter einander in der Erfahrung, als vielmehr ihr Verhältniß zur Erfahrung überhaupt, welches entweder ihre Übereinstimmung mit den formalen Bedingungen, die der Verstand erkennt, oder Zusammenhang mit dem Materialen der Sinne und der Wahrnehmung, oder beiden in einen Begriff vereinigt, folglich Möglichkeit, Wirklichkeit und Nothwendigkeit nach allgemeinen Naturgesetzen enthält; welches die physiologische Methodenlehre (Unterscheidung der Wahrheit und Hypothesen und die Grenzen der Zuverlässigkeit der letzteren) ausmachen würde. ### Section 26 § 26. Obgleich die dritte *aus der Natur des Verstandes selbst* nach kritischer Methode gezogene Tafel der Grundsätze eine Vollkommenheit an sich zeigt, darin sie sich weit über jede andre erhebt, die *von den Sachen selbst* auf dogmatische Weise, obgleich vergeblich, jemals versucht worden ist, oder nur künftig versucht werden mag: nämlich daß sie alle synthetische Grundsätze a priori vollständig enthält und nach einem Princip, nämlich dem Vermögen zu Urtheilen überhaupt, welches das Wesen der Erfahrung in Absicht auf den Verstand ausmacht, ausgeführt worden, so daß man gewiß sein kann, es gebe keine dergleichen Grundsätze mehr (eine Befriedigung, die die dogmatische Methode niemals verschaffen kann), so ist dieses doch bei weitem noch nicht ihr größtes Verdienst. Man muß auf den Beweisgrund Acht geben, der die Möglichkeit dieser Erkenntniß a priori entdeckt und alle solche Grundsätze zugleich auf eine Bedingung einschränkt, die niemals übersehen werden muß, wenn sie nicht mißverstanden und im Gebrauche weiter ausgedehnt werden soll, als der ursprüngliche Sinn, den der Verstand darin legt, es haben will: nämlich daß sie nur die Bedingungen möglicher Erfahrung überhaupt enthalten, so fern sie Gesetzen a priori unterworfen ist. So sage ich nicht: daß Dinge *an sich selbst* eine Größe, ihre Realität einen Grad, ihre Existenz Verknüpfung der Accidenzen in einer Substanz u. s. w. enthalte; denn das kann niemand beweisen, weil eine solche synthetische Verknüpfung aus bloßen Begriffen, wo alle Beziehung auf sinnliche Anschauung einerseits und alle Verknüpfung derselben in einer möglichen Erfahrung andererseits mangelt, schlechterdings unmöglich ist. Die wesentliche Einschränkung der Begriffe also in diesen Grundsätzen ist: daß alle Dinge nur *als Gegenstände der Erfahrung* unter den genannten Bedingungen nothwendig a priori stehen. Hieraus folgt denn zweitens auch eine specifisch eigenthümliche Beweisart derselben: daß die gedachte Grundsätze auch nicht geradezu auf Erscheinungen und ihr Verhältniß, sondern auf die Möglichkeit der Erfahrung, wovon Erscheinungen nur die Materie, nicht aber die Form ausmachen, d. i. auf objectiv- und allgemeingültige synthetische Sätze, worin sich eben Erfahrungsurtheile von bloßen Wahrnehmungsurtheilen unterscheiden, bezogen werden. Dieses geschieht dadurch: daß die Erscheinungen als bloße Anschauungen, *welche einen Theil von Raum und Zeit einnehmen*, unter dem Begriff der Größe stehen, welcher das Mannigfaltige derselben a priori nach Regeln synthetisch vereinigt; daß, so fern die Wahrnehmung außer der Anschauung auch Empfindung enthält, zwischen welcher und der Null, d. i. dem völligen Verschwinden derselben, jederzeit ein Übergang durch Verringerung stattfindet, das Reale der Erscheinungen einen Grad haben müsse, so fern sie nämlich selbst *keinen Theil von Raum oder Zeit einnimmt*[^309-1], aber doch der Übergang zu ihr von der leeren Zeit oder Raum nur in der Zeit möglich ist; mithin, obzwar Empfindung als die Qualität der empirischen Anschauung in Ansehung dessen, worin sie sich specifisch von andern Empfindungen unterscheidet, niemals a priori erkannt werden kann, sie dennoch in einer möglichen Erfahrung überhaupt als Größe der Wahrnehmung intensiv von jeder andern gleichartigen unterschieden werden könne; woraus denn die Anwendung der Mathematik auf Natur in Ansehung der sinnlichen Anschauung, durch welche sie uns gegeben wird, zuerst möglich gemacht und bestimmt wird. Am meisten aber muß der Leser auf die Beweisart der Grundsätze, die unter dem Namen der Analogien der Erfahrung vorkommen, aufmerksam sein. Denn weil diese nicht, so wie die Grundsätze der Anwendung der Mathematik auf Naturwissenschaft überhaupt, die Erzeugung der Anschauungen, sondern die Verknüpfung ihres Daseins in einer Erfahrung betreffen, diese aber nichts anders als die Bestimmung der Existenz in der Zeit nach nothwendigen Gesetzen sein kann, unter denen sie allein objectiv-gültig, mithin Erfahrung ist: so geht der Beweis nicht auf die synthetische Einheit in der Verknüpfung der Dinge an sich selbst, sondern der *Wahrnehmungen* und zwar dieser nicht in Ansehung ihres Inhalts, sondern der Zeitbestimmung und des Verhältnisses des Daseins in ihr nach allgemeinen Gesetzen. Diese allgemeinen Gesetze enthalten also die Nothwendigkeit der Bestimmung des Daseins in der Zeit überhaupt (folglich nach einer Regel des Verstandes a priori), wenn die empirische Bestimmung in der relativen Zeit objectiv-gültig, mithin Erfahrung sein soll. Mehr kann ich hier als in Prolegomenen nicht anführen, als nur daß ich dem Leser, welcher in der langen Gewohnheit steckt, Erfahrung für eine blos empirische Zusammensetzung der Wahrnehmungen zu halten, und daher daran gar nicht denkt, daß sie viel weiter geht, als diese reichen, nämlich empirischen Urtheilen Allgemeingültigkeit giebt und dazu einer reinen Verstandeseinheit bedarf, die a priori vorhergeht, empfehle: auf diesen Unterschied der Erfahrung von einem bloßen Aggregat von Wahrnehmungen wohl Acht zu haben und aus diesem Gesichtspunkte die Beweisart zu beurtheilen. [^309-1]: Die Wärme, das Licht ꝛc. sind im kleinen Raume (dem Grade nach) eben so groß, als in einem großen; eben so die innere Vorstellungen, der Schmerz, das Bewußtsein überhaupt nicht kleiner dem Grade nach, ob sie eine kurze oder lange Zeit hindurch dauern. Daher ist die Größe hier in einem Punkte und in einem Augenblicke eben so groß als in jedem noch so großen Raume oder Zeit. Grade sind also größer, aber nicht in der Anschauung, sondern der bloßen Empfindung nach oder auch der Größe des Grundes einer Anschauung und können nur durch das Verhältniß von 1 zu 0, d. i. dadurch daß eine jede derselben durch unendliche Zwischengrade bis zum Verschwinden, oder von der Null durch unendliche Momente des Zuwachses bis zu einer bestimmten Empfindung in einer gewissen Zeit erwachsen kann, als Größen geschätzt werden (quantitas qualitatis est gradus). ### Section 27 § 27. Hier ist nun der Ort, den *Humischen* Zweifel aus dem Grunde zu heben. Er behauptete mit Recht: daß wir die Möglichkeit der Causalität, d. i. der Beziehung des Daseins eines Dinges auf das Dasein von irgend etwas anderem, was durch jenes nothwendig gesetzt werde, durch Vernunft auf keine Weise einsehen. Ich setze noch hinzu: daß wir eben so wenig den Begriff der Subsistenz, d. i. der Nothwendigkeit, darin einsehen, daß dem Dasein der Dinge ein Subject zum Grunde liege, das selbst kein Prädicat von irgend einem anderen Dinge sein könne, ja sogar daß wir uns keinen Begriff von der Möglichkeit eines solchen Dinges machen können (obgleich wir in der Erfahrung Beispiele seines Gebrauchs aufzeigen können), imgleichen daß eben diese Unbegreiflichkeit auch die Gemeinschaft der Dinge betreffe, indem gar nicht einzusehen ist, wie aus dem Zustande eines Dinges eine Folge auf den Zustand ganz anderer Dinge außer ihm und so wechselseitig könne gezogen werden, und wie Substanzen, deren jede doch ihre eigene, abgesonderte Existenz hat, von einander und zwar nothwendig abhängen sollen. Gleichwohl bin ich weit davon entfernt, diese Begriffe als blos aus der Erfahrung entlehnt und die Nothwendigkeit, die in ihnen vorgestellt wird, als angedichtet und für bloßen Schein zu halten, den uns eine lange Gewohnheit vorspiegelt; vielmehr habe ich hinreichend gezeigt, daß sie und die Grundsätze aus denselben a priori vor aller Erfahrung fest stehen und ihre ungezweifelte objective Richtigkeit, aber freilich nur in Ansehung der Erfahrung haben. ### Section 28 § 28. Ob ich also gleich von einer solchen Verknüpfung der Dinge an sich selbst, wie sie als Substanz existiren oder als Ursache wirken oder mit andern (als Theile eines realen Ganzen) in Gemeinschaft stehen können, nicht den mindesten Begriff habe, noch weniger aber dergleichen Eigenschaften an Erscheinungen als Erscheinungen denken kann (weil jene Begriffe nichts, was in den Erscheinungen liegt, sondern was der Verstand allein denken muß, enthalten), so haben wir doch von einer solchen Verknüpfung der Vorstellungen in unserm Verstande und zwar in Urtheilen überhaupt einen dergleichen Begriff, nämlich: daß Vorstellungen in einer Art Urtheile als Subject in Beziehung auf Prädicate, in einer anderen als Grund in Beziehung auf Folge und in einer dritten als Theile, die zusammen ein ganzes mögliches Erkenntniß ausmachen, gehören. Ferner erkennen wir a priori: daß, ohne die Vorstellung eines Objects in Ansehung eines oder des andern dieser Momente als bestimmt anzusehen, wir gar keine Erkenntniß, die von dem Gegenstande gelte, haben könnten; und wenn wir uns mit dem Gegenstande an sich selbst beschäftigten, so wäre kein einziges Merkmal möglich, woran ich erkennen könnte, daß er in Ansehung eines oder des andern gedachter Momente bestimmt sei, d. i. unter den Begriff der Substanz oder der Ursache oder (im Verhältniß gegen andere Substanzen) unter den Begriff der Gemeinschaft gehöre; denn von der Möglichkeit einer solchen Verknüpfung des Daseins habe ich keinen Begriff. Es ist aber auch die Frage nicht, wie Dinge an sich, sondern wie Erfahrungserkenntniß der Dinge in Ansehung gedachter Momente der Urtheile überhaupt bestimmt sei, d. i. wie Dinge als Gegenstände der Erfahrung unter jene Verstandesbegriffe können und sollen subsumirt werden. Und da ist es klar, daß ich nicht allein die Möglichkeit, sondern auch die Nothwendigkeit, alle Erscheinungen unter diese Begriffe zu subsumiren, d. i. sie zu Grundsätzen der Möglichkeit der Erfahrung zu brauchen, vollkommen einsehe. ### Section 29 § 29. Um einen Versuch an *Humes* problematischem Begriff (dieser seiner crux metaphysicorum), nämlich dem Begriffe der Ursache, zu machen, so ist mir erstlich vermittelst der Logik die Form eines bedingten Urtheils überhaupt, nämlich ein gegebenes Erkenntniß als Grund und das andere als Folge zu gebrauchen, a priori gegeben. Es ist aber möglich, daß in der Wahrnehmung eine Regel des Verhältnisses angetroffen wird, die da sagt, daß auf eine gewisse Erscheinung eine andere (obgleich nicht umgekehrt) beständig folgt; und dieses ist ein Fall, mich des hypothetischen Urtheils zu bedienen und z. B. zu sagen: wenn ein Körper lange gnug von der Sonne beschienen ist, so wird er warm. Hier ist nun freilich noch nicht eine Nothwendigkeit der Verknüpfung, mithin der Begriff der Ursache. Allein ich fahre fort und sage: wenn obiger Satz, der blos eine subjective Verknüpfung der Wahrnehmungen ist, ein Erfahrungssatz sein soll, so muß er als nothwendig und allgemeingültig angesehen werden. Ein solcher Satz aber würde sein: Sonne ist durch ihr Licht die Ursache der Wärme. Die obige empirische Regel wird nunmehr als Gesetz angesehen und zwar nicht als geltend blos von Erscheinungen, sondern von ihnen zum Behuf einer möglichen Erfahrung, welche durchgängig und also nothwendig gültige Regeln bedarf. Ich sehe also den Begriff der Ursache als einen zur bloßen Form der Erfahrung nothwendig gehörigen Begriff und dessen Möglichkeit als einer synthetischen Vereinigung der Wahrnehmungen in einem Bewußtsein überhaupt sehr wohl ein: die Möglichkeit eines Dinges überhaupt aber als einer Ursache sehe ich gar nicht ein und zwar darum, weil der Begriff der Ursache ganz und gar keine den Dingen, sondern nur der Erfahrung anhängende Bedingung andeutet, nämlich daß diese nur eine objectiv-gültige Erkenntniß von Erscheinungen und ihrer Zeitfolge sein könne, so fern die vorhergehende mit der nachfolgenden nach der Regel hypothetischer Urtheile verbunden werden kann. ### Section 30 § 30. Daher haben auch die reine Verstandesbegriffe ganz und gar keine Bedeutung, wenn sie von Gegenständen der Erfahrung abgehen und auf Dinge an sich selbst (noumena) bezogen werden wollen. Sie dienen gleichsam nur, Erscheinungen zu buchstabiren, um sie als Erfahrung lesen zu können; die Grundsätze, die aus der Beziehung derselben auf die Sinnenwelt entspringen, dienen nur unserm Verstande zum Erfahrungsgebrauch; weiter hinaus sind es willkürliche Verbindungen ohne objective Realität, deren Möglichkeit man weder a priori erkennen, noch ihre Beziehung auf Gegenstände durch irgend ein Beispiel bestätigen, oder nur verständlich machen kann, weil alle Beispiele nur aus irgend einer möglichen Erfahrung entlehnt, mithin auch die Gegenstände jener Begriffe nirgend anders, als in einer möglichen Erfahrung angetroffen werden können. Diese vollständige, obzwar wider die Vermuthung des Urhebers ausfallende Auflösung des *Humischen* Problems rettet also den reinen Verstandesbegriffen ihren Ursprung a priori und den allgemeinen Naturgesetzen ihre Gültigkeit als Gesetzen des Verstandes, doch so, daß sie ihren Gebrauch nur auf Erfahrung einschränkt, darum weil ihre Möglichkeit blos in der Beziehung des Verstandes auf Erfahrung ihren Grund hat; nicht aber so, daß sie sich von Erfahrung, sondern daß Erfahrung sich von ihnen ableitet, welche ganz umgekehrte Art der Verknüpfung *Hume* sich niemals einfallen ließ. Hieraus fließt nun folgendes Resultat aller bisherigen Nachforschungen: „Alle synthetische Grundsätze a priori sind nichts weiter als Principien möglicher Erfahrung“ und können niemals auf Dinge an sich selbst, sondern nur auf Erscheinungen als Gegenstände der Erfahrung bezogen werden. Daher auch reine Mathematik sowohl als reine Naturwissenschaft niemals auf irgend etwas mehr als bloße Erscheinungen gehen können und nur das vorstellen, was entweder Erfahrung überhaupt möglich macht, oder was, indem es aus diesen Principien abgeleitet ist, jederzeit in irgend einer möglichen Erfahrung muß vorgestellt werden können. ### Section 31 § 31. Und so hat man denn einmal etwas Bestimmtes, und woran man sich bei allen metaphysischen Unternehmungen, die bisher kühn gnug, aber jederzeit blind über alles ohne Unterschied gegangen sind, halten kann. Dogmatische Denker haben sich es niemals einfallen lassen, daß das Ziel ihrer Bemühungen so kurz sollte ausgesteckt werden, und selbst diejenige nicht, die, trotzig auf ihre vermeinte gesunde Vernunft, mit zwar rechtmäßigen und natürlichen, aber zum bloßen Erfahrungsgebrauch bestimmten Begriffen und Grundsätzen der reinen Vernunft auf Einsichten ausgingen, für die sie keine bestimmte Grenzen kannten, noch kennen konnten, weil sie über die Natur und selbst die Möglichkeit eines solchen reinen Verstandes niemals entweder nachgedacht hatten oder nachzudenken vermochten. Mancher Naturalist der reinen Vernunft (darunter ich den verstehe, welcher sich zutraut, ohne alle Wissenschaft in Sachen der Metaphysik zu entscheiden) möchte wohl vorgeben, er habe das, was hier mit so viel Zurüstung, oder, wenn er lieber will, mit weitschweifigem pedantischen Pompe vorgetragen worden, schon längst durch den Wahrsagergeist seiner gesunden Vernunft nicht blos vermuthet, sondern auch gewußt und eingesehen: „daß wir nämlich mit aller unserer Vernunft über das Feld der Erfahrungen nie hinaus kommen können“. Allein da er doch, wenn man ihm seine Vernunftprincipien allmählig abfrägt, gestehen muß, daß darunter viele sind, die er nicht aus Erfahrung geschöpft hat, die also von dieser unabhängig und a priori gültig sind, wie und mit welchen Gründen will er denn den Dogmatiker und sich selbst in Schranken halten, der sich dieser Begriffe und Grundsätze über alle mögliche Erfahrung hinaus bedient, darum eben weil sie unabhängig von dieser erkannt werden. Und selbst er, dieser Adept der gesunden Vernunft, ist so sicher nicht, ungeachtet aller seiner angemaßten, wohlfeil erworbenen Weisheit, unvermerkt über Gegenstände der Erfahrung hinaus in das Feld der Hirngespinste zu gerathen. Auch ist er gemeiniglich tief gnug drin verwickelt, ob er zwar durch die populäre Sprache, da er alles blos für Wahrscheinlichkeit, vernünftige Vermuthung oder Analogie ausgiebt, seinen grundlosen Ansprüchen einigen Anstrich giebt. ### Section 32 § 32. Schon von den ältesten Zeiten der Philosophie her haben sich Forscher der reinen Vernunft außer den Sinnenwesen oder Erscheinungen (phaenomena), die die Sinnenwelt ausmachen, noch besondere Verstandeswesen (noumena), welche eine Verstandeswelt ausmachen sollten, gedacht, und da sie (welches einem noch unausgebildeten Zeitalter wohl zu verzeihen war) Erscheinung und Schein für einerlei hielten, den Verstandeswesen allein Wirklichkeit zugestanden. In der That, wenn wir die Gegenstände der Sinne wie billig als bloße Erscheinungen ansehen, so gestehen wir hiedurch doch zugleich, daß ihnen ein Ding an sich selbst zum Grunde liege, ob wir dasselbe gleich nicht, wie es an sich beschaffen sei, sondern nur seine Erscheinung, d. i. die Art, wie unsre Sinnen von diesem unbekannten Etwas afficirt werden, kennen. Der Verstand also, eben dadurch daß er Erscheinungen annimmt, gesteht auch das Dasein von Dingen an sich selbst zu, und so fern können wir sagen, daß die Vorstellung solcher Wesen, die den Erscheinungen zum Grunde liegen, mithin bloßer Verstandeswesen nicht allein zulässig, sondern auch unvermeidlich sei. Unsere kritische Deduction schließt dergleichen Dinge (noumena) auch keinesweges aus, sondern schränkt vielmehr die Grundsätze der Ästhetik dahin ein, daß sie sich ja nicht auf alle Dinge erstrecken sollen, wodurch alles in bloße Erscheinung verwandelt werden würde, sondern daß sie nur von Gegenständen einer möglichen Erfahrung gelten sollen. Also werden hiedurch Verstandeswesen zugelassen, nur mit Einschärfung dieser Regel, die gar keine Ausnahme leidet: daß wir von diesen reinen Verstandeswesen ganz und gar nichts Bestimmtes wissen, noch wissen können, weil unsere reine Verstandesbegriffe sowohl als reine Anschauungen auf nichts als Gegenstände möglicher Erfahrung, mithin auf bloße Sinnenwesen gehen, und, so bald man von diesen abgeht, jenen Begriffen nicht die mindeste Bedeutung mehr übrig bleibt. ### Section 33 § 33. Es ist in der That mit unseren reinen Verstandesbegriffen etwas Verfängliches in Ansehung der Anlockung zu einem transscendenten Gebrauch; denn so nenne ich denjenigen, der über alle mögliche Erfahrung hinausgeht. Nicht allein daß unsere Begriffe der Substanz, der Kraft, der Handlung, der Realität ꝛc. ganz von der Erfahrung unabhängig sind, imgleichen gar keine Erscheinung der Sinne enthalten, also in der That auf Dinge an sich selbst (noumena) zu gehen scheinen, sondern was diese Vermuthung noch bestärkt, sie enthalten eine Nothwendigkeit der Bestimmung in sich, der die Erfahrung niemals gleich kommt. Der Begriff der Ursache enthält eine Regel, nach der aus einem Zustande ein anderer nothwendiger Weise folgt; aber die Erfahrung kann uns nur zeigen, daß oft, und wenn es hoch kommt, gemeiniglich auf einen Zustand der Dinge ein anderer folge, und kann also weder strenge Allgemeinheit, noch Nothwendigkeit verschaffen ꝛc. Daher scheinen Verstandesbegriffe viel mehr Bedeutung und Inhalt zu haben, als daß der bloße Erfahrungsgebrauch ihre ganze Bestimmung erschöpfte, und so baut sich der Verstand unvermerkt an das Haus der Erfahrung noch ein viel weitläuftigeres Nebengebäude an, welches er mit lauter Gedankenwesen anfüllt, ohne es einmal zu merken, daß er sich mit seinen sonst richtigen Begriffen über die Grenzen ihres Gebrauchs verstiegen habe. ### Section 34 § 34. Es waren also zwei wichtige, ja ganz unentbehrliche, obzwar äußerst trockene Untersuchungen nöthig, welche Krit. Seite 137 ꝛc. und 235 ꝛc. angestellt worden, durch deren erstere gezeigt wurde, daß die Sinne nicht die reine Verstandesbegriffe in concreto, sondern nur das Schema zum Gebrauche derselben an die Hand geben, und der ihm gemäße Gegenstand nur in der Erfahrung (als dem Producte des Verstandes aus Materialien der Sinnlichkeit) angetroffen werde. In der zweiten Untersuchung (Krit. S. 235) wird gezeigt: daß ungeachtet der Unabhängigkeit unsrer reinen Verstandesbegriffe und Grundsätze von Erfahrung, ja selbst ihrem scheinbarlich größeren Umfange des Gebrauchs dennoch durch dieselbe außer dem Felde der Erfahrung gar nichts gedacht werden könne, weil sie nichts thun können, als blos die logische Form des Urtheils in Ansehung gegebener Anschauungen bestimmen; da es aber über das Feld der Sinnlichkeit hinaus ganz und gar keine Anschauung giebt, jenen reinen Begriffen es ganz und gar an Bedeutung fehle, indem sie durch kein Mittel in concreto können dargestellt werden, folglich alle solche Noumena zusammt dem Inbegriff derselben, einer intelligibeln[^316-1] Welt, nichts als Vorstellungen einer Aufgabe sind, deren Gegenstand an sich wohl möglich, deren Auflösung aber nach der Natur unseres Verstandes gänzlich unmöglich ist, indem unser Verstand kein Vermögen der Anschauung, sondern blos der Verknüpfung gegebener Anschauungen in einer Erfahrung ist, und daß diese daher alle Gegenstände für unsere Begriffe enthalten müsse, außer ihr aber alle Begriffe, da ihnen keine Anschauung unterlegt werden kann, ohne Bedeutung sein werden. [^316-1]: Nicht (wie man sich gemeiniglich ausdrückt) *intellectuellen* Welt. Denn *intellectuell* sind die *Erkenntnisse* durch den Verstand, und dergleichen gehen auch auf unsere Sinnenwelt; *intelligibel* aber heißen *Gegenstände*, so fern sie *blos durch den Verstand* vorgestellt werden können und auf die keine unserer sinnlichen Anschauungen gehen kann. Da aber doch jedem Gegenstande irgend eine mögliche Anschauung entsprechen muß, so würde man sich einen Verstand denken müssen, der unmittelbar Dinge anschauete; von einem solchen aber haben wir nicht den mindesten Begriff, mithin auch nicht von den *Verstandeswesen*, auf die er gehen soll. ### Section 35 § 35. Es kann der Einbildungskraft vielleicht verziehen werden, wenn sie bisweilen schwärmt, d. i. sich nicht behutsam innerhalb den Schranken der Erfahrung hält; denn wenigstens wird sie durch einen solchen freien Schwung belebt und gestärkt, und es wird immer leichter sein, ihre Kühnheit zu mäßigen, als ihrer Mattigkeit aufzuhelfen. Daß aber der Verstand, der *denken* soll, an dessen statt *schwärmt*, das kann ihm niemals verziehen werden; denn auf ihm beruht allein alle Hülfe, um der Schwärmerei der Einbildungskraft, wo es nöthig ist, Grenzen zu setzen. Er fängt es aber hiemit sehr unschuldig und sittsam an. Zuerst bringt er die Elementarerkenntnisse, die ihm vor aller Erfahrung beiwohnen, aber dennoch in der Erfahrung immer ihre Anwendung haben müssen, ins Reine. Allmählig läßt er diese Schranken weg, und was sollte ihn auch daran hindern, da der Verstand ganz frei seine Grundsätze aus sich selbst genommen hat? Und nun geht es zuerst auf neu erdachte Kräfte in der Natur, bald hernach auf Wesen außerhalb der Natur, mit einem Wort auf eine Welt, zu deren Einrichtung es uns an Bauzeug nicht fehlen kann, weil es durch fruchtbare Erdichtung reichlich herbeigeschafft und durch Erfahrung zwar nicht bestätigt, aber auch niemals widerlegt wird. Das ist auch die Ursache, weswegen junge Denker Metaphysik in ächter dogmatischer Manier so lieben und ihr oft ihre Zeit und ihr sonst brauchbares Talent aufopfern. Es kann aber gar nichts helfen, jene fruchtlose Versuche der reinen Vernunft durch allerlei Erinnerungen wegen der Schwierigkeit der Auflösung so tief verborgener Fragen, Klagen über die Schranken unserer Vernunft und Herabsetzung der Behauptungen auf bloße Muthmaßungen mäßigen zu wollen. Denn wenn die *Unmöglichkeit* derselben nicht deutlich dargethan worden, und die *Selbsterkenntniß* der Vernunft nicht wahre Wissenschaft wird, worin das Feld ihres richtigen von dem ihres nichtigen und fruchtlosen Gebrauchs, so zu sagen, mit geometrischer Gewißheit unterschieden wird, so werden jene eitle Bestrebungen niemals völlig abgestellt werden. ### Section 36 § 36. *Wie ist Natur selbst möglich?* Diese Frage, welche der höchste Punkt ist, den transscendentale Philosophie nur immer berühren mag, und zu welchem sie auch als ihrer Grenze und Vollendung geführt werden muß, enthält eigentlich zwei Fragen. **Erstlich:** Wie ist Natur in *materieller* Bedeutung, nämlich der Anschauung nach, als der Inbegriff der Erscheinungen; wie ist Raum, Zeit und das, was beide erfüllt, der Gegenstand der Empfindung, überhaupt möglich? Die Antwort ist: vermittelst der Beschaffenheit unserer Sinnlichkeit, nach welcher sie auf die ihr eigenthümliche Art von Gegenständen, die ihr an sich selbst unbekannt und von jenen Erscheinungen ganz unterschieden sind, gerührt wird. Die Beantwortung ist in dem Buche selbst in der transscendentalen Ästhetik, hier aber in den Prolegomenen durch die Auflösung der ersten Hauptfrage gegeben worden. **Zweitens:** Wie ist Natur in formeller Bedeutung, als der Inbegriff der Regeln, unter denen alle Erscheinungen stehen müssen, wenn sie in einer Erfahrung als verknüpft gedacht werden sollen, möglich? Die Antwort kann nicht anders ausfallen als: sie ist nur möglich vermittelst der Beschaffenheit unseres Verstandes, nach welcher alle jene Vorstellungen der Sinnlichkeit auf ein Bewußtsein nothwendig bezogen werden, und wodurch allererst die eigenthümliche Art unseres Denkens, nämlich durch Regeln, und vermittelst dieser die Erfahrung, welche von der Einsicht der Objecte an sich selbst ganz zu unterscheiden ist, möglich ist. Diese Beantwortung ist in dem Buche selbst in der transscendentalen Logik, hier aber in den Prolegomenen in dem Verlauf der Auflösung der zweiten Hauptfrage gegeben worden. Wie aber diese eigenthümliche Eigenschaft unsrer Sinnlichkeit selbst, oder die unseres Verstandes und der ihm und allem Denken zum Grunde liegenden nothwendigen Apperception möglich sei, läßt sich nicht weiter auflösen und beantworten, weil wir ihrer zu aller Beantwortung und zu allem Denken der Gegenstände immer wieder nöthig haben. Es sind viele Gesetze der Natur, die wir nur vermittelst der Erfahrung wissen können; aber die Gesetzmäßigkeit in Verknüpfung der Erscheinungen, d. i. die Natur überhaupt, können wir durch keine Erfahrung kennen lernen, weil Erfahrung selbst solcher Gesetze bedarf, die ihrer Möglichkeit a priori zum Grunde liegen. Die Möglichkeit der Erfahrung überhaupt ist also zugleich das allgemeine Gesetz der Natur, und die Grundsätze der erstern sind selbst die Gesetze der letztern. Denn wir kennen Natur nicht anders als den Inbegriff der Erscheinungen, d. i. der Vorstellungen in uns, und können daher das Gesetz ihrer Verknüpfung nirgend anders als von den Grundsätzen der Verknüpfung derselben in uns, d. i. den Bedingungen der nothwendigen Vereinigung in einem Bewußtsein, welche die Möglichkeit der Erfahrung ausmacht, hernehmen. Selbst der Hauptsatz, der durch diesen ganzen Abschnitt ausgeführt worden, daß allgemeine Naturgesetze a priori erkannt werden können, führt schon von selbst auf den Satz: daß die oberste Gesetzgebung der Natur in uns selbst, d. i. in unserm Verstande, liegen müsse, und daß wir die allgemeinen Gesetze derselben nicht von der Natur vermittelst der Erfahrung, sondern umgekehrt die Natur ihrer allgemeinen Gesetzmäßigkeit nach blos aus den in unserer Sinnlichkeit und dem Verstande liegenden Bedingungen der Möglichkeit der Erfahrung suchen müssen; denn wie wäre es sonst möglich, diese Gesetze, da sie nicht etwa Regeln der analytischen Erkenntniß, sondern wahrhafte synthetische Erweiterungen derselben sind, a priori zu kennen? Eine solche und zwar nothwendige Übereinstimmung der Principien möglicher Erfahrung mit den Gesetzen der Möglichkeit der Natur kann nur aus zweierlei Ursachen stattfinden: entweder diese Gesetze werden von der Natur vermittelst der Erfahrung entlehnt, oder umgekehrt, die Natur wird von den Gesetzen der Möglichkeit der Erfahrung überhaupt abgeleitet und ist mit der bloßen allgemeinen Gesetzmäßigkeit der letzteren völlig einerlei. Das erstere widerspricht sich selbst, denn die allgemeinen Naturgesetze können und müssen a priori (d. i. unabhängig von aller Erfahrung) erkannt und allem empirischen Gebrauche des Verstandes zum Grunde gelegt werden; also bleibt nur das zweite übrig.[^319-1] Wir müssen aber empirische Gesetze der Natur, die jederzeit besondere Wahrnehmungen voraussetzen, von den reinen oder allgemeinen Naturgesetzen, welche, ohne daß besondere Wahrnehmungen zum Grunde liegen, blos die Bedingungen ihrer nothwendigen Vereinigung in einer Erfahrung enthalten, unterscheiden; und in Ansehung der letztern ist Natur und mögliche Erfahrung ganz und gar einerlei; und da in dieser die Gesetzmäßigkeit auf der nothwendigen Verknüpfung der Erscheinungen in einer Erfahrung (ohne welche wir ganz und gar keinen Gegenstand der Sinnenwelt erkennen können), mithin auf den ursprünglichen Gesetzen des Verstandes beruht, so klingt es zwar anfangs befremdlich, ist aber nichts desto weniger gewiß, wenn ich in Ansehung der letztern sage: *der Verstand schöpft seine Gesetze (a priori) nicht aus der Natur, sondern schreibt sie dieser vor.* [^319-1]: *Crusius* allein wußte einen Mittelweg: daß nämlich ein Geist, der nicht irren noch betrügen kann, uns diese Naturgesetze ursprünglich eingepflanzt habe. Allein da sich doch oft auch trügliche Grundsätze einmischen, wovon das System dieses Mannes selbst nicht wenig Beispiele giebt, so sieht es bei dem Mangel sicherer Kriterien, den ächten Ursprung von dem unächten zu unterscheiden, mit dem Gebrauche eines solchen Grundsatzes sehr mißlich aus, indem man niemals sicher wissen kann, was der Geist der Wahrheit oder der Vater der Lügen uns eingeflößt haben möge. ### Section 37 § 37. Wir wollen diesen dem Anscheine nach gewagten Satz durch ein Beispiel erläutern, welches zeigen soll: daß Gesetze, die wir an Gegenständen der sinnlichen Anschauung entdecken, vornehmlich wenn sie als nothwendig erkannt worden, von uns selbst schon für solche gehalten werden, die der Verstand hinein gelegt, ob sie gleich den Naturgesetzen, die wir der Erfahrung zuschreiben, sonst in allen Stücken ähnlich sind. ### Section 38 § 38. Wenn man die Eigenschaften des Cirkels betrachtet, dadurch diese Figur so manche willkürliche Bestimmungen des Raums in ihr sofort in einer allgemeinen Regel vereinigt, so kann man nicht umhin, diesem geometrischen Dinge eine Natur beizulegen. So theilen sich nämlich zwei Linien, die sich einander und zugleich den Cirkel schneiden, nach welchem Ungefähr sie auch gezogen werden, doch jederzeit so regelmäßig, daß das Rectangel aus den Stücken einer jeden Linie dem der andern gleich ist. Nun frage ich: „Liegt dieses Gesetz im Cirkel, oder liegt es im Verstande?“ d. i. enthält diese Figur unabhängig vom Verstande den Grund dieses Gesetzes in sich, oder legt der Verstand, indem er nach seinen Begriffen (nämlich der Gleichheit der Halbmesser) die Figur selbst construirt hat, zugleich das Gesetz der einander in geometrischer Proportion schneidenden Sehnen in dieselbe hinein? Man wird bald gewahr, wenn man den Beweisen dieses Gesetzes nachgeht, daß es allein von der Bedingung, die der Verstand der Construction dieser Figur zum Grunde legte, nämlich der Gleichheit der Halbmesser, könne abgeleitet werden. Erweitern wir diesen Begriff nun, die Einheit mannigfaltiger Eigenschaften geometrischer Figuren unter gemeinschaftlichen Gesetzen noch weiter zu verfolgen, und betrachten den Cirkel als einen Kegelschnitt, der also mit andern Kegelschnitten unter eben denselben Grundbedingungen der Construction steht, so finden wir, daß alle Sehnen, die sich innerhalb der letztern, der Ellipse, der Parabel und Hyperbel, schneiden, es jederzeit so thun, daß die Rectangel aus ihren Theilen zwar nicht gleich sind, aber doch immer in gleichen Verhältnissen gegen einander stehen. Gehen wir von da noch weiter, nämlich zu den Grundlehren der physischen Astronomie, so zeigt sich ein über die ganze materielle Natur verbreitetes physisches Gesetz der wechselseitigen Attraction, deren Regel ist, daß sie umgekehrt mit dem Quadrat der Entfernungen von jedem anziehenden Punkt eben so abnimmt, wie die Kugelflächen, in die sich diese Kraft verbreitet, zunehmen, welches als nothwendig in der Natur der Dinge selbst zu liegen scheint und daher auch als a priori erkennbar vorgetragen zu werden pflegt. So einfach nun auch die Quellen dieses Gesetzes sind, indem sie blos auf dem Verhältnisse der Kugelflächen von verschiedenen Halbmessern beruhen, so ist doch die Folge davon so vortrefflich in Ansehung der Mannigfaltigkeit ihrer Zusammenstimmung und Regelmäßigkeit derselben, daß nicht allein alle mögliche Bahnen der Himmelskörper in Kegelschnitten, sondern auch ein solches Verhältniß derselben unter einander erfolgt, daß kein ander Gesetz der Attraction als das des umgekehrten Quadratverhältnisses der Entfernungen zu einem Weltsystem als schicklich erdacht werden kann. Hier ist also Natur, die auf Gesetzen beruht, welche der Verstand a priori erkennt und zwar vornehmlich aus allgemeinen Principien der Bestimmung des Raums. Nun frage ich: liegen diese Naturgesetze im Raume, und lernt sie der Verstand, indem er den reichhaltigen Sinn, der in jenem liegt, blos zu erforschen sucht, oder liegen sie im Verstande und in der Art, wie dieser den Raum nach den Bedingungen der synthetischen Einheit, darauf seine Begriffe insgesammt auslaufen, bestimmt? Der Raum ist etwas so Gleichförmiges und in Ansehung aller besondern Eigenschaften so Unbestimmtes, daß man in ihm gewiß keinen Schatz von Naturgesetzen suchen wird. Dagegen ist das, was den Raum zur Cirkelgestalt, der Figur des Kegels und der Kugel bestimmt, der Verstand, so fern er den Grund der Einheit der Construction derselben enthält. Die bloße allgemeine Form der Anschauung, die Raum heißt, ist also wohl das Substratum aller auf besondere Objecte bestimmbaren Anschauungen, und in jenem liegt freilich die Bedingung der Möglichkeit und Mannigfaltigkeit der letztern; aber die Einheit der Objecte wird doch lediglich durch den Verstand bestimmt und zwar nach Bedingungen, die in seiner eigenen Natur liegen; und so ist der Verstand der Ursprung der allgemeinen Ordnung der Natur, indem er alle Erscheinungen unter seine eigene Gesetze faßt und dadurch allererst Erfahrung (ihrer Form nach) a priori zu Stande bringt, vermöge deren alles, was nur durch Erfahrung erkannt werden soll, seinen Gesetzen nothwendig unterworfen wird. Denn wir haben es nicht mit der Natur der *Dinge an sich selbst* zu thun, die ist sowohl von Bedingungen unserer Sinnlichkeit als des Verstandes unabhängig, sondern mit der Natur als einem Gegenstande möglicher Erfahrung, und da macht es der Verstand, indem er diese möglich macht, zugleich, daß Sinnenwelt entweder gar kein Gegenstand der Erfahrung oder eine Natur ist. ### Section 39 § 39. Anhang *zur reinen Naturwissenschaft.* *Von dem System der Kategorien.* Es kann einem Philosophen nichts erwünschter sein, als wenn er das Mannigfaltige der Begriffe oder Grundsätze, die sich ihm vorher durch den Gebrauch, den er von ihnen in concreto gemacht hatte, zerstreut dargestellt hatten, aus einem Princip a priori ableiten und alles auf solche Weise in eine Erkenntniß vereinigen kann. Vorher glaubte er nur, daß, was ihm nach einer gewissen Abstraction übrig blieb und durch Vergleichung unter einander eine besondere Art von Erkenntnissen auszumachen schien, vollständig gesammlet sei, aber es war nur ein *Aggregat*; jetzt weiß er, daß gerade nur so viel, nicht mehr, nicht weniger, die Erkenntnißart ausmachen könne, und sah die Nothwendigkeit seiner Eintheilung ein, welches ein Begreifen ist, und nun hat er allererst ein *System*. Aus dem gemeinen Erkenntnisse die Begriffe heraussuchen, welche gar keine besondere Erfahrung zum Grunde liegen haben und gleichwohl in aller Erfahrungserkenntniß vorkommen, von der sie gleichsam die bloße Form der Verknüpfung ausmachen, setzte kein größeres Nachdenken oder mehr Einsicht voraus, als aus einer Sprache Regeln des wirklichen Gebrauchs der Wörter überhaupt heraussuchen und so Elemente zu einer Grammatik zusammentragen (in der That sind beide Untersuchungen einander auch sehr nahe verwandt), ohne doch eben Grund angeben zu können, warum eine jede Sprache gerade diese und keine andere formale Beschaffenheit habe, noch weniger aber, daß gerade so viel, nicht mehr noch weniger, solcher formalen Bestimmungen derselben überhaupt angetroffen werden können. *Aristoteles* hatte zehn solcher reinen Elementarbegriffe unter dem Namen der Kategorien[^323-1] zusammengetragen. Diesen, welche auch Prädicamente genannt wurden, sah er sich hernach genöthigt, noch fünf Postprädicamente beizufügen[^323-2], die doch zum Theil schon in jenen liegen (als prius, simul, motus); allein diese Rhapsodie konnte mehr für einen Wink für den künftigen Nachforscher, als für eine regelmäßig ausgeführte Idee gelten und Beifall verdienen, daher sie auch bei mehrerer Aufklärung der Philosophie als ganz unnütz verworfen worden. Bei einer Untersuchung der reinen (nichts Empirisches enthaltenden) Elemente der menschlichen Erkenntniß gelang es mir allererst nach langem Nachdenken, die reinen Elementarbegriffe der Sinnlichkeit (Raum und Zeit) von denen des Verstandes mit Zuverlässigkeit zu unterscheiden und abzusondern. Dadurch wurden nun aus jenem Register die 7te, 8te, 9te Kategorie ausgeschlossen. Die übrigen konnten mir zu nichts nutzen, weil kein Princip vorhanden war, nach welchem der Verstand völlig ausgemessen und alle Functionen desselben, daraus seine reine Begriffe entspringen, vollzählig und mit Präcision bestimmt werden könnten. Um aber ein solches Princip auszufinden, sah ich mich nach einer Verstandeshandlung um, die alle übrige enthält und sich nur durch verschiedene Modificationen oder Momente unterscheidet, das Mannigfaltige der Vorstellung unter die Einheit des Denkens überhaupt zu bringen, und da fand ich, diese Verstandeshandlung bestehe im Urtheilen. Hier lag nun schon fertige, obgleich noch nicht ganz von Mängeln freie Arbeit der Logiker vor mir, dadurch ich in den Stand gesetzt wurde, eine vollständige Tafel reiner Verstandesfunctionen, die aber in Ansehung alles Objects unbestimmt waren, darzustellen. Ich bezog endlich diese Functionen zu urtheilen auf Objecte überhaupt, oder vielmehr auf die Bedingung, Urtheile als objectiv-gültig zu bestimmen, und es entsprangen reine Verstandesbegriffe, bei denen ich außer Zweifel sein konnte, daß gerade nur diese und ihrer nur so viel, nicht mehr noch weniger, unser ganzes Erkenntniß der Dinge aus bloßem Verstande ausmachen können. Ich nannte sie wie billig nach ihrem alten Namen *Kategorien*, wobei ich mir vorbehielt, alle von diesen abzuleitende Begriffe, es sei durch Verknüpfung unter einander, oder mit der reinen Form der Erscheinung (Raum und Zeit), oder mit ihrer Materie, so fern sie noch nicht empirisch bestimmt ist (Gegenstand der Empfindung überhaupt), unter der Benennung der *Prädicabilien* vollständig hinzuzufügen, so bald ein System der transscendentalen Philosophie, zu deren Behuf ich es jetzt nur mit der Kritik der Vernunft selbst zu thun hatte, zu Stande kommen sollte. Das Wesentliche aber in diesem System der Kategorien, dadurch es sich von jener alten Rhapsodie, die ohne alles Princip fortging, unterscheidet, und warum es auch allein zur Philosophie gezählt zu werden verdient, besteht darin: daß vermittelst desselben die wahre Bedeutung der reinen Verstandesbegriffe und die Bedingung ihres Gebrauchs genau bestimmt werden konnte. Denn da zeigte sich, daß sie für sich selbst nichts als logische Functionen sind, als solche aber nicht den mindesten Begriff von einem Objecte an sich selbst ausmachen, sondern es bedürfen, daß sinnliche Anschauung zum Grunde liege, und alsdann nur dazu dienen, empirische Urtheile, die sonst in Ansehung aller Functionen zu urtheilen unbestimmt und gleichgültig sind, in Ansehung derselben zu bestimmen, ihnen dadurch Allgemeingültigkeit zu verschaffen und vermittelst ihrer *Erfahrungsurtheile* überhaupt möglich zu machen. Von einer solchen Einsicht in die Natur der Kategorien, die sie zugleich auf den bloßen Erfahrungsgebrauch einschränkte, ließ sich weder ihr erster Urheber, noch irgend einer nach ihm etwas einfallen; aber ohne diese Einsicht (die ganz genau von der Ableitung oder Deduction derselben abhängt) sind sie gänzlich unnütz und ein elendes Namenregister ohne Erklärung und Regel ihres Gebrauchs. Wäre dergleichen jemals den Alten in den Sinn gekommen, ohne Zweifel das ganze Studium der reinen Vernunfterkenntniß, welches unter dem Namen Metaphysik viele Jahrhunderte hindurch so manchen guten Kopf verdorben hat, wäre in ganz anderer Gestalt zu uns gekommen und hätte den Verstand der Menschen aufgeklärt, anstatt ihn, wie wirklich geschehen ist, in düstern und vergeblichen Grübeleien zu erschöpfen und für wahre Wissenschaft unbrauchbar zu machen. Dieses System der Kategorien macht nun alle Behandlung eines jeden Gegenstandes der reinen Vernunft selbst wiederum systematisch und giebt eine ungezweifelte Anweisung oder Leitfaden ab, wie und durch welche Punkte der Untersuchung jede metaphysische Betrachtung, wenn sie vollständig werden soll, müsse geführt werden: denn es erschöpft alle Momente des Verstandes, unter welche jeder andere Begriff gebracht werden muß. So ist auch die Tafel der Grundsätze entstanden, von deren Vollständigkeit man nur durch das System der Kategorien gewiß sein kann; und selbst in der Eintheilung der Begriffe, welche über den physiologischen Verstandesgebrauch hinausgehen sollen, (Kritik S. 344, imgleichen S. 415) ist es immer derselbe Leitfaden, der, weil er immer durch dieselbe feste, im menschlichen Verstande a priori bestimmte Punkte geführt werden muß, jederzeit einen geschlossenen Kreis bildet, der keinen Zweifel übrig läßt, daß der Gegenstand eines reinen Verstandes- oder Vernunftbegriffs, so fern er philosophisch und nach Grundsätzen a priori erwogen werden soll, auf solche Weise vollständig erkannt werden könne. Ich habe sogar nicht unterlassen können, von dieser Leitung in Ansehung einer der abstractesten ontologischen Eintheilungen, nämlich der mannigfaltigen Unterscheidung der Begriffe von Etwas und Nichts, Gebrauch zu machen und darnach eine regelmäßige und nothwendige Tafel (Kritik S. 292) zu Stande zu bringen[^325-1]. Eben dieses System zeigt seinen nicht gnug anzupreisenden Gebrauch, so wie jedes auf ein allgemeines Princip gegründete wahre System auch darin, daß es alle fremdartige Begriffe, die sich sonst zwischen jene reine Verstandesbegriffe einschleichen möchten, ausstößt und jedem Erkenntniß seine Stelle bestimmt. Diejenige Begriffe, welche ich unter dem Namen der *Reflexionsbegriffe* gleichfalls nach dem Leitfaden der Kategorien in eine Tafel gebracht hatte, mengen sich in der Ontologie ohne Vergünstigung und rechtmäßige Ansprüche unter die reinen Verstandesbegriffe, obgleich diese Begriffe der Verknüpfung und dadurch des Objects selbst, jene aber nur der bloßen Vergleichung schon gegebener Begriffe sind und daher eine ganz andere Natur und Gebrauch haben; durch meine gesetzmäßige Eintheilung (Kritik S. 260) werden sie aus diesem Gemenge geschieden. Noch viel heller aber leuchtet der Nutzen jener abgesonderten Tafel der Kategorien in die Augen, wenn wir, wie es gleich jetzt geschehen wird, die Tafel transscendentaler Vernunftbegriffe, die von ganz anderer Natur und Ursprung sind, als jene Verstandesbegriffe (daher auch eine andre Form haben muß), von jenen trennen, welche so nothwendige Absonderung doch niemals in irgend einem System der Metaphysik geschehen ist, wo daher jene Vernunftideen mit Verstandesbegriffen, als gehörten sie wie Geschwister zu einer Familie, ohne Unterschied durch einander laufen, welche Vermengung in Ermangelung eines besondern Systems der Kategorien auch niemals vermieden werden konnte. [^323-1]: 1. Substantia. 2. Qualitas. 3. Quantitas. 4. Relatio. 5. Actio. 6. Passio. 7. Quando. 8. Ubi. 9. Situs. 10. Habitus. [^323-2]: Oppositum, Prius, Simul, Motus, Habere. [^325-1]: Über eine vorgelegte Tafel der Kategorien lassen sich allerlei artige Anmerkungen machen, als 1) daß die dritte aus der ersten und zweiten in einen Begriff verbunden entspringe, 2) daß in denen von der Größe und Qualität blos ein Fortschritt von der Einheit zur Allheit, oder von dem Etwas zum Nichts (zu diesem Behuf müssen die Kategorien der Qualität so stehen: Realität, Einschränkung, völlige Negation) fortgehen, ohne correlata oder opposita, dagegen die der Relation und Modalität diese letztere bei sich führen, 3) daß so wie im *Logischen* kategorische Urtheile allen andern zum Grunde liegen, so die Kategorie der Substanz allen Begriffen von wirklichen Dingen, 4) daß, so wie die Modalität im Urtheile kein besonderes Prädicat ist, so auch die Modalbegriffe keine Bestimmung zu Dingen hinzuthun, u. s. w.; dergleichen Betrachtungen alle ihren großen Nutzen haben. Zählt man überdem alle *Prädicabilien* auf, die man ziemlich vollständig aus jeder guten Ontologie (z. E. *Baumgartens*) ziehen kann, und ordnet sie classenweise unter die Kategorien, wobei man nicht versäumen muß, eine so vollständige Vergliederung aller dieser Begriffe als möglich hinzuzufügen, so wird ein blos analytischer Theil der Metaphysik entspringen, der noch gar keinen synthetischen Satz enthält und vor dem zweiten (dem synthetischen) vorhergehen könnte und durch seine Bestimmtheit und Vollständigkeit nicht allein Nutzen, sondern vermöge des Systematischen in ihm noch überdem eine gewisse Schönheit enthalten würde. ### Third part: How is metaphysics in general possible? *Der transscendentalen Hauptfrage* **Dritter Theil.** *Wie ist Metaphysik überhaupt möglich?* ### Section 40 § 40. Reine Mathematik und reine Naturwissenschaft hätten zum *Behuf ihrer eigenen Sicherheit* und Gewißheit keiner dergleichen Deduction bedurft, als wir bisher von beiden zu Stande gebracht haben; denn die erstere stützt sich auf ihre eigene Evidenz, die zweite aber, obgleich aus reinen Quellen des Verstandes entsprungen, dennoch auf Erfahrung und deren durchgängige Bestätigung; welcher letztern Zeugniß sie darum nicht gänzlich ausschlagen und entbehren kann, weil sie mit aller ihrer Gewißheit dennoch, als Philosophie, es der Mathematik niemals gleich thun kann. Beide Wissenschaften hatten also die gedachte Untersuchung nicht für sich, sondern für eine andere Wissenschaft, nämlich Metaphysik, nöthig. Metaphysik hat es außer mit Naturbegriffen, die in der Erfahrung jederzeit ihre Anwendung finden, noch mit reinen Vernunftbegriffen zu thun, die niemals in irgend einer nur immer möglichen Erfahrung gegeben werden, mithin mit Begriffen, deren objective Realität (daß sie nicht bloße Hirngespinste sind), und mit Behauptungen, deren Wahrheit oder Falschheit durch keine Erfahrung bestätigt oder aufgedeckt werden kann; und dieser Theil der Metaphysik ist überdem gerade derjenige, welcher den wesentlichen Zweck derselben, wozu alles andre nur Mittel ist, ausmacht, und so bedarf diese Wissenschaft einer solchen Deduction um *ihrer selbst willen*. Die uns jetzt vorgelegte dritte Frage betrifft also gleichsam den Kern und das Eigenthümliche der Metaphysik, nämlich die Beschäftigung der Vernunft blos mit sich selbst und, indem sie über ihre eigene Begriffe brütet, die unmittelbar daraus vermeintlich entspringende Bekanntschaft mit Objecten, ohne dazu der Vermittelung der Erfahrung nöthig zu haben, noch überhaupt durch dieselbe dazu gelangen zu können[^327-1]. Ohne Auflösung dieser Frage thut sich Vernunft niemals selbst gnug. Der Erfahrungsgebrauch, auf welchen die Vernunft den reinen Verstand einschränkt, erfüllt nicht ihre eigene ganze Bestimmung. Jede einzelne Erfahrung ist nur ein Theil von der ganzen Sphäre ihres Gebietes, das *absolute Ganze aller möglichen Erfahrung* ist aber selbst keine Erfahrung und dennoch ein nothwendiges Problem für die Vernunft, zu dessen bloßer Vorstellung sie ganz anderer Begriffe nöthig hat, als jener reinen Verstandesbegriffe, deren Gebrauch nur immanent ist, d. i. auf Erfahrung geht, so weit sie gegeben werden kann, indessen daß Vernunftbegriffe auf die Vollständigkeit, d. i. die collective Einheit der ganzen möglichen Erfahrung, und dadurch über jede gegebne Erfahrung hinausgehen und *transscendent* werden. So wie also der Verstand der Kategorien zur Erfahrung bedurfte, so enthält die Vernunft in sich den Grund zu Ideen, worunter ich nothwendige Begriffe verstehe, deren Gegenstand gleichwohl in keiner Erfahrung gegeben werden *kann*. Die letztern sind eben sowohl in der Natur der Vernunft, als die erstere in der Natur des Verstandes gelegen, und wenn jene einen Schein bei sich führen, der leicht verleiten kann, so ist dieser Schein unvermeidlich, obzwar, „daß er nicht verführe,“ gar wohl verhütet werden kann. Da aller Schein darin besteht, daß der subjective Grund des Urtheils für objectiv gehalten wird, so wird ein Selbsterkenntniß der reinen Vernunft in ihrem transscendenten (überschwenglichen) Gebrauch das einzige Verwahrungsmittel gegen die Verirrungen sein, in welche die Vernunft geräth, wenn sie ihre Bestimmung mißdeutet und dasjenige transscendenter Weise aufs Object an sich selbst bezieht, was nur ihr eigenes Subject und die Leitung desselben in allem immanenten Gebrauche angeht. [^327-1]: Wenn man sagen kann, daß eine Wissenschaft wenigstens in der Idee aller Menschen wirklich sei, so bald es ausgemacht ist, daß die Aufgaben, die darauf führen, durch die Natur der menschlichen Vernunft jedermann vorgelegt und daher auch jederzeit darüber viele, obgleich fehlerhafte Versuche unvermeidlich sind, so wird man auch sagen müssen: Metaphysik sei *subjective* (und zwar nothwendiger Weise) wirklich, und da fragen wir also mit Recht, wie sie (*objective*) möglich sei. ### Section 41 § 41. Die Unterscheidung der *Ideen*, d. i. der reinen Vernunftbegriffe, von den Kategorien oder reinen Verstandesbegriffen, als Erkenntnissen von ganz verschiedener Art, Ursprung und Gebrauch, ist ein so wichtiges Stück zur Grundlegung einer Wissenschaft, welche das System aller dieser Erkenntnisse a priori enthalten soll, daß ohne eine solche Absonderung Metaphysik schlechterdings unmöglich oder höchstens ein regelloser, stümperhafter Versuch ist, ohne Kenntniß der Materialien, womit man sich beschäftigt, und ihrer Tauglichkeit zu dieser oder jener Absicht ein Kartengebäude zusammenzuflicken. Wenn Kritik der reinen Vernunft auch nur das allein geleistet hätte, diesen Unterschied zuerst vor Augen zu legen, so hätte sie dadurch schon mehr zur Aufklärung unseres Begriffs und der Leitung der Nachforschung im Felde der Metaphysik beigetragen, als alle fruchtlose Bemühungen den transscendenten Aufgaben der reinen Vernunft ein Gnüge zu thun, die man von je her unternommen hat, ohne jemals zu wähnen, daß man sich in einem ganz andern Felde befände als dem des Verstandes und daher Verstandes- und Vernunftbegriffe, gleich als ob sie von einerlei Art wären, in einem Striche hernannte. ### Section 42 § 42. Alle reine Verstandeserkenntnisse haben das an sich, daß sich ihre Begriffe in der Erfahrung geben und ihre Grundsätze durch Erfahrung bestätigen lassen; dagegen die transscendenten Vernunfterkenntnisse sich weder, was ihre *Ideen* betrifft, in der Erfahrung geben, noch ihre Sätze jemals durch Erfahrung bestätigen, noch widerlegen lassen; daher der dabei vielleicht einschleichende Irrthum durch nichts anders als reine Vernunft selbst aufgedeckt werden kann, welches aber sehr schwer ist, weil eben diese Vernunft vermittelst ihrer Ideen natürlicher Weise dialektisch wird, und dieser unvermeidliche Schein durch keine objective und dogmatische Untersuchungen der Sachen, sondern blos durch subjective der Vernunft selbst, als eines Quells der Ideen, in Schranken gehalten werden kann. ### Section 43 § 43. Es ist jederzeit in der Kritik mein größtes Augenmerk gewesen, wie ich nicht allein die Erkenntnißarten sorgfältig unterscheiden, sondern auch alle zu jeder derselben gehörige Begriffe aus ihrem gemeinschaftlichen Quell ableiten könnte, damit ich nicht allein dadurch, daß ich unterrichtet wäre, woher sie abstammen, ihren Gebrauch mit Sicherheit bestimmen konnte, sondern auch den noch nie vermutheten, aber unschätzbaren Vortheil hätte, die Vollständigkeit in der Aufzählung, Classificirung und Specificirung der Begriffe a priori, mithin nach Principien zu erkennen. Ohne dieses ist in der Metaphysik alles lauter Rhapsodie, wo man niemals weiß, ob dessen, was man besitzt, gnug ist, oder ob und wo noch etwas fehlen möge. Freilich kann man diesen Vortheil auch nur in der reinen Philosophie haben, von dieser aber macht derselbe auch das Wesen aus. Da ich den Ursprung der Kategorien in den vier logischen Functionen aller Urtheile des Verstandes gefunden hatte, so war es ganz natürlich, den Ursprung der Ideen in den drei Functionen der Vernunftschlüsse zu suchen; denn wenn einmal solche reine Vernunftbegriffe (transscendentale Ideen) gegeben sind, so könnten sie, wenn man sie nicht etwa für angeboren halten will, wohl nirgends anders als in derselben Vernunfthandlung angetroffen werden, welche, so fern sie blos die Form betrifft, das Logische der Vernunftschlüsse, so fern sie aber die Verstandesurtheile in Ansehung einer oder der andern Form a priori als bestimmt vorstellt, transscendentale Begriffe der reinen Vernunft ausmacht. Der formale Unterschied der Vernunftschlüsse macht die Eintheilung derselben in kategorische, hypothetische und disjunctive nothwendig. Die darauf gegründete Vernunftbegriffe enthalten also erstlich die Idee des vollständigen Subjects (Substantiale), zweitens die Idee der vollständigen Reihe der Bedingungen, drittens die Bestimmung aller Begriffe in der Idee eines vollständigen Inbegriffs des Möglichen.[^330-1] Die erste Idee war psychologisch, die zweite kosmologisch, die dritte theologisch; und da alle drei zu einer Dialektik Anlaß geben, doch jede auf ihre eigene Art, so gründete sich darauf die Eintheilung der ganzen Dialektik der reinen Vernunft in den Paralogismus, die Antinomie und endlich das Ideal derselben, durch welche Ableitung man völlig sicher gestellt wird, daß alle Ansprüche der reinen Vernunft hier ganz vollständig vorgestellt sind, und kein einziger fehlen kann, weil das Vernunftvermögen selbst, als woraus sie allen ihren Ursprung nehmen, dadurch gänzlich ausgemessen wird. [^330-1]: Im disjunctiven Urtheile betrachten wir *alle Möglichkeit* respectiv auf einen gewissen Begriff als eingetheilt. Das ontologische Princip der durchgängigen Bestimmung eines Dinges überhaupt (von allen möglichen entgegengesetzten Prädicaten kommt jedem Dinge eines zu), welches zugleich das Princip aller disjunctiven Urtheile ist, legt den Inbegriff aller Möglichkeit zum Grunde, in welchem die Möglichkeit jedes Dinges überhaupt als bestimmbar angesehen wird. Dieses dient zu einer kleinen Erläuterung des obigen Satzes: daß die Vernunfthandlung in disjunctiven Vernunftschlüssen der Form nach mit derjenigen einerlei sei, wodurch sie die Idee eines Inbegriffs aller Realität zu Stande bringt, welche das Positive aller einander entgegengesetzten Prädicate in sich enthält. ### Section 44 § 44. Es ist bei dieser Betrachtung im Allgemeinen noch merkwürdig, daß die Vernunftideen nicht etwa so wie die Kategorien uns zum Gebrauche des Verstandes in Ansehung der Erfahrung irgend etwas nutzen, sondern in Ansehung desselben völlig entbehrlich, ja wohl gar den Maximen des Vernunfterkenntnisses der Natur entgegen und hinderlich, gleichwohl aber doch in anderer, noch zu bestimmender Absicht nothwendig sind. Ob die Seele eine einfache Substanz sei, oder nicht, das kann uns zur Erklärung der Erscheinungen derselben ganz gleichgültig sein; denn wir können den Begriff eines einfachen Wesens durch keine mögliche Erfahrung sinnlich, mithin in concreto verständlich machen; und so ist er in Ansehung aller verhofften Einsicht in die Ursache der Erscheinungen ganz leer und kann zu keinem Princip der Erklärung dessen, was innere oder äußere Erfahrung an die Hand giebt, dienen. Eben so wenig können uns die kosmologischen Ideen vom Weltanfange oder der Weltewigkeit (a parte ante) dazu nutzen, um irgend eine Begebenheit in der Welt selbst daraus zu erklären. Endlich müssen wir nach einer richtigen Maxime der Naturphilosophie uns aller Erklärung der Natureinrichtung, die aus dem Willen eines höchsten Wesens gezogen worden, enthalten, weil dieses nicht mehr Naturphilosophie ist, sondern ein Geständniß, daß es damit bei uns zu Ende gehe. Es haben also diese Ideen eine ganz andere Bestimmung ihres Gebrauchs als jene Kategorien, durch die und die darauf gebauten Grundsätze Erfahrung selbst allererst möglich ward. Indessen würde doch unsre mühsame Analytik des Verstandes, wenn unsre Absicht auf nichts anders als bloße Naturerkenntniß, so wie sie in der Erfahrung gegeben werden kann, gerichtet wäre, auch ganz überflüssig sein; denn Vernunft verrichtet ihr Geschäfte sowohl in der Mathematik als Naturwissenschaft auch ohne alle diese subtile Deduction ganz sicher und gut: also vereinigt sich unsre Kritik des Verstandes mit den Ideen der reinen Vernunft zu einer Absicht, welche über den Erfahrungsgebrauch des Verstandes hinausgesetzt ist, von welchem wir doch oben gesagt haben, daß er in diesem Betracht gänzlich unmöglich und ohne Gegenstand oder Bedeutung sei. Es muß aber dennoch zwischen dem, was zur Natur der Vernunft und des Verstandes gehört, Einstimmung sein, und jene muß zur Vollkommenheit der letztern beitragen und kann sie unmöglich verwirren. Die Auflösung dieser Frage ist folgende: Die reine Vernunft hat unter ihren Ideen nicht besondere Gegenstände, die über das Feld der Erfahrung hinauslägen, zur Absicht, sondern fordert nur Vollständigkeit des Verstandesgebrauchs im Zusammenhange der Erfahrung. Diese Vollständigkeit aber kann nur eine Vollständigkeit der Principien, aber nicht der Anschauungen und Gegenstände sein. Gleichwohl, um sich jene bestimmt vorzustellen, denkt sie sich solche als die Erkenntniß eines Objects, dessen Erkenntniß in Ansehung jener Regeln vollständig bestimmt ist, welches Object aber nur eine Idee ist, um die Verstandeserkenntniß der Vollständigkeit, die jene Idee bezeichnet, so nahe wie möglich zu bringen. ### Section 45 § 45. *Vorläufige Bemerkung* *zur Dialektik der reinen Vernunft.* Wir haben oben, Paragraph 33, 34, gezeigt, daß die Reinigkeit der Kategorien von aller Beimischung sinnlicher Bestimmungen die Vernunft verleiten könne, ihren Gebrauch gänzlich über alle Erfahrung hinaus, auf Dinge an sich selbst auszudehnen, wiewohl, da sie selbst keine Anschauung finden, welche ihnen Bedeutung und Sinn in concreto verschaffen könnte, sie als blos logische Functionen zwar ein Ding überhaupt vorstellen, aber für sich allein keinen bestimmten Begriff von irgend einem Dinge geben können. Dergleichen hyperbolische Objecte sind nun die, so man *Noumena* oder reine Verstandeswesen (besser Gedankenwesen) nennt, als z. B. *Substanz*, welche aber *ohne Beharrlichkeit in der Zeit* gedacht wird, oder eine *Ursache*, die aber *nicht in der Zeit* wirkte, u. s. w.; da man ihnen denn Prädicate beilegt, die blos dazu dienen, die Gesetzmäßigkeit der Erfahrung möglich zu machen, und gleichwohl alle Bedingungen der Anschauung, unter denen allein Erfahrung möglich ist, von ihnen wegnimmt, wodurch jene Begriffe wiederum alle Bedeutung verlieren. Es hat aber keine Gefahr, daß der Verstand von selbst, ohne durch fremde Gesetze gedrungen zu sein, über seine Grenzen so ganz muthwillig in das Feld von bloßen Gedankenwesen ausschweifen werde. Wenn aber die Vernunft, die mit keinem Erfahrungsgebrauche der Verstandesregeln, als der immer noch bedingt ist, völlig befriedigt sein kann, Vollendung dieser Kette von Bedingungen fordert, so wird der Verstand aus seinem Kreise getrieben, um theils Gegenstände der Erfahrung in einer so weit erstreckten Reihe vorzustellen, dergleichen gar keine Erfahrung fassen kann, theils sogar (um sie zu vollenden) gänzlich außerhalb derselben *Noumena* zu suchen, an welche sie jene Kette knüpfen und dadurch von Erfahrungsbedingungen endlich einmal unabhängig, ihre Haltung gleichwohl vollständig machen könne. Das sind nun die transscendentalen Ideen, welche, sie mögen nun nach dem wahren, aber verborgenen Zwecke der Naturbestimmung unserer Vernunft nicht auf überschwengliche Begriffe, sondern blos auf unbegrenzte Erweiterung des Erfahrungsgebrauchs angelegt sein, dennoch durch einen unvermeidlichen Schein dem Verstande einen *transscendenten* Gebrauch ablocken, der, obzwar betrüglich, dennoch durch keinen Vorsatz innerhalb der Grenzen der Erfahrung zu bleiben, sondern nur durch wissenschaftliche Belehrung und mit Mühe in Schranken gebracht werden kann. ### Section 46 § 46. I. *Psychologische Ideen.* (Kritik S. 341 u. f.) Man hat schon längst angemerkt, daß uns an allen Substanzen das eigentliche Subject, nämlich das, was übrig bleibt, nachdem alle Accidenzen (als Prädicate) abgesondert worden, mithin das *Substantiale* selbst unbekannt sei, und über diese Schranken unsrer Einsicht vielfältig Klagen geführt. Es ist aber hiebei wohl zu merken, daß der menschliche Verstand darüber nicht in Anspruch zu nehmen sei, daß er das Substantiale der Dinge nicht kennt, d. i. für sich allein bestimmen kann, sondern vielmehr darüber, daß er es als eine bloße Idee gleich einem gegebenen Gegenstande bestimmt zu erkennen verlangt. Die reine Vernunft fordert, daß wir zu jedem Prädicate eines Dinges sein ihm zugehöriges Subject, zu diesem aber, welches nothwendiger Weise wiederum nur Prädicat ist, fernerhin sein Subject und so forthin ins Unendliche (oder so weit wir reichen) suchen sollen. Aber hieraus folgt, daß wir nichts, wozu wir gelangen können, für ein letztes Subject halten sollen, und daß das Substantiale selbst niemals von unserm noch so tief eindringenden Verstande, selbst wenn ihm die ganze Natur aufgedeckt wäre, gedacht werden könne: weil die specifische Natur unseres Verstandes darin besteht, alles discursiv, d. i. durch Begriffe, mithin auch durch lauter Prädicate zu denken, wozu also das absolute Subject jederzeit fehlen muß. Daher sind alle reale Eigenschaften, dadurch wir Körper erkennen, lauter Accidenzen, sogar die Undurchdringlichkeit, die man sich immer nur als die Wirkung einer Kraft vorstellen muß, dazu uns das Subject fehlt. Nun scheint es, als ob wir in dem Bewußtsein unserer selbst (dem denkenden Subject) dieses Substantiale haben und zwar in einer unmittelbaren Anschauung; denn alle Prädicate des innern Sinnes beziehen sich auf das Ich als Subject, und dieses kann nicht weiter als Prädicat irgend eines andern Subjects gedacht werden. Also scheint hier die Vollständigkeit in der Beziehung der gegebenen Begriffe als Prädicate auf ein Subject nicht blos Idee, sondern der Gegenstand, nämlich das *absolute Subject* selbst, in der Erfahrung gegeben zu sein. Allein diese Erwartung wird vereitelt. Denn das Ich ist gar kein Begriff[^334-1], sondern nur Bezeichnung des Gegenstandes des innern Sinnes, so fern wir es durch kein Prädicat weiter erkennen; mithin kann es zwar an sich kein Prädicat von einem andern Dinge sein, aber eben so wenig auch ein bestimmter Begriff eines absoluten Subjects, sondern nur wie in allen andern Fällen die Beziehung der innern Erscheinungen auf das unbekannte Subject derselben. Gleichwohl veranlaßt diese Idee (die gar wohl dazu dient, als regulatives Princip alle materialistische Erklärungen der innern Erscheinungen unserer Seele gänzlich zu vernichten)[^334-1] durch einen ganz natürlichen Mißverstand ein sehr scheinbares Argument, um aus diesem vermeinten Erkenntniß von dem Substantiale unseres denkenden Wesens seine Natur, so fern die Kenntniß derselben ganz außer dem Inbegriff der Erfahrung hinaus fällt, zu schließen. [^334-1]: Wäre die Vorstellung der Apperception, das Ich, ein Begriff, wodurch irgend etwas gedacht würde, so würde es auch als Prädicat von andern Dingen gebraucht werden können, oder solche Prädicate in sich enthalten. Nun ist es nichts mehr als Gefühl eines Daseins ohne den mindesten Begriff und nur Vorstellung desjenigen, worauf alles Denken in Beziehung (relatione accidentis) steht. ### Section 47 § 47. Dieses denkende Selbst (die Seele) mag nun aber auch als das letzte Subject des Denkens, was selbst nicht weiter als Prädicat eines andern Dinges vorgestellt werden kann, Substanz heißen: so bleibt dieser Begriff doch gänzlich leer und ohne alle Folgen, wenn nicht von ihm die Beharrlichkeit als das, was den Begriff der Substanzen in der Erfahrung fruchtbar macht, bewiesen werden kann. Die Beharrlichkeit kann aber niemals aus dem Begriffe einer Substanz als eines Dinges an sich, sondern nur zum Behuf der Erfahrung bewiesen werden. Dieses ist bei der ersten Analogie der Erfahrung hinreichend dargethan worden (Kritik S. 182); und will man sich diesem Beweise nicht ergeben, so darf man nur den Versuch selbst anstellen, ob es gelingen werde, aus dem Begriffe eines Subjects, was selbst nicht als Prädicat eines andern Dinges existirt, zu beweisen, daß sein Dasein durchaus beharrlich sei, und daß es weder an sich selbst, noch durch irgend eine Naturursache entstehen oder vergehen könne. Dergleichen synthetische Sätze a priori können niemals an sich selbst, sondern jederzeit nur in Beziehung auf Dinge als Gegenstände einer möglichen Erfahrung bewiesen werden. ### Section 48 § 48. Wenn wir also aus dem Begriffe der Seele als Substanz auf Beharrlichkeit derselben schließen wollen, so kann dieses von ihr doch nur zum Behuf möglicher Erfahrung und nicht von ihr als einem Dinge an sich selbst und über alle mögliche Erfahrung hinaus gelten. Nun ist die subjective Bedingung aller unserer möglichen Erfahrung das Leben: folglich kann nur auf die Beharrlichkeit der Seele im Leben geschlossen werden, denn der Tod des Menschen ist das Ende aller Erfahrung, was die Seele als einen Gegenstand derselben betrifft, wofern nicht das Gegentheil dargethan wird, als wovon eben die Frage ist. Also kann die Beharrlichkeit der Seele nur im Leben des Menschen (deren Beweis man uns wohl schenken wird), aber nicht nach dem Tode (als woran uns eigentlich gelegen ist) dargethan werden und zwar aus dem allgemeinen Grunde, weil der Begriff der Substanz, so fern er mit dem Begriff der Beharrlichkeit als nothwendig verbunden angesehen werden soll, dieses nur nach einem Grundsatze möglicher Erfahrung und also auch nur zum Behuf derselben sein kann.[^335-1] [^335-1]: Es ist in der That sehr merkwürdig, daß die Metaphysiker jederzeit so sorglos über den Grundsatz der Beharrlichkeit der Substanzen weggeschlüpft sind, ohne jemals einen Beweis davon zu versuchen; ohne Zweifel, weil sie sich, so bald sie es mit dem Begriffe Substanz anfingen, von allen Beweisthümern gänzlich verlassen sahen. Der gemeine Verstand, der gar wohl inne ward, daß ohne diese Voraussetzung keine Vereinigung der Wahrnehmungen in einer Erfahrung möglich sei, ersetzte diesen Mangel durch ein Postulat; denn aus der Erfahrung selbst konnte er diesen Grundsatz nimmermehr ziehen, theils weil sie die Materien (Substanzen) bei allen ihren Veränderungen und Auflösungen nicht so weit verfolgen kann, um den Stoff immer unvermindert anzutreffen, theils weil der Grundsatz *Nothwendigkeit* enthält, die jederzeit das Zeichen eines Princips a priori ist. Nun wandten sie diesen Grundsatz getrost auf den Begriff der Seele als einer *Substanz* an und schlossen auf eine nothwendige Fortdauer derselben nach dem Tode des Menschen (vornehmlich da die Einfachheit dieser Substanz, welche aus der Untheilbarkeit des Bewußtseins gefolgert ward, sie wegen des Unterganges durch Auflösung sicherte). Hätten sie die ächte Quelle dieses Grundsatzes gefunden, welches aber weit tiefere Untersuchungen erforderte, als sie jemals anzufangen Lust hatten, so würden sie gesehen haben: daß jenes Gesetz der Beharrlichkeit der Substanzen nur zum Behuf der Erfahrung stattfinde und daher nur auf Dinge, so fern sie in der Erfahrung erkannt und mit andern verbunden werden sollen, niemals aber von ihnen auch unangesehen aller möglichen Erfahrung, mithin auch nicht von der Seele nach dem Tode gelten könne. ### Section 49 § 49. Daß unseren äußeren Wahrnehmungen etwas Wirkliches außer uns nicht blos correspondire, sondern auch correspondiren müsse, kann gleichfalls niemals als Verknüpfung der Dinge an sich selbst, wohl aber zum Behuf der Erfahrung bewiesen werden. Dieses will so viel sagen: daß etwas auf empirische Art, mithin als Erscheinung im Raume außer uns sei, kann man gar wohl beweisen; denn mit andern Gegenständen als denen, die zu einer möglichen Erfahrung gehören, haben wir es nicht zu thun, eben darum weil sie uns in keiner Erfahrung gegeben werden können und also für uns nichts sind. Empirisch außer mir ist das, was im Raume angeschaut wird; und da dieser sammt allen Erscheinungen, die er enthält, zu den Vorstellungen gehört, deren Verknüpfung nach Erfahrungsgesetzen eben sowohl ihre objective Wahrheit beweiset, als die Verknüpfung der Erscheinungen des innern Sinnes die Wirklichkeit meiner Seele (als eines Gegenstandes des innern Sinnes), so bin ich mir vermittelst der äußern Erfahrung eben sowohl der Wirklichkeit der Körper als äußerer Erscheinungen im Raume, wie vermittelst der innern Erfahrung des Daseins meiner Seele in der Zeit bewußt, die ich auch nur als einen Gegenstand des innern Sinnes durch Erscheinungen, die einen innern Zustand ausmachen, erkennen kann, und wovon mir das Wesen an sich selbst, das diesen Erscheinungen zum Grunde liegt, unbekannt ist. Der *Cartesianische* Idealism unterscheidet also nur äußere Erfahrung vom Traume und die Gesetzmäßigkeit als ein Kriterium der Wahrheit der erstern von der Regellosigkeit und dem falschen Schein des letztern. Er setzt in beiden Raum und Zeit als Bedingungen des Daseins der Gegenstände voraus und frägt nur, ob die Gegenstände äußerer Sinne wirklich im Raum anzutreffen seien, die wir darin im Wachen setzen, so wie der Gegenstand des innern Sinnes, die Seele, wirklich in der Zeit ist, d. i. ob Erfahrung sichere Kriterien der Unterscheidung von Einbildung bei sich führe. Hier läßt sich der Zweifel nun leicht heben, und wir heben ihn auch jederzeit im gemeinen Leben dadurch, daß wir die Verknüpfung der Erscheinungen in beiden nach allgemeinen Gesetzen der Erfahrung untersuchen, und können, wenn die Vorstellung äußerer Dinge damit durchgehends übereinstimmt, nicht zweifeln, daß sie nicht wahrhafte Erfahrung ausmachen sollten. Der materiale Idealism, da Erscheinungen als Erscheinungen nur nach ihrer Verknüpfung in der Erfahrung betrachtet werden, läßt also sich sehr leicht heben; und es ist eine eben so sichere Erfahrung, daß Körper außer uns (im Raume) existiren, als daß ich selbst nach der Vorstellung des innern Sinnes (in der Zeit) da bin. Denn der Begriff: *außer uns*, bedeutet nur die Existenz im Raume. Da aber das Ich in dem Satze: *Ich bin*, nicht blos den Gegenstand der innern Anschauung (in der Zeit), sondern das Subject des Bewußtseins, so wie Körper nicht blos die äußere Anschauung (im Raume), sondern auch das Ding *an sich selbst* bedeutet, was dieser Erscheinung zum Grunde liegt: so kann die Frage, ob die Körper (als Erscheinungen des äußern Sinnes) *außer meinen Gedanken* als Körper existiren, ohne alles Bedenken in der Natur verneint werden; aber darin verhält es sich gar nicht anders mit der Frage, ob ich selbst *als Erscheinung des innern Sinnes* (Seele nach der empirischen Psychologie) außer meiner Vorstellungskraft in der Zeit existire, denn diese muß eben so wohl verneint werden. Auf solche Weise ist alles, wenn es auf seine wahre Bedeutung gebracht wird, entschieden und gewiß. Der formale Idealism (sonst von mir transscendentale genannt) hebt wirklich den materiellen oder *Cartesianischen* auf. Denn wenn der Raum nichts als eine Form meiner Sinnlichkeit ist, so ist er als Vorstellung in mir eben so wirklich als ich selbst, und es kommt nur noch auf die empirische Wahrheit der Erscheinungen in demselben an. Ist das aber nicht, sondern der Raum und Erscheinungen in ihm sind etwas außer uns Existirendes, so können alle Kriterien der Erfahrung außer unserer Wahrnehmung niemals die Wirklichkeit dieser Gegenstände außer uns beweisen. ### Section 50 § 50. II. *Kosmologische Ideen.* (Krit. S. 405 u. f.) Dieses Product der reinen Vernunft in ihrem transscendenten Gebrauch ist das merkwürdigste Phänomen derselben, welches auch unter allen am kräftigsten wirkt, die Philosophie aus ihrem dogmatischen Schlummer zu erwecken und sie zu dem schweren Geschäfte der Kritik der Vernunft selbst zu bewegen. Ich nenne diese Idee deswegen kosmologisch, weil sie ihr Object jederzeit nur in der Sinnenwelt nimmt, auch keine andere als die, deren Gegenstand ein Object der Sinne ist, braucht, mithin so fern einheimisch und nicht transscendent, folglich bis dahin noch keine Idee ist: dahingegen, die Seele sich als eine einfache Substanz denken, schon so viel heißt, als sich einen Gegenstand denken (das Einfache), dergleichen den Sinnen gar nicht vorgestellt werden können. Demungeachtet erweitert doch die kosmologische Idee die Verknüpfung des Bedingten mit seiner Bedingung (diese mag mathematisch oder dynamisch sein) so sehr, daß Erfahrung ihr niemals gleichkommen kann, und ist also in Ansehung dieses Punkts immer eine Idee, deren Gegenstand niemals adäquat in irgend einer Erfahrung gegeben werden kann. ### Section 51 § 51. Zuerst zeigt sich hier der Nutzen eines Systems der Kategorien so deutlich und unverkennbar, daß, wenn es auch nicht mehrere Beweisthümer desselben gäbe, dieser allein ihre Unentbehrlichkeit im System der reinen Vernunft hinreichend darthun würde. Es sind solcher transscendenten Ideen nicht mehr als vier, so viel als Classen der Kategorien; in jeder derselben aber gehen sie nur auf die absolute Vollständigkeit der Reihe der Bedingungen zu einem gegebenen Bedingten. Diesen kosmologischen Ideen gemäß giebt es auch nur viererlei dialektische Behauptungen der reinen Vernunft, die, da sie dialektisch sind, dadurch selbst beweisen, daß einer jeden nach eben so scheinbaren Grundsätzen der reinen Vernunft ein ihm widersprechender entgegensteht, welchen Widerstreit keine metaphysische Kunst der subtilsten Distinction verhüten kann, sondern die den Philosophen nöthigt, zu den ersten Quellen der reinen Vernunft selbst zurück zu gehen. Diese nicht etwa beliebig erdachte, sondern in der Natur der menschlichen Vernunft gegründete, mithin unvermeidliche und niemals ein Ende nehmende Antinomie enthält nun folgende vier Sätze sammt ihren Gegensätzen. | | Satz | Gegensatz | |---|---|---| | 1. | Die Welt hat der Zeit und dem Raum nach einen *Anfang* (Grenze) | Die Welt ist der Zeit und dem Raum nach *unendlich* | | 2. | Alles in der Welt besteht aus dem *Einfachen* | Es ist nichts Einfaches, sondern alles ist *zusammengesetzt* | | 3. | Es giebt in der Welt Ursachen durch *Freiheit* | Es ist keine Freiheit, sondern alles ist *Natur* | | 4. | In der Reihe der Weltursachen ist irgend ein *nothwendig* Wesen | Es ist in ihr nichts nothwendig, sondern in dieser Reihe ist *alles zufällig*. | ### Section 52 § 52. Hier ist nun das seltsamste Phänomen der menschlichen Vernunft, wovon sonst kein Beispiel in irgend einem andern Gebrauch derselben gezeigt werden kann. Wenn wir, wie es gewöhnlich geschieht, uns die Erscheinungen der Sinnenwelt als Dinge an sich selbst denken; wenn wir die Grundsätze ihrer Verbindung als allgemein von Dingen an sich selbst und nicht blos von der Erfahrung geltende Grundsätze annehmen, wie denn dieses eben so gewöhnlich, ja ohne unsre Kritik unvermeidlich ist: so thut sich ein nicht vermutheter Widerstreit hervor, der niemals auf den gewöhnlichen, dogmatischen Wege beigelegt werden kann, weil sowohl Satz als Gegensatz durch gleich einleuchtende klare und unwiderstehliche Beweise dargethan werden können — denn für die Richtigkeit aller dieser Beweise verbürge ich mich —, und die Vernunft sich also mit sich selbst entzweit sieht, ein Zustand, über den der Sceptiker frohlockt, der kritische Philosoph aber in Nachdenken und Unruhe versetzt werden muß. ### Section 52b § 52b. Man kann in der Metaphysik auf mancherlei Weise herumpfuschen, ohne eben zu besorgen, daß man auf Unwahrheit werde betreten werden. Denn wenn man sich nur nicht selbst widerspricht, welches in synthetischen, obgleich gänzlich erdichteten Sätzen gar wohl möglich ist: so können wir in allen solchen Fällen, wo die Begriffe, die wir verknüpfen, bloße Ideen sind, die gar nicht (ihrem ganzen Inhalte nach) in der Erfahrung gegeben werden können, niemals durch Erfahrung widerlegt werden. Denn wie wollten wir es durch Erfahrung ausmachen: ob die Welt von Ewigkeit her sei, oder einen Anfang habe; ob Materie ins Unendliche theilbar sei, oder aus einfachen Theilen bestehe? Dergleichen Begriffe lassen sich in keiner, auch der größtmöglichen Erfahrung geben, mithin die Unrichtigkeit des behauptenden oder verneinenden Satzes durch diesen Probirstein nicht entdecken. Der einzige mögliche Fall, da die Vernunft ihre geheime Dialektik, die sie fälschlich für Dogmatik ausgiebt, wider ihren Willen offenbarte, wäre der, wenn sie auf einen allgemein zugestandnen Grundsatz eine Behauptung gründete und aus einem andern, eben so beglaubigten mit der größten Richtigkeit der Schlußart gerade das Gegentheil folgerte. Dieser Fall ist hier nun wirklich und zwar in Ansehung vier natürlicher Vernunftideen, woraus vier Behauptungen einerseits und eben so viel Gegenbehauptungen andererseits, jede mit richtiger Consequenz aus allgemein zugestandnen Grundsätzen entspringen und dadurch den dialektischen Schein der reinen Vernunft im Gebrauch dieser Grundsätze offenbaren, der sonst auf ewig verborgen sein müßte. Hier ist also ein entscheidender Versuch, der uns nothwendig eine Unrichtigkeit entdecken muß, die in den Voraussetzungen der Vernunft verborgen liegt[^341-1]. Von zwei einander widersprechenden Sätzen können nicht alle beide falsch sein, außer wenn der Begriff selbst widersprechend ist, der beiden zum Grunde liegt; z. B. die zwei Sätze: ein viereckichter Cirkel ist rund, und: ein viereckichter Cirkel ist nicht rund, sind beide falsch. Denn was den ersten betrifft, so ist es falsch, daß der genannte Cirkel rund sei, weil er viereckicht ist; es ist aber auch falsch, daß er nicht rund, d. i. eckicht, sei, weil er ein Cirkel ist. Denn darin besteht eben das logische Merkmal der Unmöglichkeit eines Begriffs, daß unter desselben Voraussetzung zwei widersprechende Sätze zugleich falsch sein würden, mithin, weil kein Drittes zwischen ihnen gedacht werden kann, durch jenen Begriff gar nichts gedacht wird. [^341-1]: Ich wünsche daher, daß der kritische Leser sich mit dieser Antinomie hauptsächlich beschäftige, weil die Natur selbst sie aufgestellt zu haben scheint, um die Vernunft in ihren dreisten Anmaßungen stutzig zu machen und zur Selbstprüfung zu nöthigen. Jeden Beweis, den ich für die Thesis sowohl als Antithesis gegeben habe, mache ich mich anheischig zu verantworten und dadurch die Gewißheit der unvermeidlichen Antinomie der Vernunft darzuthun. Wenn der Leser nun durch diese seltsame Erscheinung dahin gebracht wird, zu der Prüfung der dabei zum Grunde liegenden Voraussetzung zurückzugehen, so wird er sich gezwungen fühlen, die erste Grundlage aller Erkenntniß der reinen Vernunft mit mir tiefer zu untersuchen. ### Section 52c § 52c. Nun liegt den zwei ersteren Antinomien, die ich mathematische nenne, weil sie sich mit der Hinzusetzung oder Theilung des Gleichartigen beschäftigen, ein solcher widersprechender Begriff zum Grunde; und daraus erkläre ich, wie es zugehe, daß Thesis sowohl als Antithesis bei beiden falsch sind. Wenn ich von Gegenständen in Zeit und Raum rede, so rede ich nicht von Dingen an sich selbst, darum weil ich von diesen nichts weiß, sondern nur von Dingen in der Erscheinung, d. i. von der Erfahrung als einer besondern Erkenntnißart der Objecte, die dem Menschen allein vergönnt ist. Was ich nun im Raume oder in der Zeit denke, von dem muß ich nicht sagen: daß es an sich selbst, auch ohne diesen meinen Gedanken, im Raume und der Zeit sei; denn da würde ich mir selbst widersprechen, weil Raum und Zeit sammt den Erscheinungen in ihnen nichts an sich selbst und außer meinen Vorstellungen Existirendes, sondern selbst nur Vorstellungsarten sind, und es offenbar widersprechend ist, zu sagen, daß eine bloße Vorstellungsart auch außer unserer Vorstellung existire. Die Gegenstände also der Sinne existiren nur in der Erfahrung; dagegen auch ohne dieselbe oder vor ihr ihnen eine eigene, für sich bestehende Existenz zu geben, heißt so viel als sich vorstellen, Erfahrung sei auch ohne Erfahrung oder vor derselben wirklich. Wenn ich nun nach der Weltgröße dem Raume und der Zeit nach frage, so ist es für alle meine Begriffe eben so unmöglich zu sagen, sie sei unendlich, als sie sei endlich. Denn keines von beiden kann in der Erfahrung enthalten sein, weil weder von einem *unendlichen* Raume oder unendlicher verflossener Zeit, noch der *Begrenzung* der Welt durch einen leeren Raum oder eine vorhergehende leere Zeit Erfahrung möglich ist; das sind nur Ideen. Also müßte diese auf die eine oder die andre Art bestimmte Größe der Welt in ihr selbst liegen, abgesondert von aller Erfahrung. Dieses widerspricht aber dem Begriffe einer Sinnenwelt, die nur ein Inbegriff der Erscheinung ist, deren Dasein und Verknüpfung nur in der Vorstellung, nämlich der Erfahrung, stattfindet, weil sie nicht Sache an sich, sondern selbst nichts als Vorstellungsart ist. Hieraus folgt, daß, da der Begriff einer für sich existirenden Sinnenwelt in sich selbst widersprechend ist, die Auflösung des Problems wegen ihrer Größe auch jederzeit falsch sein werde, man mag sie nun bejahend oder verneinend versuchen. Eben dieses gilt von der zweiten Antinomie, die die Theilung der Erscheinungen betrifft. Denn diese sind bloße Vorstellungen, und die Theile existiren blos in der Vorstellung derselben, mithin in der Theilung, d. i. in einer möglichen Erfahrung, darin sie gegeben werden, und jene geht daher nur so weit, als diese reicht. Anzunehmen, daß eine Erscheinung, z. B. die des Körpers, alle Theile vor aller Erfahrung an sich selbst enthalte, zu denen nur immer mögliche Erfahrung gelangen kann, heißt: einer bloßen Erscheinung, die nur in der Erfahrung existiren kann, doch zugleich eine eigene, vor Erfahrung vorhergehende Existenz geben, oder zu sagen, daß bloße Vorstellungen da sind, ehe sie in der Vorstellungskraft angetroffen werden, welches sich widerspricht, und mithin auch jede Auflösung der mißverstandnen Aufgabe, man mag darin behaupten, die Körper bestehen an sich aus unendlich viel Theilen, oder einer endlichen Zahl einfacher Theile. ### Section 53 § 53. In der ersten Classe der Antinomie (der mathematischen) bestand die Falschheit der Voraussetzung darin: daß, was sich widerspricht (nämlich Erscheinung als Sache an sich selbst), als vereinbar in einem Begriffe vorgestellt würde. Was aber die zweite, nämlich dynamische, Classe der Antinomie betrifft, so besteht die Falschheit der Voraussetzung darin: daß, was vereinbar ist, als widersprechend vorgestellt wird, folglich, da im ersteren Falle alle beide einander entgegengesetzte Behauptungen falsch waren, hier wiederum solche, die durch bloßen Mißverstand einander entgegengesetzt werden, alle beide wahr sein können. Die mathematische Verknüpfung nämlich setzt nothwendig Gleichartigkeit des Verknüpften (im Begriffe der Größe) voraus, die dynamische erfordert dieses keinesweges. Wenn es auf die Größe des Ausgedehnten ankommt, so müssen alle Theile unter sich und mit dem Ganzen gleichartig sein; dagegen in der Verknüpfung der Ursache und Wirkung kann zwar auch Gleichartigkeit angetroffen werden, aber sie ist nicht nothwendig; denn der Begriff der Causalität (vermittelst dessen durch Etwas etwas ganz davon Verschiedenes gesetzt wird) erfordert sie wenigstens nicht. Würden die Gegenstände der Sinnenwelt für Dinge an sich selbst genommen und die oben angeführte Naturgesetze für Gesetze der Dinge an sich selbst, so wäre der Widerspruch unvermeidlich. Eben so wenn das Subject der Freiheit gleich den übrigen Gegenständen als bloße Erscheinung vorgestellt würde, so könnte eben so wohl der Widerspruch nicht vermieden werden; denn es würde eben dasselbe von einerlei Gegenstande in derselben Bedeutung zugleich bejaht und verneint werden. Ist aber Naturnothwendigkeit blos auf Erscheinungen bezogen und Freiheit blos auf Dinge an sich selbst, so entspringt kein Widerspruch, wenn man gleich beide Arten von Causalität annimmt oder zugiebt, so schwer oder unmöglich es auch sein möchte, die von der letzteren Art begreiflich zu machen. In der Erscheinung ist jede Wirkung eine Begebenheit oder etwas, das in der Zeit geschieht; vor ihr muß nach dem allgemeinen Naturgesetze eine Bestimmung der Causalität ihrer Ursache (ein Zustand derselben) vorhergehen, worauf sie nach einem beständigen Gesetze folgt. Aber diese Bestimmung der Ursache zur Causalität muß auch etwas sein, was sich eräugnet oder *geschieht*; die Ursache muß *angefangen* haben zu *handeln*, denn sonst ließe sich zwischen ihr und der Wirkung keine Zeitfolge denken. Die Wirkung wäre immer gewesen, so wie die Causalität der Ursache. Also muß unter Erscheinungen die *Bestimmung* der Ursache *zum Wirken* auch entstanden und mithin eben so wohl als ihre Wirkung eine Begebenheit sein, die wiederum ihre Ursache haben muß u. s. w., und folglich Naturnothwendigkeit die Bedingung sein, nach welcher die wirkende Ursachen bestimmt werden. Soll dagegen Freiheit eine Eigenschaft gewisser Ursachen der Erscheinungen sein, so muß sie respective auf die letztere als Begebenheiten ein Vermögen sein, sie *von selbst* (sponte) anzufangen, d. i. ohne daß die Causalität der Ursache selbst anfangen dürfte und daher keines andern, ihren Anfang bestimmenden Grundes benöthigt wäre. Alsdann aber müßte die *Ursache* ihrer Causalität nach nicht unter Zeitbestimmungen ihres Zustandes stehen, d. i. gar nicht *Erscheinung* sein, d. i. sie müßte als ein Ding an sich selbst, die *Wirkungen* aber allein als *Erscheinungen* angenommen werden[^344-1]. Kann man einen solchen Einfluß der Verstandeswesen auf Erscheinungen ohne Widerspruch denken, so wird zwar aller Verknüpfung der Ursache und Wirkung in der Sinnenwelt Naturnothwendigkeit anhängen, dagegen doch derjenigen Ursache, die selbst keine Erscheinung ist (obzwar ihr zum Grunde liegt), Freiheit zugestanden, Natur also und Freiheit eben demselben Dinge, aber in verschiedener Beziehung, einmal als Erscheinung, das andre Mal als einem Dinge an sich selbst, ohne Widerspruch beigelegt werden können. Wir haben in uns ein Vermögen, welches nicht blos mit seinen subjectiv bestimmenden Gründen, welche die Naturursachen seiner Handlungen sind, in Verknüpfung steht und so fern das Vermögen eines Wesens ist, das selbst zu den Erscheinungen gehört, sondern auch auf objective Gründe, die blos Ideen sind, bezogen wird, so fern sie dieses Vermögen bestimmen können, welche Verknüpfung durch *Sollen* ausgedrückt wird. Dieses Vermögen heißt *Vernunft*; und so fern wir ein Wesen (den Menschen) lediglich nach dieser objectiv bestimmbaren Vernunft betrachten, kann es nicht als ein Sinnenwesen betrachtet werden, sondern die gedachte Eigenschaft ist die Eigenschaft eines Dinges an sich selbst, deren Möglichkeit, wie nämlich das *Sollen*, was doch noch nie geschehen ist, die Thätigkeit desselben bestimme und Ursache von Handlungen sein könne, deren Wirkung Erscheinung in der Sinnenwelt ist, wir gar nicht begreifen können. Indessen würde doch die Causalität der Vernunft in Ansehung der Wirkungen in der Sinnenwelt Freiheit sein, so fern *objective Gründe*, die selbst Ideen sind, in Ansehung ihrer als bestimmend angesehen werden. Denn ihre Handlung hinge alsdann nicht an subjectiven, mithin auch keinen Zeitbedingungen und also auch nicht vom Naturgesetze ab, das diese zu bestimmen dient, weil Gründe der Vernunft allgemein, aus Principien, ohne Einfluß der Umstände der Zeit oder des Orts Handlungen die Regel geben. Was ich hier anführe, gilt nur als Beispiel zur Verständlichkeit und gehört nicht nothwendig zu unserer Frage, welche unabhängig von Eigenschaften, die wir in der wirklichen Welt antreffen, aus bloßen Begriffen entschieden werden muß. Nun kann ich ohne Widerspruch sagen: alle Handlungen vernünftiger Wesen, so fern sie Erscheinungen sind (in irgend einer Erfahrung angetroffen werden), stehen unter der Naturnothwendigkeit; eben dieselbe Handlungen aber blos respective auf das vernünftige Subject und dessen Vermögen nach bloßer Vernunft zu handeln sind frei. Denn was wird zur Naturnothwendigkeit erfordert? Nichts weiter als die Bestimmbarkeit jeder Begebenheit der Sinnenwelt nach beständigen Gesetzen, mithin eine Beziehung auf Ursache in der Erscheinung, wobei das Ding an sich selbst, was zum Grunde liegt, und dessen Causalität unbekannt bleibt. Ich sage aber: das *Naturgesetz bleibt*, es mag nun das vernünftige Wesen aus Vernunft, mithin durch Freiheit Ursache der Wirkungen der Sinnenwelt sein, oder es mag diese auch nicht aus Vernunftgründen bestimmen. Denn ist das erste, so geschieht die Handlung nach Maximen, deren Wirkung in der Erscheinung jederzeit beständigen Gesetzen gemäß sein wird; ist das zweite, und die Handlung geschieht nicht nach Principien der Vernunft, so ist sie den empirischen Gesetzen der Sinnlichkeit unterworfen, und in beiden Fällen hängen die Wirkungen nach beständigen Gesetzen zusammen; mehr verlangen wir aber nicht zur Naturnothwendigkeit, ja mehr kennen wir an ihr auch nicht. Aber im ersten Falle ist Vernunft die Ursache dieser Naturgesetze und ist also frei, im zweiten Falle laufen die Wirkungen nach bloßen Naturgesetzen der Sinnlichkeit, darum weil die Vernunft keinen Einfluß auf sie ausübt: sie, die Vernunft, wird aber darum nicht selbst durch die Sinnlichkeit bestimmt (welches unmöglich ist) und ist daher auch in diesem Falle frei. Die Freiheit hindert also nicht das Naturgesetz der Erscheinungen, so wenig wie dieses der Freiheit des praktischen Vernunftgebrauchs, der mit Dingen an sich selbst als bestimmenden Gründen in Verbindung steht, Abbruch thut. Hiedurch wird also die praktische Freiheit, nämlich diejenige, in welcher die Vernunft nach objectiv-bestimmenden Gründen Causalität hat, gerettet, ohne daß der Naturnothwendigkeit in Ansehung eben derselben Wirkungen als Erscheinungen der mindeste Eintrag geschieht. Eben dieses kann auch zur Erläuterung desjenigen, was wir wegen der transscendentalen Freiheit und deren Vereinbarung mit Naturnothwendigkeit (in demselben Subjecte, aber nicht in einer und derselben Beziehung genommen) zu sagen hatten, dienlich sein. Denn was diese betrifft, so ist ein jeder Anfang der Handlung eines Wesens aus objectiven Ursachen respective auf diese bestimmende Gründe immer ein erster Anfang, obgleich dieselbe Handlung in der Reihe der Erscheinungen nur ein *subalterner Anfang* ist, vor welchem ein Zustand der Ursache vorhergehen muß, der sie bestimmt und selbst eben so von einer noch vorhergehenden bestimmt wird: so daß man sich an vernünftigen Wesen oder überhaupt an Wesen, so fern ihre Causalität in ihnen als Dingen an sich selbst bestimmt wird, ohne in Widerspruch mit Naturgesetzen zu gerathen, ein Vermögen denken kann, eine Reihe von Zuständen von selbst anzufangen. Denn das Verhältniß der Handlung zu objectiven Vernunftgründen ist kein Zeitverhältniß: hier geht das, was die Causalität bestimmt, nicht der Zeit nach vor der Handlung vorher, weil solche bestimmende Gründe nicht Beziehung der Gegenstände auf Sinne, mithin nicht auf Ursachen in der Erscheinung, sondern bestimmende Ursachen als Dinge an sich selbst, die nicht unter Zeitbedingungen stehen, vorstellen. So kann die Handlung in Ansehung der Causalität der Vernunft als ein erster Anfang, in Ansehung der Reihe der Erscheinungen aber doch zugleich als ein blos subordinirter Anfang angesehen und ohne Widerspruch in jenem Betracht als frei, in diesem (da sie blos Erscheinung ist) als der Naturnothwendigkeit unterworfen angesehen werden. Was die *vierte* Antinomie betrifft, so wird sie auf die ähnliche Art gehoben, wie der Widerstreit der Vernunft mit sich selbst in der dritten. Denn wenn die *Ursache in der Erscheinung* nur von der *Ursache der Erscheinungen*, so fern sie als *Ding an sich selbst* gedacht werden kann, unterschieden wird, so können beide Sätze wohl neben einander bestehen, nämlich daß von der Sinnenwelt überall keine Ursache (nach ähnlichen Gesetzen der Causalität) stattfinde, deren Existenz schlechthin nothwendig sei, imgleichen andererseits, daß diese Welt dennoch mit einem nothwendigen Wesen als ihrer Ursache (aber von anderer Art und nach einem andern Gesetze) verbunden sei; welcher zwei Sätze Unverträglichkeit lediglich auf dem Mißverstande beruht, das, was blos von Erscheinungen gilt, über Dinge an sich selbst auszudehnen und überhaupt beide in einem Begriffe zu vermengen. [^344-1]: Die Idee der Freiheit findet lediglich in dem Verhältnisse des *Intellectuellen* als Ursache zur *Erscheinung* als Wirkung statt. Daher können wir der Materie in Ansehung ihrer unaufhörlichen Handlung, dadurch sie ihren Raum erfüllt, nicht Freiheit beilegen, obschon diese Handlung aus innerem Princip geschieht. Eben so wenig können wir für reine Verstandeswesen, z. B. Gott, so fern seine Handlung immanent ist, keinen Begriff von Freiheit angemessen finden. Denn seine Handlung, obzwar unabhängig von äußeren bestimmenden Ursachen, ist dennoch in seiner ewigen Vernunft, mithin der göttlichen *Natur* bestimmt. Nur wenn durch eine Handlung *etwas anfangen* soll, mithin die Wirkung in der Zeitreihe, folglich der Sinnenwelt anzutreffen sein soll (z. B. Anfang der Welt), da erhebt sich die Frage, ob die Causalität der Ursache selbst auch anfangen müsse, oder ob die Ursache eine Wirkung anheben könne, ohne daß ihre Causalität selbst anfängt. Im ersteren Falle ist der Begriff dieser Causalität ein Begriff der Naturnothwendigkeit, im zweiten der Freiheit. Hieraus wird der Leser ersehen, daß, da ich Freiheit als das Vermögen eine Begebenheit von selbst anzufangen erklärte, ich genau den Begriff traf, der das Problem der Metaphysik ist. ### Section 54 § 54. Dies ist nun die Aufstellung und Auflösung der ganzen Antinomie, darin sich die Vernunft bei der Anwendung ihrer Principien auf die Sinnenwelt verwickelt findet, und wovon auch jene (die bloße Aufstellung) sogar allein schon ein beträchtliches Verdienst um die Kenntniß der menschlichen Vernunft sein würde, wenn gleich die Auflösung dieses Widerstreits den Leser, der hier einen natürlichen Schein zu bekämpfen hat, welcher ihm nur neuerlich als ein solcher vorgestellt worden, nachdem er ihn bisher immer für wahr gehalten, hiedurch noch nicht völlig befriedigt haben sollte. Denn eine Folge hievon ist doch unausbleiblich, nämlich daß, weil es ganz unmöglich ist, aus diesem Widerstreit der Vernunft mit sich selbst herauszukommen, so lange man die Gegenstände der Sinnenwelt für Sachen an sich selbst nimmt und nicht für das, was sie in der That sind, nämlich bloße Erscheinungen, der Leser dadurch genöthigt werde, die Deduction aller unsrer Erkenntniß a priori und die Prüfung derjenigen, die ich davon gegeben habe, nochmals vorzunehmen, um darüber zur Entscheidung zu kommen. Mehr verlange ich jetzt nicht; denn wenn er sich bei dieser Beschäftigung nur allererst tief gnug in die Natur der reinen Vernunft hinein gedacht hat, so werden die Begriffe, durch welche die Auflösung des Widerstreits der Vernunft allein möglich ist, ihm schon geläufig sein, ohne welchen Umstand ich selbst von dem aufmerksamsten Leser völligen Beifall nicht erwarten kann. ### Section 55 § 55. III. *Theologische Idee.* (Kritik S. 571 u. f.) Die dritte transscendentale Idee, die zu dem allerwichtigsten, aber, wenn er blos speculativ betrieben wird, überschwenglichen (transscendenten) und eben dadurch dialektischen Gebrauch der Vernunft Stoff giebt, ist das Ideal der reinen Vernunft. Da die Vernunft hier nicht, wie bei der psychologischen und kosmologischen Idee von der Erfahrung anhebt und durch Steigerung der Gründe wo möglich zur absoluten Vollständigkeit ihrer Reihe zu trachten verleitet wird, sondern gänzlich abbricht und aus bloßen Begriffen von dem, was die absolute Vollständigkeit eines Dinges überhaupt ausmachen würde, mithin vermittelst der Idee eines höchst vollkommnen Urwesens zur Bestimmung der Möglichkeit, mithin auch der Wirklichkeit aller andern Dinge herabgeht: so ist hier die bloße Voraussetzung eines Wesens, welches, obzwar nicht in der Erfahrungsreihe, dennoch zum Behuf der Erfahrung um der Begreiflichkeit der Verknüpfung, Ordnung und Einheit der letzteren willen gedacht wird, d. i. die Idee, von dem Verstandesbegriffe leichter wie in den vorigen Fällen zu unterscheiden. Daher konnte hier der dialektische Schein, welcher daraus entspringt, daß wir die subjective Bedingungen unseres Denkens für objective Bedingungen der Sachen selbst und eine nothwendige Hypothese zur Befriedigung unserer Vernunft für ein Dogma halten, leicht vor Augen gestellt werden; und ich habe daher nichts weiter über die Anmaßungen der transscendentalen Theologie zu erinnern, da das, was die Kritik hierüber sagt, faßlich, einleuchtend und entscheidend ist. ### Section 56 § 56. *Allgemeine Anmerkung* *zu den transscendentalen Ideen.* Die Gegenstände, welche uns durch Erfahrung gegeben werden, sind uns in vielerlei Absicht unbegreiflich, und es können viele Fragen, auf die uns das Naturgesetz führt, wenn sie bis zu einer gewissen Höhe, aber immer diesen Gesetzen gemäß getrieben werden, gar nicht aufgelöset werden, z. B. woher Materien einander anziehen. Allein wenn wir die Natur ganz und gar verlassen, oder im Fortgange ihrer Verknüpfung alle mögliche Erfahrung übersteigen, mithin uns in bloße Ideen vertiefen, alsdann können wir nicht sagen, daß uns der Gegenstand unbegreiflich sei, und die Natur der Dinge uns unauflösliche Aufgaben vorlege; denn wir haben es alsdann gar nicht mit der Natur oder überhaupt mit gegebenen Objecten, sondern blos mit Begriffen zu thun, die in unserer Vernunft lediglich ihren Ursprung haben, und mit bloßen Gedanken-Wesen, in Ansehung deren alle Aufgaben, die aus dem Begriffe derselben entspringen, müssen aufgelöset werden können, weil die Vernunft von ihrem eigenen Verfahren allerdings vollständige Rechenschaft geben kann und muß[^349-1]. Da die psychologische, kosmologische und theologische Ideen lauter reine Vernunftbegriffe sind, die in keiner Erfahrung gegeben werden können, so sind uns die Fragen, die uns die Vernunft in Ansehung ihrer vorlegt, nicht durch die Gegenstände, sondern durch bloße Maximen der Vernunft um ihrer Selbstbefriedigung willen aufgegeben und müssen insgesammt hinreichend beantwortet werden können; welches auch dadurch geschieht, daß man zeigt, daß sie Grundsätze sind, unsern Verstandesgebrauch zur durchgängigen Einhelligkeit, Vollständigkeit und synthetischen Einheit zu bringen, und so fern blos von der Erfahrung, aber im *Ganzen* derselben gelten. Obgleich aber ein absolutes Ganze der Erfahrung unmöglich ist, so ist doch die Idee eines Ganzen der Erkenntniß nach Principien überhaupt dasjenige, was ihr allein eine besondere Art der Einheit, nämlich die von einem System, verschaffen kann, ohne die unser Erkenntniß nichts als Stückwerk ist und zum höchsten Zwecke (der immer nur das System aller Zwecke ist) nicht gebraucht werden kann; ich verstehe aber hier nicht blos den praktischen, sondern auch den höchsten Zweck des speculativen Gebrauchs der Vernunft. Die transscendentale Ideen drücken also die eigenthümliche Bestimmung der Vernunft aus, nämlich als eines Princips der systematischen Einheit des Verstandesgebrauchs. Wenn man aber diese Einheit der Erkenntnißart dafür ansieht, als ob sie dem Objecte der Erkenntniß anhänge; wenn man sie, die eigentlich blos *regulativ* ist, für *constitutiv* hält und sich überredet, man könne vermittelst dieser Ideen seine Kenntniß weit über alle mögliche Erfahrung, mithin auf transscendente Art erweitern, da sie doch blos dazu dient, Erfahrung in ihr selbst der Vollständigkeit so nahe wie möglich zu bringen, d. i. ihren Fortgang durch nichts einzuschränken, was zur Erfahrung nicht gehören kann: so ist dieses ein bloßer Mißverstand in Beurtheilung der eigentlichen Bestimmung unserer Vernunft und ihrer Grundsätze und eine Dialektik, die theils den Erfahrungsgebrauch der Vernunft verwirrt, theils die Vernunft mit sich selbst entzweiet. *Beschluß von der Grenzbestimmung der reinen Vernunft.* [^349-1]: Herr *Platner* in seinen Aphorismen sagt daher mit Scharfsinnigkeit § 728, 729: „Wenn die Vernunft ein Kriterium ist, so kann kein Begriff möglich sein, welcher der menschlichen Vernunft unbegreiflich ist. — In dem Wirklichen allein findet Unbegreiflichkeit statt. Hier entsteht die Unbegreiflichkeit aus der Unzulänglichkeit der erworbenen Ideen.“ — Es klingt also nur paradox und ist übrigens nicht befremdlich, zu sagen, in der Natur sei uns vieles unbegreiflich (z. B. das Zeugungsvermögen), wenn wir aber noch höher steigen und selbst über die Natur hinaus gehen, so werde uns wieder alles begreiflich; denn wir verlassen alsdann ganz die *Gegenstände*, die uns gegeben werden können, und beschäftigen uns blos mit Ideen, bei denen wir das Gesetz, welches die Vernunft durch sie dem Verstande zu seinem Gebrauch in der Erfahrung vorschreibt, gar wohl begreifen können, weil es ihr eigenes Product ist. ### Section 57 § 57. Nach den allerklärsten Beweisen, die wir oben gegeben haben, würde es Ungereimtheit sein, wenn wir von irgend einem Gegenstande mehr zu erkennen hofften, als zur möglichen Erfahrung desselben gehört, oder auch von irgend einem Dinge, wovon wir annehmen, es sei nicht ein Gegenstand möglicher Erfahrung, nur auf das mindeste Erkenntniß Anspruch machten, es nach seiner Beschaffenheit, wie es an sich selbst ist, zu bestimmen; denn wodurch wollen wir diese Bestimmung verrichten, da Zeit, Raum und alle Verstandesbegriffe, vielmehr aber noch die durch empirische Anschauung oder *Wahrnehmung* in der Sinnenwelt gezogene Begriffe keinen andern Gebrauch haben, noch haben können, als blos Erfahrung möglich zu machen, und lassen wir selbst von den reinen Verstandesbegriffen diese Bedingung weg, sie alsdann ganz und gar kein Object bestimmen und überall keine Bedeutung haben. Es würde aber andererseits eine noch größere Ungereimtheit sein, wenn wir gar keine Dinge an sich selbst einräumen, oder unsere Erfahrung für die einzig mögliche Erkenntnißart der Dinge, mithin unsre Anschauung in Raum und Zeit für die allein mögliche Anschauung, unsern discursiven Verstand aber für das Urbild von jedem möglichen Verstande ausgeben wollten, mithin Principien der Möglichkeit der Erfahrung für allgemeine Bedingungen der Dinge an sich selbst wollten gehalten wissen. Unsere Principien, welche den Gebrauch der Vernunft blos auf mögliche Erfahrung einschränken, könnten demnach selbst *transscendent* werden und die Schranken unsrer Vernunft für Schranken der Möglichkeit der Dinge selbst ausgeben, wie davon *Humes* Dialogen zum Beispiel dienen können, wenn nicht eine sorgfältige Kritik die Grenzen unserer Vernunft auch in Ansehung ihres empirischen Gebrauchs bewachte und ihren Anmaßungen ihr Ziel setzte. Der Scepticism ist uranfänglich aus der Metaphysik und ihrer polizeilosen Dialektik entsprungen. Anfangs mochte er wohl blos zu Gunsten des Erfahrungsgebrauchs der Vernunft alles, was diesen übersteigt, für nichtig und betrüglich ausgeben; nach und nach aber, da man inne ward, daß es doch eben dieselbe Grundsätze a priori sind, deren man sich bei der Erfahrung bedient, die unvermerkt und, wie es schien, mit eben demselben Rechte noch weiter führten, als Erfahrung reicht, so fing man an, selbst in Erfahrungsgrundsätze einen Zweifel zu setzen. Hiemit hat es nun wohl keine Noth; denn der gesunde Verstand wird hierin wohl jederzeit seine Rechte behaupten; allein es entsprang doch eine besondere Verwirrung in der Wissenschaft, die nicht bestimmen kann, wie weit und warum nur bis dahin und nicht weiter der Vernunft zu trauen sei; dieser Verwirrung aber kann nur durch förmliche und aus Grundsätzen gezogene Grenzbestimmung unseres Vernunftgebrauchs abgeholfen und allem Rückfall auf künftige Zeit vorgebeugt werden. Es ist wahr: wir können über alle mögliche Erfahrung hinaus von dem, was Dinge an sich selbst sein mögen, keinen bestimmten Begriff geben. Wir sind aber dennoch nicht frei vor der Nachfrage nach diesen, uns gänzlich derselben zu enthalten; denn Erfahrung thut der Vernunft niemals völlig Gnüge; sie weiset uns in Beantwortung der Fragen immer weiter zurück und läßt uns in Ansehung des völligen Aufschlusses derselben unbefriedigt, wie jedermann dieses aus der Dialektik der reinen Vernunft, die eben darum ihren guten subjectiven Grund hat, hinreichend ersehen kann. Wer kann es wohl ertragen, daß wir von der Natur unserer Seele bis zum klaren Bewußtsein des Subjects und zugleich der Überzeugung gelangen, daß seine Erscheinungen nicht *materialistisch* können erklärt werden, ohne zu fragen, was denn die Seele eigentlich sei, und wenn kein Erfahrungsbegriff hiezu zureicht, allenfalls einen Vernunftbegriff (eines einfachen immateriellen Wesens) blos zu diesem Behuf anzunehmen, ob wir gleich seine objective Realität gar nicht darthun können? Wer kann sich bei der bloßen Erfahrungserkenntniß in allen kosmologischen Fragen von der Weltdauer und Größe, der Freiheit oder Naturnothwendigkeit befriedigen, da, wir mögen es anfangen, wie wir wollen, eine jede nach Erfahrungsgrundsätzen gegebene Antwort immer eine neue Frage gebiert, die eben so wohl beantwortet sein will und dadurch die Unzulänglichkeit aller physischen Erklärungsarten zur Befriedigung der Vernunft deutlich darthut? Endlich, wer sieht nicht bei der durchgängigen Zufälligkeit und Abhängigkeit alles dessen, was er nur nach Erfahrungsprincipien denken und annehmen mag, die Unmöglichkeit, bei diesen stehen zu bleiben, und fühlt sich nicht nothgedrungen, unerachtet alles Verbots, sich nicht in transscendente Ideen zu verlieren, dennoch über alle Begriffe, die er durch Erfahrung rechtfertigen kann, noch in dem Begriffe eines Wesens Ruhe und Befriedigung zu suchen, davon die Idee zwar an sich selbst der Möglichkeit nach nicht eingesehen, obgleich auch nicht widerlegt werden kann, weil sie ein bloßes Verstandeswesen betrifft, ohne die aber die Vernunft auf immer unbefriedigt bleiben müßte. Grenzen (bei ausgedehnten Wesen) setzen immer einen Raum voraus, der außerhalb einem gewissen bestimmten Platze angetroffen wird und ihn einschließt; Schranken bedürfen dergleichen nicht, sondern sind bloße Verneinungen, die eine Größe afficiren, so fern sie nicht absolute Vollständigkeit hat. Unsre Vernunft aber sieht gleichsam um sich einen Raum für die Erkenntniß der Dinge an sich selbst, ob sie gleich von ihnen niemals bestimmte Begriffe haben kann und nur auf Erscheinungen eingeschränkt ist. So lange die Erkenntniß der Vernunft gleichartig ist, lassen sich von ihr keine bestimmte Grenzen denken. In der Mathematik und Naturwissenschaft erkennt die menschliche Vernunft zwar Schranken, aber keine Grenzen, d. i. zwar daß etwas außer ihr liege, wohin sie niemals gelangen kann, aber nicht daß sie selbst in ihrem innern Fortgange irgendwo vollendet sein werde. Die Erweiterung der Einsichten in der Mathematik und die Möglichkeit immer neuer Erfindungen geht ins Unendliche; eben so die Entdeckung neuer Natureigenschaften, neuer Kräfte und Gesetze durch fortgesetzte Erfahrung und Vereinigung derselben durch die Vernunft. Aber Schranken sind hier gleichwohl nicht zu verkennen, denn Mathematik geht nur auf *Erscheinungen*, und was nicht ein Gegenstand der sinnlichen Anschauung sein kann, als die Begriffe der Metaphysik und Moral, das liegt ganz außerhalb ihrer Sphäre, und dahin kann sie niemals führen; sie bedarf aber derselben auch gar nicht. Es ist also kein continuirlicher Fortgang und Annäherung zu diesen Wissenschaften und gleichsam ein Punkt oder Linie der Berührung. Naturwissenschaft wird uns niemals das Innere der Dinge, d. i. dasjenige, was nicht Erscheinung ist, aber doch zum obersten Erklärungsgrunde der Erscheinungen dienen kann, entdecken; aber sie braucht dieses auch nicht zu ihren physischen Erklärungen; ja wenn ihr auch dergleichen anderweitig angeboten würde (z. B. Einfluß immaterieller Wesen), so soll sie es doch ausschlagen und gar nicht in den Fortgang ihrer Erklärungen bringen, sondern diese jederzeit nur auf das gründen, was als Gegenstand der Sinne zu Erfahrung gehören und mit unsern wirklichen Wahrnehmungen nach Erfahrungsgesetzen in Zusammenhang gebracht werden kann. Allein Metaphysik führt uns in den dialektischen Versuchen der reinen Vernunft (die nicht willkürlich oder muthwilliger Weise angefangen werden, sondern dazu die Natur der Vernunft selbst treibt) auf Grenzen; und die transscendentale Ideen, eben dadurch daß man ihrer nicht Umgang haben kann, daß sie sich gleichwohl niemals wollen realisiren lassen, dienen dazu, nicht allein uns wirklich die Grenzen des reinen Vernunftgebrauchs zu zeigen, sondern auch die Art, solche zu bestimmen; und das ist auch der Zweck und Nutzen dieser Naturanlage unserer Vernunft, welche Metaphysik als ihr Lieblingskind ausgeboren hat, dessen Erzeugung so wie jede andere in der Welt nicht dem ungefähren Zufalle, sondern einem ursprünglichen Keime zuzuschreiben ist, welcher zu großen Zwecken weislich organisirt ist. Denn Metaphysik ist vielleicht mehr, wie irgend eine andere Wissenschaft durch die Natur selbst ihren Grundzügen nach in uns gelegt und kann gar nicht als das Product einer beliebigen Wahl, oder als zufällige Erweiterung beim Fortgange der Erfahrungen (von denen sie sich gänzlich abtrennt) angesehen werden. Die Vernunft, durch alle ihre Begriffe und Gesetze des Verstandes, die ihr zum empirischen Gebrauche, mithin innerhalb der Sinnenwelt hinreichend sind, findet doch von sich dabei keine Befriedigung; denn durch ins Unendliche immer wiederkommende Fragen wird ihr alle Hoffnung zur vollendeten Auflösung derselben benommen. Die transscendentale Ideen, welche diese Vollendung zur Absicht haben, sind solche Probleme der Vernunft. Nun sieht sie klärlich, daß die Sinnenwelt diese Vollendung nicht enthalten könne, mithin eben so wenig auch alle jene Begriffe, die lediglich zum Verständnisse derselben dienen: Raum und Zeit und alles, was wir unter dem Namen der reinen Verstandesbegriffe angeführt haben. Die Sinnenwelt ist nichts als eine Kette nach allgemeinen Gesetzen verknüpfter Erscheinungen, sie hat also kein Bestehen für sich, sie ist eigentlich nicht das Ding an sich selbst und bezieht sich also nothwendig auf das, was den Grund dieser Erscheinung enthält, auf Wesen, die nicht blos als Erscheinung, sondern als Dinge an sich selbst erkannt werden können. In der Erkenntniß derselben kann Vernunft allein hoffen, ihr Verlangen nach Vollständigkeit im Fortgange vom Bedingten zu dessen Bedingungen einmal befriedigt zu sehen. Oben (§ 33, 34) haben wir Schranken der Vernunft in Ansehung aller Erkenntniß bloßer Gedankenwesen angezeigt; jetzt, da uns die transscendentale Ideen dennoch den Fortgang bis zu ihnen nothwendig machen und nur also gleichsam bis zur Berührung des vollen Raumes (der Erfahrung) mit dem leeren (wovon wir nichts wissen können, den Noumenis) geführt haben, können wir auch die Grenzen der reinen Vernunft bestimmen; denn in allen Grenzen ist auch etwas Positives (z. B. Fläche ist die Grenze des körperlichen Raumes, indessen doch selbst ein Raum, Linie ein Raum, der die Grenze der Fläche ist, Punkt die Grenze der Linie, aber doch noch immer ein Ort im Raume), dahingegen Schranken bloße Negationen enthalten. Die im angeführten §ph angezeigte Schranken sind noch nicht genug, nachdem wir gefunden haben, daß noch über dieselbe etwas (ob wir es gleich, was es an sich selbst sei, niemals erkennen werden) hinausliege. Denn nun frägt sich: wie verhält sich unsere Vernunft bei dieser Verknüpfung dessen, was wir kennen, mit dem, was wir nicht kennen und auch niemals kennen werden? Hier ist eine wirkliche Verknüpfung des Bekannten mit einem völlig Unbekannten (was es auch jederzeit bleiben wird), und wenn dabei das Unbekannte auch nicht im Mindesten bekannter werden sollte — wie denn das in der That auch nicht zu hoffen ist —, so muß doch der Begriff von dieser Verknüpfung bestimmt und zur Deutlichkeit gebracht werden können. Wir sollen uns denn also ein immaterielles Wesen, eine Verstandeswelt und ein höchstes aller Wesen (lauter Noumena) denken, weil die Vernunft nur in diesen als Dingen an sich selbst Vollendung und Befriedigung antrifft, die sie in der Ableitung der Erscheinungen aus ihren gleichartigen Gründen niemals hoffen kann, und weil diese sich wirklich auf etwas von ihnen Unterschiedenes (mithin gänzlich Ungleichartiges) beziehen, indem Erscheinungen doch jederzeit eine Sache an sich selbst voraussetzen und also darauf Anzeige thun, man mag sie nun näher erkennen, oder nicht. Da wir nun aber diese Verstandeswesen nach dem, was sie an sich selbst sein mögen, d. i. bestimmt, niemals erkennen können, gleichwohl aber solche im Verhältniß auf die Sinnenwelt dennoch annehmen und durch die Vernunft damit verknüpfen müssen: so werden wir doch wenigstens diese Verknüpfung vermittelst solcher Begriffe denken können, die ihr Verhältniß zur Sinnenwelt ausdrücken. Denn denken wir das Verstandeswesen durch nichts als reine Verstandesbegriffe, so denken wir uns dadurch wirklich nichts Bestimmtes, mithin ist unser Begriff ohne Bedeutung; denken wir es uns durch Eigenschaften, die von der Sinnenwelt entlehnt sind, so ist es nicht mehr Verstandeswesen, es wird als eines von den Phänomenen gedacht und gehört zur Sinnenwelt. Wir wollen ein Beispiel vom Begriffe des höchsten Wesens hernehmen. Der *deistische* Begriff ist ein ganz reiner Vernunftbegriff, welcher aber nur ein Ding, das alle Realität enthält, vorstellt, ohne deren eine einzige bestimmen zu können, weil dazu das Beispiel aus der Sinnenwelt entlehnt werden müßte, in welchem Falle ich es immer nur mit einem Gegenstande der Sinne, nicht aber mit etwas ganz Ungleichartigem, was gar nicht ein Gegenstand der Sinne sein kann, zu thun haben würde. Denn ich würde ihm z. B. Verstand beilegen; ich habe aber gar keinen Begriff von einem Verstande als dem, der so ist wie der meinige, nämlich ein solcher, dem durch Sinne Anschauungen müssen gegeben werden, und der sich damit beschäftigt, sie unter Regeln der Einheit des Bewußtseins zu bringen. Aber alsdann würden die Elemente meines Begriffs immer in der Erscheinung liegen; ich wurde aber eben durch die Unzulänglichkeit der Erscheinungen genöthigt, über dieselbe hinaus zum Begriffe eines Wesens zu gehen, was gar nicht von Erscheinungen abhängig, oder damit als Bedingungen seiner Bestimmung verflochten ist. Sondere ich aber den Verstand von der Sinnlichkeit ab, um einen reinen Verstand zu haben: so bleibt nichts als die bloße Form des Denkens ohne Anschauung übrig, wodurch allein ich nichts Bestimmtes, also keinen Gegenstand erkennen kann. Ich müßte mir zu dem Ende einen andern Verstand denken, der die Gegenstände anschauete, wovon ich aber nicht den mindesten Begriff habe, weil der menschliche discursiv ist und nur durch allgemeine Begriffe erkennen kann. Eben das widerfährt mir auch, wenn ich dem höchsten Wesen einen Willen beilege: denn ich habe diesen Begriff nur, indem ich ihn aus meiner innern Erfahrung ziehe, dabei aber meiner Abhängigkeit der Zufriedenheit von Gegenständen, deren Existenz wir bedürfen, und also Sinnlichkeit zum Grunde liegt, welches dem reinen Begriffe des höchsten Wesens gänzlich widerspricht. Die Einwürfe des *Hume* wider den Deismus sind schwach und treffen niemals etwas mehr als die Beweisthümer, niemals aber den Satz der deistischen Behauptung selbst. Aber in Ansehung des Theismus, der durch eine nähere Bestimmung unseres dort blos transscendenten Begriffs vom höchsten Wesen zu Stande kommen soll, sind sie sehr stark und, nachdem man diesen Begriff einrichtet, in gewissen (in der That allen gewöhnlichen) Fällen unwiderleglich. Hume hält sich immer daran: daß durch den bloßen Begriff eines Urwesens, dem wir keine andere als ontologische Prädicate (Ewigkeit, Allgegenwart, Allmacht) beilegen, wir wirklich gar nichts Bestimmtes denken, sondern es müßten Eigenschaften hinzukommen, die einen Begriff in concreto abgeben können; es sei nicht genug, zu sagen: er sei Ursache, sondern: wie seine Causalität beschaffen sei, etwa durch Verstand und Willen; und da fangen seine Angriffe der Sache selbst, nämlich des Theismus, an, da er vorher nur die Beweisgründe des Deismus gestürmt hatte, welches keine sonderliche Gefahr nach sich zieht. Seine gefährlichen Argumente beziehen sich insgesammt auf den Anthropomorphismus, von dem er dafür hält, er sei von dem Theism unabtrennlich und mache ihn in sich selbst widersprechend, ließe man ihn aber weg, so fiele dieser hiemit auch, und es bliebe nichts als ein Deism übrig, aus dem man nichts machen, der uns zu nichts nützen und zu gar keinen Fundamenten der Religion und Sitten dienen kann. Wenn diese Unvermeidlichkeit des Anthropomorphismus gewiß wäre, so möchten die Beweise vom Dasein eines höchsten Wesens sein, welche sie wollen, und alle eingeräumt werden, der Begriff von diesem Wesen würde doch niemals von uns bestimmt werden können, ohne uns in Widersprüche zu verwickeln. Wenn wir mit dem Verbot, alle transscendente Urtheile der reinen Vernunft zu vermeiden, das damit dem Anschein nach streitende Gebot, bis zu Begriffen, die außerhalb dem Felde des immanenten (empirischen) Gebrauchs liegen, hinauszugehen, verknüpfen: so werden wir inne, daß beide zusammen bestehen können, aber nur gerade auf der *Grenze* alles erlaubten Vernunftgebrauchs; denn diese gehört eben so wohl zum Felde der Erfahrung, als dem der Gedankenwesen; und wir werden dadurch zugleich belehrt, wie jene so merkwürdige Ideen lediglich zur Grenzbestimmung der menschlichen Vernunft dienen, nämlich einerseits Erfahrungserkenntniß nicht unbegrenzt auszudehnen, so daß gar nichts mehr als blos Welt von uns zu erkennen übrig bliebe, und andererseits dennoch nicht über die Grenze der Erfahrung hinauszugehen und von Dingen außerhalb derselben als Dingen an sich selbst urtheilen zu wollen. Wir halten uns aber auf dieser Grenze, wenn wir unser Urtheil blos auf das Verhältniß einschränken, welches die Welt zu einem Wesen haben mag, dessen Begriff selbst außer aller Erkenntniß liegt, deren wir innerhalb der Welt fähig sind. Denn alsdann eignen wir dem höchsten Wesen keine von den Eigenschaften *an sich selbst* zu, durch die wir uns Gegenstände der Erfahrung denken, und vermeiden dadurch den *dogmatischen* Anthropomorphismus; wir legen sie aber dennoch dem Verhältnisse desselben zur Welt bei und erlauben uns einen *symbolischen* Anthropomorphism, der in der That nur die Sprache und nicht das Object selbst angeht. Wenn ich sage: wir sind genöthigt, die Welt so anzusehen, *als ob* sie das Werk eines höchsten Verstandes und Willens sei, so sage ich wirklich nichts mehr als: wie sich verhält eine Uhr, ein Schiff, ein Regiment zum Künstler, Baumeister, Befehlshaber, so die Sinnenwelt (oder alles das, was die Grundlage dieses Inbegriffs von Erscheinungen ausmacht) zu dem Unbekannten, das ich also hiedurch zwar nicht nach dem, was es an sich selbst ist, aber doch nach dem, was es für mich ist, nämlich in Ansehung der Welt, davon ich ein Theil bin, erkenne. ### Section 58 § 58. Eine solche Erkenntniß ist die *nach der Analogie*, welche nicht etwa, wie man das Wort gemeiniglich nimmt, eine unvollkommene Ähnlichkeit zweier Dinge, sondern eine vollkommne Ähnlichkeit zweier Verhältnisse zwischen ganz unähnlichen Dingen bedeutet[^357-1]. Vermittelst dieser Analogie bleibt doch ein für uns hinlänglich bestimmter Begriff von dem höchsten Wesen übrig, ob wir gleich alles weggelassen haben, was ihn schlechthin und *an sich selbst bestimmen* könnte; denn wir bestimmen ihn doch respectiv auf die Welt und mithin auf uns, und mehr ist uns auch nicht nöthig. Die Angriffe, welche *Hume* auf diejenigen thut, welche diesen Begriff absolut bestimmen wollen, indem sie die Materialien dazu von sich selbst und der Welt entlehnen, treffen uns nicht; auch kann er uns nicht vorwerfen, es bleibe uns gar nichts übrig, wenn man uns den objectiven Anthropomorphism von dem Begriffe des höchsten Wesens wegnähme. Denn wenn man uns nur anfangs (wie es auch Hume in der Person des Philo gegen den Kleanth in seinen Dialogen thut) als eine nothwendige Hypothese den *deistischen* Begriff des Urwesens einräumt, in welchem man sich das Urwesen durch lauter ontologische Prädikate der Substanz, Ursache ꝛc. denkt (*welches man thun muß*, weil die Vernunft, in der Sinnenwelt durch lauter Bedingungen, die immer wiederum bedingt sind, getrieben, ohne das gar keine Befriedigung haben kann; und *welches man auch füglich thun kann*, ohne in den Anthropomorphism zu gerathen, der Prädicate aus der Sinnenwelt auf ein von der Welt ganz unterschiedenes Wesen überträgt, indem jene Prädicate bloße Kategorien sind, die zwar keinen bestimmten, aber auch eben dadurch keinen auf Bedingungen der Sinnlichkeit eingeschränkten Begriff desselben geben): so kann uns nichts hindern von diesem Wesen eine *Causalität durch Vernunft* in Ansehung der Welt zu prädiciren und so zum Theismus überzuschreiten, ohne eben genöthigt zu sein, ihm diese Vernunft *an ihm selbst* als eine ihm anklebende Eigenschaft beizulegen. Denn was das Erste betrifft, so ist es der einzige mögliche Weg, den Gebrauch der Vernunft in Ansehung aller möglichen Erfahrung in der Sinnenwelt durchgängig mit sich einstimmig auf den höchsten Grad zu treiben, wenn man selbst wiederum eine höchste Vernunft als eine Ursache aller Verknüpfungen in der Welt annimmt: ein solches Princip muß ihr durchgängig vortheilhaft sein, kann ihr aber nirgend in ihrem Naturgebrauche schaden. *Zweitens* aber wird dadurch doch die Vernunft nicht als Eigenschaft auf das Urwesen *an sich selbst* übertragen, sondern nur *auf das Verhältniß* desselben zur Sinnenwelt und also der Anthropomorphism gänzlich vermieden. Denn hier wird nur die *Ursache* der Vernunftform betrachtet, die in der Welt allenthalben angetroffen wird, und dem höchsten Wesen, so fern es den Grund dieser Vernunftform der Welt enthält, zwar Vernunft beigelegt, aber nur nach der Analogie, d. i. so fern dieser Ausdruck nur das Verhältniß anzeigt, was die uns unbekannte oberste Ursache zur Welt hat, um darin alles im höchsten Grade vernunftmäßig zu bestimmen. Dadurch wird nun verhütet, daß wir uns der Eigenschaft der Vernunft nicht bedienen, um Gott, sondern um die Welt vermittelst derselben so zu denken, als es nothwendig ist, um den größtmöglichen Vernunftgebrauch in Ansehung dieser nach einem Princip zu haben. Wir gestehen dadurch: daß uns das höchste Wesen nach demjenigen, was es an sich selbst sei, gänzlich unerforschlich und *auf bestimmte Weise* sogar undenkbar sei; und werden dadurch abgehalten, nach unseren Begriffen, die wir von der Vernunft als einer wirkenden Ursache (vermittelst des Willens) haben, keinen transscendenten Gebrauch zu machen, um die göttliche Natur durch Eigenschaften, die doch immer nur von der menschlichen Natur entlehnt sind, zu bestimmen und uns in grobe oder schwärmerische Begriffe zu verlieren, andererseits aber auch nicht die Weltbetrachtung nach unseren auf Gott übertragenden Begriffen von der menschlichen Vernunft mit hyperphysischen Erklärungsarten zu überschwemmen und von ihrer eigentlichen Bestimmung abzubringen, nach der sie ein Studium der bloßen Natur durch die Vernunft und nicht eine vermessene Ableitung ihrer Erscheinungen von einer höchsten Vernunft sein soll. Der unseren schwachen Begriffen angemessene Ausdruck wird sein: daß wir uns die Welt so denken, *als ob* sie von einer höchsten Vernunft ihrem Dasein und inneren Bestimmung nach abstamme, wodurch wir theils die Beschaffenheit, die ihr, der Welt, selbst zukommt, erkennen, ohne uns doch anzumaßen, die ihrer Ursache an sich selbst bestimmen zu wollen, theils andererseits in *das Verhältniß* der obersten Ursache zur Welt den Grund dieser Beschaffenheit (der Vernunftform in der Welt) legen, ohne die Welt dazu für sich selbst zureichend zu finden[^360-1]. Auf solche Weise verschwinden die Schwierigkeiten, die dem Theismus zu widerstehen scheinen, dadurch: daß man mit dem Grundsatze des *Hume*, den Gebrauch der Vernunft nicht über das Feld aller möglichen Erfahrung dogmatisch hinaus zu treiben, einen anderen Grundsatz verbindet, den *Hume* gänzlich übersah, nämlich: das Feld möglicher Erfahrung nicht für dasjenige, was in den Augen unserer Vernunft sich selbst begrenzte, anzusehen. Kritik der Vernunft bezeichnet hier den wahren Mittelweg zwischen dem Dogmatism, den *Hume* bekämpfte, und dem Scepticism, den er dagegen einführen wollte, einen Mittelweg, der nicht wie andere Mittelwege, die man gleichsam mechanisch (etwas von einem und etwas von dem andern) sich selbst zu bestimmen anräth, und wodurch kein Mensch eines Besseren belehrt wird, sondern einen solchen, den man nach Principien genau bestimmen kann. [^357-1]: So ist eine Analogie zwischen dem rechtlichen Verhältnisse menschlicher Handlungen und dem mechanischen Verhältnisse der bewegenden Kräfte: ich kann gegen einen andern niemals etwas thun, ohne ihm ein Recht zu geben, unter den nämlichen Bedingungen eben dasselbe gegen mich zu thun; eben so wie kein Körper auf einen andern mit seiner bewegenden Kraft wirken kann, ohne dadurch zu verursachen, daß der andre ihm eben so viel entgegen wirke. Hier sind Recht und bewegende Kraft ganz unähnliche Dinge, aber in ihrem Verhältnisse ist doch völlige Ähnlichkeit. Vermittelst einer solchen Analogie kann ich daher einen Verhältnißbegriff von Dingen, die mir absolut unbekannt sind, geben. Z. B. wie sich verhält die Beförderung des Glücks der Kinder = a zu der Liebe der Eltern = b, so die Wohlfahrt des menschlichen Geschlechts = c zu dem Unbekannten in Gott = x, welches wir Liebe nennen; nicht als wenn es die mindeste Ähnlichkeit mit irgend einer menschlichen Neigung hätte, sondern weil wir das Verhältniß desselben zur Welt demjenigen ähnlich setzen können, was Dinge der Welt unter einander haben. Der Verhältnißbegriff aber ist hier eine bloße Kategorie, nämlich der Begriff der Ursache, der nichts mit Sinnlichkeit zu thun hat. [^360-1]: Ich werde sagen: die Causalität der obersten Ursache ist dasjenige in Ansehung der Welt, was menschliche Vernunft in Ansehung ihrer Kunstwerke ist. Dabei bleibt mir die Natur der obersten Ursache selbst unbekannt; ich vergleiche nur ihre mir bekannte Wirkung (die Weltordnung) und deren Vernunftmäßigkeit mit den mir bekannten Wirkungen menschlicher Vernunft und nenne daher jene eine Vernunft, ohne darum eben dasselbe, was ich am Menschen unter diesem Ausdruck verstehe, oder sonst etwas mir Bekanntes ihr als ihre Eigenschaft beizulegen. ### Section 59 § 59. Ich habe mich zu Anfange dieser Anmerkung des Sinnbildes einer *Grenze* bedient, um die Schranken der Vernunft in Ansehung ihres ihr angemessenen Gebrauchs festzusetzen. Die Sinnenwelt enthält blos Erscheinungen, die doch nicht Dinge an sich selbst sind, welche letztere (Noumena) also der Verstand, eben darum weil er die Gegenstände der Erfahrung für bloße Erscheinungen erkennt, annehmen muß. In unserer Vernunft sind beide zusammen befaßt, und es frägt sich: wie verfährt Vernunft, den Verstand in Ansehung beider Felder zu begrenzen? Erfahrung, welche alles, was zur Sinnenwelt gehört, enthält, begrenzt sich nicht selbst: sie gelangt von jedem Bedingten immer nur auf ein anderes Bedingte. Das, was sie begrenzen soll, muß gänzlich außer ihr liegen, und dieses ist das Feld der reinen Verstandeswesen. Dieses aber ist für uns ein leerer Raum, so fern es auf die *Bestimmung* der Natur dieser Verstandeswesen ankommt, und so fern können wir, wenn es auf dogmatisch-bestimmte Begriffe angesehen ist, nicht über das Feld möglicher Erfahrung hinaus kommen. Da aber eine Grenze selbst etwas Positives ist, welches sowohl zu dem gehört, was innerhalb derselben, als zum Raume, der außer einem gegebenen Inbegriff liegt, so ist es doch eine wirkliche positive Erkenntniß, deren die Vernunft blos dadurch theilhaftig wird, daß sie sich bis zu dieser Grenze erweitert, so doch, daß sie nicht über diese Grenze hinaus zu gehen versucht, weil sie daselbst einen leeren Raum vor sich findet, in welchem sie zwar Formen zu Dingen, aber keine Dinge selbst denken kann. Aber die *Begrenzung* des Erfahrungsfeldes durch etwas, was ihr sonst unbekannt ist, ist doch eine Erkenntniß, die der Vernunft in diesem Standpunkte noch übrig bleibt, dadurch sie nicht innerhalb der Sinnenwelt beschlossen, auch nicht außer derselben schwärmend, sondern so, wie es einer Kenntniß der Grenze zukommt, sich blos auf das Verhältniß desjenigen, was außerhalb derselben liegt, zu dem, was innerhalb enthalten ist, einschränkt. Die natürliche Theologie ist ein solcher Begriff auf der Grenze der menschlichen Vernunft, da sie sich genöthigt sieht, zu der Idee eines höchsten Wesens (und in praktischer Beziehung auch auf die einer intelligibelen Welt) hinauszusehen, nicht um in Ansehung dieses bloßen Verstandeswesens, mithin außerhalb der Sinnenwelt etwas zu bestimmen, sondern nur um ihren eigenen Gebrauch innerhalb derselben nach Principien der größtmöglichen (theoretischen sowohl als praktischen) Einheit zu leiten und zu diesem Behuf sich der Beziehung derselben auf eine selbständige Vernunft, als der Ursache aller dieser Verknüpfungen, zu bedienen, hiedurch aber nicht etwa sich blos ein Wesen zu *erdichten*, sondern, da außer der Sinnenwelt nothwendig Etwas, was nur der reine Verstand denkt, anzutreffen sein muß, dieses nur auf solche Weise, obwohl freilich blos nach der Analogie, zu bestimmen. Auf solche Weise bleibt unser obiger Satz, der das Resultat der ganzen Kritik ist: „daß uns Vernunft durch alle ihre Principien a priori niemals etwas mehr, als lediglich Gegenstände möglicher Erfahrung und auch von diesen nichts mehr, als was in der Erfahrung erkannt werden kann, lehre“; aber diese Einschränkung hindert nicht, daß sie uns nicht bis zur objectiven *Grenze* der Erfahrung, nämlich der *Beziehung* auf etwas, was selbst nicht Gegenstand der Erfahrung, aber doch der oberste Grund aller derselben sein muß, führe, ohne uns doch von demselben etwas an sich, sondern nur in Beziehung auf ihren eigenen vollständigen und auf die höchsten Zwecke gerichteten Gebrauch im Felde möglicher Erfahrung zu lehren. Dieses ist aber auch aller Nutzen, den man vernünftiger Weise hiebei auch nur wünschen kann, und mit welchem man Ursache hat zufrieden zu sein. ### Section 60 § 60. So haben wir Metaphysik, wie sie wirklich in der *Naturanlage* der menschlichen Vernunft gegeben ist, und zwar in demjenigen, was den wesentlichen Zweck ihrer Bearbeitung ausmacht, nach ihrer subjectiven Möglichkeit ausführlich dargestellt. Da wir indessen doch fanden, daß dieser *blos natürliche* Gebrauch einer solchen Anlage unserer Vernunft, wenn keine Disciplin derselben, welche nur durch wissenschaftliche Kritik möglich ist, sie zügelt und in Schranken setzt, sie in übersteigende, theils blos scheinbare, theils unter sich sogar strittige *dialektische* Schlüsse verwickelt, und überdem diese vernünftelnde Metaphysik zur Beförderung der Naturerkenntniß entbehrlich, ja wohl gar ihr nachtheilig ist: so bleibt es noch immer eine der Nachforschung würdige Aufgabe, die *Naturzwecke*, worauf diese Anlage zu transscendenten Begriffen in unserer Natur abgezielt sein mag, auszufinden, weil alles, was in der Natur liegt, doch auf irgend eine nützliche Absicht ursprünglich angelegt sein muß. Eine solche Untersuchung ist in der That mißlich; auch gestehe ich, daß es nur Muthmaßung sei wie alles, was die ersten Zwecke der Natur betrifft, was ich hievon zu sagen weiß, welches mir auch in diesem Fall allein erlaubt sein mag, da die Frage nicht die objective Gültigkeit metaphysischer Urtheile, sondern die Naturanlage zu denselben angeht und also außer dem System der Metaphysik in der Anthropologie liegt. Wenn ich alle transscendentale Ideen, deren Inbegriff die eigentliche Aufgabe der natürlichen reinen Vernunft ausmacht, welche sie nöthigt, die bloße Naturbetrachtung zu verlassen und über alle mögliche Erfahrung hinauszugehen und in dieser Bestrebung das Ding (es sei Wissen oder Vernünfteln), was Metaphysik heißt, zu Stande zu bringen, zusammennehme: so glaube ich gewahr zu werden, daß diese Naturanlage dahin abgezielt sei, unseren Begriff von den Fesseln der Erfahrung und den Schranken der bloßen Naturbetrachtung so weit loszumachen, daß er wenigstens ein Feld vor sich eröffnet sehe, was blos Gegenstände für den reinen Verstand enthält, die keine Sinnlichkeit erreichen kann, zwar nicht in der Absicht, um uns mit diesen speculativ zu beschäftigen (weil wir keinen Boden finden, worauf wir Fuß fassen können), sondern damit praktische Principien, die, ohne einen solchen Raum für ihre nothwendige Erwartung und Hoffnung vor sich zu finden, sich nicht zu der Allgemeinheit ausbreiten könnten, deren die Vernunft in moralischer Absicht unumgänglich bedarf. Da finde ich nun, daß die *psychologische* Idee, ich mag dadurch auch noch so wenig von der reinen und über alle Erfahrungsbegriffe erhabenen Natur der menschlichen Seele einsehen, doch wenigstens die Unzulänglichkeit der letzteren deutlich gnug zeige und mich dadurch vom Materialism, als einem zu keiner Naturerklärung tauglichen und überdem die Vernunft in praktischer Absicht verengenden psychologischen Begriffe, abführe. So dienen die *kosmologische* Ideen durch die sichtbare Unzulänglichkeit aller möglichen Naturerkenntniß, die Vernunft in ihrer rechtmäßigen Nachfrage zu befriedigen, uns vom Naturalism, der die Natur für sich selbst gnugsam ausgeben will, abzuhalten. Endlich da alle Naturnothwendigkeit in der Sinnenwelt jederzeit bedingt ist, indem sie immer Abhängigkeit der Dinge von andern voraussetzt, und die unbedingte Nothwendigkeit nur in der Einheit einer von der Sinnenwelt unterschiedenen Ursache gesucht werden muß, die Causalität derselben aber wiederum, wenn sie blos Natur wäre, niemals das Dasein des Zufälligen als seine Folge begreiflich machen könnte, so macht sich die Vernunft vermittelst der *theologischen* Idee vom Fatalism los, sowohl einer blinden Naturnothwendigkeit in dem Zusammenhange der Natur selbst ohne erstes Princip, als auch in der Causalität dieses Princips selbst, und führt auf den Begriff einer Ursache durch Freiheit, mithin einer obersten Intelligenz. So dienen die transscendentale Ideen, wenn gleich nicht dazu, uns positiv zu belehren, doch die freche und das Feld der Vernunft verengende Behauptungen des *Materialismus*, *Naturalismus* und *Fatalismus* aufzuheben und dadurch den moralischen Ideen außer dem Felde der Speculation Raum zu verschaffen; und dieses würde, dünkt mich, jene Naturanlage einigermaßen erklären. Der praktische Nutzen, den eine blos speculative Wissenschaft haben mag, liegt außerhalb den Grenzen dieser Wissenschaft, kann also blos als ein Scholion angesehen werden und gehört wie alle Scholien nicht als ein Theil zur Wissenschaft selbst. Gleichwohl liegt diese Beziehung doch wenigstens innerhalb den Grenzen der Philosophie, vornehmlich derjenigen, welche aus reinen Vernunftquellen schöpft, wo der speculative Gebrauch der Vernunft in der Metaphysik mit dem praktischen in der Moral nothwendig Einheit haben muß. Daher die unvermeidliche Dialektik der reinen Vernunft in einer Metaphysik als Naturanlage betrachtet, nicht blos als ein Schein, der aufgelöset zu werden bedarf, sondern auch als Naturanstalt seinem Zwecke nach, wenn man kann, erklärt zu werden verdient, wiewohl dieses Geschäfte als überverdienstlich der eigentlichen Metaphysik mit Recht nicht zugemuthet werden darf. Für ein zweites, aber mehr mit dem Inhalte der Metaphysik verwandtes Scholion müßte die Auflösung der Fragen gehalten werden, die in der Kritik von Seite 642 bis 668 fortgehen. Denn da werden gewisse Vernunftprincipien vorgetragen, die die Naturordnung, oder vielmehr den Verstand, der ihre Gesetze durch Erfahrung suchen soll, a priori bestimmen. Sie scheinen constitutiv und gesetzgebend in Ansehung der Erfahrung zu sein, da sie doch aus bloßer Vernunft entspringen, welche nicht so wie Verstand als ein Princip möglicher Erfahrung angesehen werden darf. Ob nun diese Übereinstimmung darauf beruhe, daß, so wie Natur den Erscheinungen oder ihrem Quell, der Sinnlichkeit, nicht an sich selbst anhängt, sondern nur in der Beziehung der letzteren auf den Verstand angetroffen wird, so diesem Verstande die durchgängige Einheit seines Gebrauchs zum Behuf einer gesammten möglichen Erfahrung (in einem System) nur mit Beziehung auf die Vernunft zukommen könne, mithin auch Erfahrung mittelbar unter der Gesetzgebung der Vernunft stehe: mag von denen, welche der Natur der Vernunft auch außer ihrem Gebrauch in der Metaphysik sogar in den allgemeinen Principien eine Naturgeschichte überhaupt systematisch zu machen, nachspüren wollen, weiter erwogen werden; denn diese Aufgabe habe ich in der Schrift selbst zwar als wichtig vorgestellt, aber ihre Auflösung nicht versucht[^364-1]. Und so endige ich die analytische Auflösung der von mir selbst aufgestellten Hauptfrage: Wie ist Metaphysik überhaupt möglich?, indem ich von demjenigen, wo ihr Gebrauch wirklich, wenigstens in den Folgen gegeben ist, zu den Gründen ihrer Möglichkeit hinaufstieg. [^364-1]: Es ist mein immerwährender Vorsatz durch die Kritik gewesen, nichts zu versäumen, was die Nachforschung der Natur der reinen Vernunft zur Vollständigkeit bringen könnte, ob es gleich noch so tief verborgen liegen möchte. Es steht nachher in jedermanns Belieben, wie weit er seine Untersuchung treiben will, wenn ihm nur angezeigt worden, welche noch anzustellen sein möchten; denn dieses kann man von demjenigen billig erwarten, der es sich zum Geschäfte gemacht hat, dieses ganze Feld zu übermessen, um es hernach zum künftigen Anbau und beliebigen Austheilung andern zu überlassen. Dahin gehören auch die beiden Scholien, welche sich durch ihre Trockenheit Liebhabern wohl schwerlich empfehlen dürften und daher nur für Kenner hingestellt worden. ### Solution: How is metaphysics possible as a science? *Auflösung der allgemeinen Frage der Prolegomenen.* *Wie ist Metaphysik als Wissenschaft möglich?* Metaphysik als Naturanlage der Vernunft ist wirklich, aber sie ist auch für sich allein (wie die analytische Auflösung der dritten Hauptfrage bewies) dialektisch und trüglich. Aus dieser also die Grundsätze hernehmen wollen und in dem Gebrauche derselben dem zwar natürlichen, nichts destoweniger aber falschen Scheine folgen, kann niemals Wissenschaft, sondern nur eitele dialektische Kunst hervorbringen, darin es eine Schule der andern zuvorthun, keine aber jemals einen rechtmäßigen und dauernden Beifall erwerben kann. Damit sie nun als Wissenschaft nicht blos auf trügliche Überredung, sondern auf Einsicht und Überzeugung Anspruch machen könne, so muß eine Kritik der Vernunft selbst den ganzen Vorrath der Begriffe a priori, die Eintheilung derselben nach den verschiedenen Quellen, der Sinnlichkeit, dem Verstande und der Vernunft, ferner eine vollständige Tafel derselben und die Zergliederung aller dieser Begriffe mit allem, was daraus gefolgert werden kann, darauf aber vornehmlich die Möglichkeit des synthetischen Erkenntnisses a priori vermittelst der Deduction dieser Begriffe, die Grundsätze ihres Gebrauchs, endlich auch die Grenzen desselben, alles aber in einem vollständigen System darlegen. Also enthält Kritik und auch sie ganz allein den ganzen wohlgeprüften und bewährten Plan, ja sogar alle Mittel der Vollziehung in sich, wornach Metaphysik als Wissenschaft zu Stande gebracht werden kann; durch andere Wege und Mittel ist sie unmöglich. Es frägt sich also hier nicht sowohl, wie dieses Geschäfte möglich, sondern nur wie es in Gang zu bringen und gute Köpfe von der bisherigen verkehrten und fruchtlosen zu einer untrüglichen Bearbeitung zu bewegen seien, und wie eine solche Vereinigung auf den gemeinschaftlichen Zweck am füglichsten gelenkt werden könne. So viel ist gewiß: wer einmal Kritik gekostet hat, den ekelt auf immer alles dogmatische Gewäsche, womit er vorher aus Noth vorlieb nahm, weil seine Vernunft etwas bedurfte und nichts Besseres zu ihrer Unterhaltung finden konnte. Die Kritik verhält sich zur gewöhnlichen Schulmetaphysik gerade wie *Chemie* zur *Alchymie*, oder wie *Astronomie* zur wahrsagenden *Astrologie*. Ich bin dafür gut, daß Niemand, der die Grundsätze der Kritik auch nur in diesen Prolegomenen durchgedacht und gefaßt hat, jemals wieder zu jener alten und sophistischen Scheinwissenschaft zurückkehren werde; vielmehr wird er mit einem gewissen Ergötzen auf eine Metaphysik hinaussehen, die nunmehr allerdings in seiner Gewalt ist, auch keiner vorbereitenden Entdeckungen mehr bedarf, und die zuerst der Vernunft daurende Befriedigung verschaffen kann. Denn das ist ein Vorzug, auf welchen unter allen möglichen Wissenschaften Metaphysik allein mit Zuversicht rechnen kann, nämlich daß sie zur Vollendung und in den beharrlichen Zustand gebracht werden kann, da sie sich weiter nicht verändern darf, auch keiner Vermehrung durch neue Entdeckungen fähig ist: weil die Vernunft hier die Quellen ihrer Erkenntniß nicht in den Gegenständen und ihrer Anschauung (durch die sie nicht ferner eines Mehreren belehrt werden kann), sondern in sich selbst hat, und, wenn sie die Grundgesetze ihres Vermögens vollständig und gegen alle Mißdeutung bestimmt dargestellt hat, nichts übrig bleibt, was reine Vernunft a priori erkennen, ja auch nur, was sie mit Grunde fragen könnte. Die sichere Aussicht auf ein so bestimmtes und geschlossenes Wissen hat einen besondern Reiz bei sich, wenn man gleich allen Nutzen (von welchem ich hernach noch reden werde) bei Seite setzt. Alle falsche Kunst, alle eitele Weisheit dauert ihre Zeit; denn endlich zerstört sie sich selbst, und die höchste Cultur derselben ist zugleich der Zeitpunkt ihres Unterganges. Daß in Ansehung der Metaphysik diese Zeit jetzt da sei, beweiset der Zustand, in welchen sie bei allem Eifer, womit sonst Wissenschaften aller Art bearbeitet werden, unter allen gelehrten Völkern verfallen ist. Die alte Einrichtung der Universitätsstudien erhält noch ihren Schatten, eine einzige Akademie der Wissenschaften bewegt noch dann und wann durch ausgesetzte Preise, einen und anderen Versuch darin zu machen; aber unter gründliche Wissenschaften wird sie nicht mehr gezählt, und man mag selbst urtheilen, wie etwa ein geistreicher Mann, den man einen großen Metaphysiker nennen wollte, diesen wohlgemeinten, aber kaum von jemanden beneideten Lobspruch aufnehmen würde. Ob aber gleich die Zeit des Verfalls aller dogmatischen Metaphysik ungezweifelt da ist, so fehlt doch noch manches dran, um sagen zu können, daß die Zeit ihrer Wiedergeburt vermittelst einer gründlichen und vollendeten Kritik der Vernunft dagegen schon erschienen sei. Alle Übergänge von einer Neigung zu der ihr entgegengesetzten gehen durch den Zustand der Gleichgültigkeit, und dieser Zeitpunkt ist der gefährlichste für einen Verfasser, aber, wie mich dünkt, doch der günstigste für die Wissenschaft. Denn wenn durch gänzliche Trennung vormaliger Verbindungen der Parteigeist erloschen ist, so sind die Gemüther in der besten Verfassung, um allmählig Vorschläge zur Verbindung nach einem anderen Plane anzuhören. Wenn ich sage, daß ich von diesen Prolegomenen hoffe, sie werden die Nachforschung im Felde der Kritik vielleicht rege machen und dem allgemeinen Geiste der Philosophie, dem es im speculativen Theile an Nahrung zu fehlen scheint, einen neuen und viel versprechenden Gegenstand der Unterhaltung darreichen, so kann ich mir schon zum voraus vorstellen: daß jedermann, den die dornichten Wege, die ich ihn in der Kritik geführt habe, unwillig und überdrüssig gemacht haben, mich fragen werde, worauf ich wohl diese Hoffnung gründe. Ich antworte: *auf das unwiderstehliche Gesetz der Nothwendigkeit*. Daß der Geist des Menschen metaphysische Untersuchungen einmal gänzlich aufgeben werde, ist eben so wenig zu erwarten, als daß wir, um nicht immer unreine Luft zu schöpfen, das Athemholen einmal lieber ganz und gar einstellen würden. Es wird also in der Welt jederzeit, und was noch mehr, bei jedem, vornehmlich dem nachdenkenden Menschen Metaphysik sein, die in Ermangelung eines öffentlichen Richtmaßes jeder sich nach seiner Art zuschneiden wird. Nun kann das, was bis daher Metaphysik geheißen hat, keinem prüfenden Kopfe ein Gnüge thun, ihr aber gänzlich zu entsagen, ist doch auch unmöglich: also muß endlich eine Kritik der reinen Vernunft selbst *versucht*, oder, wenn eine da ist, *untersucht* und in allgemeine Prüfung gezogen werden, weil es sonst kein Mittel giebt, diesem dringenden Bedürfniß, welches noch etwas mehr als bloße Wißbegierde ist, abzuhelfen. Seitdem ich Kritik kenne, habe ich am Ende des Durchlesens einer Schrift metaphysischen Inhalts, die mich durch Bestimmung ihrer Begriffe, durch Mannigfaltigkeit und Ordnung und einen leichten Vortrag sowohl unterhielt, als auch cultivirte, mich nicht entbrechen können, zu fragen: *hat dieser Autor wohl die Metaphysik um einen Schritt weiter gebracht?* Ich bitte die gelehrte Männer um Vergebung, deren Schriften mir in anderer Absicht genutzt und immer zur Cultur der Gemüthskräfte beigetragen haben, weil ich gestehe, daß ich weder in ihren noch in meinen geringeren Versuchen (denen doch Eigenliebe zum Vortheil spricht) habe finden können, daß dadurch die Wissenschaft im mindesten weiter gebracht worden, und dieses zwar aus dem ganz natürlichen Grunde, weil die Wissenschaft noch nicht existirte und auch nicht stückweise zusammengebracht werden kann, sondern ihr Keim in der Kritik vorher völlig präformirt sein muß. Man muß aber, um alle Mißdeutung zu verhüten, sich aus dem vorigen wohl erinnern, daß durch analytische Behandlung unserer Begriffe zwar dem Verstande allerdings recht viel genutzt, die Wissenschaft (der Metaphysik) aber dadurch nicht im mindesten weiter gebracht werde: weil jene Zergliederungen der Begriffe nur Materialien sind, daraus allererst Wissenschaft gezimmert werden soll. So mag man den Begriff von Substanz und Accidens noch so schön zergliedern und bestimmen; das ist recht gut als Vorbereitung zu irgend einem künftigen Gebrauche. Kann ich aber gar nicht beweisen, daß in allem, was da ist, die Substanz beharre, und nur die Accidenzen wechseln, so war durch alle jene Zergliederung die Wissenschaft nicht im mindesten weiter gebracht. Nun hat Metaphysik weder diesen Satz, noch den Satz des zureichenden Grundes, viel weniger irgend einen zusammengesetztern, als z. B. einen zur Seelenlehre oder Kosmologie gehörigen, und überall gar keinen synthetischen Satz bisher a priori gültig beweisen können: also ist durch alle jene Analysis nichts ausgerichtet, nichts geschafft und gefördert worden, und die Wissenschaft ist nach so viel Gewühl und Geräusch noch immer da, wo sie zu *Aristoteles’* Zeiten war, obzwar die Veranstaltungen dazu, wenn man nur erst den Leitfaden zu synthetischen Erkenntnissen gefunden hätte, unstreitig viel besser wie sonst getroffen worden. Glaubt jemand sich hiedurch beleidigt, so kann er diese Beschuldigung leicht zu nichte machen, wenn er nur einen einzigen synthetischen, zur Metaphysik gehörigen Satz anführen will, den er auf dogmatische Art a priori zu beweisen sich erbietet; denn nur dann, wenn er dieses leistet, werde ich ihm einräumen, daß er wirklich die Wissenschaft weiter gebracht habe: sollte dieser Satz auch sonst durch die gemeine Erfahrung genug bestätigt sein. Keine Forderung kann gemäßigter und billiger sein und im (unausbleiblich gewissen) Fall der Nichtleistung kein Ausspruch gerechter, als der: daß Metaphysik als Wissenschaft bisher noch gar nicht existirt habe. Nur zwei Dinge muß ich im Fall, daß die Ausforderung angenommen wird, verbitten: erstlich das Spielwerk von *Wahrscheinlichkeit* und Muthmaßung, welches der Metaphysik eben so schlecht ansteht als der Geometrie; zweitens die Entscheidung vermittelst der Wünschelruthe des so genannten *gesunden Menschenverstandes*, die nicht jedermann schlägt, sondern sich nach persönlichen Eigenschaften richtet. Denn was *das erstere anlangt*, so kann wohl nichts Ungereimteres gefunden werden, als in einer Metaphysik, einer Philosophie aus reiner Vernunft, seine Urtheile auf Wahrscheinlichkeit und Muthmaßung gründen zu wollen. Alles, was a priori erkannt werden soll, wird eben dadurch für apodiktisch gewiß ausgegeben und muß also auch so bewiesen werden. Man könnte eben so gut eine Geometrie oder Arithmetik auf Muthmaßungen gründen wollen; denn was den calculus probabilium der letzteren betrifft, so enthält er nicht wahrscheinliche, sondern ganz gewisse Urtheile über den Grad der Möglichkeit gewisser Fälle unter gegebenen gleichartigen Bedingungen, die in der Summe aller möglichen Fälle ganz unfehlbar der Regel gemäß zutreffen müssen, ob diese gleich in Ansehung jedes einzelnen Zufalles nicht gnug bestimmt ist. Nur in der empirischen Naturwissenschaft können Muthmaßungen (vermittelst der Induction und Analogie) gelitten werden, doch so, daß wenigstens die Möglichkeit dessen, was ich annehme, völlig gewiß sein muß. Mit der *Berufung auf den gesunden Menschenverstand*, wenn von Begriffen und Grundsätzen, nicht so fern sie in Ansehung der Erfahrung gültig sein sollen, sondern so fern sie auch außer den Bedingungen der Erfahrung für geltend ausgegeben werden wollen, die Rede ist, ist es wo möglich noch schlechter bewandt. Denn was ist der *gesunde Verstand*? Es ist der *gemeine Verstand*, so fern er richtig urtheilt. Und was ist nun der gemeine Verstand? Er ist das Vermögen der Erkenntniß und des Gebrauchs der Regeln in concreto zum Unterschiede des *speculativen Verstandes*, welcher ein Vermögen der Erkenntniß der Regeln in abstracto ist. So wird der gemeine Verstand die Regel, daß alles, was geschieht, vermittelst seiner Ursache bestimmt sei, kaum verstehen, niemals aber so im allgemeinen einsehen können. Er fordert daher ein Beispiel aus Erfahrung, und wenn er hört, daß dieses nichts anders bedeute, als was er jederzeit gedacht hat, wenn ihm eine Fensterscheibe zerbrochen oder ein Hausrath verschwunden war, so versteht er den Grundsatz und räumt ihn auch ein. Gemeiner Verstand hat also weiter keinen Gebrauch, als so fern er seine Regeln (obgleich dieselben ihm wirklich a priori beiwohnen) in der Erfahrung bestätigt sehen kann; mithin sie a priori und unabhängig von der Erfahrung einzusehen, gehört vor den speculativen Verstand und liegt ganz außer dem Gesichtskreise des gemeinen Verstandes. Metaphysik hat es ja aber lediglich mit der letzteren Art Erkenntniß zu thun; und es ist gewiß ein schlechtes Zeichen eines gesunden Verstandes, sich auf jenen Gewährsmann zu berufen, der hier gar kein Urtheil hat, und den man sonst wohl nur über die Achsel ansieht, außer wenn man sich im Gedränge sieht und sich in seiner Speculation weder zu rathen, noch zu helfen weiß. Es ist eine gewöhnliche Ausflucht, deren sich diese falsche Freunde des gemeinen Menschenverstandes (die ihn gelegentlich hoch preisen, gemeiniglich aber verachten) zu bedienen pflegen, daß sie sagen: Es müssen doch endlich einige Sätze sein, die unmittelbar gewiß sind, und von denen man nicht allein keinen Beweis, sondern auch überall keine Rechenschaft zu geben brauche, weil man sonst mit den Gründen seiner Urtheile niemals zu Ende kommen würde; aber zum Beweise dieser Befugniß können sie (außer dem Satze des Widerspruchs, der aber die Wahrheit synthetischer Urtheile darzuthun nicht hinreichend ist) niemals etwas anderes Ungezweifeltes, was sie dem gemeinen Menschenverstande unmittelbar beimessen dürfen, anführen, als mathematische Sätze: z. B. daß zweimal zwei vier ausmachen, daß zwischen zwei Punkten nur eine gerade Linie sei, u. a. m. Das sind aber Urtheile, die von denen der Metaphysik himmelweit unterschieden sind. Denn in der Mathematik kann ich alles das durch mein Denken selbst machen (construiren), was ich mir durch einen Begriff als möglich vorstelle: ich thue zu einer Zwei die andere Zwei nach und nach hinzu und mache selbst die Zahl vier, oder ziehe in Gedanken von einem Punkte zum andern allerlei Linien und kann nur eine einzige ziehen, die sich in allen ihren Theilen (gleichen sowohl als ungleichen) ähnlich ist. Aber ich kann aus dem Begriffe eines Dinges durch meine ganze Denkkraft nicht den Begriff von etwas anderem, dessen Dasein nothwendig mit dem ersteren verknüpft ist, herausbringen, sondern muß die Erfahrung zu rathe ziehen; und obgleich mir mein Verstand a priori (doch immer nur in Beziehung auf mögliche Erfahrung) den Begriff von einer solchen Verknüpfung (der Causalität) an die Hand giebt, so kann ich ihn doch nicht, wie die Begriffe der Mathematik a priori in der Anschauung darstellen und also seine Möglichkeit a priori darlegen; sondern dieser Begriff sammt den Grundsätzen seiner Anwendung bedarf immer, wenn er a priori gültig sein soll — wie es doch in der Metaphysik verlangt wird —, eine Rechtfertigung und Deduction seiner Möglichkeit, weil man sonst nicht weiß, wie weit er gültig sei, und ob er nur in der Erfahrung oder auch außer ihr gebraucht werden könne. Also kann man sich in der Metaphysik, als einer speculativen Wissenschaft der reinen Vernunft, niemals auf den gemeinen Menschenverstand berufen, aber wohl, wenn man genöthigt ist, sie zu verlassen und auf alles reine speculative Erkenntniß, welches jederzeit ein Wissen sein muß, mithin auch auf Metaphysik selbst und deren Belehrung (bei gewissen Angelegenheiten) Verzicht zu thun, und ein vernünftiger Glaube uns allein möglich, zu unserm Bedürfniß auch hinreichend (vielleicht gar heilsamer als das Wissen selbst) befunden wird. Denn alsdann ist die Gestalt der Sache ganz verändert. Metaphysik muß Wissenschaft sein, nicht allein im Ganzen, sondern auch allen ihren Theilen, sonst ist sie gar nichts, weil sie als Speculation der reinen Vernunft sonst nirgends Haltung hat als an allgemeinen Einsichten. Außer ihr aber können Wahrscheinlichkeit und gesunder Menschenverstand gar wohl ihren nützlichen und rechtmäßigen Gebrauch haben, aber nach ganz eigenen Grundsätzen, deren Gewicht immer von der Beziehung aufs Praktische abhängt. Das ist es, was ich zur Möglichkeit einer Metaphysik als Wissenschaft zu fordern mich berechtigt halte. ### Appendix *Anhang* *von dem, was geschehen kann, um Metaphysik als Wissenschaft wirklich zu machen.* Da alle Wege, die man bisher eingeschlagen ist, diesen Zweck nicht erreicht haben, auch außer einer vorhergehenden Kritik der reinen Vernunft ein solcher wohl niemals erreicht werden wird, so scheint die Zumuthung nicht unbillig, den Versuch, der hievon jetzt vor Augen gelegt ist, einer genauen und sorgfältigen Prüfung zu unterwerfen, wofern man es nicht für noch rathsamer hält, lieber alle Ansprüche auf Metaphysik gänzlich aufzugeben, in welchem Falle, wenn man seinem Vorsatze nur treu bleibt, nichts dawider einzuwenden ist. Wenn man den Lauf der Dinge nimmt, wie er wirklich geht, nicht wie er gehen sollte, so giebt es zweierlei Urtheile: ein *Urtheil, das vor der Untersuchung vorhergeht*, und dergleichen ist in unserm Falle dasjenige, wo der Leser aus seiner Metaphysik über die Kritik der reinen Vernunft (die allererst die Möglichkeit derselben untersuchen soll) ein Urtheil fällt; und dann ein anderes *Urtheil, welches auf die Untersuchung folgt*, wo der Leser die Folgerungen aus den kritischen Untersuchungen, die ziemlich stark wider seine sonst angenommene Metaphysik verstoßen dürften, ein Zeit lang bei Seite zu setzen vermag und allererst die Gründe prüft, woraus jene Folgerungen abgeleitet sein mögen. Wäre das, was gemeine Metaphysik vorträgt, ausgemacht gewiß (etwa wie Geometrie), so würde die erste Art zu urtheilen gelten; denn wenn die Folgerungen gewisser Grundsätze ausgemachten Wahrheiten widerstreiten, so sind jene Grundsätze falsch und ohne alle weitere Untersuchung zu verwerfen. Verhält es sich aber nicht so, daß Metaphysik von unstreitig gewissen (synthetischen) Sätzen einen Vorrath habe, und vielleicht gar so, daß ihrer eine Menge, die, eben so scheinbar als die besten unter ihnen, gleichwohl in ihren Folgerungen selbst unter sich streitig sind, überall aber ganz und gar kein sicheres Kriterium der Wahrheit eigentlich-metaphysischer (synthetischer) Sätze in ihr anzutreffen ist: so kann die vorhergehende Art zu urtheilen nicht Statt haben, sondern die Untersuchung der Grundsätze der Kritik muß vor allem Urtheile über ihren Werth oder Unwerth vorhergehen. *Probe eines Urtheils über die Kritik, das vor der Untersuchung vorhergeht.* Dergleichen Urtheil ist in den Göttingischen gelehrten Anzeigen, der Zugabe dritten Stück vom 19. Jenner 1782 Seite 40 u. f. anzutreffen. Wenn ein Verfasser, der mit dem Gegenstande seines Werks wohl bekannt ist, der durchgängig eigenes Nachdenken in die Bearbeitung desselben zu legen beflissen gewesen, einem Recensenten in die Hände fällt, der seinerseits scharfsichtig gnug ist, die Momente auszuspähen, auf die der Werth oder Unwerth der Schrift eigentlich beruht, nicht an Worten hängt, sondern den Sachen nachgeht und nicht bloß die Principien, von denen der Verfasser ausging, sichtet und prüft: so mag dem letzteren zwar die Strenge des Urtheils mißfallen, das Publicum ist dagegen gleichgültig, denn es gewinnt dabei; und der Verfasser selbst kann zufrieden sein, daß er Gelegenheit bekommt, seine von einem Kenner frühzeitig geprüfte Aufsätze zu berichtigen oder zu erläuteren und auf solche Weise, wenn er im Grunde Recht zu haben glaubt, den Stein des Anstoßes, der seiner Schrift in der Folge nachtheilig werden könnte, bei Zeiten wegzuräumen. Ich befinde mich mit meinem Recensenten in einer ganz anderen Lage. Er scheint gar nicht einzusehen, worauf es bei der Untersuchung, womit ich mich (glücklich oder unglücklich) beschäftigte, eigentlich ankam, und es sei nun Ungeduld ein weitläuftig Werk durchzudenken, oder verdrießliche Laune über eine angedrohte Reform einer Wissenschaft, bei der er schon längstens alles ins Reine gebracht zu haben glaubte, oder, welches ich ungern vermuthe, ein wirklich eingeschränkter Begriff daran Schuld, dadurch er sich über seine Schulmetaphysik niemals hinauszudenken vermag; kurz, er geht mit Ungestüm eine lange Reihe von Sätzen durch, bei denen man, ohne ihre Prämissen zu kennen, gar nichts denken kann, streut hin und wieder seinen Tadel aus, von welchem der Leser eben so wenig den Grund sieht, als er die Sätze versteht, dawider derselbe gerichtet sein soll, und kann also weder dem Publicum zur Nachricht nützen, noch mir im Urtheile der Kenner das mindeste schaden; daher ich diese Beurtheilung gänzlich übergangen sein würde, wenn sie mir nicht zu einigen Erläuterungen Anlaß gäbe, die den Leser dieser Prolegomenen in einigen Fällen vor Mißdeutung bewahren könnten. Damit Recensent aber doch einen Gesichtspunkt fasse, aus dem er am leichtesten auf eine dem Verfasser unvortheilhafte Art das ganze Werk vor Augen stellen könne, ohne sich mit irgend einer besondern Untersuchung bemühen zu dürfen, so fängt er damit an und endigt auch damit, daß er sagt: „Dies Werk ist ein System des transscendentellen (oder, wie er es übersetzt, des höheren)[^373-1] Idealismus.“ Beim Anblicke dieser Zeile sah ich bald, was für eine Recension da herauskommen würde, ungefähr so, als wenn jemand, der niemals von Geometrie etwas gehört oder gesehen hätte, einen *Euklid* fände und ersucht würde, sein Urtheil darüber zu fällen, nachdem er beim Durchblättern auf viel Figuren gestoßen, etwa sagte: „Das Buch ist eine systematische Anweisung zum Zeichnen: der Verfasser bedient sich einer besondern Sprache, um dunkele, unverständliche Vorschriften zu geben, die am Ende doch nichts mehr ausrichten können, als was jeder durch ein gutes natürliches Augenmaß zu Stande bringen kann ꝛc.“ Laßt uns indessen doch zusehen, was denn das für ein Idealism sei, der durch mein ganzes Werk geht, obgleich bei weitem noch nicht die Seele des Systems ausmacht. Der Satz aller ächten Idealisten von der Eleatischen Schule an bis zum Bischof *Berkeley* ist in dieser Formel enthalten: „Alle Erkenntniß durch Sinne und Erfahrung ist nichts als lauter Schein, und nur in den Ideen des reinen Verstandes und Vernunft ist Wahrheit.“ Der Grundsatz, der meinen Idealism durchgängig regiert und bestimmt, ist dagegen: „Alles Erkenntniß von Dingen aus bloßem reinen Verstande oder reiner Vernunft ist nichts als lauter Schein, und nur in der Erfahrung ist Wahrheit.“ Das ist ja aber gerade das Gegentheil von jenem eigentlichen Idealism; wie kam ich denn dazu, mich dieses Ausdrucks zu einer ganz entgegengesetzten Absicht zu bedienen, und wie der Recensent, ihn allenthalben zu sehen? Die Auflösung dieser Schwierigkeit beruht auf etwas, was man sehr leicht aus dem Zusammenhange der Schrift hätte einsehen können, wenn man gewollt hätte. Raum und Zeit sammt allem, was sie in sich enthalten, sind nicht die Dinge oder deren Eigenschaften an sich selbst, sondern gehören bloß zu Erscheinungen derselben; bis dahin bin ich mit jenen Idealisten auf einem Bekenntnisse. Allein diese und unter ihnen vornehmlich *Berkeley* sahen den Raum für eine bloße empirische Vorstellung an, die eben so wie die Erscheinungen in ihm uns nur vermittelst der Erfahrung oder Wahrnehmung zusammt allen seinen Bestimmungen bekannt würde; ich dagegen zeige zuerst: daß der Raum (und eben so die Zeit, auf welche *Berkeley* nicht Acht hatte) sammt allen seinen Bestimmungen a priori von uns erkannt werden könne, weil er sowohl als die Zeit uns vor aller Wahrnehmung oder Erfahrung als reine Form unserer Sinnlichkeit beiwohnt und alle Anschauung derselben, mithin auch alle Erscheinungen möglich macht. Hieraus folgt: daß, da Wahrheit auf allgemeinen und nothwendigen Gesetzen als ihren Kriterien beruht, die Erfahrung bei *Berkeley* keine Kriterien der Wahrheit haben könne, weil den Erscheinungen derselben (von ihm) nichts a priori zum Grunde gelegt ward, woraus denn folgte, daß sie nichts als lauter Schein sei, dagegen bei uns Raum und Zeit (in Verbindung mit den reinen Verstandesbegriffen) a priori aller möglichen Erfahrung ihr Gesetz vorschreiben, welches zugleich das sichere Kriterium abgiebt, in ihr Wahrheit von Schein zu unterscheiden.[^375-1] Mein so genannter (eigentlich kritischer) Idealism ist also von ganz eigenthümlicher Art, nämlich so, daß er den gewöhnlichen umstürzt, daß durch ihn alle Erkenntniß a priori, selbst die der Geometrie, zuerst objective Realität bekommt, welche ohne diese meine bewiesene Idealität des Raumes und der Zeit selbst von den eifrigsten Realisten gar nicht behauptet werden könnte. Bei solcher Bewandtniß der Sachen wünschte ich, um allen Mißverstand zu verhüten, daß ich diesen meinen Begriff anders benennen könnte; aber ihn ganz abzuändern will sich nicht wohl thun lassen. Es sei mir also erlaubt, ihn künftig, wie oben schon angeführt worden, den formalen, besser noch den kritischen Idealism zu nennen, um ihn vom dogmatischen des *Berkeley* und vom sceptischen des *Cartesius* zu unterscheiden. Weiter finde ich in der Beurtheilung dieses Buchs nichts Merkwürdiges. Der Verfasser derselben urtheilt durch und durch en gros, eine Manier, die klüglich gewählt ist, weil man dabei sein eigen Wissen oder Nichtwissen nicht verräth: ein einziges ausführliches Urtheil en detail würde, wenn es wie billig die Hauptfrage betroffen hätte, vielleicht meinen Irrthum, vielleicht auch das Maß der Einsicht des Recensenten in dieser Art von Untersuchungen aufgedeckt haben. Es war auch kein übelausgedachter Kunstgriff, um Lesern, welche sich nur aus Zeitungsnachrichten von Büchern einen Begriff zu machen gewohnt sind, die Lust zum Lesen des Buchs selbst frühzeitig zu benehmen, eine Menge von Sätzen, die, außer dem Zusammenhange mit ihren Beweisgründen und Erläuterungen gerissen, (vornehmlich so antipodisch, wie diese in Ansehung aller Schulmetaphysik sind) nothwendig widersinnisch lauten müssen, in einem Athem hinter einander her zu jagen, die Geduld des Lesers bis zum Ekel zu bestürmen und dann, nachdem man mich mit dem sinnreichen Satze, daß beständiger Schein Wahrheit sei, bekannt gemacht hat, mit der derben, doch väterlichen Lection zu schließen: Wozu denn der Streit wider die gemein angenommene Sprache, wozu denn und woher die idealistische Unterscheidung? Ein Urtheil, welches alles Eigenthümliche meines Buchs, da es vorher metaphysisch-ketzerisch sein sollte, zuletzt in einer bloßen Sprachneuerung setzt und klar beweist, daß mein angemaßter Richter auch nicht das mindeste davon und obenein sich selbst nicht recht verstanden habe.[^376-1] Recensent spricht indessen wie ein Mann, der sich wichtiger und vorzüglicher Einsichten bewußt sein muß, die er aber noch verborgen hält; denn mir ist in Ansehung der Metaphysik neuerlich nichts bekannt geworden, was zu einem solchen Tone berechtigen könnte. Daran thut er aber sehr unrecht, daß er der Welt seine Entdeckungen vorenthält; denn es geht ohne Zweifel noch mehreren so wie mir, daß sie bei allem Schönen, was seit langer Zeit in diesem Fache geschrieben worden, doch nicht finden konnten, daß die Wissenschaft dadurch um einen Finger breit weiter gebracht worden. Sonst Definitionen anspitzen, lahme Beweise mit neuen Krücken versehen, dem Cento der Metaphysik neue Lappen oder einen veränderten Zuschnitt geben, das findet man noch wohl, aber das verlangt die Welt nicht. Metaphysischer Behauptungen ist die Welt satt: man will die Möglichkeit dieser Wissenschaft, die Quellen, aus denen Gewißheit in derselben abgeleitet werden könne, und sichere Kriterien, den dialektischen Schein der reinen Vernunft von der Wahrheit zu unterscheiden. Hiezu muß der Recensent den Schlüssel besitzen, sonst würde er nimmermehr aus so hohem Tone gesprochen haben. Aber ich gerathe auf den Verdacht, daß ihm ein solches Bedürfniß der Wissenschaft vielleicht niemals in Gedanken gekommen sein mag; denn sonst würde er seine Beurtheilung auf diesen Punkt gerichtet und selbst ein fehlgeschlagener Versuch in einer so wichtigen Angelegenheit Achtung bei ihm erworben haben. Wenn das ist, so sind wir wieder gute Freunde. Er mag sich so tief in seine Metaphysik hinein denken, als ihm gut dünkt, daran soll ihn Niemand hindern; nur über das, was außer der Metaphysik liegt, die in der Vernunft befindliche Quelle derselben, kann er nicht urtheilen. Daß mein Verdacht aber nicht ohne Grund sei, beweise ich dadurch, daß er von der Möglichkeit der synthetischen Erkenntniß a priori, welche die eigentliche Aufgabe war, auf deren Auflösung das Schicksal der Metaphysik gänzlich beruht, und worauf meine Kritik (eben so wie hier meine Prolegomena) ganz und gar hinauslief, nicht ein Wort erwähnte. Der Idealism, auf den er stieß, und an welchem er auch hängen blieb, war nur als das einige Mittel jene Aufgabe aufzulösen in den Lehrbegriff aufgenommen worden (wiewohl er denn auch noch aus andern Gründen seine Bestätigung erhielt); und da hätte er zeigen müssen, daß entweder jene Aufgabe die Wichtigkeit nicht habe, die ich ihr (wie auch jetzt in den Prolegomenen) beilege, oder daß sie durch meinen Begriff von Erscheinungen gar nicht, oder auch auf andere Art besser könne aufgelöset werden: davon aber finde ich in der Recension kein Wort. Der Recensent verstand also nichts von meiner Schrift und vielleicht auch nichts von dem Geist und dem Wesen der Metaphysik selbst, wofern nicht vielmehr, welches ich lieber annehme, Recensenteneilfertigkeit, über die Schwierigkeit, sich durch so viel Hindernisse durchzuarbeiten, entrüstet, einen nachtheiligen Schatten auf das vor ihm liegende Werk warf und es ihm in seinen Grundzügen unkenntlich machte. Es fehlt noch sehr viel daran, daß eine gelehrte Zeitung, ihre Mitarbeiter mögen auch mit noch so guter Wahl und Sorgfalt ausgesucht werden, ihr sonst verdientes Ansehen im Felde der Metaphysik eben so wie anderwärts behaupten könne. Andere Wissenschaften und Kenntnisse haben doch ihren Maßstab. Mathematik hat ihren in sich selbst, Geschichte und Theologie in weltlichen oder heiligen Büchern, Naturwissenschaft und Arzneikunst in Mathematik und Erfahrung, Rechtsgelehrsamkeit in Gesetzbüchern und sogar Sachen des Geschmacks in Mustern der Alten. Allein zur Beurtheilung des Dinges, das Metaphysik heißt, soll erst der Maßstab gefunden werden (ich habe einen Versuch gemacht, ihn sowohl als seinen Gebrauch zu bestimmen). Was ist nun so lange, bis dieser ausgemittelt wird, zu thun, wenn doch über Schriften dieser Art geurtheilt werden muß? Sind sie von dogmatischer Art, so mag man es halten, wie man will: lange wird keiner hierin über den andern den Meister spielen, ohne daß sich einer findet, der es ihm wieder vergilt. Sind sie aber von kritischer Art und zwar nicht in Absicht auf andere Schriften, sondern auf die Vernunft selbst, so daß der Maßstab der Beurtheilung nicht schon angenommen werden kann, sondern allererst gesucht wird: so mag Einwendung und Tadel unverbeten sein, aber Verträglichkeit muß dabei doch zum Grunde liegen, weil das Bedürfniß gemeinschaftlich ist, und der Mangel benöthigter Einsicht ein richterlich-entscheidendes Ansehen unstatthaft macht. Um aber diese meine Vertheidigung zugleich an das Interesse des philosophirenden gemeinen Wesens zu knüpfen, schlage ich einen Versuch vor, der über die Art, wie alle metaphysische Untersuchungen auf ihren gemeinschaftlichen Zweck gerichtet werden müssen, entscheidend ist. Dieser ist nichts anders, als was sonst wohl Mathematiker gethan haben, um in einem Wettstreit den Vorzug ihrer Methoden auszumachen, nämlich eine Ausforderung an meinen Recensenten, nach seiner Art irgend einen einzigen von ihm behaupteten wahrhaftig metaphysischen, d. i. synthetischen und a priori aus Begriffen erkannten, allenfalls auch einen der unentbehrlichsten, als z. B. den Grundsatz der Beharrlichkeit der Substanz, oder der nothwendigen Bestimmung der Weltbegebenheiten durch ihre Ursache, aber, wie es sich gebührt, durch Gründe a priori zu erweisen. Kann er dies nicht (Stillschweigen aber ist Bekenntniß), so muß er einräumen: daß, da Metaphysik ohne apodiktische Gewißheit der Sätze dieser Art ganz und gar nichts ist, die Möglichkeit oder Unmöglichkeit derselben vor allen Dingen zuerst in einer Kritik der reinen Vernunft ausgemacht werden müsse; mithin ist er verbunden, entweder zu gestehen, daß meine Grundsätze der Kritik richtig sind, oder ihre Ungültigkeit zu beweisen. Da ich aber schon zum voraus sehe, daß, so unbesorgt er sich auch bisher auf die Gewißheit seiner Grundsätze verlassen hat, dennoch, da es auf eine strenge Probe ankommt, er in dem ganzen Umfange der Metaphysik auch nicht einen einzigen auffinden werde, mit dem er dreust auftreten könne, so will ich ihm die vortheilhafteste Bedingung bewilligen, die man nur in einem Wettstreite erwarten kann, nämlich ihm das onus probandi abnehmen und es mir auflegen lassen. Er findet nämlich in diesen Prolegomenen und in meiner Kritik S. 426–461 acht Sätze, deren zwei und zwei immer einander widerstreiten, jeder aber nothwendig zur Metaphysik gehört, die ihn entweder annehmen oder widerlegen muß (wiewohl kein einziger derselben ist, der nicht zu seiner Zeit von irgend einem Philosophen wäre angenommen worden). Nun hat er die Freiheit, sich einen von diesen acht Sätzen nach Wohlgefallen auszusuchen und ihn ohne Beweis, den ich ihm schenke, anzunehmen, aber nur einen (denn ihm wird Zeitverspillerung eben so wenig dienlich sein wie mir), und alsdann meinen Beweis des Gegensatzes anzugreifen. Kann ich nun diesen gleichwohl retten und auf solche Art zeigen, daß nach Grundsätzen, die jede dogmatische Metaphysik nothwendig anerkennen muß, das Gegentheil des von ihm adoptirten Satzes gerade eben so klar bewiesen werden könne, so ist dadurch ausgemacht, daß in der Metaphysik ein Erbfehler liege, der nicht erklärt, viel weniger gehoben werden kann, als wenn man bis zu ihrem Geburtsort, der reinen Vernunft selbst, hinaufsteigt; und so muß meine Kritik entweder angenommen, oder an ihrer Statt eine bessere gesetzt, sie also wenigstens studirt werden; welches das einzige ist, das ich jetzt nur verlange. Kann ich dagegen meinen Beweis nicht retten, so steht ein synthetischer Satz a priori aus dogmatischen Grundsätzen auf der Seite meines Gegners fest, meine Beschuldigung der gemeinen Metaphysik war darum ungerecht, und ich erbiete mich, seinen Tadel meiner Kritik (obgleich das lange noch nicht die Folge sein dürfte) für rechtmäßig zu erkennen. Hiezu aber würde es, dünkt mich, nöthig sein, aus dem Incognito zu treten, weil ich nicht absehe, wie es sonst zu verhüten wäre, daß ich nicht statt einer Aufgabe von ungenannten und doch unberufenen Gegnern mit mehreren beehrt oder bestürmt würde. *Vorschlag zu einer Untersuchung der Kritik, auf welche das Urtheil folgen kann.* Ich bin dem gelehrten Publicum auch für das Stillschweigen verbunden, womit es eine geraume Zeit hindurch meine Kritik beehrt hat; denn dieses beweiset doch einen Aufschub des Urtheils und also einige Vermuthung, daß in einem Werke, was alle gewohnte Wege verläßt und einen neuen einschlägt, in den man sich nicht sofort finden kann, doch vielleicht etwas liegen möge, wodurch ein wichtiger, aber jetzt abgestorbener Zweig menschlicher Erkenntnisse neues Leben und Fruchtbarkeit bekommen könne, mithin eine Behutsamkeit, durch kein übereiltes Urtheil das noch zarte Pfropfreis abzubrechen und zu zerstören. Eine Probe eines aus solchen Gründen verspäteten Urtheils kommt mir nur eben jetzt in der Gothaischen gelehrten Zeitung vor Augen, dessen Gründlichkeit (ohne mein hiebei verdächtiges Lob in Betracht zu ziehen) aus der faßlichen und unverfälschten Vorstellung eines zu den ersten Principien meines Werks gehörigen Stücks jeder Leser von selbst wahrnehmen wird. Und nun schlage ich vor, da ein weitläuftig Gebäude unmöglich durch einen flüchtigen Überschlag sofort im Ganzen beurtheilt werden kann, es von seiner Grundlage an Stück für Stück zu prüfen und hiebei gegenwärtige Prolegomena als einen allgemeinen Abriß zu brauchen, mit welchem dann gelegentlich das Werk selbst verglichen werden könnte. Dieses Ansinnen, wenn es nichts weiter als meine Einbildung von Wichtigkeit, die die Eitelkeit gewöhnlicher Maßen allen eigenen Producten leiht, zum Grunde hätte, wäre unbescheiden und verdiente mit Unwillen abgewiesen zu werden. Nun aber stehen die Sachen der ganzen speculativen Philosophie so, daß sie auf dem Punkte sind, völlig zu erlöschen, obgleich die menschliche Vernunft an ihnen mit nie erlöschender Neigung hängt, die nur darum, weil sie unaufhörlich getäuscht wird, es jetzt, obgleich vergeblich versucht, sich in Gleichgültigkeit zu verwandeln. In unserm denkenden Zeitalter läßt sich nicht vermuthen, daß nicht viele verdiente Männer jede gute Veranlassung benutzen sollten, zu dem gemeinschaftlichen Interesse der sich immer mehr aufklärenden Vernunft mit zu arbeiten, wenn sich nur einige Hoffnung zeigt, dadurch zum Zweck zu gelangen. Mathematik, Naturwissenschaft, Gesetze, Künste, selbst Moral ꝛc. füllen die Seele noch nicht gänzlich aus; es bleibt immer noch ein Raum in ihr übrig, der für die bloße reine und speculative Vernunft abgestochen ist, und dessen Leere uns zwingt, in Fratzen oder Tändelwerk oder auch Schwärmerei dem Scheine nach Beschäftigung und Unterhaltung, im Grunde aber nur Zerstreuung zu suchen, um den beschwerlichen Ruf der Vernunft zu übertäuben, die ihrer Bestimmung gemäß etwas verlangt, was sie für sich selbst befriedige und nicht bloß zum Behuf anderer Absichten oder zum Interesse der Neigungen in Geschäftigkeit versetze. Daher hat eine Betrachtung, die sich bloß mit diesem Umfange der für sich selbst bestehenden Vernunft beschäftigt, darum weil eben in demselben alle andere Kenntnisse, sogar Zwecke zusammenstoßen und sich in ein Ganzes vereinigen müssen, wie ich mit Grunde vermuthe, für jedermann, der es nur versucht hat, seine Begriffe so zu erweitern, einen großen Reiz, und ich darf wohl sagen, einen größeren als jedes andere theoretische Wissen, welches man gegen jenes nicht leichtlich eintauschen würde. Ich schlage aber darum diese Prolegomena zum Plane und Leitfaden der Untersuchung vor und nicht das Werk selbst, weil ich mit diesem zwar, was den Inhalt, die Ordnung und Lehrart und die Sorgfalt betrifft, die auf jeden Satz gewandt worden, um ihn genau zu wägen und zu prüfen, ehe ich ihn hinstellte, auch noch jetzt ganz wohl zufrieden bin (denn es haben Jahre dazu gehört, mich nicht allein von dem Ganzen, sondern bisweilen auch nur von einem einzigen Satze in Ansehung seiner Quellen völlig zu befriedigen), aber mit meinem Vortrage in einigen Abschnitten der Elementarlehre, z. B. der Deduction der Verstandesbegriffe oder dem von den Paralogismen der reinen Vernunft, nicht völlig zufrieden bin, weil eine gewisse Weitläuftigkeit in denselben die Deutlichkeit hindert, an deren statt man das, was hier die Prolegomenen in Ansehung dieser Abschnitte sagen, zum Grunde der Prüfung legen kann. Man rühmt von den Deutschen, daß, wozu Beharrlichkeit und anhaltender Fleiß erforderlich sind, sie es darin weiter als andere Völker bringen können. Wenn diese Meinung gegründet ist, so zeigt sich hier nun eine Gelegenheit, ein Geschäfte, an dessen glücklichem Ausgange kaum zu zweifeln ist, und woran alle denkende Menschen gleichen Antheil nehmen, welches doch bisher nicht gelungen war, zur Vollendung zu bringen und jene vortheilhafte Meinung zu bestätigen; vornehmlich da die Wissenschaft, welche es betrifft, von so besonderer Art ist, daß sie auf einmal zu ihrer ganzen Vollständigkeit und in denjenigen *beharrlichen* Zustand gebracht werden kann, da sie nicht im mindesten weiter gebracht und durch spätere Entdeckung weder vermehrt, noch auch nur verändert werden kann (den Ausputz durch hin und wieder vergrößerte Deutlichkeit oder angehängten Nutzen in allerlei Absicht rechne ich hieher nicht), ein Vortheil, den keine andere Wissenschaft hat, noch haben kann, weil keine ein so völlig isolirtes, von andern unabhängiges und mit ihnen unvermengtes Erkenntnißvermögen betrifft. Auch scheint dieser meiner Zumuthung der jetzige Zeitpunkt nicht ungünstig zu sein, da man jetzt in Deutschland fast nicht weiß, womit man sich außer den so genannten nützlichen Wissenschaften noch sonst beschäftigen könne, so daß es doch nicht bloßes Spiel, sondern zugleich Geschäfte sei, wodurch ein bleibender Zweck erreicht wird. Wie die Bemühungen der Gelehrten zu einem solchen Zweck vereinigt werden könnten, dazu die Mittel zu ersinnen, muß ich andern überlassen. Indessen ist meine Meinung nicht, irgend jemanden eine bloße Befolgung meiner Sätze zuzumuthen, oder mir auch nur mit der Hoffnung derselben zu schmeicheln, sondern es mögen sich, wie es zutrifft, Angriffe, Wiederholungen, Einschränkungen, oder auch Bestätigung, Ergänzung und Erweiterung dabei zutragen: wenn die Sache nur von Grund aus untersucht wird, so kann es jetzt nicht mehr fehlen, daß nicht ein Lehrgebäude, wenn gleich nicht das meinige, dadurch zu Stande komme, was ein Vermächtniß für die Nachkommenschaft werden kann, dafür sie Ursache haben wird, dankbar zu sein. Was, wenn man nur allererst mit den Grundsätzen der Kritik in Richtigkeit ist, für eine Metaphysik ihr zu Folge könne erwartet werden und wie diese keinesweges dadurch, daß man ihr die falsche Federn abgezogen, armselig und zu einer nur kleinen Figur herabgesetzt erscheinen dürfe, sondern in anderer Absicht reichlich und anständig ausgestattet erscheinen könne, würde hier zu zeigen zu weitläuftig sein; allein andere große Nutzen, die eine solche Reform nach sich ziehen würde, fallen sofort in die Augen. Die gemeine Metaphysik schaffte dadurch doch schon Nutzen, daß sie die Elementarbegriffe des reinen Verstandes aufsuchte, um sie durch Zergliederung deutlich und durch Erklärungen bestimmt zu machen. Dadurch ward sie eine Cultur für die Vernunft, wohin diese sich auch nachher zu wenden gut finden möchte. Allein das war auch alles Gute, was sie that. Denn dieses ihr Verdienst vernichtete sie dadurch wieder, daß sie durch waghälsige Behauptungen den Eigendünkel, durch subtile Ausflüchte und Beschönigung die Sophisterei und durch die Leichtigkeit, über die schwersten Aufgaben mit ein wenig Schulweisheit wegzukommen, die Seichtigkeit begünstigte, welche desto verführerischer ist, je mehr sie einerseits etwas von der Sprache der Wissenschaft, andererseits von der Popularität anzunehmen die Wahl hat und dadurch allen Alles, in der That aber überall nichts ist. Durch Kritik dagegen wird unserem Urtheil der Maßstab zugetheilt, wodurch Wissen von Scheinwissen mit Sicherheit unterschieden werden kann, und diese gründet dadurch, daß sie in der Metaphysik in ihre volle Ausübung gebracht wird, eine Denkungsart, die ihren wohlthätigen Einfluß nachher auf jeden andern Vernunftgebrauch erstreckt und zuerst den wahren philosophischen Geist einflößt. Aber auch der Dienst, den sie der Theologie leistet, indem sie solche von dem Urtheil der dogmatischen Speculation unabhängig macht und sie eben dadurch wider alle Angriffe solcher Gegner völlig in Sicherheit stellt, ist gewiß nicht gering zu schätzen. Denn gemeine Metaphysik, ob sie gleich jener viel Vorschub verhieß, konnte doch dieses Versprechen nachher nicht erfüllen und hatte noch überdem dadurch, daß sie speculative Dogmatik zu ihrem Beistand aufgeboten, nichts anders gethan, als Feinde wider sich selbst zu bewaffnen. Schwärmerei, die in einem aufgeklärten Zeitalter nicht aufkommen kann, als nur wenn sie sich hinter einer Schulmetaphysik verbirgt, unter deren Schutz sie es wagen darf, gleichsam mit Vernunft zu rasen, wird durch kritische Philosophie aus diesem ihrem letzten Schlupfwinkel vertrieben, und über das Alles kann es doch einem Lehrer der Metaphysik nicht anders als wichtig sein, einmal mit allgemeiner Beistimmung sagen zu können, daß, was er vorträgt, nun endlich auch *Wissenschaft* sei, und dadurch dem gemeinen Wesen wirklicher Nutzen geleistet werde. [^373-1]: Bei Leibe nicht der *höhere*. Hohe Thürme und die ihnen ähnliche metaphysisch-große Männer, um welche beide gemeiniglich viel Wind ist, sind nicht für mich. Mein Platz ist das fruchtbare *Bathos* der Erfahrung, und das Wort transscendental, dessen so vielfältig von mir angezeigte Bedeutung vom Recensenten nicht einmal gefaßt worden (so flüchtig hat er alles angesehen), bedeutet nicht etwas, das über alle Erfahrung hinausgeht, sondern was vor ihr (a priori) zwar vorhergeht, aber doch zu nichts mehrerem bestimmt ist, als lediglich Erfahrungserkenntniß möglich zu machen. Wenn diese Begriffe die Erfahrung überschreiten, dann heißt ihr Gebrauch transscendent, welcher von dem immanenten, d. i. auf Erfahrung eingeschränkten, Gebrauch unterschieden wird. Allen Mißdeutungen dieser Art ist in dem Werke hinreichend vorgebeugt worden: allein der Recensent fand seinen Vortheil bei Mißdeutungen. [^375-1]: Der eigentliche Idealismus hat jederzeit eine schwärmerische Absicht und kann auch keine andre haben; der meinige aber ist lediglich dazu, um die Möglichkeit unserer Erkenntniß a priori von Gegenständen der Erfahrung zu begreifen, welches ein Problem ist, das bisher noch nicht aufgelöset, ja nicht einmal aufgeworfen worden. Dadurch fällt nun der ganze schwärmerische Idealism, der immer (wie auch schon aus dem *Plato* zu ersehen) aus unseren Erkenntnissen a priori (selbst denen der Geometrie) auf eine andere (nämlich intellectuelle) Anschauung als die der Sinne schloß, weil man sich gar nicht einfallen ließ, daß Sinne auch a priori anschauen sollten. [^376-1]: Der Recensent schlägt sich mehrentheils mit seinem eigenen Schatten. Wenn ich die Wahrheit der Erfahrung dem Traum entgegensetze, so denkt er gar nicht daran, daß hier nur von dem bekannten somnio objective sumto der *Wolffischen* Philosophie die Rede sei, der bloß formal ist, und wobei es auf den Unterschied des Schlafens und Wachens gar nicht angesehen ist und in einer Transscendentalphilosophie auch nicht gesehen werden kann. Übrigens nennt er meine Deduction der Kategorien und die Tafel der Verstandesgrundsätze: „gemein bekannte Grundsätze der Logik und Ontologie, auf idealistische Art ausgedrückt.“ Der Leser darf nur darüber diese Prolegomenen nachsehen, um sich zu überzeugen, daß ein elenderes und selbst historisch unrichtigeres Urtheil gar nicht könne gefällt werden. ## Translation — Mouseia > 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. ### Title Prolegomena to Any Future Metaphysics That Will Be Able to Present Itself *as a Science* by Immanuel Kant. ### Preface These prolegomena are intended not for apprentices but for future teachers, and even for them their purpose is not to arrange the presentation of a science already in existence, but to discover that science in the first place. There are scholars for whom the history of philosophy, ancient as well as modern, is itself their philosophy; these prolegomena were not written for them. They must wait until those who seek to draw from the springs of reason itself have settled their business; then it will be their turn to inform the world of what has happened. Otherwise nothing can be said that, in their opinion, has not been said before. Indeed, this may count as an infallible prediction of everything to come: since the human understanding has ranged fancifully over countless objects in many different ways for many centuries, something old can scarcely fail to be found that bears some resemblance to every novelty. My purpose is to convince all those who think metaphysics worth pursuing that they must necessarily suspend their work for the present, regard everything done so far as undone, and, before all else, first raise the question: is anything like metaphysics possible at all? If it is a science, how is it that, unlike other sciences, it cannot win universal and lasting acceptance? If it is not, how does it keep boasting under the guise of a science, holding out to the human understanding hopes that never die but are never fulfilled? Whether we demonstrate our knowledge or our ignorance, something certain must at last be established about the nature of this self-appointed science; it cannot possibly go on as it has. It seems almost laughable that, while every other science advances without cease, this one—which claims to be wisdom itself, whose oracle everyone consults—keeps turning on the same spot without advancing a single step. Its adherents, too, have dwindled greatly. Those who feel strong enough to shine in other sciences show no wish to risk their reputations in this one, where everyone, however ignorant in all other matters, presumes to pronounce a decisive judgment, because in this realm there are still no reliable weights and measures for distinguishing sound work from shallow chatter. Yet it is hardly unheard of that, after long work on a science, when people imagine how wonderfully far they have advanced, someone should at last think to ask whether and how such a science is possible at all. Human reason is so eager to build that it has repeatedly erected its tower, only to tear it down again afterward to see what its foundations might be like. It is never too late to become reasonable and wise; but when insight comes late, putting it to work is always harder. To ask whether a science is even possible presupposes doubt about its actuality. Such doubt, however, offends anyone whose entire possessions may consist of this supposed jewel; and whoever lets such a doubt escape must be prepared for resistance from every side. Some, proudly conscious of their ancient—and therefore, they believe, rightful—possession, will look down on him with contempt, their metaphysical compendiums in hand. Others, who see nothing anywhere except what is identical with something they have already seen elsewhere, will not understand him. For a time everything will remain as though nothing had happened to make anyone fear or hope for an approaching change. Nevertheless, I venture to predict that the reader of these prolegomena who thinks for himself will not merely doubt his previous science. He will eventually become entirely convinced that no such science can exist unless the requirements stated here, on which its possibility depends, are fulfilled; and since this has never yet happened, that there is as yet no metaphysics at all. Yet the demand for it can never disappear,[^257-1] because the interests of universal human reason are too intimately bound up with it. He will therefore acknowledge that a complete reform—or rather a new birth according to a plan wholly unknown until now—is inevitably approaching, however much people may resist it for a while. Since the attempts of *Locke* and *Leibniz*, or rather since the origin of metaphysics as far back as its history reaches, nothing has happened that could have been more decisive for the fate of this science than *David Hume’s* attack on it. He brought no light to this kind of cognition, but he did strike a spark from which a light might well have been kindled, had it met receptive tinder whose glow was carefully tended and increased. *Hume* began chiefly with a single but important concept of metaphysics, namely the *connection of cause and effect* (and thus also its derivative concepts of force and action, and so forth). He challenged reason, which claims to have brought this concept forth from its own womb, to account for the right by which it thinks that something can be so constituted that, once it is posited, something else must necessarily be posited through it as well; for this is what the concept of a cause says. He proved beyond dispute that reason is wholly incapable of thinking such a connection a priori and from concepts, since it contains necessity; yet there is simply no seeing why, because something is, something else must necessarily be as well, or how the concept of such a connection can therefore be introduced a priori. From this he concluded that reason is entirely deceived in this concept, falsely taking it for its own child when it is nothing but a bastard of the imagination: made pregnant by experience, the imagination has brought certain representations under the law of association and passed off the subjective necessity arising from this—that is, habit—as an objective necessity arising from insight. From this he concluded that reason has no capacity even to think such connections in general, since its concepts would then be mere inventions; and all its allegedly a priori cognitions would be nothing but ordinary experiences bearing a counterfeit stamp. This amounts to saying that there is no metaphysics at all, nor can there be any.[^258-1] However hasty and mistaken his conclusion, it was at least founded on an investigation; and that investigation well deserved the combined efforts of the able minds of his time to solve the problem, if possible more successfully, in the sense in which he presented it. A complete reform of the science would then soon have had to follow. But the fate that has always been unkind to metaphysics decreed that no one should understand him. It is painful to see how completely his opponents—*Reid*, *Oswald*, *Beattie*, and finally *Priestley*—missed the point of his problem. They constantly took for granted precisely what he doubted, while vehemently and, for the most part, with great immodesty proving what he had never thought of doubting. They so misunderstood his hint toward improvement that everything remained as it had been, as though nothing had happened. The question was not whether the concept of a cause was correct, useful, and indispensable for the whole cognition of nature; *Hume* had never doubted this. It was whether this concept is thought a priori through reason and thus possesses an inner truth independent of all experience, and consequently a usefulness that may extend further and is not confined merely to objects of experience. This was the point on which *Hume* expected clarification. The issue was solely the origin of the concept, not its indispensability in use. Once the former had been established, the conditions of its use and the range within which it could be valid would have followed of themselves. To do justice to the problem, however, the celebrated man’s opponents would have had to penetrate very deeply into the nature of reason insofar as it concerns itself solely with pure thinking, and this did not suit them. They therefore invented a more convenient means of defiance without any insight at all: an appeal to *common sense*. It is indeed a great gift of heaven to possess a sound—or, as it has recently been called, plain—human understanding. But one must prove it by deeds, by what is considered and reasonable in one’s thoughts and words, not by appealing to it as an oracle whenever one has nothing intelligent to offer in one’s defense. To appeal to common sense when insight and science run out, and not before, is one of the subtle inventions of modern times, allowing the most insipid chatterer confidently to confront the most thorough thinker and hold his own. Yet as long as even a small remnant of insight remains, one will take care not to resort to this expedient. Seen clearly, this appeal is nothing but an appeal to the judgment of the crowd: applause that makes the philosopher blush, while the popular wit triumphs and puts on a bold face. I should have thought, however, that *Hume* could claim a sound understanding just as well as *Beattie*, and something more that the latter certainly did not possess: a critical reason that keeps common understanding within bounds, preventing it from soaring into speculation or, where speculation alone is at issue, from claiming to decide anything, since it cannot justify its own principles. Only in this way will it remain a sound understanding. Chisel and mallet may serve very well for working a piece of timber, but copper engraving requires an etching needle. Sound understanding and speculative understanding are both useful, then, but each in its own way: the former when judgments find their immediate application in experience, the latter when judgments are to be made universally, from mere concepts—for example in metaphysics, where what calls itself sound understanding, often per antiphrasin, or by an opposite name, has no judgment whatever. I freely confess: it was *David Hume’s* reminder that first, many years ago, interrupted my dogmatic slumber and gave my investigations in speculative philosophy an entirely different direction. I was far from accepting his conclusions, which arose only because he did not represent his problem to himself as a whole but struck upon just one part of it, a part that cannot provide an answer unless the whole is taken into account. If we begin with a well-founded though undeveloped thought left to us by another, continued reflection may give us hope of advancing further than the penetrating thinker to whom we owed the first spark of this light. I therefore first tried to see whether *Hume’s* objection could be represented universally, and soon found that the concept of the connection of cause and effect was far from the only one through which the understanding thinks connections of things a priori; rather, metaphysics consists entirely of such concepts. I sought to ascertain their number, and when I had succeeded as I wished—namely, from a single principle—I proceeded to the deduction of these concepts, of which I was now certain that they were not derived from experience, as *Hume* had feared, but sprang from pure understanding. This deduction, which seemed impossible to my acute predecessor, and which no one but he had even thought of, although everyone confidently used these concepts without asking what grounded their objective validity—this, I say, was the most difficult undertaking that could ever have been attempted on behalf of metaphysics. Worst of all, metaphysics, whatever of it might exist anywhere, could not give me the slightest help, since that deduction must first establish the possibility of metaphysics. Once I had succeeded in solving Hume’s problem, not merely in one particular case but with respect to the entire faculty of pure reason, I could take secure, though still only slow, steps toward finally determining the whole extent of pure reason, both its boundaries and its content, completely and according to universal principles. This was precisely what metaphysics needed in order to construct its system according to a reliable plan. But I fear that the *development* of Hume’s problem in its greatest possible extension—namely, the Critique of Pure Reason—may meet the same fate as the *problem* itself when it was first presented. People will judge it wrongly because they do not understand it; they will not understand it because they are inclined to leaf through the book, but not to think it through; and they will not wish to expend this effort because the work is dry, obscure, contrary to all their accustomed concepts, and lengthy besides. Now I confess that I did not expect to hear a philosopher complain of a lack of popular appeal, entertainment, and ease when the issue is the very existence of a celebrated cognition indispensable to humanity, which can be established only by the strictest rules of scholarly precision. Popular appeal may follow these in time, but must never come first. Yet the complaint about a certain obscurity is just: this arises in part from the breadth of the plan, which makes it difficult to survey the principal points on which the investigation turns. I shall remedy this through the present *Prolegomena*. That work, which presents the faculty of pure reason in its entire extent and within its boundaries, nevertheless remains the foundation; the Prolegomena relate to it only as preliminary exercises. For that critique must stand as a science, systematic and complete down to its smallest parts, before anyone can think of bringing metaphysics forward, or even entertain a distant hope of it. People have long been accustomed to seeing old, worn-out cognitions dressed up anew by removing them from their former connections and fitting them with a systematic costume, arbitrarily cut but bearing new titles; and most readers will expect nothing else from that critique in advance. These prolegomena, however, will lead them to see that it is an entirely new science, of which no one had previously formed even the thought, whose very idea was unknown, and for which nothing previously available could be used except the hint supplied by *Hume’s* doubts. He likewise had no inkling of such a possible formal science; to bring his ship to safety, he ran it onto the beach—skepticism—where it may lie and rot. My concern, by contrast, is to give it a pilot who, equipped with a complete sea chart and a compass, can steer the ship safely wherever he sees fit, according to reliable principles of navigation drawn from knowledge of the globe. To approach a new science, entirely isolated and unique of its kind, with the prejudice that one can judge it by the supposed knowledge one has already acquired elsewhere—even though the reality of precisely that knowledge must first be called wholly into doubt—achieves nothing but the belief that one sees everywhere what one already knew, perhaps because the expressions sound similar. Everything must then appear utterly distorted, absurd, and gibberish, since one is taking as one’s basis not the author’s thoughts but always one’s own way of thinking, which long habit has made second nature. Yet the length of the work, insofar as it is grounded in the science itself rather than its presentation, together with its unavoidable dryness and scholastic precision, are qualities that may be exceedingly advantageous to the subject itself but must certainly be detrimental to the book. Not everyone, admittedly, is gifted with the ability to write as subtly and yet as engagingly as *David Hume*, or as thoroughly and yet as elegantly as *Moses Mendelssohn*. But I could, I flatter myself, have given my presentation popular appeal had I been concerned merely to sketch a plan and commend its execution to others, and had the welfare of the science that occupied me so long not been close to my heart. Apart from that, it took much perseverance and no little self-denial to set aside the lure of an early, favorable reception for the prospect of an acceptance that, though late, would endure. *Making plans* is often a lavish, boastful occupation of the mind, by which one gives oneself the air of creative genius: demanding what one cannot oneself accomplish, censuring what one cannot improve, and proposing what one does not oneself know where to find. Even a competent plan for a general critique of reason, however, would have required somewhat more than might be supposed, unless it was to become, as usual, a mere declamation of pious wishes. But pure reason is a sphere so separate, so thoroughly interconnected within itself, that no part of it can be touched without touching all the rest, and nothing can be accomplished without first determining each part’s place and its influence on the others. For there is nothing outside it that could correct our judgment within it, and so the validity and use of every part depend on its relation to the others within reason itself, just as in the structure of an organized body the purpose of each member can be derived only from the complete concept of the whole. We may therefore say of such a critique that it is never reliable unless it is *whole* and *completed* down to the least elements of pure reason, and that in the sphere of this faculty we must determine and establish either *everything* or *nothing*. Although a mere plan preceding the *Critique* of Pure Reason would be unintelligible, unreliable, and useless, a plan following it is all the more useful. It enables us to survey the whole, examine separately the principal points on which this science turns, and arrange much in the presentation better than was possible in the first execution of the work. Here, then, is such a *plan* after the completion of the work. It may now be laid out according to the *analytic method*, whereas the work itself had to be composed entirely according to the *synthetic manner of exposition*, so that the science would display all its articulations in their natural connection, as the structure of a quite distinctive faculty of cognition. Anyone who finds even this plan obscure, which I put forward as prolegomena to all future metaphysics, should consider that not everyone needs to study metaphysics; that many a talent thrives in rigorous and even profound sciences that lie closer to intuition, yet cannot succeed in investigations conducted through abstract concepts alone, and in that case must devote its powers of mind to another object; but that whoever undertakes to judge metaphysics, or indeed to compose one, must satisfy the demands made here in full, whether by accepting my solution or by thoroughly refuting it and putting another in its place—for he cannot dismiss them. Finally, he should consider that the much-decried obscurity—a customary cloak for one’s own indolence or shortsightedness—also has its use. All those who maintain a cautious silence about every other science speak with mastery and decide boldly on questions of metaphysics, because here their ignorance does not, to be sure, stand out clearly against the knowledge of others; but it does stand out against genuine critical principles, of which we may therefore boast: ignavum, fucos, pecus a praesepibus arcent. *Virg.* They drive the drones, that idle herd, away from the hives. [^257-1]: Rusticus exspectat, dum defluat amnis; at ille Labitur et labetur in omne volubilis aevum. *Horat.* The countryman waits for the river to finish flowing; but it flows and will flow, rolling on through all eternity. [^258-1]: Yet *Hume* himself called this very destructive philosophy metaphysics, and set a high value on it. “Metaphysics and morals,” he says (Essays, Part 4, page 214, German translation), “are the most important branches of science; mathematics and natural science are not worth half as much.” Here, however, this acute thinker saw only the negative benefit that moderating the excessive claims of speculative reason would have in putting a complete end to so many endless, persecuting disputes that trouble the human race. In doing so, he lost sight of the positive harm that results when reason is deprived of its most important prospects, which alone enable it to set before the will the highest goal of all its endeavors. ### Preliminary reminder Prolegomena. Preliminary Reminder Concerning *What Is Distinctive in All Metaphysical Cognition.* ### Section 1 § 1. *On the Sources of Metaphysics.* If we wish to present a cognition as a *science*, we must first be able to determine precisely what distinguishes it, what it shares with no other, and what is therefore *distinctive* to it. Otherwise the boundaries of all the sciences run together, and none can be treated thoroughly according to its nature. Whether this distinctive character consists in a difference of *object*, of *sources of cognition*, of *kind of cognition*, or of some if not all of these together, it is on this that the idea of the possible science and its territory first depends. First, as regards the *sources* of metaphysical cognition, its very concept already entails that they cannot be empirical. Its principles—which include not only its fundamental propositions but also its fundamental concepts—must therefore never be drawn from experience; for it is to be not physical but metaphysical cognition, that is, cognition lying beyond experience. Hence neither outer experience, the source of physics proper, nor inner experience, the foundation of empirical psychology, will underlie it. It is therefore cognition a priori, or cognition from pure understanding and pure reason. In this, however, it would have nothing to distinguish it from pure mathematics; it must therefore be called *pure philosophical* cognition. For the meaning of this expression I refer to the Critique of Pure Reason, page 712 and following, where the distinction between these two ways of using reason has been presented clearly and satisfactorily.—So much for the sources of metaphysical cognition. ### Section 2 § 2. *On the Kind of Cognition That Alone Can Be Called Metaphysical.* a) *On the Distinction between Synthetic and Analytic Judgments in General.* Metaphysical cognition must contain nothing but judgments a priori: the distinctive character of its sources requires this. Yet whatever their origin or logical form, judgments differ in content in a way that makes them either merely *explicative*, adding nothing to the content of cognition, or *ampliative*, increasing the given cognition. The former may be called *analytic*, the latter *synthetic* judgments. Analytic judgments say nothing in the predicate except what was already actually thought in the concept of the subject, though not so clearly or with the same awareness. If I say, “All bodies are extended,” I have not enlarged my concept of body in the least, but have merely analyzed it: even before the judgment, extension was actually thought in that concept, though not expressly stated. The judgment is therefore analytic. By contrast, the proposition “Some bodies are heavy” contains something in its predicate that is not actually thought in the general concept of body. It therefore increases my cognition by adding something to my concept, and must accordingly be called a synthetic judgment. b) *The Common Principle of All Analytic Judgments Is the Principle of Contradiction.* All analytic judgments rest entirely on the principle of contradiction and are by their nature cognitions a priori, whether or not the concepts serving as their material are empirical. For since the predicate of an affirmative analytic judgment is already thought in the concept of the subject, it cannot be denied of that subject without contradiction. Likewise, in an analytic but negative judgment, its opposite is necessarily denied of the subject, again by the principle of contradiction. This is how matters stand with the propositions “Every body is extended” and “No body is unextended (simple).” For this very reason all analytic propositions are also judgments a priori, even when their concepts are empirical: for example, “Gold is a yellow metal.” To know this I need no further experience beyond my concept of gold, which contained the thought that this body is yellow and is a metal. This constituted my very concept, and I needed to do nothing but analyze it, without looking beyond it for anything else. c) *Synthetic Judgments Require a Principle Other Than the Principle of Contradiction.* There are synthetic judgments a posteriori, whose origin is empirical; but there are also synthetic judgments that are certain a priori and spring from pure understanding and reason. Both agree, however, in never being able to arise from the principle of analysis—the principle of contradiction—alone. They require yet another, quite different principle, although, from whatever principle they are derived, they must always be derived *in accordance with* the principle of contradiction. For nothing may contradict this principle, although not everything can be derived from it. I shall first divide synthetic judgments into classes. 1) *Judgments of experience* are always synthetic. It would be absurd to base an analytic judgment on experience, since I need not go beyond my concept at all to make the judgment, and therefore need no testimony from experience. That a body is extended is a proposition established a priori, not a judgment of experience. Before I turn to experience, I already have all the conditions for my judgment in the concept, from which I need only extract the predicate according to the principle of contradiction. In doing so I can also become conscious of the judgment’s *necessity*, which experience could not even teach me. 2) *Mathematical judgments* are all synthetic. This proposition seems hitherto to have wholly escaped the notice of those who analyze human reason, and indeed to run directly counter to all their conjectures, although it is indisputably certain and of great importance for what follows. Because it was found that all the inferences of mathematicians proceed according to the principle of contradiction—as the nature of all apodictic certainty requires—people persuaded themselves that the fundamental propositions too were known from the principle of contradiction. In this they were greatly mistaken. A synthetic proposition can indeed be discerned according to the principle of contradiction, but only if another synthetic proposition is presupposed from which it can be inferred; never in itself. First we must note that properly mathematical propositions are always judgments a priori, not empirical, because they carry necessity with them, and this cannot be drawn from experience. If anyone refuses to grant me this, then I shall restrict my assertion to *pure mathematics*, whose very concept entails that it contains not empirical cognition but only pure cognition a priori. At first one might well think that the proposition 7 + 5 = 12 is merely analytic, following from the concept of a sum of seven and five according to the principle of contradiction. On closer examination, however, we find that the concept of the sum of 7 and 5 contains nothing more than the union of the two numbers into a single one; it does not in the least specify what this single number embracing both is. The concept of twelve is by no means already thought when I merely think of that union of seven and five. However long I analyze my concept of such a possible sum, I shall never find twelve in it. We must go beyond these concepts by calling on the intuition corresponding to one of the two—say, our five fingers, or five points, as *Segner* does in his arithmetic—and thus gradually add the units of the five given in intuition to the concept of seven. We therefore really enlarge our concept through the proposition 7 + 5 = 12, adding to the first concept a new one not thought in it at all. That is, an arithmetical proposition is always synthetic. This becomes all the clearer if we take somewhat larger numbers: it is then evident that, however we turn and twist our concept, we could never find the sum by merely analyzing our concepts without calling on intuition. Nor is any fundamental proposition of pure geometry analytic. That the straight line between two points is the shortest is a synthetic proposition. For my concept of straightness contains nothing of magnitude, but only a quality. The concept of the shortest is therefore added in its entirety, and cannot be drawn from the concept of the straight line by any analysis. Here, then, we must call on intuition, through which alone the synthesis is possible. Some other fundamental propositions presupposed by geometers are indeed analytic and rest on the principle of contradiction; but, like identical propositions, they serve only as links in the method, not as principles: for example, a = a, the whole is equal to itself, or (a + b) > a, that is, the whole is greater than its part. Yet even these, although valid according to mere concepts, are admitted in mathematics only because they can be exhibited in intuition. What commonly makes us believe here that the predicate of such apodictic judgments already lies in our concept, and that the judgment is therefore analytic, is merely an ambiguity of expression. We *ought*, namely, to add a certain predicate in thought to a given concept, and this necessity already attaches to the concepts. But the question is not what we *ought to think in addition* to the given concept, but what we *actually*, even if only obscurely, *think within it*. Here it becomes clear that the predicate attaches to that concept necessarily indeed, but not immediately: rather, through an intuition that must be added. ### Section 3 § 3. Remark *On the General Division of Judgments into Analytic and Synthetic.* This division is indispensable for the critique of the human understanding, and therefore deserves to become *canonical* within it; otherwise I know of no considerable use it would have anywhere else. Here, too, I find the reason why dogmatic philosophers—who always sought the sources of metaphysical judgments only within metaphysics itself, and not outside it in the laws of pure reason in general—neglected this division, which seems to offer itself of its own accord. They could thus, like the celebrated *Wolff* or the acute *Baumgarten* who followed in his footsteps, seek in the principle of contradiction a proof of the principle of sufficient reason, which is plainly synthetic. In *Locke’s* Essay Concerning Human Understanding, by contrast, I already find a hint of this division. For in Book Four, Chapter Three, § 9 and following—having already spoken earlier of the different connections of representations in judgments and of their sources, placing one in identity or contradiction (analytic judgments) and the other in the existence of representations in a subject (synthetic judgments)—he admits in § 10 that our cognition (a priori) of the latter is very limited, and almost nothing. Yet what he says of this kind of cognition is so little determined or brought under rules that we should not be surprised that no one, notably not even Hume, took it as an occasion to reflect on propositions of this kind. For principles that are at once so universal and so determinate are not readily learned from others who had only a dim intimation of them. We must first have arrived at them through our own reflection; afterward we find them elsewhere too, where we certainly would not have encountered them first, since the authors themselves did not even know that such an idea underlay their own observations. Those who never think for themselves nevertheless possess the acumen to ferret out everything, once it has been shown to them, in what was said before, where no one could previously see it. ### General questions The General Question of the Prolegomena: Is Metaphysics Possible at All? ### Section 4 § 4. If a metaphysics capable of maintaining its standing as a science actually existed; if we could say, “Here is metaphysics: you need only learn it, and it will convince you of its truth irresistibly and permanently,” then this question would be needless. Only the further question would remain, one that concerns a test of our acumen rather than a proof of the thing’s existence: *how it is possible*, and how reason sets about attaining it. But in this case human reason has not been so fortunate. We cannot point to a single book, as we might present a work of *Euclid*, and say: this is metaphysics; here you will find the chief aim of this science, cognition of a highest being and a future world, demonstrated from principles of pure reason. We can indeed be shown many propositions that are apodictically certain and have never been disputed; but they are all analytic, and concern the materials and building supplies for metaphysics rather than the enlargement of cognition that ought to be our proper purpose in pursuing it (§ 2, letter c). Even if you also present synthetic propositions—for example, the principle of sufficient reason—which you have never proved from reason alone, hence a priori, as was your duty, but which people nevertheless readily grant you, you still fall into such untenable and uncertain assertions when you try to use them for your principal aim that at all times one metaphysics has contradicted another, either in its assertions themselves or in their proofs, and thereby destroyed its own claim to lasting acceptance. Indeed, attempts to bring such a science into being were undoubtedly the first cause of skepticism’s early appearance: a way of thinking in which reason acts so violently against itself that it could have arisen only in complete despair of satisfying its most important aims. For long before people began to question nature methodically, they questioned only their own isolated reason, which ordinary experience had already exercised to some extent. They did so because reason is always present to us, whereas the laws of nature must generally be sought with difficulty. And so metaphysics floated on the surface like foam: as soon as what had been scooped up dissolved, more appeared on the surface, always eagerly gathered by some, while others, instead of seeking in the depths the cause of this appearance, thought themselves wise for laughing at the former’s fruitless labor. The essential feature distinguishing pure *mathematical* cognition from all other cognition a priori is that it must proceed *not from concepts* at all, but always only through the construction of concepts (Critique, page 713). Since in its propositions it must therefore go beyond the concept to what is contained in the intuition corresponding to it, its propositions can never—and should never—arise through the analysis of concepts, that is, analytically; hence they are all synthetic. I cannot, however, refrain from noting the harm done to philosophy by neglecting this otherwise easy and seemingly insignificant observation. When *Hume* felt the calling, worthy of a philosopher, to cast his eyes over the whole field of pure cognition a priori, in which the human understanding lays claim to such vast possessions, he thoughtlessly cut off an entire province—indeed, the most important one—namely pure mathematics. He imagined that its nature and, so to speak, its constitution rested on quite different principles, namely on the principle of contradiction alone. Although he did not divide propositions as formally or universally, or under the same names, as I do here, it amounted to his saying: pure mathematics contains only *analytic* propositions, but metaphysics contains synthetic propositions a priori. In this he was greatly mistaken, and the mistake had decisively harmful consequences for his whole conception. Had he not made it, he would have extended his question about the origin of our synthetic judgments far beyond his metaphysical concept of causality, to include the possibility of mathematics a priori; for he would have had to acknowledge mathematics as synthetic as well. But then he could by no means have grounded his metaphysical propositions on mere experience, since he would otherwise have subjected the axioms of pure mathematics to experience too, and he had far too much insight to do that. The good company in which metaphysics would then have found itself would have secured it against the danger of contemptuous mistreatment: the blows intended for metaphysics would also have had to strike mathematics, which was not, and could not have been, his intention. Thus this acute thinker would have been drawn into reflections that would have had to resemble those occupying us now, but which would have gained immeasurably from his inimitably beautiful presentation. *Properly metaphysical* judgments are all synthetic. We must distinguish judgments *belonging to metaphysics* from judgments that are properly *metaphysical*. Among the former, very many are analytic; but they constitute only the means to metaphysical judgments, toward which the science’s aim is wholly directed, and which are always synthetic. For if concepts belong to metaphysics—for example, the concept of substance—then judgments arising from their mere analysis necessarily belong to metaphysics as well: for example, substance is that which exists only as subject, and so forth. By means of several such analytic judgments we seek to approach the definition of the concepts. But the analysis of a pure concept of the understanding, such as metaphysics contains, proceeds no differently from the analysis of any other concept, including an empirical one not belonging to metaphysics—for example, air is an elastic fluid whose elasticity is not abolished by any known degree of cold. Thus the concept, but not the analytic judgment, is distinctively metaphysical. For this science has something special and distinctive in the generation of its cognitions a priori, which must therefore be distinguished from what it shares with all other cognitions of the understanding. Thus, for example, the proposition “Everything in things that is substance is permanent” is a synthetic and distinctively metaphysical proposition. Once the concepts a priori that constitute the material and building supplies of metaphysics have been assembled according to certain principles, the analysis of these concepts is of great value. It can also be presented separately as a special part—a kind of philosophia definitiva, or philosophy of definitions—containing only analytic propositions belonging to metaphysics, apart from all the synthetic propositions that constitute metaphysics itself. For in fact these analyses have no considerable use anywhere except in metaphysics, that is, with a view to the synthetic propositions that are to be generated from those concepts after their initial analysis. The conclusion of this section, then, is that metaphysics is properly concerned with synthetic propositions a priori, and these alone constitute its aim. For this it certainly needs many analyses of its concepts, hence analytic judgments; but the procedure here is no different from that in any other kind of cognition where we seek merely to make our concepts distinct by analyzing them. The generation of cognition a priori, however, both in intuition and in concepts, and finally also of synthetic propositions a priori—and specifically within philosophical cognition—constitutes the essential content of metaphysics. Weary, then, of a dogmatism that teaches us nothing, and at the same time of a skepticism that promises us nothing at all, not even the repose of a permissible ignorance; summoned by the importance of the cognition we need, and made distrustful by long experience of every cognition we believe we possess, or that offers itself under the title of pure reason, we are left with only one critical question, whose answer can guide our future conduct: *Is metaphysics possible at all?* But this question must be answered not through skeptical objections to particular assertions of an actual metaphysics—for at present we acknowledge none—but from the still merely *problematic* concept of such a science. In the *Critique of Pure Reason* I proceeded synthetically with regard to this question: I investigated pure reason itself and sought, in this very source, to determine according to principles both the elements and the laws of its pure use. This work is difficult. It requires a resolute reader to think his way gradually into a system that takes nothing as given for its foundation except reason itself, and thus seeks, without relying on any fact, to unfold cognition from its original seeds. *Prolegomena*, by contrast, are to be preliminary exercises; they should indicate what must be done to bring a science into actuality if possible, rather than present the science itself. They must therefore rest on something already known to be reliable, from which we can confidently set out and ascend to sources not yet known, whose discovery will not only explain what we knew but also display a range of many cognitions all springing from those same sources. The methodical procedure of prolegomena, particularly those intended to prepare for a future metaphysics, will therefore be *analytic*. Fortunately, however, although we cannot assume that metaphysics as a science is *actual*, we can confidently say that certain pure synthetic cognitions a priori are actual and given: namely, *pure mathematics* and *pure natural science*. Both contain propositions universally acknowledged, some as apodictically certain through reason alone, others through the universal agreement of experience, yet throughout as independent of experience. We therefore possess at least some *undisputed* synthetic cognition a priori, and need not ask whether it is possible—for it is actual—but only *how it is possible*, so that from the principle of the possibility of what is given we may also derive the possibility of all the rest. Prolegomena. General Question: How Is Cognition from Pure Reason Possible? ### Section 5 § 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. ### First part: How is pure mathematics possible? The Principal Transcendental Question First Part. How Is Pure Mathematics Possible? ### Section 6 § 6. Here, then, is a great and proven body of cognition, already admirable in extent and promising boundless expansion in the future. Throughout, it carries apodictic certainty—that is, absolute necessity—and thus rests on no grounds of experience; it is therefore a pure product of reason, yet is also synthetic through and through. “How, then, is it possible for human reason to bring about such cognition wholly a priori?” Since this capacity does not and cannot rest on experiences, does it not presuppose some ground of cognition a priori, lying deeply hidden, which might reveal itself through these effects if only we diligently traced them to their first beginnings? ### Section 7 § 7. We find, however, that all mathematical cognition has this distinctive feature: it must first exhibit its concept *in intuition*, and do so a priori, hence in an intuition that is not empirical but pure; without this means it cannot take a single step. Its judgments are therefore always *intuitive*, whereas philosophy must content itself with *discursive* judgments, *from mere concepts*, and may indeed elucidate its apodictic teachings through intuition, but can never derive them from it. This observation about the nature of mathematics already guides us to the first and highest condition of its possibility: some *pure intuition* must underlie it, in which it can exhibit all its concepts in concreto and yet a priori—or, as it is called, *construct* them.[^281-1] If we can discover this pure intuition and the possibility of such an intuition, it will be easy to explain how synthetic propositions a priori are possible in pure mathematics, and thus how this science itself is possible. For just as empirical intuition makes it possible without difficulty for us synthetically to enlarge, in experience, the concept we form of an object of intuition by adding new predicates supplied by intuition itself, so pure intuition will do the same. There is only this difference: in the latter case the synthetic judgment will be certain a priori and apodictic, whereas in the former it will be certain only a posteriori and empirically. For the empirically certain judgment contains only what is encountered in contingent empirical intuition, whereas the a priori judgment contains what must necessarily be encountered in pure intuition, since this intuition, as intuition a priori, is inseparably connected with the concept *before all experience* or particular perception. [^281-1]: See Critique, page 713. ### Section 8 § 8. Yet at this step the difficulty seems to grow rather than diminish. For the question now reads: *how is it possible to intuit something a priori?* Intuition is a representation of the sort that would depend immediately on the presence of the object. It therefore seems impossible to intuit a priori *originally*, since the intuition would then have to occur without an object, present either previously or now, to which it referred, and thus could not be intuition. Concepts, to be sure, are such that we can quite well form some of them a priori—namely, those containing only the thinking of an object in general—without standing in an immediate relation to the object: for example, the concept of magnitude, of cause, and so forth. But even these require a certain use in concreto to give them significance and sense: that is, application to some intuition through which an object corresponding to them is given to us. Yet how can intuition of the object precede the object itself? ### Section 9 § 9. If our intuition had to be of such a kind that it represented things *as they are in themselves*, no intuition a priori could occur; it would always be empirical. For I can know what is contained in the object in itself only when it is present and given to me. Even then, admittedly, it is incomprehensible how the intuition of a present thing should allow me to cognize it as it is in itself, since its properties cannot migrate into my power of representation. But even granting this possibility, such intuition would not occur a priori, that is, before the object was represented to me; for without this no ground could be conceived for my representation’s relation to it, unless that representation rested on inspiration. There is therefore only one way in which my intuition can precede the actuality of the object and occur as cognition a priori: *if it contains nothing but the form of sensibility, which in me as subject precedes all the actual impressions through which I am affected by objects*. For I can know a priori that objects of the senses can be intuited only in accordance with this form of sensibility. It follows that propositions concerning this form of sensible intuition alone will be possible and valid for objects of the senses; and conversely, that intuitions possible a priori can never concern anything other than objects of our senses. ### Section 10 § 10. It is therefore only through the form of sensible intuition that we can intuit things a priori; but through it we cognize objects only as they can *appear* to us—to our senses—not as they may be in themselves. This presupposition is absolutely necessary if synthetic propositions a priori are to be admitted as possible or, where they are actually found, if their possibility is to be understood and determined in advance. Now space and time are the intuitions on which pure mathematics bases all its cognitions and judgments, which at the same time present themselves as apodictic and necessary. For mathematics must first exhibit all its concepts in intuition, and pure mathematics in pure intuition—that is, it must construct them. Without this, it is impossible for mathematics to take a step, because it can proceed not analytically, by analyzing concepts, but synthetically: so long as it lacks pure intuition, it lacks the sole medium in which the material for synthetic judgments a priori can be given. Geometry takes the pure intuition of space as its foundation. Arithmetic itself produces its concepts of number by the successive addition of units in time; above all, pure mechanics can produce its concepts of motion only by means of the representation of time. Both representations, however, are merely intuitions. For if we leave out of the empirical intuitions of bodies and their alterations—motion—everything empirical, namely what belongs to sensation, space and time still remain. They are therefore pure intuitions, underlying those empirical intuitions a priori, and for this reason can never themselves be left out. But precisely because they are pure intuitions a priori, they prove that they are mere forms of our sensibility, which must precede all empirical intuition, that is, the perception of actual objects, and in accordance with which objects can be cognized a priori—though, to be sure, only as they appear to us. ### Section 11 § 11. The problem of the present division is therefore solved. Pure mathematics is possible as synthetic cognition a priori only because it concerns no objects other than mere objects of the senses, whose empirical intuition is grounded a priori in a pure intuition—of space and time. This intuition can underlie it because it is nothing but the mere form of sensibility, which precedes the actual appearance of objects by first making that appearance possible. Yet this capacity to intuit a priori concerns not the matter of appearance—that is, what in it is sensation, for this constitutes the empirical—but only its form, space and time. If anyone wished to doubt in the slightest that these are determinations attaching not at all to things in themselves, but merely to their relation to sensibility, I should like to know how he can find it possible to know a priori, and thus before all acquaintance with things, before they are given to us, what their intuition must be like—as is the case here with space and time. But this is entirely comprehensible as soon as the two are taken to be nothing more than formal conditions of our sensibility, and the objects merely appearances. For then the form of appearance, that is, pure intuition, can indeed be represented from ourselves, that is, a priori. ### Section 12 § 12. To add something by way of elucidation and confirmation, we need only consider the usual and indispensably necessary procedure of geometers. All proofs of the complete equality of two given figures—where one can be substituted for the other in every respect—ultimately come down to their coinciding with one another. This is plainly nothing but a synthetic proposition resting on immediate intuition; and this intuition must be pure and given a priori, for otherwise that proposition could not count as apodictically certain, but would have only empirical certainty. It would mean only that this is always observed to be so, and would hold only as far as our perception had hitherto extended. That complete space—which is itself no longer the boundary of another space—has three dimensions, and that space in general can have no more, is based on the proposition that no more than three lines can intersect at right angles at a single point. But this proposition cannot be demonstrated from concepts at all: it rests immediately on intuition, indeed pure intuition a priori, because it is apodictically certain. That we can demand that a line be drawn on indefinitely (in indefinitum), or that a series of alterations—for example, spaces traversed by motion—be continued indefinitely, presupposes a representation of space and time that can attach only to intuition, namely insofar as intuition is in itself bounded by nothing; for it could never be inferred from concepts. Pure intuitions a priori therefore really do underlie mathematics and make possible its synthetic propositions with their apodictic validity. Hence our transcendental deduction of the concepts in space and time also explains the possibility of a pure mathematics. Without such a deduction, and without our assuming that “everything that may be given to our senses—to the outer senses in space, to the inner sense in time—is intuited by us only as it appears to us, not as it is in itself,” that possibility could indeed be conceded, but by no means understood. ### Section 13 § 13. Those who still cannot free themselves from the notion that space and time are actual characteristics attaching to things in themselves may exercise their acumen on the following paradox. If their attempts to solve it prove fruitless, they may, free of prejudice for at least a few moments, suspect that there might after all be grounds for demoting space and time to mere forms of our sensible intuition. If two things are completely identical in every respect that can be cognized in each taken by itself—in all determinations belonging to magnitude and quality—it must surely follow that one can be substituted for the other in every case and relation without this exchange causing the slightest discernible difference. This is indeed so with plane figures in geometry. But various spherical figures, despite that complete inner agreement, nevertheless show such a difference in their external relation that one cannot be substituted for the other at all. For example, two spherical triangles in opposite hemispheres, sharing an arc of the equator as their common base, can be completely equal in both sides and angles, so that a complete description of either one taken alone contains nothing that is not also contained in the description of the other; and yet one cannot be substituted for the other, namely in the opposite hemisphere. Here, then, there is an inner difference between the two triangles that no understanding can specify as internal, and that reveals itself only through their external relation in space. But I shall offer more familiar cases that can be drawn from everyday life. What could be more like my hand or my ear, and more equal to it in every respect, than its image in a mirror? Yet I cannot substitute such a hand as is seen in the mirror for its original. For if the original was a right hand, the one in the mirror is a left; and the image of the right ear is a left ear, which can never take the former’s place. Now here there are no inner differences that any understanding could so much as think; yet the differences are internal, as far as the senses teach, since the left hand, despite all the equality and similarity on both sides, cannot be enclosed within the same boundaries as the right—they cannot be congruent. The glove of one hand cannot be used on the other. What, then, is the solution? These objects are not representations of things as they are in themselves and as pure understanding would cognize them. Rather, they are sensible intuitions, that is, appearances whose possibility rests on the relation of certain things, unknown in themselves, to something else, namely our sensibility. Space is the form of this sensibility’s outer intuition, and the inner determination of any particular space is possible only through the determination of its external relation to the whole space of which it is a part—its relation to outer sense. That is, the part is possible only through the whole, which never occurs with things in themselves as objects of mere understanding, but does occur with mere appearances. Hence we cannot make the difference between things that are similar and equal yet incongruent—for example, spiral shells winding in opposite directions—intelligible through any single concept, but only through their relation to the right and left hands, which refers immediately to intuition. *Remark I.* Pure mathematics, and pure geometry in particular, can have objective reality only on the sole condition that it concerns merely objects of the senses. With regard to these objects, however, the principle stands firm: our sensible representation is by no means a representation of things in themselves, but only of the way they appear to us. It follows that the propositions of geometry are not determinations of a mere creature of our inventive imagination, incapable of being reliably related to actual objects. Rather, they necessarily hold of space, and therefore also of everything that may be encountered in space, because space is nothing but the form of all outer appearances, under which alone objects of the senses can be given to us. Sensibility, whose form geometry takes as its foundation, is that on which the possibility of outer appearances rests; these can therefore never contain anything other than what geometry prescribes to them. It would be quite otherwise if the senses had to represent objects as they are in themselves. For then it would by no means follow, from the representation of space with all the various properties the geometer assumes a priori as his foundation, that all this, together with what is inferred from it, must be just so in nature. The geometer’s space would be considered a mere invention, and no objective validity would be credited to it, since there is simply no seeing why things must necessarily agree with the image we form of them by ourselves and in advance. But if this image—or rather this formal intuition—is the essential property of our sensibility, through which alone objects are given to us, and if this sensibility represents not things in themselves but only their appearances, then it is quite easy to understand, and at the same time indisputably proved, that all outer objects of our sensible world must necessarily agree with the propositions of geometry with perfect exactness. For sensibility, through its form of outer intuition—space—with which the geometer concerns himself, first makes these objects themselves possible as mere appearances. It will always remain a remarkable phenomenon in the history of philosophy that there was a time when even mathematicians who were also philosophers began to doubt, not indeed the correctness of their geometrical propositions insofar as these concerned space alone, but the objective validity and application to nature of this concept itself and all its geometrical determinations. They feared that a line in nature might after all consist of physical points, and hence that true space in the object might consist of simple parts, although the space the geometer has in mind cannot possibly consist of them. They failed to recognize that this space in thought makes physical space—that is, the extension of matter itself—possible; that space is not a characteristic of things in themselves at all, but only a form of our sensible power of representation; that all objects in space are mere appearances, that is, not things in themselves but representations of our sensible intuition; and that, since space as the geometer thinks it is precisely the form of sensible intuition that we find in ourselves a priori, containing the ground of the possibility of all outer appearances as regards their form, these appearances must necessarily and with the greatest precision agree with the geometer’s propositions. He draws these not from an invented concept but from the subjective foundation of all outer appearances, namely sensibility itself. In this way, and no other, the geometer can be secured against all the quibbles of a shallow metaphysics concerning the undoubted objective reality of his propositions, however strange they must appear to that metaphysics, since it does not return to the sources of its concepts. *Remark II.* Everything that is to be given to us as an object must be given to us in intuition. But all our intuition takes place only through the senses; the understanding intuits nothing, but only reflects. Now, according to what has just been proved, the senses never and in no respect allow us to cognize things in themselves, but only their appearances; and these are mere representations of sensibility. “Thus all bodies, together with the space in which they are situated, must be held to be nothing but mere representations in us, and exist nowhere else but merely in our thoughts.” Is this not manifest idealism? Idealism consists in the assertion that there are none but thinking beings; the other things we believe we perceive in intuition are only representations in thinking beings, to which in fact no object outside these beings corresponds. I, by contrast, say: things are given to us as objects of our senses situated outside us, but of what they may be in themselves we know nothing; we know only their appearances, that is, the representations they produce in us by affecting our senses. Accordingly, I certainly admit that there are bodies outside us: that is, things which, though entirely unknown to us as regards what they may be in themselves, we know through the representations that their influence on our sensibility supplies us, and to which we give the name of a body. This word therefore signifies merely the appearance of that object, unknown to us but nonetheless actual. Can this really be called idealism? It is precisely the opposite. That we can say of a multitude of the predicates of outer things, without detriment to their actual existence, that these predicates belong not to the things in themselves but only to their appearances, and have no existence of their own outside our representation, is something that was generally accepted and conceded long before *Locke’s* time, but especially afterward. Heat, color, taste, and so forth belong here. But for important reasons I also count among mere appearances, in addition to these, the remaining qualities of bodies that are called primary: extension, place, and space in general, with everything attaching to it—impenetrability or materiality, shape, and so forth. Not the slightest ground for denying the legitimacy of this can be offered. Someone who regards colors not as properties attaching to the object in itself, but only as modifications attaching to the sense of sight, cannot be called an idealist on that account; no more can my doctrine be called idealistic merely because I find that still more, indeed *all properties constituting the intuition of a body*, belong only to its appearance. For the existence of the thing that appears is not thereby abolished, as it is in genuine idealism; it is only shown that through the senses we cannot cognize it at all as it is in itself. I should like to know what my assertions would have to be like, then, in order not to contain an idealism. No doubt I would have to say that the representation of space is not merely in perfect accord with the relation our sensibility bears to objects—for I have said this—but is even completely similar to the object. This is an assertion to which I can attach no sense, any more than to the assertion that the sensation of red bears a similarity to the property of cinnabar that arouses this sensation in me. *Remark III.* We can now easily dismiss an objection that is easy to foresee but groundless: “that through the ideality of space and time the entire sensible world would be transformed into sheer illusion.” First, all philosophical insight into the nature of sensible cognition had been spoiled by making sensibility consist merely in a confused way of representing things, according to which we still cognized things as they are, only without the capacity to bring everything in this representation of ours to clear consciousness. We, by contrast, have proved that sensibility consists not in this logical distinction of clarity or obscurity, but in the genetic distinction concerning the origin of cognition itself: sensible cognition does not represent things as they are at all, but only the way they affect our senses; thus through it only appearances, not the things themselves, are given to the understanding for reflection. After this necessary correction, an objection arises from an inexcusable and almost deliberate misinterpretation, as though my doctrine transformed all things in the sensible world into sheer illusion. When an appearance is given to us, we remain entirely free as to how we wish to judge the thing on its basis. The former, namely appearance, rested on the senses, but this judgment rests on the understanding; the only question is whether there is truth in the determination of the object or not. The difference between truth and dream, however, is not settled by the character of the representations referred to objects, since these are the same in both cases. It is settled by their connection according to the rules determining the coherence of representations in the concept of an object, and by the extent to which they can or cannot coexist in one experience. Here it is not the fault of appearances if our cognition takes illusion for truth: that is, if the intuition through which an object is given to us is taken for a concept of the object, or even of its existence, which the understanding alone can think. The senses represent the planets’ course to us now as forward, now as backward, and in this there is neither falsehood nor truth. As long as we recognize that for the present this is only appearance, we make no judgment at all about the objective character of their motion. But since a false judgment can easily arise if the understanding does not take proper care to prevent this subjective way of representing from being taken as objective, we say that they seem to move backward. Yet the illusion is not charged to the senses, but to the understanding, to which alone it belongs to make an objective judgment from the appearance. In this way, even if we did not reflect at all on the origin of our representations, but connected our sensible intuitions, whatever they might contain, in space and time according to the rules governing the coherence of all cognition in one experience, either deceptive illusion or truth could arise, depending on whether we were careless or cautious. This concerns solely the use of sensible representations in the understanding, not their origin. Likewise, if I regard all representations of the senses, together with their form, namely space and time, as nothing but appearances, and the latter as a mere form of sensibility not encountered in objects outside it, and if I use these representations only in relation to possible experience, then my regarding them as mere appearances contains not the slightest inducement to error or illusion. For despite this they can cohere correctly in experience according to the rules of truth. In this way all geometrical propositions about space hold equally of all objects of the senses, hence of all possible experience, whether I regard space as a mere form of sensibility or as something attaching to things themselves. Yet only in the former case can I understand how it is possible to know those propositions a priori of all objects of outer intuition. Otherwise, as regards all possible experience, everything remains just as though I had never undertaken this departure from common opinion. But if I venture with my concepts of space and time beyond all possible experience, an important error can arise that rests on an illusion. Such a venture is unavoidable if I declare them to be characteristics attaching to things in themselves—for what would then prevent me from holding that they apply to those very things, even if my senses were differently constituted, whether or not they were suited to those things? The error arises when I declare universally valid what was a condition of the intuition of things attaching merely to me as subject, and certainly valid for all objects of the senses, hence for all possible experience, because I refer it to things in themselves rather than restrict it to conditions of experience. Thus, far from making the whole sensible world into mere illusion, my doctrine of the ideality of space and time is rather the only means of securing the application to actual objects of one of the most important kinds of cognition of all—namely, what mathematics sets forth a priori—and preventing it from being taken for mere illusion. Without this observation it would be wholly impossible to determine whether the intuitions of space and time, which we borrow from no experience and which nevertheless lie a priori in our representation, were merely self-made fantasies to which no object corresponded, at least not adequately, and whether geometry itself was therefore a mere illusion. By contrast, we have been able to demonstrate its indisputable validity for all objects of the sensible world precisely because these are mere appearances. Second, far from transforming the truth of experience into mere illusion by making representations of the senses into appearances, these principles of mine are rather the only means of preventing the transcendental illusion by which metaphysics has always been deceived and thereby led into the childish endeavor to catch soap bubbles, because appearances, which are mere representations, were taken for things in themselves. From this arose all those remarkable episodes of the antinomy of reason that I shall mention later. This antinomy is resolved by the single observation that appearance, so long as it is used in experience, produces truth; but as soon as it passes beyond experience’s boundary and becomes transcendent, it produces nothing but sheer illusion. Since, then, I leave their actuality to the things we represent to ourselves through the senses, and restrict only our sensible intuition of these things by saying that in no respect—not even in the pure intuitions of space and time—does it represent anything more than mere appearance of those things, never their character as they are in themselves, this is no pervasive illusion I have fictitiously ascribed to nature. My protest against every imputation of idealism is so conclusive and clear that it would even seem superfluous if there were not unauthorized judges ready to substitute their own delusion for well-determined concepts and thereby twist and disfigure them: judges who would like an old name for every departure from their mistaken though common opinion, and who never judge the spirit of philosophical terms but cling merely to the letter. For my having given this theory of mine the name of transcendental idealism cannot entitle anyone to confuse it with the empirical idealism of *Descartes*—although his was only a problem, whose insolubility, in *Descartes’s* opinion, left everyone free to deny the existence of the corporeal world, since it could never be answered satisfactorily—or with the mystical and visionary idealism of *Berkeley*, against which, and other similar fantasies, our Critique in fact contains the proper antidote. For this idealism, as I called it, concerned not the existence of things—doubting this is what really constitutes idealism in the received meaning, and it never occurred to me to doubt it—but only the sensible representation of things, to which space and time belong first and foremost. Of these, and thus of all appearances in general, I have shown only that they are not things, but mere ways of representing, nor determinations belonging to things in themselves. The word “transcendental,” however, which for me never signifies a relation of our cognition to things, but only to the *faculty of cognition*, was meant to prevent this misinterpretation. Rather than let it occasion the same misunderstanding in the future, I prefer to withdraw this designation and wish my idealism to be called *critical*. But if it is indeed a reprehensible idealism to transform actual things—not appearances—into mere representations, what name shall we give the idealism that, conversely, makes mere representations into things? I think it can be called *dreaming* idealism, as distinct from the former, which may be called *visionary*; both, through my otherwise should have been kept from so-called transcendental, or better, *critical* idealism. ### Second part: How is pure natural science possible? *The Main Transcendental Question* **Second Part.** *How is pure natural science possible?* ### Section 14 § 14. *Nature* is the *existence* of things insofar as it is determined according to universal laws. If nature meant the existence of things in themselves, we could never cognize it, either a priori or a posteriori. Not a priori: for how could we know what belongs to things in themselves, since this can never be accomplished by analyzing our concepts (through analytic propositions)? What I want to know is not what is contained in my concept of a thing—for that belongs to its logical essence—but what, in the thing's actuality, is added to this concept, and what determines the thing itself in its existence outside my concept. My understanding, and the conditions under which alone it can connect the determinations of things in their existence, prescribe no rule to the things themselves. They do not conform to my understanding; rather, my understanding would have to conform to them. They would therefore have to be given to me beforehand, so that I could derive these determinations from them; but then they would not be cognized a priori. Nor would such cognition of the nature of things in themselves be possible a posteriori. For if experience is to teach me the *laws* governing the existence of things, then these laws, insofar as they concern things in themselves, would have to belong to them *necessarily*, even beyond my experience. Now experience does indeed teach me what exists and how it is, but never that it must necessarily be so and not otherwise. Thus it can never teach me the nature of things in themselves. ### Section 15 § 15. Yet we do in fact possess a pure natural science, which sets forth a priori, and with all the necessity required of apodictic propositions, laws to which nature is subject. I need only call as witness here that preliminary study of nature which, under the title of general natural science, precedes all physics (which is grounded on empirical principles). In it we find mathematics applied to appearances, as well as merely discursive principles (drawn from concepts), which constitute the philosophical part of pure cognition of nature. Yet it also contains much that is not entirely pure and independent of experiential sources: the concepts of *motion*, of *impenetrability* (on which the empirical concept of matter rests), of *inertia*, and others, which prevent it from being called a wholly pure natural science. Moreover, it concerns only the objects of the outer senses, and so provides no example of a universal natural science in the strict sense. For such a science must bring nature in general under universal laws, whether it concerns the object of the outer senses or that of inner sense—the object of physics as well as of psychology. But among the principles of that general physics there are some that really do possess the universality we require: for instance, the proposition *that substance remains* and persists; *that everything that happens* is always *determined beforehand* *by a cause* according to constant laws; and so on. These are truly universal laws of nature, which hold entirely a priori. Thus there actually is a pure natural science, and the question now is: *how is it possible?* ### Section 16 § 16. The word nature also takes on another meaning, one that determines the *object*, whereas in the meaning given above it indicated only the *lawfulness* of the determinations of the existence of things in general. Nature, then, considered materialiter, is the *sum total of all objects of experience*. This alone concerns us here. For if things that can never become objects of experience were to be cognized as to their nature, they would compel us to employ concepts whose meaning could never be given in concreto (in any example of a possible experience). Of their nature, therefore, we would have to form nothing but concepts whose reality—that is, whether they actually refer to objects or are mere things of thought—could not be decided at all. Cognition of what cannot be an object of experience would be hyperphysical. Here we are not concerned with anything of that kind, but with cognition of nature whose reality can be confirmed through experience, even though it is possible a priori and precedes all experience. ### Section 17 § 17. The *formal* aspect of nature in this narrower sense is therefore the lawfulness of all objects of experience and, insofar as it is cognized a priori, their *necessary* lawfulness. But we have just shown that the laws of nature can never be cognized a priori in objects insofar as these are considered not in relation to possible experience, but as things in themselves. Here, however, we are not concerned with things in themselves—we leave their properties undecided—but only with things as objects of a possible experience; and it is properly their sum total that we here call nature. I now ask whether, when the possibility of an a priori cognition of nature is at issue, it is better to frame the problem this way: how is it possible to cognize a priori the necessary lawfulness *of things* as objects of experience? Or this way: how is it possible to cognize a priori the necessary lawfulness *of experience* itself with respect to all its objects in general? On closer examination, the answer to the question, however it is framed, will come to precisely the same thing as regards pure cognition of nature, which is the real point at issue. For the subjective laws under which alone an experiential cognition of things is possible also hold for those things as objects of a possible experience—though certainly not for them as things in themselves, which are not under consideration here. It is entirely the same whether I say: without the law that, whenever an event is perceived, it must always be referred to something preceding it, upon which it follows according to a universal rule, no judgment of perception can ever count as experience; or whether I express myself this way: everything that experience teaches us to occur must have a cause. Nevertheless, it is more appropriate to choose the former formulation. For we can indeed have cognition, a priori and prior to all given objects, of the conditions under which alone an experience of them is possible, but never of the laws to which they may be subject in themselves, without reference to possible experience. We can therefore study the nature of things a priori only by investigating the conditions and universal (though subjective) laws under which alone such cognition is possible as experience, in respect of its mere form, and by then determining accordingly the possibility of things as objects of experience. For if I chose the second mode of expression and sought the a priori conditions under which nature is possible as an *object* of experience, I could easily fall into a misunderstanding and imagine that I was speaking of nature as a thing in itself. I would then be driven fruitlessly through endless efforts to seek laws for things of which nothing is given to me. Here, then, we shall concern ourselves solely with experience and the universal, a priori given conditions of its possibility, and from these determine nature as the entire object of all possible experience. I trust it will be understood that I do not mean here the rules for *observing* a nature already given: these already presuppose experience. Thus I do not mean how we can learn nature's laws from it through experience, for these would then not be a priori laws and would yield no pure natural science. I mean, rather, how the a priori conditions of the possibility of experience are at the same time the sources from which all universal laws of nature must be derived. ### Section 18 § 18. We must therefore first observe that although all judgments of experience are empirical—that is, grounded in the immediate perception of the senses—the converse does not hold: not all empirical judgments are therefore judgments of experience. Rather, beyond the empirical element, and in general beyond what is given to sensible intuition, special concepts must also be added. These have their origin entirely a priori in the pure understanding; every perception must first be subsumed under them, and can then, by their means, be transformed into experience. *Empirical judgments, insofar as they have objective validity*, are **judgments of experience**; those, however, *that are only subjectively valid* I call mere **judgments of perception**. The latter require no pure concept of the understanding, but only the logical connection of perceptions in a thinking subject. The former, however, always require, beyond the representations of sensible intuition, special *concepts originally generated in the understanding*, which are precisely what make the judgment of experience *objectively valid*. All our judgments are at first mere judgments of perception: they hold only for us, that is, for our subject. Only afterward do we give them a new relation, namely to an object, and intend that the judgment should hold for us at all times, and equally for everyone else. For if a judgment agrees with an object, all judgments concerning that same object must also agree with one another; thus the objective validity of a judgment of experience means nothing other than its necessary universal validity. Conversely, if we find reason to regard a judgment as necessarily universally valid—which never rests on perception, but on the pure concept of the understanding under which the perception is subsumed—we must also regard it as objective: that is, as expressing not merely a relation of perception to a subject, but a property of the object. For there would be no reason why the judgments of others should necessarily agree with mine, were it not for the unity of the object to which they all refer, with which they agree, and through which they must therefore all agree with one another. ### Section 19 § 19. Objective validity and necessary universal validity (for everyone) are therefore interchangeable concepts. Although we do not know the object in itself, when we regard a judgment as universally valid and hence necessary, it is precisely its objective validity that we mean. Through this judgment we cognize the object—even if it otherwise remains unknown as it may be in itself—through the universally valid and necessary connection of the given perceptions. Since this is the case with all objects of the senses, judgments of experience will derive their objective validity not from immediate cognition of the object (for this is impossible), but solely from the condition for the universal validity of empirical judgments. As has been said, this condition never rests on empirical, or indeed on any sensible, conditions, but on a pure concept of the understanding. The object in itself always remains unknown; but when the connection of the representations it gives to our sensibility is determined as universally valid through the concept of the understanding, the object is determined through this relation, and the judgment is objective. Let us make this clear: that the room is warm, the sugar sweet, the wormwood disagreeable[^299-1]—these are merely subjectively valid judgments. I do not at all demand that I should always find it so, or that everyone else should find it as I do. They express only a relation of two sensations to the same subject, namely myself, and only in my present state of perception; they are therefore not intended to hold of the object. I call such judgments *judgments of perception*. A judgment of experience is quite another matter. What experience teaches me under certain circumstances, it must teach me at all times, and everyone else as well; its validity is not confined to the subject or to the subject's state at the time. I therefore pronounce all such judgments to be objectively valid. For example, when I say that air is elastic, this judgment is initially only a judgment of perception: I merely relate two sensations in my senses to one another. If I intend it to be called a judgment of experience, I demand that this connection stand under a condition that makes it universally valid. Thus I require that I, at all times, and everyone else too, must necessarily connect the same perception under the same circumstances. [^299-1]: I readily admit that these examples do not present judgments of perception that could ever become judgments of experience, even if a concept of the understanding were added. For they refer merely to feeling, which everyone recognizes as merely subjective and which therefore must never be ascribed to the object; thus they can never become objective. For the moment I wished only to give an example of a judgment that is merely subjectively valid and contains in itself no ground for necessary universal validity and thereby for a relation to the object. An example of judgments of perception that become judgments of experience through the addition of a concept of the understanding follows in the next note. ### Section 20 § 20. We must therefore analyze experience in general, to see what is contained in this product of the senses and the understanding, and how the judgment of experience itself is possible. At its foundation lies the intuition of which I am conscious, that is, perception (perceptio), which belongs solely to the senses. But, second, judging also belongs to it, and this belongs solely to the understanding. This judging can take two forms: either I merely compare the perceptions and connect them in a consciousness of my state, or I connect them in a consciousness in general. The first judgment is merely a judgment of perception and as such has only subjective validity; it is merely a connection of perceptions in my mental state, without reference to the object. Hence, contrary to the common supposition, it is not enough for experience to compare perceptions and connect them in one consciousness by means of judging. This gives rise to no universal validity and necessity of the judgment, through which alone it can be objectively valid and be experience. Thus a quite different judgment must precede, before perception can become experience. The given intuition must be subsumed under a concept that determines the form of judging in general with respect to the intuition, connects the empirical consciousness of the latter in a consciousness in general, and thereby gives empirical judgments universal validity. Such a concept is a pure a priori concept of the understanding, which does nothing but determine for an intuition, in general, the way in which it can serve for judgments. Let such a concept be the concept of cause. It then determines the intuition subsumed under it—for example, that of air—with respect to judging in general: namely, that the concept of air, with respect to expansion, should serve in the relation of antecedent to consequent in a hypothetical judgment. The concept of cause is therefore a pure concept of the understanding, entirely distinct from any possible perception, serving only to determine the representation contained under it with respect to judging in general, and thus to make a universally valid judgment possible. Now before a judgment of perception can become a judgment of experience, the first requirement is that the perception be subsumed under such a concept of the understanding. For example, air falls under the concept of cause, which determines the judgment concerning it, with respect to expansion, as hypothetical.[^301-1] This expansion is thereby represented not as belonging merely to my perception of air in my present state, or in several of my states, or in the perceptual state of others, but as belonging to it *necessarily*. The judgment that air is elastic becomes universally valid, and only thereby a judgment of experience, because certain judgments precede it that subsume the intuition of air under the concept of cause and effect. They thereby determine the perceptions not merely in relation to one another in my subject, but with respect to the form of judging in general—here the hypothetical form—and in this way make the empirical judgment universally valid. If we analyze all our synthetic judgments insofar as they hold objectively, we find that they never consist of mere intuitions connected in a judgment simply by comparison, as is commonly supposed. Rather, they would be impossible unless, beyond the concepts abstracted from intuition, a pure concept of the understanding had also been added, under which those concepts were subsumed and thus first connected in an objectively valid judgment. Even the judgments of pure mathematics in its simplest axioms are not exempt from this condition. The principle that a straight line is the shortest between two points presupposes that the line is subsumed under the concept of magnitude. This is certainly no mere intuition, but has its seat solely in the understanding, and serves to determine the intuition of the line, with regard to the judgments that may be made about it, as to their quantity—namely plurality (as iudicia plurativa).[^302-1] For these judgments mean that a given intuition contains many things of the same kind. [^301-1]: For an example that is easier to grasp, consider the following: when the sun shines on the stone, it grows warm. This judgment is a mere judgment of perception and contains no necessity, however often I and others may have perceived this; the perceptions are only habitually found connected in this way. But if I say that the sun *warms* the stone, then beyond the perception the understanding's concept of cause is added, which *necessarily* connects the concept of sunshine with that of heat. The synthetic judgment becomes necessarily universally valid, and hence objective, and is transformed from a perception into experience. [^302-1]: I would prefer this name for the judgments that in logic are called particularia. For the latter expression already contains the thought that they are not universal. But if I begin with unity (in singular judgments) and thus proceed to totality, I cannot yet introduce any reference to totality: I think only plurality without totality, not an exception to it. This is necessary if the logical moments are to be made the basis of the pure concepts of the understanding; in logical use, matters may be left as they are. ### Section 21 § 21. To set forth the possibility of experience insofar as it rests on pure a priori concepts of the understanding, we must therefore first present in a complete table what belongs to judging in general, and the various moments of the understanding within it. For the pure concepts of the understanding will correspond to them in exact parallel: these are nothing more than concepts of intuitions in general, insofar as the intuitions are in themselves determined for judging with respect to one or another of these moments, and hence necessarily and with universal validity. This will also determine precisely the a priori principles of the possibility of all experience as objectively valid empirical cognition. For these are nothing other than propositions that subsume all perception, according to certain universal conditions of intuition, under those pure concepts of the understanding. **Logical Table** *of Judgments.* | | 1. | | |---|---|---| | | *By Quantity* | | | | Universal | | | | Particular | | | | Singular | | | 2. | | 3. | | *By Quality* | | *By Relation* | | Affirmative | | Categorical | | Negative | | Hypothetical | | Infinite | | Disjunctive | | | 4. | | | | *By Modality* | | | | Problematic | | | | Assertoric | | | | Apodictic. | | **Transcendental Table** *of the Concepts of the Understanding.* | | 1. | | |---|---|---| | | *By Quantity* | | | | Unity (measure) | | | | Plurality (magnitude) | | | | Totality (the whole) | | | 2. | | 3. | | *By Quality* | | *By Relation* | | Reality | | Substance | | Negation | | Cause | | Limitation | | Community | | | 4. | | | | *By Modality* | | | | Possibility | | | | Existence | | | | Necessity. | | **Pure Physiological Table** of Universal Principles of Natural Science. | | 1. | | |---|---|---| | | *Axioms* | | | | of Intuition | | | 2. | | 3. | | *Anticipations* | | *Analogies* | | of Perception | | of Experience | | | 4. | | | | *Postulates* | | | | of Empirical Thinking | | | | in General. | | ### Section 21a § 21. [a]. To gather everything said so far into a single concept, it is first necessary to remind readers that we are concerned here not with how experience comes into being, but with what lies within it. The former belongs to empirical psychology, and even there it could never be properly developed without the latter, which belongs to the critique of cognition and especially of the understanding. Experience consists of intuitions, which belong to sensibility, and of judgments, which are solely the business of the understanding. Yet the judgments the understanding makes merely from sensible intuitions are still very far from being judgments of experience. For in the one case the judgment would connect perceptions only as they are given in sensible intuition; in the latter case, however, the judgments are to state what experience in general contains, and hence not what mere perception contains, whose validity is only subjective. Beyond sensible intuition and its logical connection in a judgment—after it has been made universal through comparison—the judgment of experience must therefore add something that determines the synthetic judgment as necessary, and thereby as universally valid. This can be nothing other than the concept that represents the intuition as determined in itself with respect to one form of judgment rather than another: that is, a concept of that synthetic unity of intuitions which can be represented only through a given logical function of judgments. ### Section 22 § 22. The sum of the matter is this: the business of the senses is to intuit; that of the understanding, to think. Thinking, however, is uniting representations in one consciousness. This unification either arises merely in relation to the subject and is contingent and subjective, or takes place unconditionally and is necessary or objective. The unification of representations in one consciousness is judgment. Thus thinking is the same as judging, or referring representations to judgments in general. Judgments are therefore either merely subjective, when representations are referred solely to a consciousness in one subject and united in it, or objective, when they are united in a consciousness in general, that is, necessarily united in it. The logical moments of all judgments are so many possible ways of uniting representations in one consciousness. But when these very same moments serve as concepts, they are concepts of the *necessary* unification of representations in one consciousness, and hence principles of objectively valid judgments. This unification in one consciousness is either analytic, through identity, or synthetic, through the combination and addition of different representations to one another. Experience consists in the synthetic connection of appearances (perceptions) in one consciousness, insofar as this connection is necessary. Thus pure concepts of the understanding are those under which all perceptions must first be subsumed before they can serve for judgments of experience, in which the synthetic unity of perceptions is represented as necessary and universally valid.[^305-1] [^305-1]: But how does this proposition, that judgments of experience must contain necessity in the synthesis of perceptions, agree with the proposition I have repeatedly stressed above, that experience, as a posteriori cognition, can yield only contingent judgments? When I say that experience teaches me something, I always mean only the perception contained in it: for example, that whenever the sun shines on the stone, warmth follows. To that extent the experiential proposition is always contingent. That this warming follows necessarily from the sun's illumination is indeed contained in the judgment of experience, by virtue of the concept of cause; but I do not learn this through experience. On the contrary, experience is first generated by this addition of the understanding's concept (of cause) to perception. As to how perception comes to receive this addition, one must consult the Critique, in the section on the transcendental power of judgment, page 137 and following. ### Section 23 § 23. Judgments, insofar as they are considered merely as the condition for uniting given representations in one consciousness, are rules. These rules, insofar as they represent the unification as necessary, are a priori rules; and insofar as no higher rules stand above them from which they are derived, they are principles. Now with respect to the possibility of all experience, if we consider in it merely the form of thinking, there are no conditions of judgments of experience beyond those that bring appearances, according to the different form of their intuition, under pure concepts of the understanding, which make the empirical judgment objectively valid. These, therefore, are the a priori principles of possible experience. The principles of possible experience are at the same time universal laws of nature, which can be cognized a priori. Thus the problem contained in our present second question—*How is pure natural science possible?*—is solved. For the systematic character required for the form of a science is fully present here. Beyond the formal conditions of all judgments in general, and hence of all rules in general, which logic supplies, no others are possible. These conditions constitute a logical system; the concepts grounded on them, which contain the a priori conditions for all synthetic and necessary judgments, constitute, for that very reason, a transcendental system; and finally the principles through which all appearances are subsumed under these concepts constitute a physiological system, that is, a system of nature. This precedes all empirical cognition of nature, first makes it possible, and can therefore be called universal and pure natural science in the proper sense. ### Section 24 § 24. The first[^306-1] of those physiological principles subsumes all appearances, as intuitions in space and time, under the concept of *magnitude*, and is to that extent a principle for applying mathematics to experience. The second does not directly subsume the properly empirical element—namely sensation, which designates the real in intuitions—under the concept of *magnitude*. For sensation is no intuition that *contains* space or time, although it places its corresponding object in both. Yet in time there is a difference that has a magnitude between reality (the representation of sensation) and zero, that is, the complete emptiness of intuition. For between any given degree of light and darkness, between any degree of warmth and complete cold, any degree of weight and absolute lightness, any degree of the filling of space and entirely empty space, still smaller degrees can always be thought; just as, even between a consciousness and complete unconsciousness (psychological darkness), still smaller degrees always occur. Hence no perception is possible that would prove an absolute absence: for example, no psychological darkness that could not be regarded as a consciousness merely outweighed by another, stronger consciousness; and so it is in every case of sensation. The understanding can therefore anticipate even sensations, which constitute the proper quality of empirical representations (appearances), by means of the principle that all of them, and hence the real in every appearance, have degrees. This is the second application of mathematics (mathesis intensorum) to natural science. [^306-1]: These three consecutive sections will scarcely be properly understood unless one has at hand what the Critique says about the principles. They may, however, serve to make their general character easier to survey and to draw attention to their main moments. ### Section 25 § 25. As regards the relation of appearances, and solely with respect to their existence, the determination of this relation is not mathematical but dynamical. It can never be objectively valid, and hence fit to constitute experience, unless it stands under a priori principles that first make experiential cognition of appearances possible in this respect. Appearances must therefore be subsumed under the concept of substance, which underlies every determination of existence as a concept of the thing itself; or, second, insofar as a succession in time is found among appearances—that is, an event—under the concept of an effect in relation to a cause; or, insofar as coexistence is to be cognized objectively, that is, through a judgment of experience, under the concept of community (reciprocal interaction). Thus a priori principles underlie objectively valid, though empirical, judgments: that is, the possibility of experience insofar as it is to connect objects in nature with respect to their existence. These principles are the laws of nature in the proper sense, which may be called dynamical. Finally, judgments of experience also involve cognition of agreement and connection: not so much that of appearances with one another in experience, as their relation to experience in general. This relation contains either their agreement with the formal conditions that the understanding cognizes, or their connection with the material element of the senses and of perception, or both united in one concept; hence it contains possibility, actuality, and necessity according to universal laws of nature. This would constitute the physiological doctrine of method: the distinction between truth and hypotheses, and the limits of the reliability of the latter. ### Section 26 § 26. The third table, that of principles, drawn by the critical method *from the nature of the understanding itself*, displays a perfection that raises it far above any other ever attempted, or that may yet be attempted, dogmatically *from the things themselves*, though always in vain. It contains all synthetic a priori principles in their completeness, and has been developed according to a single principle: the faculty of judging in general, which constitutes the essence of experience with respect to the understanding. We can therefore be certain that there are no further principles of this kind—a satisfaction the dogmatic method can never provide. Yet this is far from being its greatest merit. We must attend to the ground of proof that reveals the possibility of this a priori cognition, and at the same time restricts all such principles to a condition that must never be overlooked, if it is not to be misunderstood and extended in use beyond the original meaning the understanding gives it: namely, that these principles contain only the conditions of possible experience in general, insofar as it is subject to a priori laws. Thus I do not say that things *in themselves* contain a magnitude, their reality a degree, their existence a connection of accidents in a substance, and so on. For no one can prove this, since such a synthetic connection from mere concepts, where all reference to sensible intuition on the one hand, and all connection of it in a possible experience on the other, are lacking, is simply impossible. The essential restriction of the concepts in these principles is therefore this: that all things stand necessarily a priori under the stated conditions only *as objects of experience*. From this, second, follows their specifically distinctive manner of proof: the principles in question are not referred directly to appearances and their relation, but to the possibility of experience, of which appearances constitute only the matter, not the form—that is, to objectively and universally valid synthetic propositions, in which precisely judgments of experience differ from mere judgments of perception. This is accomplished as follows. Appearances, as mere intuitions *that occupy a part of space and time*, stand under the concept of magnitude, which synthetically unites their manifold a priori according to rules. Insofar as perception contains sensation as well as intuition, and a transition through diminution always takes place between sensation and zero—that is, its complete disappearance—the real in appearances must have a degree. For sensation itself *occupies no part of space or time*,[^309-1] yet the transition to it from empty time or space is possible only in time. Consequently, although sensation as the quality of empirical intuition can never be cognized a priori with respect to that in which it differs specifically from other sensations, it can nevertheless, in a possible experience in general, be distinguished intensively, as a magnitude of perception, from every other sensation of the same kind. This first makes possible and determines the application of mathematics to nature with respect to the sensible intuition through which nature is given to us. Above all, however, the reader must attend to the manner of proof of the principles that bear the name analogies of experience. Unlike the principles for applying mathematics to natural science in general, these concern not the generation of intuitions, but the connection of their existence in one experience. This can be nothing other than the determination of existence in time according to necessary laws, under which alone it is objectively valid, and hence experience. The proof therefore concerns not synthetic unity in the connection of things in themselves, but of *perceptions*, and these not with respect to their content, but with respect to their determination in time and the relation of existence in time according to universal laws. These universal laws thus contain the necessity of determining existence in time in general—and consequently according to an a priori rule of the understanding—if the empirical determination in relative time is to be objectively valid, and hence experience. Here, in prolegomena, I can offer no more than this recommendation to the reader who has long been accustomed to regard experience as a merely empirical combination of perceptions, and therefore never considers that it goes much further than they reach: namely, that it gives empirical judgments universal validity, and for this requires a pure unity of the understanding that precedes them a priori. I recommend paying close attention to this distinction between experience and a mere aggregate of perceptions, and judging the manner of proof from this standpoint. [^309-1]: Heat, light, etc., are just as great in a small space, in degree, as in a large one; likewise inner representations, pain, and consciousness in general are no smaller in degree whether they last a short or a long time. Thus magnitude here is just as great at a point and in an instant as in any space or time, however large. Degrees, then, are greater not in intuition, but in mere sensation, or in the magnitude of the ground of an intuition. They can be estimated as magnitudes only through the relation of 1 to 0: that is, because each can diminish through infinitely many intermediate degrees until it disappears, or grow from zero through infinitely many moments of increase to a determinate sensation within a certain time (quantitas qualitatis est gradus). ### Section 27 § 27. Here is the place to remove *Hume's* doubt from its very foundation. He rightly maintained that through reason we can in no way comprehend the possibility of causality: that is, the relation of the existence of one thing to the existence of something else necessarily posited by it. I add that we have just as little insight into the concept of subsistence: that is, into the necessity that the existence of things be grounded in a subject that cannot itself be a predicate of any other thing. Indeed, we cannot even form a concept of the possibility of such a thing, although we can point to examples of its use in experience. Likewise, this very same incomprehensibility concerns the community of things. For we cannot at all comprehend how, from the state of one thing, an inference can be drawn to the state of entirely different things outside it, and reciprocally; or how substances, each of which has its own separate existence, are to depend on one another, and necessarily so. Nevertheless, I am far from regarding these concepts as merely borrowed from experience, and the necessity represented in them as invented and as a mere illusion conjured up by long habit. Rather, I have sufficiently shown that these concepts, and the principles drawn from them, stand firm a priori before all experience and have their undoubted objective correctness—though, to be sure, only with respect to experience. ### Section 28 § 28. Thus, although I have not the slightest concept of such a connection of things in themselves—how they can exist as substance, act as cause, or stand in community with others as parts of a real whole—and still less can I think such properties in appearances as appearances (since those concepts contain nothing that lies in appearances, but only what the understanding alone must think), we nevertheless have such a concept of a connection of representations in our understanding, specifically in judgments in general. Namely, in one kind of judgment representations stand as subject in relation to predicates; in another, as ground in relation to consequence; and in a third, as parts that together make up a whole possible cognition. Further, we cognize a priori that unless we regard the representation of an object as determined with respect to one or another of these moments, we can have no cognition at all that holds of the object. If we were occupied with the object in itself, however, not a single mark would be possible by which I could cognize that it was determined with respect to one or another of the moments just mentioned: that is, that it fell under the concept of substance, or of cause, or, in relation to other substances, under the concept of community. For I have no concept of the possibility of such a connection of existence. But the question is not how things in themselves are determined, but how experiential cognition of things is determined with respect to these moments of judgments in general: that is, how things as objects of experience can and must be subsumed under those concepts of the understanding. Here it is clear that I fully comprehend not only the possibility but also the necessity of subsuming all appearances under these concepts: that is, of using them as principles of the possibility of experience. ### Section 29 § 29. To make a trial with *Hume's* problematic concept—his crux metaphysicorum—namely the concept of cause: first, through logic, the form of a conditional judgment in general is given to me a priori, namely the use of one given cognition as ground and another as consequence. Now it is possible to find in perception a rule of relation stating that one appearance constantly follows a certain other appearance, though not conversely. This is a case for employing the hypothetical judgment and saying, for example: if a body has been exposed to the sun long enough, it grows warm. Here, to be sure, there is as yet no necessity of connection, and hence no concept of cause. But I proceed and say: if the foregoing proposition, which is merely a subjective connection of perceptions, is to be a proposition of experience, it must be regarded as necessary and universally valid. Such a proposition would be: the sun, through its light, is the cause of the warmth. The foregoing empirical rule is now regarded as a law, and as holding not merely of appearances, but of them for the sake of a possible experience, which requires rules that hold throughout and therefore necessarily. Thus I can very well comprehend the concept of cause as a concept necessarily belonging to the mere form of experience, and its possibility as a synthetic unification of perceptions in a consciousness in general. But I cannot at all comprehend the possibility of a thing in general as a cause, precisely because the concept of cause indicates no condition attaching to things, but only one attaching to experience: namely, that experience can be an objectively valid cognition of appearances and their succession in time only insofar as the preceding appearance can be connected with the following one according to the rule of hypothetical judgments. ### Section 30 § 30. Hence the pure concepts of the understanding have no meaning whatsoever if they depart from objects of experience and seek to refer to things in themselves (noumena). They serve, as it were, only to spell out appearances so that we can read them as experience. The principles arising from their reference to the world of sense serve our understanding only for use in experience. Beyond that, they are arbitrary combinations without objective reality, whose possibility we cannot cognize a priori, and whose reference to objects we cannot confirm, or even make intelligible, through any example. For all examples can be borrowed only from some possible experience, and consequently the objects of those concepts can be found nowhere else than in a possible experience. This complete solution of *Hume's* problem, though contrary to its author's expectation, thus preserves for the pure concepts of the understanding their a priori origin, and for the universal laws of nature their validity as laws of the understanding. Yet it does so by restricting their use to experience, because their possibility is grounded solely in the relation of the understanding to experience—not in such a way that they are derived from experience, but that experience is derived from them. This completely reversed kind of connection never occurred to *Hume*. From this follows the result of all our investigations so far: “All synthetic a priori principles are nothing more than principles of possible experience,” and can never be referred to things in themselves, but only to appearances as objects of experience. Hence pure mathematics and pure natural science alike can never concern anything more than mere appearances, and represent only what either makes experience in general possible, or, being derived from these principles, must always be capable of being represented in some possible experience. ### Section 31 § 31. Thus at last we have something definite to hold to in all metaphysical undertakings, which until now have swept boldly enough, but always blindly, over everything without distinction. It never occurred to dogmatic thinkers that the goal of their efforts should be set within such narrow bounds. Nor did it occur even to those who, defiantly trusting their supposed sound reason, used concepts and principles of pure reason that were indeed legitimate and natural, but intended merely for use in experience, to pursue insights for which they knew no definite limits, nor could know any, since they had never reflected, or been able to reflect, on the nature and even the possibility of such a pure understanding. Many a naturalist of pure reason—by which I mean someone who trusts himself to decide questions of metaphysics without any science—might well claim that what has been presented here with so much preparation, or, if he prefers, with such long-winded pedantic pomp, was something he had long since not merely suspected, but known and understood through the prophetic spirit of his sound reason: “that with all our reason we can never pass beyond the field of experiences.” Yet if one gradually questions him about his principles of reason, he must admit that among them are many he has not drawn from experience, and that are therefore independent of it and valid a priori. How, then, and on what grounds will he keep within bounds the dogmatist, and himself, who employs these concepts and principles beyond all possible experience precisely because they are cognized independently of it? Even he, this adept of sound reason, is not so secure, despite all his pretended wisdom so cheaply acquired, against slipping unnoticed beyond objects of experience into the realm of fantasies. Indeed, he is usually entangled deeply enough in it, though he gives his groundless claims a certain gloss through popular language, presenting everything as mere probability, reasonable conjecture, or analogy. ### Section 32 § 32. Since the earliest days of philosophy, investigators of pure reason have thought, besides beings of sense or appearances (phaenomena), which make up the world of sense, of special beings of the understanding (noumena), which were supposed to make up a world of the understanding. And since they took appearance and illusion to be the same—a mistake readily pardonable in an age still undeveloped—they granted actuality to beings of the understanding alone. In fact, if we rightly regard the objects of the senses as mere appearances, we thereby admit at the same time that a thing in itself underlies them, although we do not know it as it is in itself, but know only its appearance: that is, the way in which our senses are affected by this unknown something. Thus the understanding, precisely by admitting appearances, also admits the existence of things in themselves. To that extent we can say that the representation of such beings, which underlie appearances, and hence of mere beings of the understanding, is not only permissible but unavoidable. Our critical deduction in no way excludes such things (noumena), but rather restricts the principles of the aesthetic so that they must not extend to all things—which would transform everything into mere appearance—but must hold only of objects of a possible experience. Beings of the understanding are thus admitted, but with emphasis on this rule, which allows no exception: that of these pure beings of the understanding we know, and can know, absolutely nothing determinate. For our pure concepts of the understanding, as well as our pure intuitions, concern nothing but objects of possible experience, and hence mere beings of sense; and as soon as we depart from these, not the slightest meaning remains for those concepts. ### Section 33 § 33. There is indeed something ensnaring about our pure concepts of the understanding in the way they entice us toward a transcendent use—for that is what I call a use that goes beyond all possible experience. Not only are our concepts of substance, force, action, reality, etc., wholly independent of experience, and likewise contain no appearance of the senses, so that they really seem to concern things in themselves (noumena); but, strengthening this supposition, they contain a necessity of determination that experience can never match. The concept of cause contains a rule according to which one state necessarily follows from another; but experience can show us only that one state of things is followed by another often, or at best usually, and thus can provide neither strict universality nor necessity, etc. Concepts of the understanding therefore seem to have far more meaning and content than their mere use in experience could exhaust. And so, without noticing it, the understanding adds to the house of experience a much more extensive adjoining structure, which it fills with nothing but beings of thought, without even realizing that, with its otherwise correct concepts, it has ventured beyond the limits of their use. ### Section 34 § 34. Two important, indeed wholly indispensable, though extremely dry investigations were therefore needed. These were undertaken in the Critique, pages 137 ff. and 235 ff. The first showed that the senses supply not the pure concepts of the understanding in concreto, but only the schema for their use, and that the object corresponding to it is found only in experience, as the product the understanding makes from materials of sensibility. The second investigation (Critique, p. 235) shows that despite the independence of our pure concepts and principles of the understanding from experience, and even their apparently wider range of use, nothing at all can be thought through them outside the field of experience. For they can do nothing but determine the logical form of judgment with respect to given intuitions. But since beyond the field of sensibility there is no intuition whatsoever, those pure concepts lack all meaning, because there is no means of presenting them in concreto. Consequently, all such noumena, together with their sum total, an intelligible[^316-1] world, are nothing but representations of a problem whose object may well be possible in itself, but whose solution is wholly impossible given the nature of our understanding. For our understanding is no faculty of intuition, but only of connecting given intuitions in one experience; experience must therefore contain all objects for our concepts, while outside it all concepts will be without meaning, since no intuition can be placed under them. [^316-1]: Not, as is commonly said, an *intellectual* world. For *cognitions* through the understanding are *intellectual*, and these also concern our world of sense; but *objects* are called *intelligible* insofar as they can be represented *solely through the understanding* and none of our sensible intuitions can concern them. Yet since some possible intuition must correspond to every object, one would have to think of an understanding that immediately intuited things. But of such an understanding we have not the slightest concept, and hence none either of the *beings of the understanding* it is supposed to concern. ### Section 35 § 35. The imagination may perhaps be forgiven if it sometimes indulges in fantasy, that is, fails to remain cautiously within the bounds of experience. For at least such a free flight enlivens and strengthens it, and it will always be easier to restrain its boldness than to remedy its languor. But that the understanding, which ought to *think*, should instead *fantasize*, can never be forgiven it; for it alone provides all the help we have in setting limits, where necessary, to the imagination's extravagance. Yet it begins this business most innocently and modestly. First it puts in order the elementary cognitions it possesses before all experience, but which must nevertheless always find their application in experience. Gradually it drops these restrictions—and what should prevent it, since the understanding has taken its principles quite freely from itself? Now it turns first to newly invented forces in nature, soon afterward to beings outside nature: in a word, to a world for whose construction we cannot lack building materials, since fertile invention supplies them in abundance, and experience, though it does not confirm them, never refutes them either. This is also why young thinkers are so fond of metaphysics in the true dogmatic manner, and often sacrifice to it their time and their otherwise useful talent. But it is of no use to try to moderate those fruitless attempts of pure reason through all manner of reminders about the difficulty of solving such deeply hidden questions, complaints about the limits of our reason, and reductions of assertions to mere conjectures. For unless their *impossibility* has been clearly demonstrated, and reason's *self-cognition* becomes a true science, in which the field of its rightful use is distinguished from that of its empty and fruitless use with, so to speak, geometrical certainty, those vain endeavors will never be entirely put to rest. ### Section 36 § 36. *How is nature itself possible?* This question, which is the highest point transcendental philosophy can ever touch, and to which it must also be led as its limit and completion, actually contains two questions. **First:** How is nature possible in the *material* sense, namely with respect to intuition, as the sum total of appearances? How are space, time, and what fills them both—the object of sensation—possible in general? The answer is: through the constitution of our sensibility, according to which it is affected, in its own distinctive way, by objects that are unknown to it in themselves and entirely distinct from those appearances. This answer was given in the book itself in the transcendental aesthetic, and here in the Prolegomena through the solution of the first main question. **Second:** How is nature possible in the formal sense, as the sum total of the rules under which all appearances must stand if they are to be thought as connected in one experience? The answer can only be this: it is possible solely through the constitution of our understanding, according to which all those representations of sensibility are necessarily referred to one consciousness. Through this alone, our distinctive way of thinking—namely through rules—and by means of these, experience, which must be entirely distinguished from insight into objects in themselves, become possible. This answer was given in the book itself in the transcendental logic, and here in the Prolegomena in the course of solving the second main question. But how this distinctive property of our sensibility itself is possible, or that of our understanding and of the necessary apperception underlying it and all thinking, admits of no further solution or answer, because we always need them in turn for every answer and for all thinking about objects. There are many laws of nature that we can know only through experience. But the lawfulness in the connection of appearances, that is, nature in general, can be learned from no experience, because experience itself requires such laws, which underlie its possibility a priori. The possibility of experience in general is therefore at the same time the universal law of nature, and the principles of the former are themselves the laws of the latter. For we know nature only as the sum total of appearances, that is, of representations in us, and can therefore take the law of their connection from nowhere but the principles of their connection in us: that is, from the conditions of their necessary unification in one consciousness, which constitutes the possibility of experience. Even the main proposition developed throughout this entire part—that universal laws of nature can be cognized a priori—already leads of itself to the proposition that nature's supreme legislation must lie within ourselves, that is, in our understanding. We must seek its universal laws not from nature through experience, but conversely, nature in its universal lawfulness solely from the conditions of the possibility of experience lying in our sensibility and understanding. For how else would it be possible to know these laws a priori, since they are not rules of analytic cognition, but genuine synthetic extensions of it? Such an agreement, and a necessary one, between the principles of possible experience and the laws of the possibility of nature can arise from only two causes: either these laws are borrowed from nature through experience, or conversely, nature is derived from the laws of the possibility of experience in general, and is completely identical with the mere universal lawfulness of experience. The former contradicts itself, for universal laws of nature can and must be cognized a priori—that is, independently of all experience—and laid at the foundation of every empirical use of the understanding. Thus only the second alternative remains.[^319-1] But we must distinguish empirical laws of nature, which always presuppose particular perceptions, from pure or universal laws of nature, which, without any particular perceptions as their basis, contain merely the conditions for their necessary unification in one experience. With respect to the latter, nature and possible experience are entirely one and the same. And since in experience lawfulness rests on the necessary connection of appearances in one experience—without which we can cognize no object at all of the world of sense—and hence on the original laws of the understanding, it sounds strange at first, yet is nonetheless certain, when I say with respect to these universal laws: *the understanding does not draw its laws (a priori) from nature, but prescribes them to it.* [^319-1]: *Crusius* alone knew a middle way: that a spirit who can neither err nor deceive originally implanted these laws of nature in us. Yet since deceptive principles also often intrude, of which this man's own system supplies not a few examples, and since there are no secure criteria for distinguishing the genuine origin from the spurious, the use of such a principle is a very precarious matter. For we can never know with certainty what the spirit of truth or the father of lies may have instilled in us. ### Section 37 § 37. Let us illustrate this seemingly venturesome proposition through an example, which is to show that laws we discover in objects of sensible intuition, especially when they have been cognized as necessary, are already regarded by us as laws the understanding has placed there, although in all other respects they resemble the laws of nature that we attribute to experience. ### Section 38 § 38. If we consider the properties of the circle, through which this figure at once unites so many arbitrary determinations of space within itself under a universal rule, we cannot help ascribing a nature to this geometrical thing. Thus two lines that intersect one another and also the circle, however randomly they may be drawn, always divide each other with such regularity that the rectangle formed from the segments of either line is equal to that formed from the segments of the other. Now I ask: “Does this law lie in the circle, or does it lie in the understanding?” That is, does this figure contain the ground of this law within itself independently of the understanding, or does the understanding, having constructed the figure itself according to its concepts—namely, the equality of the radii—at the same time place in it the law of chords intersecting one another in geometrical proportion? If we follow the proofs of this law, we soon discover that it can be derived solely from the condition the understanding laid at the foundation of the construction of this figure: namely, the equality of the radii. If we now extend this concept, pursuing still further the unity of the manifold properties of geometrical figures under common laws, and consider the circle as a conic section, which therefore stands with other conic sections under the very same fundamental conditions of construction, we find that all chords intersecting within the latter—the ellipse, parabola, and hyperbola—always do so in such a way that the rectangles formed from their parts, though not equal, nevertheless always stand in equal ratios to one another. If we go still further, namely to the fundamental doctrines of physical astronomy, a physical law of reciprocal attraction appears, extending throughout all material nature. Its rule is that attraction decreases inversely as the square of the distance from each attracting point, just as the spherical surfaces over which this force spreads increase. This seems to lie necessarily in the nature of things themselves, and is therefore usually presented as cognizable a priori. Now simple as the sources of this law are, resting merely on the relation of spherical surfaces of different radii, its consequence is so excellent in the manifold harmony and regularity it produces that not only do all possible orbits of celestial bodies follow conic sections, but such a relation among them also results that no law of attraction other than that of the inverse square of the distances can be conceived as suitable for a world system. Here, then, is nature resting on laws that the understanding cognizes a priori, chiefly from universal principles of the determination of space. Now I ask: do these laws of nature lie in space, and does the understanding learn them by merely seeking to explore the rich meaning lying within it? Or do they lie in the understanding and in the way it determines space according to the conditions of synthetic unity, toward which all its concepts converge? Space is something so uniform, and so undetermined with respect to all particular properties, that we shall certainly seek no treasure of natural laws within it. Rather, what determines space into the shape of a circle, or the figure of a cone and a sphere, is the understanding, insofar as it contains the ground of the unity of their construction. The mere universal form of intuition called space is indeed the substratum of all intuitions determinable with respect to particular objects, and in it certainly lies the condition of their possibility and manifoldness. But the unity of objects is determined solely by the understanding, and according to conditions that lie in its own nature. Thus the understanding is the source of the universal order of nature, by bringing all appearances under its own laws and thereby first producing experience, as to its form, a priori. By virtue of this, everything that is to be cognized only through experience is necessarily subjected to its laws. For we are not concerned with the nature of *things in themselves*, which is independent of the conditions both of our sensibility and of our understanding, but with nature as an object of possible experience. Here the understanding, in making experience possible, at the same time brings it about that the world of sense is either no object of experience at all, or a nature. ### Section 39 § 39. Appendix *to Pure Natural Science.* *On the System of Categories.* Nothing can be more welcome to a philosopher than to be able to derive from a single a priori principle the manifold of concepts or principles that had previously appeared scattered through the use he had made of them in concreto, and in this way unite everything in a single cognition. Previously he merely believed that what remained to him after a certain abstraction, and seemed through mutual comparison to constitute a particular kind of cognition, had been collected in its entirety; but it was only an *aggregate*. Now he knows that precisely this much, neither more nor less, can constitute that kind of cognition, and sees the necessity of his division, which is comprehension; only now does he have a *system*. To seek out from common cognition the concepts that have no particular experience at their foundation, yet occur in all experiential cognition, of which they constitute, as it were, the mere form of connection, required no greater reflection or insight than to seek out from a language the rules of the actual use of words in general, and so collect elements for a grammar. In fact, the two investigations are very closely related. Yet this does not enable us to give a reason why each language has precisely this formal constitution and no other, still less why precisely this many formal determinations of it, neither more nor less, can be found in general. *Aristotle* had collected ten such pure elementary concepts under the name of categories.[^323-1] To these, which were also called predicaments, he later found himself compelled to add five postpredicaments,[^323-2] some of which are already contained in the former (such as prius, simul, motus). But this rhapsody could count, and deserve approval, more as a hint for future investigators than as a systematically developed idea; hence, with the further enlightenment of philosophy, it was rejected as entirely useless. In investigating the pure elements of human cognition—those containing nothing empirical—I succeeded, only after long reflection, in reliably distinguishing and separating the pure elementary concepts of sensibility (space and time) from those of the understanding. This excluded the 7th, 8th, and 9th categories from that list. The remainder could be of no use to me, because there was no principle by which the understanding could be fully measured and all its functions, from which its pure concepts arise, could be determined completely and precisely. To discover such a principle, however, I looked for an act of the understanding that contains all the others and differs only through various modifications or moments in bringing the manifold of representation under the unity of thinking in general. I found that this act of the understanding consists in judging. Here the logicians' work lay ready before me, though not yet entirely free of defects, and enabled me to present a complete table of pure functions of the understanding, which were, however, undetermined with respect to any object. Finally I referred these functions of judging to objects in general, or rather to the condition for determining judgments as objectively valid, and pure concepts of the understanding arose. Of these I could be certain beyond doubt that precisely these concepts, and only this many, neither more nor less, can constitute our entire cognition of things from the understanding alone. I rightly called them by their old name, *categories*, reserving the task of adding in full, under the name of *predicables*, all concepts derivable from them: whether by combination with one another, with the pure form of appearance (space and time), or with its matter insofar as this is not yet empirically determined (the object of sensation in general). This would be done as soon as a system of transcendental philosophy should be established, in preparation for which I was now concerned only with the critique of reason itself. But what is essential in this system of categories, what distinguishes it from that old rhapsody, which proceeded without any principle, and why it alone deserves to be counted as philosophy, consists in this: by its means, the true meaning of the pure concepts of the understanding and the condition of their use could be precisely determined. For it became evident that in themselves they are nothing but logical functions, and as such do not constitute the slightest concept of an object in itself. Rather, they require sensible intuition as their foundation, and then serve only to determine empirical judgments with respect to the functions of judging—judgments otherwise undetermined and indifferent with respect to all these functions—thereby giving them universal validity and by their means making *judgments of experience* in general possible. Neither their first author nor anyone after him ever conceived of such an insight into the nature of the categories, which at the same time restricted them to mere use in experience. But without this insight, which depends entirely upon their derivation or deduction, they are wholly useless: a wretched list of names without explanation or rule for their use. Had anything of the kind ever occurred to the ancients, the entire study of pure rational cognition, which under the name of metaphysics has ruined so many fine minds over many centuries, would doubtless have come down to us in quite a different form. It would have enlightened human understanding instead of exhausting it, as actually happened, in dark and fruitless brooding, and making it unfit for true science. This system of categories in turn makes every treatment of any object of pure reason itself systematic, and supplies an indubitable direction or guiding thread as to how, and through which points of investigation, every metaphysical inquiry must be conducted if it is to be complete. For it exhausts all the moments of the understanding under which every other concept must be brought. This is also how the table of principles arose, of whose completeness we can be certain only through the system of categories. Even in the division of concepts that are supposed to go beyond the physiological use of the understanding (Critique, p. 344, and likewise p. 415), it is always the same guiding thread. Since it must always be carried through the same fixed points, determined a priori in the human understanding, it always forms a closed circle, leaving no doubt that the object of a pure concept of the understanding or of reason, insofar as it is to be considered philosophically and according to a priori principles, can be completely cognized in this way. I could not even refrain from using this guidance with respect to one of the most abstract ontological divisions, namely the manifold distinction between the concepts of Something and Nothing, and from producing accordingly a systematic and necessary table (Critique, p. 292).[^325-1] This same system, like every true system grounded on a universal principle, also displays its usefulness, which cannot be praised enough, in excluding all alien concepts that might otherwise slip in among those pure concepts of the understanding, and assigning every cognition its place. The concepts that I likewise arranged in a table, following the guiding thread of the categories, under the name of *concepts of reflection*, mingle in ontology, without permission or legitimate claim, among the pure concepts of the understanding. Yet the latter are concepts of connection, and thereby of the object itself, whereas the former are concepts only of the mere comparison of concepts already given, and hence have an entirely different nature and use. Through my division according to law (Critique, p. 260), they are separated from this mixture. But the advantage of that separate table of categories becomes still more striking when, as will happen immediately, we distinguish from it the table of transcendental concepts of reason, which are of an entirely different nature and origin from those concepts of the understanding, and must therefore also have a different form. Yet this indispensable separation has never been made in any system of metaphysics, where those ideas of reason consequently run indiscriminately together with concepts of the understanding, as though they belonged like siblings to one family. In the absence of a distinct system of categories, this confusion could never have been avoided. [^323-1]: 1. Substantia. 2. Qualitas. 3. Quantitas. 4. Relatio. 5. Actio. 6. Passio. 7. Quando. 8. Ubi. 9. Situs. 10. Habitus. [^323-2]: Oppositum, Prius, Simul, Motus, Habere. [^325-1]: All manner of interesting observations can be made about a presented table of categories. For instance: 1) that the third arises from the first and second combined in one concept; 2) that in the categories of magnitude and quality there is merely a progression from unity to totality, or from Something to Nothing—for this purpose the categories of quality must stand in this order: reality, limitation, complete negation—without correlata or opposita, whereas those of relation and modality carry these latter with them; 3) that just as in *logic* categorical judgments underlie all the others, so the category of substance underlies all concepts of actual things; 4) that just as modality is no special predicate in a judgment, so the modal concepts add no determination to things; and so on. All such observations have their great usefulness. If, moreover, we enumerate all the *predicables*, which can be drawn fairly completely from any good ontology—for example, *Baumgarten's*—and arrange them by classes under the categories, taking care to add as complete an articulation of all these concepts as possible, a merely analytic part of metaphysics will arise. This will contain no synthetic proposition as yet and could precede the second, synthetic part; through its definiteness and completeness it would possess not only usefulness, but, by virtue of its systematic character, a certain beauty as well. ### Third part: How is metaphysics in general possible? *The Main Transcendental Question* **Third Part.** *How Is Metaphysics in General Possible?* ### Section 40 § 40. For *their own security* and certainty, pure mathematics and pure natural science would have needed no such deduction as we have so far accomplished for both. The former rests on its own self-evidence; the latter, although it springs from pure sources of the understanding, nevertheless rests on experience and its thoroughgoing confirmation. It cannot entirely reject or dispense with the testimony of experience, because, for all its certainty, it can never, as philosophy, equal mathematics. Neither science, then, needed the investigation in question for itself, but for another science, namely metaphysics. Besides concepts of nature, which always find their application in experience, metaphysics also deals with pure concepts of reason, which are never given in any experience whatever that could possibly occur. It therefore deals with concepts whose objective reality—that they are not mere fantasies—cannot be established by experience, and with assertions whose truth or falsity cannot be confirmed or exposed by any experience. Moreover, this part of metaphysics is precisely what constitutes its essential purpose, to which everything else is only a means; and so this science needs such a deduction *for its own sake*. The third question now before us thus concerns, as it were, the very heart and distinctive character of metaphysics: reason's occupation solely with itself and, as it broods over its own concepts, the acquaintance with objects that supposedly arises directly from them, without needing experience to mediate it or indeed being able to reach it through experience.[^327-1] Without a solution to this question, reason can never satisfy itself. The use within experience to which reason confines the pure understanding does not fulfill reason's own entire vocation. Each individual experience is only a part of the whole sphere of its domain; but the *absolute whole of all possible experience* is not itself an experience, and is nevertheless a necessary problem for reason. Merely to represent it, reason needs concepts quite different from those pure concepts of the understanding whose use is only immanent—that is, directed toward experience as far as it can be given. Concepts of reason, by contrast, aim at completeness, that is, the collective unity of the whole of possible experience, and thereby go beyond every given experience and become *transcendent*. Just as the understanding needed the categories for experience, so reason contains within itself the ground of ideas, by which I mean necessary concepts whose object nevertheless *cannot* be given in any experience. These lie in the nature of reason just as the former lie in the nature of the understanding; and if they carry with them an illusion that can easily mislead, this illusion is unavoidable, although it is quite possible to prevent it "from deceiving us." Since all illusion consists in taking the subjective ground of a judgment to be objective, pure reason's self-knowledge in its transcendent (extravagant) use will be the sole safeguard against the errors into which reason falls when it misconstrues its vocation and refers, in a transcendent way, to the object in itself what concerns only its own subject and the guidance of that subject in every immanent use. [^327-1]: If we can say that a science actually exists at least in everyone's idea as soon as it is established that the problems leading to it are posed to everyone by the nature of human reason, and that many attempts at them, though mistaken, are therefore always unavoidable, then we must also say that metaphysics is *subjectively* (and indeed necessarily) actual; and so we rightly ask how it is (*objectively*) possible. ### Section 41 § 41. Distinguishing *ideas*, that is, pure concepts of reason, from the categories, or pure concepts of the understanding, as cognitions of entirely different kinds, origins, and uses, is so important to laying the foundations of a science meant to contain the system of all these a priori cognitions that, without such a separation, metaphysics is simply impossible, or at most a disorderly, bungling attempt to patch together a house of cards without knowing the materials one is handling or their fitness for this or that purpose. Even if the Critique of Pure Reason had accomplished nothing but first bringing this distinction to light, it would already have contributed more to clarifying our concept and guiding inquiry in the field of metaphysics than all the fruitless efforts ever undertaken to satisfy the transcendent problems of pure reason. Those efforts never so much as suspected that they stood in an entirely different field from that of the understanding, and therefore named concepts of the understanding and concepts of reason in one breath, as though they were of the same kind. ### Section 42 § 42. All pure cognitions of the understanding have this characteristic: their concepts can be given in experience, and their principles confirmed by experience. Transcendent cognitions of reason, by contrast, can neither be given in experience as regards their *ideas*, nor can their propositions ever be confirmed or refuted by experience. Any error that may creep into them can therefore be exposed by nothing but pure reason itself. This, however, is very difficult, because precisely this reason becomes naturally dialectical through its ideas, and this unavoidable illusion can be kept within bounds not by any objective and dogmatic investigations of things, but only by subjective investigations of reason itself as a source of ideas. ### Section 43 § 43. Throughout the Critique, my chief concern has been not only to distinguish the kinds of cognition carefully, but also to derive all the concepts belonging to each from their common source. In this way, knowing where they originate would not only enable me to determine their use with certainty; I would also gain the previously unsuspected but inestimable advantage of knowing, a priori and thus from principles, that the enumeration, classification, and specification of these concepts were complete. Without this, everything in metaphysics is mere rhapsody: one never knows whether what one possesses is enough, or whether and where something may still be missing. Admittedly, this advantage is possible only in pure philosophy; but it also constitutes the very essence of that philosophy. Since I had found the origin of the categories in the four logical functions of all judgments of the understanding, it was entirely natural to seek the origin of ideas in the three functions of syllogistic inferences. For once such pure concepts of reason (transcendental ideas) are given, then, unless one chooses to regard them as innate, they can surely be found nowhere but in that same act of reason which, insofar as it concerns form alone, constitutes the logical element in syllogistic inferences, but which, insofar as it represents judgments of the understanding as determined a priori with regard to one or another form, constitutes transcendental concepts of pure reason. The formal difference between syllogistic inferences requires their division into categorical, hypothetical, and disjunctive. The concepts of reason founded on them therefore contain, first, the idea of the complete subject (the substantial); second, the idea of the complete series of conditions; and third, the determination of all concepts in the idea of a complete totality of the possible.[^330-1] The first idea was psychological, the second cosmological, the third theological; and since all three occasion a dialectic, though each in its own way, this grounded the division of the whole dialectic of pure reason into its paralogism, its antinomy, and finally its ideal. This derivation provides complete assurance that all the claims of pure reason are presented here in their entirety, and that not a single one can be missing, because reason itself, the faculty from which they all originate, is thereby measured out in full. [^330-1]: In a disjunctive judgment we consider *all possibility* as divided in relation to a certain concept. The ontological principle of the thoroughgoing determination of a thing in general—of all possible opposed predicates, one belongs to every thing—which is also the principle of all disjunctive judgments, takes as its basis the totality of all possibility, within which the possibility of every thing in general is regarded as determinable. This serves to explain briefly the proposition above: that reason's act in disjunctive syllogisms is, in form, the same as the act by which it brings about the idea of a totality of all reality, containing within itself the positive element of all mutually opposed predicates. ### Section 44 § 44. In considering this matter generally, it is also worth noting that ideas of reason do not, like the categories, offer us any aid in using the understanding with regard to experience. They are entirely dispensable for that purpose, and may even oppose and obstruct the maxims of rational cognition of nature; yet they are necessary in another respect, still to be determined. Whether or not the soul is a simple substance can be entirely indifferent to us in explaining its appearances. For no possible experience allows us to make the concept of a simple being sensible, and thus intelligible in concreto; so, with regard to any hoped-for insight into the cause of appearances, it is entirely empty and cannot serve as a principle for explaining what inner or outer experience provides. Nor can the cosmological ideas of the world's beginning or its eternity (a parte ante) help us explain any event in the world itself. Finally, a sound maxim of natural philosophy requires us to refrain from any explanation of nature's arrangement drawn from the will of a supreme being, because this is no longer natural philosophy, but an admission that we have reached the end of it. These ideas therefore have an entirely different intended use from those categories through which, together with the principles built on them, experience itself first became possible. Yet our laborious analytic of the understanding would also be entirely superfluous if our aim were nothing but the mere cognition of nature as it can be given in experience. For reason performs its business quite securely and well, both in mathematics and in natural science, without all this subtle deduction. Our critique of the understanding thus joins with the ideas of pure reason in an aim that lies beyond the understanding's use in experience, although we have said above that, in this regard, such a use is entirely impossible and without object or meaning. Nevertheless, there must be agreement between what belongs to the nature of reason and what belongs to that of the understanding; and the former must contribute to the perfection of the latter and cannot possibly confuse it. The solution to this question is as follows. In its ideas, pure reason does not aim at particular objects lying beyond the field of experience; it demands only completeness in the use of the understanding within the connected whole of experience. This completeness, however, can only be a completeness of principles, not of intuitions and objects. Yet to represent that completeness determinately, reason thinks of it as the cognition of an object whose cognition is completely determined with regard to those rules. This object, however, is only an idea, intended to bring cognition by the understanding as close as possible to the completeness that the idea designates. ### Section 45 § 45. *Preliminary Remark* *on the Dialectic of Pure Reason.* We have shown above, in §§ 33 and 34, that the categories' purity from any admixture of sensible determinations can tempt reason to extend their use entirely beyond all experience, to things in themselves. Yet because the categories themselves find no intuition that could give them meaning and sense in concreto, they can, as merely logical functions, represent a thing in general, but cannot by themselves supply a determinate concept of any thing. Such extravagant objects are what are called *noumena*, or pure beings of the understanding (better: beings of thought): for example, *substance* thought *without permanence in time*, or a *cause* that acted *outside time*, and so on. We thus attribute predicates to them that serve only to make the lawfulness of experience possible, while removing from them all the conditions of intuition under which alone experience is possible; as a result, those concepts once again lose all meaning. There is, however, no danger that the understanding, of its own accord and without being driven by laws foreign to it, will stray so wantonly beyond its bounds into the field of mere beings of thought. But when reason, which cannot be fully satisfied with any experiential use of the understanding's rules because that use is always still conditioned, demands the completion of this chain of conditions, the understanding is driven beyond its own sphere. It is driven partly to represent objects of experience in a series so extended that no experience can encompass it, and partly even, in order to complete that series, to seek *noumena* entirely outside experience. To these reason may attach that chain and so at last become independent of the conditions of experience, while still making the chain's support complete. These, then, are the transcendental ideas. Although the true but hidden purpose of our reason's natural constitution may have designed them not for extravagant concepts, but solely for an unlimited extension of the use of experience, they nevertheless, through an unavoidable illusion, entice the understanding into a *transcendent* use. Deceptive though this use is, it can be restrained not by any resolve to stay within the bounds of experience, but only by scientific instruction, and with difficulty. ### Section 46 § 46. I. *Psychological Ideas.* (Critique, p. 341 ff.) It has long been observed that in all substances the actual subject—that which remains after all accidents, as predicates, have been separated off, and thus the *substantial* itself—is unknown to us; and there have been many complaints about these limits to our insight. But we should note here that the human understanding is not to be called to account for not knowing the substantial in things, that is, for being unable to determine it by itself, but rather for demanding to cognize it determinately, although it is a mere idea, as though it were a given object. Pure reason demands that for every predicate of a thing we seek the subject belonging to it; and for this subject, which in turn is necessarily only a predicate, we seek its subject, and so on to infinity, or as far as we can reach. But it follows that we must regard nothing we can reach as an ultimate subject, and that the substantial itself can never be thought by our understanding, however deeply it penetrates, even if all of nature were laid open to it. For the specific nature of our understanding consists in thinking everything discursively, that is, through concepts, and thus through nothing but predicates, for which the absolute subject must therefore always be lacking. Hence all the real properties through which we cognize bodies are merely accidents, even impenetrability, which we must always represent to ourselves only as the effect of a force for which we lack the subject. Now it seems as though in our consciousness of ourselves (the thinking subject) we possess this substantial, and indeed in an immediate intuition. For all predicates of inner sense refer to the I as subject, and this can no longer be thought as the predicate of any other subject. Here, then, completeness in the reference of given concepts as predicates to a subject seems to be not merely an idea: the object, namely the *absolute subject* itself, seems to be given in experience. But this expectation is frustrated. For the I is not a concept at all,[^334-1] but only a designation of the object of inner sense insofar as we do not cognize it further through any predicate. Thus, although in itself it cannot be a predicate of another thing, neither can it be a determinate concept of an absolute subject; rather, as in all other cases, it can only be the reference of inner appearances to their unknown subject. Nevertheless, through a wholly natural misunderstanding, this idea—which serves very well, as a regulative principle, to destroy entirely all materialistic explanations of the inner appearances of our soul[^334-1]—gives rise to a highly plausible argument for inferring, from this supposed cognition of the substantial in our thinking being, its nature insofar as knowledge of that nature lies wholly beyond the compass of experience. [^334-1]: If the representation of apperception, the I, were a concept through which anything were thought, it could also be used as a predicate of other things, or contain such predicates within itself. But it is nothing more than the feeling of an existence without the slightest concept, and only the representation of that to which all thinking stands in relation (relatione accidentis). ### Section 47 § 47. Now this thinking self (the soul), as the ultimate subject of thought, which cannot itself be represented in turn as the predicate of another thing, may indeed be called substance. Yet this concept remains entirely empty and without consequences unless permanence can be proved of it, as that which makes the concept of substances fruitful in experience. But permanence can never be proved from the concept of a substance as a thing in itself, only for the sake of experience. This has been sufficiently demonstrated in the first analogy of experience (Critique, p. 182); and anyone unwilling to accept that proof need only try for themselves whether it is possible, from the concept of a subject that does not itself exist as the predicate of another thing, to prove that its existence is absolutely permanent, and that it can neither arise nor perish, either of itself or through any natural cause. Such synthetic a priori propositions can never be proved in themselves, but always only in relation to things as objects of a possible experience. ### Section 48 § 48. If, then, we wish to infer the soul's permanence from its concept as substance, this can hold of it only for the sake of possible experience, not of it as a thing in itself and beyond all possible experience. Now the subjective condition of all our possible experience is life. We can therefore infer only the permanence of the soul during life; for a human being's death is the end of all experience as regards the soul as an object of experience, unless the contrary is demonstrated—and that is precisely what is in question. Thus the soul's permanence can be demonstrated only during a human being's life—a proof we may surely be excused from giving—but not after death, which is what actually concerns us. The general ground for this is that the concept of substance, insofar as it is to be regarded as necessarily connected with the concept of permanence, can be so connected only according to a principle of possible experience, and thus only for the sake of experience.[^335-1] [^335-1]: It is indeed very remarkable that metaphysicians have always slipped so heedlessly past the principle of the permanence of substances without ever attempting a proof of it; doubtless because, as soon as they began with the concept of substance, they found themselves wholly deserted by all grounds of proof. Common understanding, which was well aware that without this presupposition no unification of perceptions in one experience was possible, supplied the deficiency with a postulate. For it could never have drawn this principle from experience itself, partly because experience cannot follow materials (substances) through all their changes and dissolutions far enough to find the matter always undiminished, and partly because the principle contains *necessity*, which is always the mark of an a priori principle. They now confidently applied this principle to the concept of the soul as a *substance* and inferred its necessary continuance after a human being's death—especially since the simplicity of this substance, inferred from the indivisibility of consciousness, secured it against destruction through dissolution. Had they found the genuine source of this principle, which would have required far deeper investigations than they ever cared to begin, they would have seen that this law of the permanence of substances holds only for the sake of experience, and therefore can apply only to things insofar as they are to be cognized in experience and connected with others. It can never apply to them apart from all possible experience, and hence cannot apply to the soul after death. ### Section 49 § 49. That something actual outside us not only corresponds, but must correspond, to our outer perceptions can likewise never be proved as a connection of things in themselves, though it can be proved for the sake of experience. This means that we can very well prove that something is outside us in an empirical manner, and thus as an appearance in space. For we are concerned with no objects other than those belonging to a possible experience, precisely because other objects cannot be given to us in any experience and are therefore nothing for us. What is intuited in space is empirically outside me. And since space, together with all the appearances it contains, belongs to representations whose connection according to laws of experience proves their objective truth just as the connection of the appearances of inner sense proves the actuality of my soul as an object of inner sense, I am, through outer experience, just as conscious of the actuality of bodies as outer appearances in space as I am, through inner experience, of the existence of my soul in time. I can cognize this soul, too, only as an object of inner sense through appearances that constitute an inner state, while its essence in itself, underlying these appearances, remains unknown to me. *Cartesian* idealism thus distinguishes only outer experience from dreaming, and lawfulness, as a criterion of the truth of the former, from the disorder and deceptive illusion of the latter. In both, it presupposes space and time as conditions of the existence of objects, and asks only whether the objects of the outer senses are actually to be found in the space in which we place them while awake, just as the object of inner sense, the soul, is actually in time; that is, whether experience carries with it reliable criteria for distinguishing it from imagination. Here the doubt is easily removed, and in ordinary life we always remove it by examining the connection of appearances in both according to universal laws of experience. If the representation of outer things agrees throughout with these laws, we cannot doubt that they constitute genuine experience. Material idealism can thus very easily be removed when appearances, as appearances, are considered only according to their connection in experience; and it is just as certain an experience that bodies exist outside us (in space) as that I myself exist according to the representation of inner sense (in time). For the concept *outside us* means only existence in space. But since the I in the proposition *I am* means not merely the object of inner intuition (in time), but the subject of consciousness, just as body means not merely outer intuition (in space), but also the thing *in itself* that underlies this appearance, the question whether bodies, as appearances of outer sense, exist as bodies *outside my thoughts* can be denied without any hesitation in the study of nature. Yet the matter is no different with the question whether I myself, *as an appearance of inner sense*—the soul according to empirical psychology—exist in time outside my power of representation; for this must likewise be denied. In this way everything, once brought to its true meaning, is settled and certain. Formal idealism, which I have elsewhere called transcendental idealism, actually abolishes material or *Cartesian* idealism. For if space is nothing but a form of my sensibility, then as a representation it is just as actual in me as I myself am, and only the empirical truth of the appearances in it remains at issue. But if that is not so, and space and the appearances in it are something existing outside us, then all the criteria of experience can never prove the actuality of these objects outside us, apart from our perception. ### Section 50 § 50. II. *Cosmological Ideas.* (Critique, p. 405 ff.) This product of pure reason in its transcendent use is its most remarkable phenomenon, and of all phenomena it is the most powerful in awakening philosophy from its dogmatic slumber and moving it to the difficult business of critiquing reason itself. I call this idea cosmological because it always takes its object only from the world of sense, and needs no idea other than one whose object belongs to the senses. To that extent it is at home, not transcendent, and thus is not yet an idea. By contrast, to think of the soul as a simple substance already means thinking of an object—the simple—the like of which cannot be represented to the senses at all. Nevertheless, the cosmological idea extends the connection of the conditioned with its condition, whether mathematical or dynamical, so far that experience can never equal it; and in this respect it is always an idea whose object can never be adequately given in any experience. ### Section 51 § 51. Here, first of all, the usefulness of a system of categories appears so clearly and unmistakably that, even if there were no further proofs of it, this alone would sufficiently demonstrate their indispensability in the system of pure reason. There are no more than four such transcendent ideas, as many as there are classes of categories; but in each class they aim only at the absolute completeness of the series of conditions for a given conditioned thing. In accordance with these cosmological ideas, there are also only four kinds of dialectical assertions of pure reason. Since they are dialectical, they themselves prove that each is opposed by an assertion contradicting it on equally plausible principles of pure reason. No metaphysical art of the subtlest distinction can prevent this conflict; rather, it forces the philosopher to return to the first sources of pure reason itself. This antinomy, which is not arbitrarily invented but grounded in the nature of human reason, and is therefore unavoidable and never-ending, contains the following four theses together with their antitheses. | | Thesis | Antithesis | |---|---|---| | 1. | The world has a *beginning* (boundary) in time and space | The world is *infinite* in time and space | | 2. | Everything in the world consists of the *simple* | Nothing is simple; everything is *composite* | | 3. | There are causes in the world through *freedom* | There is no freedom; everything is *nature* | | 4. | In the series of causes in the world there is some *necessary* being | Nothing in it is necessary; in this series *everything is contingent*. | ### Section 52 § 52. Here, then, is the strangest phenomenon of human reason, of which no example can be shown in any other use of it. If, as usually happens, we think of the appearances of the world of sense as things in themselves; if we assume the principles of their connection to be principles holding universally of things in themselves, and not merely of experience—as is equally usual and indeed unavoidable without our critique—then an unsuspected conflict emerges that can never be settled by the ordinary dogmatic path. For both thesis and antithesis can be demonstrated by equally evident, clear, and irresistible proofs—I vouch for the correctness of all these proofs—and reason thus finds itself divided against itself: a state in which the skeptic exults, but which must trouble the critical philosopher and set him thinking. ### Section 52b § 52b. One can bungle metaphysics in all sorts of ways without having much to fear from being caught in a falsehood. For so long as we do not contradict ourselves—which is quite possible in synthetic propositions even when they are entirely invented—we can never be refuted by experience in any case where the concepts we connect are mere ideas that cannot, in their full content, be given in experience at all. For how could we decide through experience whether the world has existed from eternity or had a beginning; whether matter is divisible to infinity or consists of simple parts? Such concepts cannot be given in any experience, even the greatest possible; hence this touchstone cannot reveal the incorrectness of either the affirmative or the negative proposition. The only possible case in which reason would reveal against its will its secret dialectic, which it falsely passes off as dogmatics, would be if it grounded an assertion on a universally accepted principle and, from another equally accredited principle, inferred precisely the opposite with the utmost correctness of inference. This case is now actual here, and indeed in regard to four natural ideas of reason, from which four assertions on one side and as many counterassertions on the other arise, each as a correct consequence of universally accepted principles. They thereby reveal the dialectical illusion of pure reason in the use of these principles, which would otherwise have to remain hidden forever. Here, then, is a decisive test that must necessarily disclose an error hidden in reason's presuppositions.[^341-1] Of two mutually contradictory propositions, both cannot be false unless the concept underlying both is itself contradictory. For example, the two propositions "a square circle is round" and "a square circle is not round" are both false. As regards the first, it is false that the circle in question is round, because it is square; but it is also false that it is not round, that is, angular, because it is a circle. For the logical mark of a concept's impossibility consists precisely in this: on the presupposition of that concept, two contradictory propositions would be false at once. Thus, since no third possibility can be thought between them, nothing at all is thought through that concept. [^341-1]: I therefore wish the critical reader to concern themselves chiefly with this antinomy, because nature itself seems to have set it up to make reason pause in its bold pretensions and compel it to examine itself. I undertake to answer for every proof I have given, both of the thesis and of the antithesis, and thereby to demonstrate the certainty of reason's unavoidable antinomy. If this strange phenomenon now leads the reader back to examine the presupposition underlying it, they will feel compelled to join me in investigating more deeply the first foundation of all cognition of pure reason. ### Section 52c § 52c. Now just such a contradictory concept underlies the first two antinomies, which I call mathematical because they deal with adding or dividing the homogeneous; and from this I explain how it happens that both thesis and antithesis are false in each. When I speak of objects in time and space, I am not speaking of things in themselves, because I know nothing of them, but only of things in appearance—that is, of experience as a particular kind of cognition of objects, the only kind granted to human beings. Of what I think in space or in time, I must not say that in itself, even without this thought of mine, it is in space and time. For then I would contradict myself, because space and time, together with the appearances in them, are not something existing in itself and outside my representations, but are themselves only modes of representation; and it is plainly contradictory to say that a mere mode of representation also exists outside our representation. Objects of the senses thus exist only in experience; to give them, by contrast, an existence of their own, subsisting by itself even without experience or before it, means as much as to represent experience as actual even without or before experience. If I now ask about the magnitude of the world in space and time, it is, for all my concepts, just as impossible to say that it is infinite as that it is finite. For neither can be contained in experience, because no experience is possible either of an *infinite* space or of an infinitely elapsed time, or of the world's *limitation* by an empty space or by a preceding empty time. These are only ideas. Thus this magnitude of the world, determined in one way or the other, would have to lie in the world itself, apart from all experience. But this contradicts the concept of a world of sense, which is only a totality of appearance, whose existence and connection occur only in representation, namely experience, because it is not a thing in itself but itself nothing but a mode of representation. It follows that, since the concept of a world of sense existing by itself is self-contradictory, any solution to the problem of its magnitude will always be false, whether attempted affirmatively or negatively. The very same holds of the second antinomy, which concerns the division of appearances. For these are mere representations, and the parts exist only in the representation of them, hence in the division—that is, in a possible experience in which they are given—and the division therefore goes only as far as this experience reaches. To assume that an appearance, for example that of a body, contains in itself, before all experience, all the parts that any possible experience can ever reach, means giving a mere appearance, which can exist only in experience, an existence of its own that precedes experience; or saying that mere representations exist before they are encountered in the power of representation. This is self-contradictory, and so too is every solution of the misunderstood problem, whether one asserts that bodies in themselves consist of infinitely many parts or of a finite number of simple parts. ### Section 53 § 53. In the first class of antinomy, the mathematical, the falsity of the presupposition consisted in representing what is contradictory—namely appearance as a thing in itself—as compatible within one concept. In the second, or dynamical, class of antinomy, however, the falsity of the presupposition consists in representing what is compatible as contradictory. Consequently, whereas in the first case both opposed assertions were false, here, by contrast, assertions opposed to each other through mere misunderstanding can both be true. Mathematical connection necessarily presupposes homogeneity in what is connected, in the concept of magnitude; dynamical connection by no means requires this. When the magnitude of something extended is at issue, all its parts must be homogeneous with one another and with the whole. In the connection of cause and effect, by contrast, homogeneity may indeed also be found, but it is not necessary; for the concept of causality, through which something entirely different is posited by something, at least does not require it. If the objects of the world of sense were taken for things in themselves, and the laws of nature cited above for laws of things in themselves, the contradiction would be unavoidable. Likewise, if the subject of freedom were represented, like the other objects, as mere appearance, the contradiction could no more be avoided; for the very same thing would be both affirmed and denied of one and the same object in the same sense. But if natural necessity is referred solely to appearances, and freedom solely to things in themselves, no contradiction arises even if both kinds of causality are assumed or admitted, however difficult or impossible it might be to make the latter kind comprehensible. In appearance, every effect is an event, or something that happens in time. According to the universal law of nature, a determination of the causality of its cause—a state of that cause—must precede it, upon which it follows according to a constant law. But this determination of the cause to causality must also be something that occurs or *happens*: the cause must have *begun* to *act*, for otherwise no succession in time could be thought between it and the effect. The effect would always have existed, just like the causality of the cause. Thus, among appearances, the *determination* of the cause *to act* must also have arisen, and must therefore, just like its effect, be an event that must in turn have its cause, and so on. Natural necessity must consequently be the condition according to which efficient causes are determined. If, on the other hand, freedom is to be a property of certain causes of appearances, then in relation to these appearances as events it must be a capacity to begin them *of itself* (sponte), that is, without the causality of the cause itself having to begin, and hence without its needing any other ground to determine its beginning. But in that case the *cause*, with regard to its causality, would have to stand outside temporal determinations of its state; that is, it could not be an *appearance* at all. It would have to be assumed as a thing in itself, and only its *effects* as *appearances*.[^344-1] If such an influence of beings of the understanding on appearances can be thought without contradiction, then natural necessity will indeed attach to every connection of cause and effect in the world of sense. Yet freedom can be conceded to the cause that is itself no appearance, although it underlies appearance; thus nature and freedom can be attributed without contradiction to the very same thing, but in different relations: in the one case as appearance, in the other as a thing in itself. We possess within us a faculty that is connected not only with its subjectively determining grounds, which are the natural causes of its actions, and in this respect is the faculty of a being that itself belongs among appearances. It is also referred to objective grounds that are mere ideas, insofar as they can determine this faculty—a connection expressed by *ought*. This faculty is called *reason*; and insofar as we consider a being, a human being, solely according to this objectively determinable reason, it cannot be regarded as a sensible being. Rather, the property in question is the property of a thing in itself. We cannot at all comprehend its possibility: how, namely, the *ought*, which has never yet happened, could determine its activity and be the cause of actions whose effect is appearance in the world of sense. Nevertheless, the causality of reason with regard to effects in the world of sense would be freedom, insofar as *objective grounds*, which are themselves ideas, are regarded as determining it. For its action would then depend on no subjective conditions, hence on no temporal conditions either, and thus not on the law of nature that serves to determine these. For grounds of reason give actions their rule universally, from principles, without the influence of circumstances of time or place. What I offer here serves only as an example to aid understanding, and does not necessarily belong to our question, which must be decided from mere concepts, independently of properties we encounter in the actual world. Now I can say without contradiction: all actions of rational beings, insofar as they are appearances, encountered in any experience, stand under natural necessity; yet these very same actions, considered solely in relation to the rational subject and its capacity to act from reason alone, are free. For what is required for natural necessity? Nothing more than the determinability of every event in the world of sense according to constant laws, hence a relation to a cause in appearance, while the underlying thing in itself and its causality remain unknown. But I say: the *law of nature remains*, whether the rational being is the cause of effects in the world of sense from reason, hence through freedom, or whether it does not determine these effects from rational grounds. In the first case, the action occurs according to maxims whose effect in appearance will always accord with constant laws; in the second, where the action does not occur according to principles of reason, it is subject to the empirical laws of sensibility. In both cases the effects are connected according to constant laws; but we demand nothing more for natural necessity, and indeed know nothing more about it. In the first case, however, reason is the cause of these laws of nature and is therefore free; in the second, the effects proceed according to mere natural laws of sensibility, because reason exercises no influence on them. But reason is not itself thereby determined by sensibility—which is impossible—and is therefore free in this case as well. Freedom thus does not hinder the natural law of appearances, any more than this law impairs the freedom of reason's practical use, which stands in connection with things in themselves as determining grounds. In this way practical freedom—namely the freedom in which reason has causality according to objectively determining grounds—is preserved without the slightest detriment to natural necessity with regard to these very same effects as appearances. This can also serve to explain what we had to say about transcendental freedom and its compatibility with natural necessity, taken in the same subject but not in one and the same relation. For as regards this freedom, every beginning of a being's action from objective causes is always a first beginning relative to these determining grounds, although the very same action in the series of appearances is only a *subordinate beginning*, which must be preceded by a state of the cause that determines it and is itself likewise determined by a still earlier state. Thus we can think, without contradicting laws of nature, of a capacity in rational beings—or in beings generally insofar as their causality is determined in them as things in themselves—to begin a series of states spontaneously. For the relation of the action to objective grounds of reason is not a temporal relation: here what determines causality does not precede the action in time, because such determining grounds represent not a relation of objects to the senses, hence not to causes in appearance, but determining causes as things in themselves, which do not stand under temporal conditions. Thus the action can be regarded as a first beginning with respect to the causality of reason, yet at the same time as a merely subordinate beginning with respect to the series of appearances; and without contradiction it can be regarded in the former respect as free, and in the latter, where it is mere appearance, as subject to natural necessity. As for the *fourth* antinomy, it is removed in a way similar to the conflict of reason with itself in the third. For if only the *cause in appearance* is distinguished from the *cause of appearances*, insofar as this can be thought as a *thing in itself*, both propositions can very well stand together: that nowhere in the world of sense is there a cause, according to similar laws of causality, whose existence is absolutely necessary; and, on the other hand, that this world is nevertheless connected with a necessary being as its cause, but of another kind and according to another law. The incompatibility of these two propositions rests solely on the misunderstanding of extending to things in themselves what holds only of appearances, and of generally confusing both in one concept. [^344-1]: The idea of freedom applies solely in the relation of the *intellectual* as cause to *appearance* as effect. Hence we cannot attribute freedom to matter with regard to its ceaseless action by which it fills its space, even though this action occurs from an inner principle. Nor can we find an appropriate concept of freedom for pure beings of the understanding, for example God, insofar as his action is immanent. For his action, although independent of external determining causes, is nevertheless determined in his eternal reason, hence in the divine *nature*. Only when *something is to begin* through an action, and the effect is therefore to be found in the temporal series, hence in the world of sense—for example, the beginning of the world—does the question arise whether the causality of the cause must itself also begin, or whether the cause can initiate an effect without its causality itself beginning. In the first case, the concept of this causality is a concept of natural necessity; in the second, of freedom. From this the reader will see that, in explaining freedom as the capacity to begin an event of itself, I hit precisely upon the concept that is the problem of metaphysics. ### Section 54 § 54. This, then, is the exposition and resolution of the entire antinomy in which reason finds itself entangled when applying its principles to the world of sense. The former—the mere exposition—would even by itself be a considerable contribution to knowledge of human reason, although this resolution of the conflict may not yet have fully satisfied the reader, who here has to combat a natural illusion only recently presented as such, after always having taken it for truth. For one consequence of this is nevertheless inevitable. Since it is entirely impossible to escape this conflict of reason with itself so long as one takes the objects of the world of sense for things in themselves, and not for what they actually are, namely mere appearances, the reader will be compelled to undertake once again the deduction of all our a priori cognition, and the examination of the deduction I have given, in order to reach a decision about it. I ask no more for now. For once the reader, in this occupation, has first thought deeply enough into the nature of pure reason, the concepts through which alone the resolution of reason's conflict is possible will already be familiar; without this, I cannot expect full assent even from the most attentive reader. ### Section 55 § 55. III. *Theological Idea.* (Critique, p. 571 ff.) The third transcendental idea, which supplies material for the most important use of reason—a use that, if pursued merely speculatively, is extravagant (transcendent), and for that very reason dialectical—is the ideal of pure reason. Here reason does not, as with the psychological and cosmological ideas, begin from experience and find itself tempted, by ascending through grounds, to strive if possible for the absolute completeness of their series. Instead, it breaks off entirely and descends from mere concepts of what would constitute the absolute completeness of a thing in general—hence by means of the idea of a supremely perfect original being—to determine the possibility, and thus also the actuality, of all other things. Here, therefore, the mere presupposition of a being that is thought, though not in the series of experience, nevertheless for the sake of experience, to make its connection, order, and unity comprehensible—that is, the idea—is easier to distinguish from a concept of the understanding than in the previous cases. Hence the dialectical illusion could easily be brought to light here: it arises from our taking the subjective conditions of our thinking for objective conditions of things themselves, and a hypothesis necessary to satisfy our reason for a dogma. I therefore have nothing further to observe about the pretensions of transcendental theology, since what the Critique says on this subject is intelligible, evident, and decisive. ### Section 56 § 56. *General Remark* *on the Transcendental Ideas.* The objects given to us through experience are incomprehensible to us in many respects. Many questions to which the law of nature leads us cannot be solved at all when pursued to a certain height, though always in accordance with these laws—for example, why materials attract one another. But when we leave nature entirely, or in pursuing its connections exceed all possible experience, and thus immerse ourselves in mere ideas, we cannot then say that the object is incomprehensible to us, and that the nature of things poses insoluble problems for us. For then we are not dealing with nature or with given objects at all, but solely with concepts originating only in our reason, and with mere beings of thought. As regards these, all problems arising from their concept must be capable of solution, because reason certainly can and must give a complete account of its own procedure.[^349-1] Since the psychological, cosmological, and theological ideas are nothing but pure concepts of reason that cannot be given in any experience, the questions reason poses about them are set for us not by the objects, but by mere maxims of reason for the sake of its own satisfaction; and they must all be capable of a sufficient answer. This is indeed achieved by showing that they are principles for bringing our use of the understanding to thoroughgoing harmony, completeness, and synthetic unity, and that to this extent they hold only of experience, but of experience as a *whole*. Although an absolute whole of experience is impossible, the idea of a whole of cognition according to principles in general is nevertheless what alone can give cognition a particular kind of unity, namely that of a system. Without this, our cognition is nothing but fragments and cannot be used for the highest purpose, which is always only the system of all purposes. Here, however, I mean not merely the practical purpose, but also the highest purpose of the speculative use of reason. The transcendental ideas thus express reason's distinctive vocation, namely as a principle of the systematic unity of the use of the understanding. But if we regard this unity of the mode of cognition as though it attached to the object of cognition; if we take what is properly only *regulative* to be *constitutive*, and persuade ourselves that by means of these ideas we can extend our knowledge far beyond all possible experience, hence in a transcendent manner, although that unity serves only to bring experience within itself as close as possible to completeness—that is, to restrict its progress by nothing that cannot belong to experience—then this is a mere misunderstanding in judging the proper vocation of our reason and its principles. It is a dialectic that partly confuses reason's use in experience and partly divides reason against itself. *Conclusion: Determining the Boundaries of Pure Reason.* [^349-1]: Thus Mr. *Platner* says acutely in his Aphorisms, §§ 728, 729: "If reason is a criterion, then no concept can be possible that is incomprehensible to human reason.—Only in the actual is there incomprehensibility. Here incomprehensibility arises from the inadequacy of the ideas acquired."—It therefore only sounds paradoxical, and is otherwise not surprising, to say that much in nature is incomprehensible to us—for example, the power of generation—but that if we ascend still higher and even go beyond nature, everything becomes comprehensible to us again. For then we entirely leave behind the *objects* that can be given to us, and concern ourselves solely with ideas. With these we can very well comprehend the law that reason, through them, prescribes to the understanding for its use in experience, because this law is reason's own product. ### Section 57 § 57. After the clearest proofs we have given above, it would be absurd to hope to cognize more of any object than belongs to possible experience of it, or to claim even the slightest cognition of anything we assume not to be an object of possible experience, in order to determine it according to its constitution as it is in itself. For by what means would we make this determination? Time, space, and all concepts of the understanding—and still more the concepts drawn from empirical intuition or *perception* in the world of sense—have, and can have, no use other than to make experience possible. If we remove this condition even from the pure concepts of the understanding, they then determine no object at all and have no meaning whatever. On the other hand, it would be a still greater absurdity to admit no things in themselves at all, or to pass off our experience as the only possible way of cognizing things, hence our intuition in space and time as the only possible intuition, and our discursive understanding as the archetype of every possible understanding; thus to demand that principles of the possibility of experience be taken as universal conditions of things in themselves. Our principles, which restrict the use of reason solely to possible experience, could therefore themselves become *transcendent*, and pass off the limits of our reason as limits of the possibility of things themselves—as *Hume's* Dialogues can serve to illustrate—unless a careful critique guarded the boundaries of our reason even with regard to its empirical use and set a limit to its pretensions. Skepticism originally sprang from metaphysics and its unpoliced dialectic. At first it may well have declared everything that exceeds reason's use in experience null and deceptive merely for the benefit of that use. But gradually, when it became apparent that the very same a priori principles employed in experience led imperceptibly, and seemingly with the very same right, still further than experience reaches, people began to cast doubt even on principles of experience. There is, to be sure, no danger in this, for sound understanding will surely always maintain its rights here. Yet a peculiar confusion arose in science, which cannot determine how far reason is to be trusted, and why only that far and no further. This confusion can be remedied, and any future relapse prevented, only by a formal determination, drawn from principles, of the boundaries of our use of reason. It is true: beyond all possible experience, we can give no determinate concept of what things in themselves may be. Yet we are not free to refrain entirely from asking about them. For experience never fully satisfies reason; in answering questions, it always refers us further back and leaves us dissatisfied with regard to their complete elucidation, as anyone can sufficiently see from the dialectic of pure reason, which has its sound subjective ground for precisely this reason. Who can bear our arriving, as regards the nature of our soul, at a clear consciousness of the subject and at the same time the conviction that its appearances cannot be explained *materialistically*, without asking what the soul actually is, and, if no concept of experience suffices for this, at least assuming a concept of reason—that of a simple immaterial being—solely for this purpose, although we cannot demonstrate its objective reality at all? Who can be satisfied with mere experiential cognition in all cosmological questions about the world's duration and magnitude, about freedom or natural necessity, when, however we proceed, every answer given according to principles of experience always gives birth to a new question that equally demands an answer, and thereby clearly demonstrates the inadequacy of all physical modes of explanation to satisfy reason? Finally, who does not see, in the thoroughgoing contingency and dependence of everything they can think and assume merely according to principles of experience, the impossibility of stopping there? And who does not feel compelled, despite every prohibition against losing themselves in transcendent ideas, to seek rest and satisfaction beyond all concepts they can justify through experience in the concept of a being whose idea cannot indeed be understood in itself as regards its possibility, though neither can it be refuted, because it concerns a mere being of the understanding—yet without which reason would have to remain forever unsatisfied? Boundaries, in extended beings, always presuppose a space found outside a certain determinate place and enclosing it; limits require no such thing, but are mere negations affecting a magnitude insofar as it lacks absolute completeness. Our reason, however, sees around itself, as it were, a space for the cognition of things in themselves, although it can never have determinate concepts of them and is restricted solely to appearances. So long as reason's cognition is homogeneous, no determinate boundaries of it can be thought. In mathematics and natural science human reason does indeed recognize limits, but no boundaries: it recognizes that something lies outside it, where it can never reach, but not that it will itself ever reach completion anywhere in its internal progress. The expansion of mathematical insight and the possibility of ever new discoveries extend to infinity; so too does the discovery, through continuing experience, of new properties of nature, new forces and laws, and their unification through reason. Yet limits here are nonetheless unmistakable. For mathematics concerns only *appearances*, and whatever cannot be an object of sensible intuition, such as the concepts of metaphysics and morals, lies entirely outside its sphere, and it can never lead there; nor does it need these concepts at all. There is therefore no continuous progress and approach toward these sciences, and, as it were, no point or line of contact. Natural science will never disclose to us the interior of things, that is, what is not appearance but can nevertheless serve as the supreme ground for explaining appearances; but neither does it need this for its physical explanations. Indeed, even if something of this sort were offered to it from elsewhere—for example, the influence of immaterial beings—it should reject it and not introduce it at all into the progress of its explanations. Instead, it should always ground these only on what, as an object of the senses, can belong to experience and be connected with our actual perceptions according to laws of experience. Metaphysics, however, leads us to boundaries in the dialectical attempts of pure reason—attempts not begun arbitrarily or wantonly, but driven by the very nature of reason. And the transcendental ideas, precisely because we cannot avoid them yet they can never be realized, serve not only to show us the boundaries of the use of pure reason, but also the way to determine them. This, too, is the purpose and benefit of this natural predisposition of our reason, which has brought forth metaphysics as its favorite child. Its generation, like any other in the world, must be ascribed not to chance accident, but to an original germ wisely organized for great purposes. For metaphysics, perhaps more than any other science, is implanted in us in its basic features by nature itself, and cannot be regarded at all as the product of an arbitrary choice or as a chance extension in the course of experiences, from which it separates itself entirely. With all its concepts and laws of the understanding, which suffice for empirical use and thus within the world of sense, reason nevertheless finds no satisfaction on its own. For questions recurring endlessly rob it of all hope of a complete solution. The transcendental ideas, which aim at this completion, are such problems of reason. Now reason sees clearly that the world of sense cannot contain this completion; nor, therefore, can all those concepts that serve solely to understand it: space and time and everything we have listed under the name of pure concepts of the understanding. The world of sense is nothing but a chain of appearances connected according to universal laws. It therefore has no subsistence of its own, is not properly the thing in itself, and thus necessarily refers to what contains the ground of this appearance: to beings that can be cognized not merely as appearance but as things in themselves. In cognition of them alone can reason hope at last to see its demand for completeness satisfied in the progression from the conditioned to its conditions. Above (§§ 33, 34) we indicated the limits of reason with regard to all cognition of mere beings of thought. Now that the transcendental ideas nevertheless make our progression to them necessary, and have thus led us, as it were, only to the contact between full space, experience, and empty space, the noumena of which we can know nothing, we can also determine the boundaries of pure reason. For in all boundaries there is also something positive—for example, a surface is the boundary of corporeal space, yet is itself a space; a line is a space that is the boundary of a surface; a point is the boundary of a line, yet is still a place in space—whereas limits contain mere negations. The limits indicated in the sections cited are no longer enough once we have found that something lies beyond them, although we shall never cognize what it is in itself. For the question now is: how does our reason proceed in this connection of what we know with what we do not know and shall never know? Here there is an actual connection of the known with something wholly unknown, which it will always remain; and even if the unknown should not thereby become in the least better known—as indeed there is no hope that it will—the concept of this connection must nevertheless be capable of determination and clarification. We are therefore to think of an immaterial being, an intelligible world, and a supreme being—all of them noumena—because only in these, as things in themselves, does reason find the completion and satisfaction it can never hope for in deriving appearances from grounds homogeneous with them. And appearances do actually refer to something distinct from them, hence wholly heterogeneous, since appearances always presuppose a thing in itself and therefore point to it, whether or not we cognize it more closely. But since we can never cognize these beings of the understanding according to what they may be in themselves—that is, determinately—and yet must nevertheless assume them in relation to the world of sense and connect them with it through reason, we can at least think this connection by means of concepts expressing their relation to the world of sense. For if we think a being of the understanding through nothing but pure concepts of the understanding, we actually think nothing determinate, and our concept is therefore without meaning. If we think it through properties borrowed from the world of sense, it is no longer a being of the understanding; it is thought as one of the phenomena and belongs to the world of sense. Let us take an example from the concept of the supreme being. The *deistic* concept is an entirely pure concept of reason, but it represents only a thing containing all reality, without being able to determine any single reality. For to do that, an example would have to be borrowed from the world of sense, in which case I would always be dealing only with an object of the senses, not with something wholly heterogeneous that cannot be an object of the senses at all. For example, I would attribute understanding to it; but I have no concept whatever of an understanding except one like my own, namely one to which intuitions must be given through the senses, and which occupies itself with bringing them under rules of the unity of consciousness. But then the elements of my concept would always lie in appearance, whereas it was precisely the inadequacy of appearances that compelled me to go beyond them to the concept of a being not at all dependent on appearances, or intertwined with them as conditions of its determination. But if I separate understanding from sensibility in order to have a pure understanding, nothing remains but the mere form of thinking without intuition, through which alone I can cognize nothing determinate, hence no object. For that purpose I would have to think of another understanding that intuited objects, of which, however, I have not the slightest concept, because human understanding is discursive and can cognize only through universal concepts. The same happens if I attribute a will to the supreme being. For I have this concept only by drawing it from my inner experience; but underlying it is the dependence of my satisfaction on objects whose existence we need, and thus sensibility, which entirely contradicts the pure concept of the supreme being. *Hume's* objections to deism are weak and never touch anything more than the proofs, never the proposition of the deistic assertion itself. But as regards theism, which is supposed to arise through a closer determination of our concept of the supreme being, there merely transcendent, they are very strong and, depending on how this concept is framed, irrefutable in certain cases—indeed, in all the usual ones. Hume always insists on this: through the mere concept of an original being to which we attribute only ontological predicates—eternity, omnipresence, omnipotence—we actually think nothing determinate at all; properties must be added that can yield a concept in concreto. It is not enough to say that it is a cause; one must say what its causality is like, perhaps through understanding and will. Here his attacks on the matter itself, namely theism, begin, whereas before he had merely stormed the grounds of proof for deism, which entails no particular danger. His dangerous arguments all concern anthropomorphism, which he holds to be inseparable from theism and to make it self-contradictory; but if anthropomorphism were omitted, theism would fall with it, leaving nothing but a deism from which nothing can be made, which can be of no use to us and serve as no foundation whatever for religion and morals. If this inevitability of anthropomorphism were certain, then, whatever the proofs of the existence of a supreme being might be, and even if they were all conceded, we could still never determine the concept of this being without entangling ourselves in contradictions. If we combine the prohibition requiring us to avoid all transcendent judgments of pure reason with the command, apparently conflicting with it, to go out as far as concepts lying outside the field of immanent, or empirical, use, we become aware that both can coexist, but only precisely on the *boundary* of every permissible use of reason. For this boundary belongs just as much to the field of experience as to that of beings of thought. At the same time, we are thereby taught how those remarkable ideas serve solely to determine the boundaries of human reason: on the one hand, not to extend experiential cognition without limit, so that nothing but the world would remain for us to cognize; and on the other, nevertheless not to go beyond the boundary of experience and seek to judge things outside it as things in themselves. We keep to this boundary, however, when we restrict our judgment solely to the relation the world may have to a being whose very concept lies outside all cognition of which we are capable within the world. For then we attribute to the supreme being none of the properties *in themselves* through which we think objects of experience, and thereby avoid *dogmatic* anthropomorphism. Yet we do attribute them to its relation to the world, and allow ourselves a *symbolic* anthropomorphism, which in fact concerns only language and not the object itself. When I say that we are compelled to regard the world *as if* it were the work of a supreme understanding and will, I actually say no more than this: as a watch, a ship, a regiment stands to its maker, builder, commander, so the world of sense—or everything that constitutes the foundation of this totality of appearances—stands to the unknown. Through this I cognize the unknown not indeed according to what it is in itself, but according to what it is for me, namely with regard to the world of which I am a part. ### Section 58 § 58. Such cognition is cognition *by analogy*. This does not mean, as the word is commonly taken, an imperfect similarity between two things, but a perfect similarity between two relations in wholly dissimilar things.[^357-1] Through this analogy there still remains a concept of the supreme being sufficiently determinate for us, although we have omitted everything that could determine it absolutely and *in itself*. For we do determine it in relation to the world, and thus to ourselves, and we need nothing more. *Hume's* attacks on those who wish to determine this concept absolutely, borrowing the materials for it from themselves and the world, do not touch us. Nor can he reproach us with having nothing left once objective anthropomorphism is removed from our concept of the supreme being. For if we are only granted at the outset—as Hume, too, speaking as Philo against Cleanthes in his Dialogues, allows—the *deistic* concept of the original being as a necessary hypothesis, in which the original being is thought through purely ontological predicates of substance, cause, etc., then nothing can prevent us from predicating of this being a *causality through reason* with regard to the world, and thus passing over to theism, without being compelled to attribute this reason to it *in itself* as an inherent property. *We must make this concession*, because reason, driven through the world of sense by nothing but conditions that are always conditioned in turn, can have no satisfaction at all without it; and *we can quite properly make it*, without falling into anthropomorphism, which transfers predicates from the world of sense to a being entirely distinct from the world. For those predicates are mere categories, which yield no determinate concept of it, but for that very reason no concept restricted to conditions of sensibility either. As regards the first point, assuming in turn a supreme reason as a cause of all connections in the world is the only possible way to advance the use of reason with regard to all possible experience in the world of sense to the highest degree, in thoroughgoing agreement with itself. Such a principle must benefit reason throughout, and can nowhere harm it in its use in nature. *Secondly*, however, reason is not thereby transferred as a property to the original being *in itself*, but only to *its relation* to the world of sense, and anthropomorphism is thus entirely avoided. For here we consider only the *cause* of the rational form found everywhere in the world, and attribute reason to the supreme being insofar as it contains the ground of this rational form of the world, but only by analogy. That is, this expression indicates only the relation that the supreme cause, unknown to us, has to the world, so as to determine everything in it in the highest degree according to reason. This ensures that we use the property of reason not to think God, but to think the world through it in the way necessary to obtain the greatest possible use of reason with regard to the world according to one principle. We thereby acknowledge that the supreme being, as regards what it is in itself, is entirely inscrutable to us and even unthinkable *in a determinate way*. We are thus restrained from making a transcendent use of our concepts of reason as an efficient cause, through the will, in order to determine the divine nature through properties that are, after all, always borrowed only from human nature, and from losing ourselves in crude or visionary concepts. On the other hand, we are also restrained from flooding our contemplation of the world with hyperphysical modes of explanation according to our concepts of human reason transferred to God, and from diverting it from its proper vocation: to be a study of mere nature through reason, not a presumptuous derivation of its appearances from a supreme reason. The expression suited to our weak concepts will be that we think the world *as if*, in its existence and inner determination, it originated from a supreme reason. In this way we partly cognize the constitution belonging to the world itself, without presuming to determine that of its cause in itself; and partly, on the other hand, place the ground of this constitution, the rational form in the world, in *the relation* of the supreme cause to the world, without finding the world sufficient for this on its own.[^360-1] In this way the difficulties that seem to oppose theism disappear: we combine *Hume's* principle of not driving the use of reason dogmatically beyond the field of all possible experience with another principle that *Hume* entirely overlooked, namely, not to regard the field of possible experience as something that bounds itself in the eyes of our reason. Here the critique of reason marks the true middle way between the dogmatism that *Hume* opposed and the skepticism he wished to introduce in its place. This is not like other middle ways that one is advised to determine for oneself more or less mechanically—a little from one side and a little from the other—by which no one is taught anything better; rather, it is a middle way that can be determined precisely according to principles. [^357-1]: Thus there is an analogy between the juridical relation of human actions and the mechanical relation of motive forces: I can never do anything to another without giving them a right to do the very same to me under the same conditions, just as no body can act on another with its motive force without thereby causing the other to react against it with an equal force. Here right and motive force are entirely dissimilar things, yet in their relation there is complete similarity. Through such an analogy I can therefore give a relational concept of things that are absolutely unknown to me. For example, as the promotion of children's happiness = a stands to parental love = b, so the welfare of the human race = c stands to the unknown in God = x, which we call love. This is not as though it had the slightest similarity to any human inclination, but because we can regard its relation to the world as similar to the relation things in the world have to one another. The relational concept here, however, is a mere category, namely the concept of cause, which has nothing to do with sensibility. [^360-1]: I will say: the causality of the supreme cause is, with regard to the world, what human reason is with regard to the works of its skill. The nature of the supreme cause itself thereby remains unknown to me; I compare only its effect known to me—the order of the world—and the rationality of that effect with the effects of human reason known to me, and therefore call the supreme cause a reason, without thereby attributing to it as its property precisely what I understand by this expression in human beings, or anything else known to me. ### Section 59 § 59. At the beginning of this remark I used the image of a *boundary* to establish the limits of reason with regard to the use appropriate to it. The world of sense contains only appearances, which are not things in themselves; the understanding must therefore assume the latter, noumena, precisely because it cognizes the objects of experience as mere appearances. Both are comprehended together in our reason, and the question is: how does reason proceed to bound the understanding with regard to both fields? Experience, which contains everything belonging to the world of sense, does not bound itself: from every conditioned thing it always reaches only another conditioned thing. What is to bound it must lie entirely outside it, and this is the field of pure beings of the understanding. But for us this is an empty space insofar as the *determination* of the nature of these beings is concerned; and to this extent, if dogmatically determinate concepts are intended, we cannot go beyond the field of possible experience. Yet since a boundary is itself something positive, belonging both to what is within it and to the space lying outside a given totality, there is indeed an actual positive cognition in which reason shares merely by extending itself to this boundary, though without attempting to go beyond it. For there reason finds before it an empty space in which it can indeed think forms for things, but no things themselves. Yet the *bounding* of the field of experience by something otherwise unknown to it is still a cognition that remains to reason from this standpoint. Through this, reason is neither confined within the world of sense nor wandering fancifully outside it, but, as befits knowledge of the boundary, restricts itself solely to the relation of what lies outside it to what is contained within. Natural theology is such a concept on the boundary of human reason, where reason sees itself compelled to look out toward the idea of a supreme being—and in a practical relation also toward that of an intelligible world—not to determine anything with regard to this mere being of the understanding, and thus outside the world of sense, but only to guide its own use within that world according to principles of the greatest possible unity, both theoretical and practical. For this purpose it employs the world's relation to an independent reason as the cause of all these connections, not thereby merely to *invent* a being, but, since outside the world of sense there must necessarily be something that only the pure understanding thinks, to determine this something only in such a way, though admittedly solely by analogy. In this way our earlier proposition, the result of the whole Critique, remains: "that reason, through all its a priori principles, never teaches us anything more than objects of possible experience alone, and even of these nothing more than can be cognized in experience." But this restriction does not prevent reason from leading us to the objective *boundary* of experience, namely its *relation* to something that is not itself an object of experience but must nevertheless be the supreme ground of all experience. Yet it teaches us nothing about this something in itself, but only in relation to reason's own complete use, directed toward the highest purposes, in the field of possible experience. This, however, is also all the benefit one can reasonably even wish for here, and with which one has cause to be satisfied. ### Section 60 § 60. We have thus presented metaphysics in detail as regards its subjective possibility, as it is actually given in the *natural predisposition* of human reason, and specifically in what constitutes the essential purpose of its cultivation. We found, however, that this *merely natural* use of such a predisposition of our reason, unless a discipline possible only through scientific critique curbs it and sets limits to it, entangles reason in extravagant *dialectical* inferences, partly merely specious and partly even in conflict with one another. Moreover, this sophistical metaphysics is dispensable for advancing cognition of nature, and may even be detrimental to it. It therefore still remains a problem worthy of investigation to discover the *natural purposes* at which this predisposition to transcendent concepts in our nature may be aimed, since everything found in nature must originally be designed for some useful purpose. Such an investigation is indeed precarious. I also confess that what I can say about it is only conjecture, like everything concerning nature's primary purposes; and this alone may be allowed me in this case, since the question concerns not the objective validity of metaphysical judgments but our natural predisposition to them, and thus lies outside the system of metaphysics, in anthropology. If I take together all the transcendental ideas, whose totality constitutes the proper task of natural pure reason, compelling it to leave mere contemplation of nature and go beyond all possible experience, and in this endeavor to bring about the thing—whether knowledge or sophistry—called metaphysics, I believe I discern the following aim of this natural predisposition. It is to free our concept from the fetters of experience and the limits of mere contemplation of nature far enough for it at least to see a field opened before it containing only objects for the pure understanding, which no sensibility can reach. This is not intended to occupy us speculatively with these objects, since we find no ground on which to gain a footing, but to enable practical principles to extend to the universality that reason indispensably requires for moral purposes: without finding such a space before them for their necessary expectation and hope, they could not so extend. Now I find that the *psychological* idea, however little insight it gives me into the pure nature of the human soul, exalted above all concepts of experience, at least shows clearly enough the inadequacy of those concepts. It thereby leads me away from materialism, as a psychological concept unfit for any explanation of nature and, moreover, one that narrows reason in a practical respect. Likewise, through the visible inadequacy of all possible cognition of nature to satisfy reason in its rightful inquiries, the *cosmological* ideas serve to keep us from naturalism, which would present nature as sufficient in itself. Finally, all natural necessity in the world of sense is always conditioned, since it always presupposes the dependence of things on others; unconditioned necessity must be sought only in the unity of a cause distinct from the world of sense. Yet the causality of this cause, in turn, if it were mere nature, could never make the existence of the contingent comprehensible as its consequence. Through the *theological* idea, therefore, reason frees itself from fatalism, both that of a blind natural necessity in the connection of nature itself without a first principle, and that of a blind necessity in the causality of this principle itself; it leads to the concept of a cause through freedom, hence of a supreme intelligence. Thus the transcendental ideas, even if they do not serve to instruct us positively, nevertheless serve to abolish the presumptuous assertions of *materialism*, *naturalism*, and *fatalism*, which narrow the field of reason, and thereby to make room for moral ideas outside the field of speculation. This, it seems to me, would to some extent explain that natural predisposition. The practical benefit that a merely speculative science may have lies outside the boundaries of that science. It can therefore be regarded only as a scholium and, like all scholia, does not belong to the science itself as a part of it. Yet this relation lies at least within the boundaries of philosophy, especially philosophy drawn from pure sources of reason, where the speculative use of reason in metaphysics must necessarily be united with its practical use in morals. Hence the unavoidable dialectic of pure reason, considered in a metaphysics as a natural predisposition, deserves not merely to be resolved as an illusion but also, if possible, to be explained in its purpose as an arrangement of nature. Yet this task, as a work beyond its strict obligation, cannot rightly be required of metaphysics proper. A second scholium, more closely related to the content of metaphysics, would be the solution of the questions pursued in the Critique from pages 642 to 668. For certain principles of reason are set forth there that determine a priori the order of nature, or rather the understanding that is to seek its laws through experience. They seem to be constitutive and legislative with regard to experience, although they spring from reason alone, which, unlike the understanding, must not be regarded as a principle of possible experience. Whether this agreement rests on the following may be considered further by those who wish to investigate the nature of reason even outside its use in metaphysics, indeed in the general principles for making a natural history as such systematic: just as nature does not attach in itself to appearances or their source, sensibility, but is found only in the relation of the latter to the understanding, so the thoroughgoing unity of this understanding's use for the sake of the entirety of possible experience in a system may belong to it only in relation to reason, so that experience too stands indirectly under the legislation of reason. For in the work itself I did indeed present this problem as important, but did not attempt its solution.[^364-1] And so I conclude the analytic solution of the principal question I posed myself—How is metaphysics possible at all?—by ascending from that in which its use is actually given, at least in its consequences, to the grounds of its possibility. [^364-1]: Throughout the Critique it has been my constant intention to omit nothing that could bring the investigation of the nature of pure reason to completeness, however deeply hidden it might lie. Everyone is afterward free to decide how far to pursue their investigation, provided they have been shown what investigations might still be undertaken. For this can fairly be expected of someone who has made it their task to survey this whole field, so as afterward to leave it to others for future cultivation and division as they choose. The two scholia also belong here; their dryness will hardly commend them to amateurs, and they have therefore been set out only for connoisseurs. ### Solution: How is metaphysics possible as a science? *Solution of the general question of the Prolegomena.* *How is metaphysics possible as a science?* Metaphysics exists as a natural predisposition of reason, but on its own it is also dialectical and deceptive, as the analytic solution of the third principal question showed. To draw principles from this predisposition, then, and in using them to follow an illusion that is natural but nonetheless false, can never produce science. It can produce only a vain dialectical art, in which one school may outdo another, but none can ever win rightful and lasting assent. If metaphysics as a science is to claim not merely deceptive persuasion but insight and conviction, a critique of reason itself must set out the entire stock of a priori concepts; their division according to their various sources (sensibility, understanding, and reason); a complete table of them; and the analysis of all these concepts, together with everything that can be inferred from them. Above all, it must then establish, through the deduction of these concepts, the possibility of synthetic a priori cognition, the principles of their use, and finally the limits of that use—all within a complete system. Critique, and critique alone, thus contains the entire thoroughly tested and substantiated plan, indeed all the means of carrying it out, by which metaphysics can be brought into being as a science. By any other paths or means it is impossible. The question here, then, is not so much how this undertaking is possible, but how to set it in motion, how to turn capable minds from their previous misguided and fruitless work toward an unerring approach, and how best to direct their united efforts toward the common goal. This much is certain: anyone who has once tasted critique will forever be disgusted by all the dogmatic babble he previously made do with out of necessity, because his reason needed something and could find nothing better to occupy it. Critique stands to ordinary school metaphysics exactly as *chemistry* stands to *alchemy*, or *astronomy* to fortune-telling *astrology*. I will vouch for this: no one who has thought through and grasped the principles of critique, even as presented in these Prolegomena alone, will ever return to that old and sophistical sham science. Rather, he will look forward with a certain delight to a metaphysics now entirely within his power, requiring no further preliminary discoveries, and capable for the first time of giving reason lasting satisfaction. For among all possible sciences, metaphysics alone can confidently count on this advantage: it can be brought to completion and to a stable condition, in which it need undergo no further change and can receive no additions through new discoveries. Here reason has the sources of its cognition not in objects and their intuition, from which it cannot learn anything further, but within itself. Once it has set out the fundamental laws of its capacity completely and with a precision that excludes every misinterpretation, nothing remains that pure reason could cognize a priori, or even ask about with justification. The assured prospect of knowledge so definite and self-contained has a particular attraction, even if we set aside all its usefulness, of which I will speak later. Every spurious art, every hollow wisdom, has its day; eventually it destroys itself, and its highest cultivation is also the moment of its downfall. That this time has now come for metaphysics is shown by the condition into which it has fallen among all learned nations, despite the zeal otherwise devoted to sciences of every kind. The old organization of university studies still preserves its shadow; a single academy of sciences still occasionally encourages one attempt or another by offering prizes. Yet it is no longer counted among the solid sciences, and anyone may judge for himself how an intelligent man would receive the well-meant but scarcely envied compliment of being called a great metaphysician. But although the time of decline for all dogmatic metaphysics is undoubtedly here, much is still lacking before we can say that the time of its rebirth through a thorough and complete critique of reason has arrived in turn. Every transition from one inclination to its opposite passes through a state of indifference, and this moment is the most dangerous for an author, but, it seems to me, the most favorable for the science. For when partisan feeling has died out through the complete dissolution of former ties, minds are in the best condition to listen gradually to proposals for an alliance on a different plan. When I say that I hope these Prolegomena may stimulate inquiry in the field of critique and offer the general spirit of philosophy, which seems to lack nourishment in its speculative part, a new and promising subject to engage it, I can already imagine the question everyone will ask who grew impatient and weary on the thorny paths along which I led him in the Critique: On what do I base this hope? I answer: *on the irresistible law of necessity*. We can no more expect the human mind ever to abandon metaphysical inquiries altogether than expect ourselves to stop breathing altogether rather than keep drawing impure air. There will therefore always be metaphysics in the world, and, what is more, in every person, especially every reflective person. In the absence of a public standard, each will fashion it to suit himself. Now, what has so far been called metaphysics can satisfy no examining mind, yet renouncing it altogether is also impossible. In the end, then, a critique of pure reason itself must be *attempted*, or, if one already exists, *examined* and submitted to general scrutiny. There is otherwise no means of meeting this pressing need, which is something more than mere curiosity. Since becoming acquainted with critique, whenever I have finished reading a work of metaphysical content that entertained and cultivated me through the precision of its concepts, its variety and order, and its easy presentation, I have been unable to refrain from asking: *Has this author advanced metaphysics even a single step?* I ask the forgiveness of those learned men whose writings have benefited me in other respects and have always helped cultivate my mental powers. For I confess that neither in their attempts nor in my own more modest ones, for which self-love nonetheless pleads, have I been able to find the slightest advance in the science. The reason is quite natural: the science did not yet exist, and it cannot be assembled piecemeal; its germ must first be fully preformed in critique. To prevent any misunderstanding, however, we must recall from what has gone before that although the analytic treatment of our concepts certainly does much to benefit the understanding, it does not advance the science of metaphysics in the least. These analyses of concepts are only the materials from which the science must first be built. Thus one may analyze and define the concept of substance and accident as beautifully as one likes; that is all very well as preparation for some future use. But if I cannot prove that in everything that exists substance persists and only accidents change, all that analysis has not advanced the science in the least. So far metaphysics has been unable to give a valid a priori proof of this proposition, or of the principle of sufficient reason, much less of any more complex proposition, such as one belonging to the doctrine of the soul or cosmology, or indeed of any synthetic proposition whatever. All that analysis has therefore accomplished, created, and advanced nothing; after so much bustle and noise, the science still stands where it stood in *Aristotle’s* day, although the preparations for it have unquestionably been made much better than before, if only the guiding thread to synthetic cognitions had first been found. If anyone feels offended by this, he can easily refute the charge. Let him cite just one synthetic proposition belonging to metaphysics that he offers to prove a priori in a dogmatic manner. Only when he achieves this will I grant that he has really advanced the science, even if ordinary experience already confirms the proposition sufficiently. No demand could be more moderate or fair; and in the inevitably certain event of failure, no verdict could be more just than this: metaphysics as a science has never yet existed at all. Should the challenge be accepted, I must ask that only two things be excluded: first, the plaything of *probability* and conjecture, which suits metaphysics as poorly as it does geometry; second, decisions reached by the divining rod of so-called *common sense*, which does not respond for everyone but follows personal characteristics. *As for the first*, nothing could be more absurd than to seek to ground one’s judgments on probability and conjecture in metaphysics, a philosophy of pure reason. Whatever is to be cognized a priori is thereby declared apodictically certain, and must therefore be proved as such. One might just as well try to found geometry or arithmetic on conjectures. For the latter’s calculus probabilium contains not probable judgments but entirely certain judgments about the degree of possibility of certain cases under given conditions of the same kind. In the sum of all possible cases these must accord with the rule without fail, although the rule is not sufficiently determinate for each individual occurrence. Only in empirical natural science may conjectures be admitted, through induction and analogy, and even there the possibility of what I assume must at least be entirely certain. The *appeal to common sense* is, if possible, in an even worse position when the issue is concepts and principles, not insofar as they are to hold for experience, but insofar as they are claimed to hold even outside the conditions of experience. For what is *sound understanding*? It is *ordinary understanding* insofar as it judges correctly. And what, then, is ordinary understanding? It is the capacity to cognize and use rules in concreto, as distinguished from *speculative understanding*, which is the capacity to cognize rules in abstracto. Ordinary understanding will thus scarcely understand the rule that everything that happens is determined through its cause, and will never be able to discern it in this general form. It therefore asks for an example from experience. When it hears that this means nothing other than what it has always thought whenever a windowpane was broken or a household object went missing, it understands the principle and accepts it as well. Ordinary understanding thus has no further use than where it can see its rules confirmed in experience, although they really do belong to it a priori. To discern them a priori and independently of experience, then, is a matter for speculative understanding and lies entirely outside the horizon of ordinary understanding. Yet metaphysics is concerned solely with the latter kind of cognition. It is surely a poor sign of sound understanding to appeal to that authority, which has no judgment at all in this matter and which one otherwise tends to look down on, except when one is hard pressed and at a loss for guidance or help in one’s speculation. There is a familiar evasion regularly employed by these false friends of ordinary common sense, who praise it highly on occasion but generally despise it. They say: There must, after all, be some propositions that are immediately certain, for which not only is no proof required, but no account at all; otherwise we could never reach an end to the grounds of our judgments. But to justify this license they can never cite anything indubitable that they may immediately attribute to ordinary common sense—apart from the principle of contradiction, which is insufficient to establish the truth of synthetic judgments—except mathematical propositions: for example, that twice two makes four, that between two points there is only one straight line, and so on. But these judgments are worlds apart from those of metaphysics. For in mathematics I can myself make, or construct, through my thinking everything that I represent to myself through a concept as possible. I successively add one two to another two and myself make the number four; or in thought I draw all sorts of lines from one point to another, and can draw only a single one that is similar to itself in all its parts, equal as well as unequal. But from the concept of one thing, all my powers of thought cannot extract the concept of something else whose existence is necessarily connected with the first. I must consult experience. And although my understanding supplies me a priori—yet always only in relation to possible experience—with the concept of such a connection, that of causality, I still cannot present it a priori in intuition, as I can the concepts of mathematics, and thus demonstrate its possibility a priori. Rather, if this concept is to be valid a priori—as metaphysics requires—then, together with the principles of its application, it always needs a justification and a deduction of its possibility. Otherwise we do not know how far it is valid, or whether it can be used only within experience or also beyond it. In metaphysics, then, as a speculative science of pure reason, we can never appeal to ordinary common sense. We can, however, appeal to it when we are compelled to leave metaphysics behind and renounce all pure speculative cognition, which must always be knowledge, and therefore also renounce metaphysics itself and its instruction in certain matters; when a rational faith is found to be the only thing possible for us and sufficient for our needs, perhaps even more beneficial than knowledge itself. For then the situation is entirely changed. Metaphysics must be science, not only as a whole but in all its parts, or else it is nothing at all; as speculation of pure reason, it has nowhere to find support except in universal insights. Outside it, however, probability and common sense may well have their useful and rightful employment, but according to quite distinct principles, whose weight always depends on their relation to the practical. This is what I consider myself entitled to require for the possibility of metaphysics as a science. ### Appendix *Appendix* *On what can be done to make metaphysics actual as a science.* Since none of the paths taken so far has reached this goal, and it will probably never be reached without a prior critique of pure reason, it seems no unreasonable demand that the attempt now laid before us be subjected to an exact and careful examination—unless one considers it more advisable to abandon all claims to metaphysics altogether. In that case there is no objection, provided one remains faithful to the resolution. If we take the course of things as it actually runs, not as it ought to run, there are two kinds of judgment: a *judgment that precedes investigation*, such as, in our case, the reader’s judgment of the Critique of Pure Reason from the standpoint of his own metaphysics, although the Critique is first to investigate the very possibility of metaphysics; and another *judgment that follows investigation*, in which the reader can for a time set aside the consequences of the critical inquiries, which may clash sharply with the metaphysics he has hitherto accepted, and first examine the grounds from which those consequences may have been derived. If what ordinary metaphysics teaches were established as certain, as geometry is, for example, the first manner of judging would hold good. For if the consequences of certain principles contradict established truths, those principles are false and must be rejected without any further investigation. But if metaphysics has no stock of indisputably certain synthetic propositions; if, perhaps, many of its propositions, as plausible as the best of them, nevertheless conflict with one another in their consequences; and if no secure criterion at all for the truth of properly metaphysical, synthetic propositions is to be found in it, then the preceding manner of judging is inadmissible. An investigation of the principles of critique must precede every judgment of its worth or worthlessness. *An example of a judgment of the Critique that precedes investigation.* A judgment of this kind is to be found in the Göttingische gelehrte Anzeigen, supplement, third issue, January 19, 1782, page 40 and following. When an author thoroughly acquainted with the subject of his work, who has endeavored throughout to bring his own reflection to its treatment, falls into the hands of a reviewer perceptive enough to discern the points on which the work’s worth or worthlessness really depends, who does not cling to words but pursues the matters themselves, and does not merely sift and test the principles from which the author began, the author may indeed dislike the severity of the judgment. The public is indifferent to this, for it gains by it. And the author himself can be satisfied that he has an opportunity to correct or clarify his arguments, examined early on by someone who understands the subject, and thus, if he believes himself fundamentally right, to remove in good time a stumbling block that might later harm his work. With my reviewer I find myself in an entirely different position. He seems not to understand at all what really mattered in the inquiry on which I was engaged, successfully or not. Whether the fault lies in impatience at thinking through an extensive work, in a peevish mood at the threatened reform of a science in which he believed he had long since settled everything, or—what I am reluctant to suspect—in a genuinely narrow conception that prevents him from ever thinking beyond his school metaphysics: in short, he rushes headlong through a long series of propositions from which, without knowing their premises, one can make no sense at all. Here and there he scatters his censure, whose grounds the reader can no more discern than he can understand the propositions against which it is supposed to be directed. He can therefore neither help inform the public nor do me the slightest harm in the judgment of those who understand the subject. I would accordingly have passed over this review altogether, had it not given me occasion for a few clarifications that may in some cases protect the reader of these Prolegomena from misinterpretation. Yet in order to seize on a standpoint from which he can most easily present the entire work in a light unfavorable to its author, without troubling himself with any particular investigation, the reviewer begins and ends by saying: “This work is a system of transcendental—or, as he translates it, higher[^373-1]—idealism.” At the sight of this line I soon knew what sort of review would emerge. It was much as if someone who had never heard of or seen anything of geometry found a copy of *Euclid* and was asked to judge it. Having come across many figures as he leafed through it, he might say: “The book is a systematic guide to drawing. The author uses a peculiar language to give obscure, unintelligible instructions, which in the end can accomplish no more than anyone can achieve with a good natural eye, etc.” Let us nevertheless see what sort of idealism it is that runs throughout my work, although it is far from constituting the soul of the system. The proposition of all genuine idealists, from the Eleatic school to Bishop *Berkeley*, is contained in this formula: “All cognition through the senses and experience is nothing but sheer illusion, and only in the ideas of pure understanding and reason is there truth.” The principle that governs and determines my idealism throughout, by contrast, is this: “All cognition of things from mere pure understanding or pure reason is nothing but sheer illusion, and only in experience is there truth.” But this is precisely the opposite of that idealism proper. How, then, did I come to use this expression for an entirely opposite purpose, and how did the reviewer come to see it everywhere? The solution to this difficulty rests on something that could very easily have been discerned from the context of the work, had one wished to do so. Space and time, together with everything they contain, are not things or their properties in themselves, but belong only to their appearances. Up to this point I profess the same belief as those idealists. But they, and *Berkeley* above all, regarded space as a merely empirical representation, which, like the appearances within it, becomes known to us with all its determinations only through experience or perception. I, by contrast, first show that space, and likewise time, to which *Berkeley* paid no attention, can be cognized by us a priori together with all its determinations. For space, as well as time, belongs to us before all perception or experience as the pure form of our sensibility, making possible all its intuition and therefore all appearances as well. It follows that, since truth rests on universal and necessary laws as its criteria, experience for *Berkeley* can have no criteria of truth: he laid nothing a priori at the foundation of its appearances, from which it followed that experience was nothing but sheer illusion. For us, by contrast, space and time, in conjunction with the pure concepts of the understanding, prescribe a priori the law for all possible experience, which at the same time provides the secure criterion for distinguishing truth from illusion within it.[^375-1] My so-called idealism, properly speaking critical idealism, is therefore of an entirely distinctive kind: it overturns ordinary idealism, and through it all a priori cognition, even that of geometry, first acquires objective reality. Without the ideality of space and time that I have proved, not even the most zealous realists could assert that reality. Under these circumstances, I wish I could give this conception of mine another name, to prevent every misunderstanding; but changing the term altogether is scarcely feasible. Let me therefore be permitted in the future, as already mentioned above, to call it formal, or better still critical, idealism, to distinguish it from the dogmatic idealism of *Berkeley* and the skeptical idealism of *Descartes*. I find nothing else remarkable in the review of this book. Its author judges throughout en gros, a manner shrewdly chosen, since it betrays neither one’s own knowledge nor one’s ignorance. A single detailed judgment en détail, had it properly addressed the principal question, might have exposed my error, or perhaps the measure of the reviewer’s insight into inquiries of this kind. Nor was it an ill-conceived stratagem for depriving readers who habitually form their idea of books solely from journal notices of any desire to read the book itself: to rush through a mass of propositions in a single breath, propositions that, torn from their connection with their proofs and explanations, must necessarily sound absurd, especially when they are so diametrically opposed to all school metaphysics as these are; to batter the reader’s patience to the point of disgust; and then, after acquainting me with the ingenious proposition that enduring illusion is truth, to close with the blunt but paternal lesson: “Why this quarrel with commonly accepted language? Why, and from what source, this idealistic distinction?” A judgment that in the end reduces everything distinctive in my book to a mere innovation in language, after declaring it metaphysical heresy, and clearly proves that my self-appointed judge has understood not the slightest thing about it, and has not properly understood even himself.[^376-1] The reviewer nevertheless speaks like a man who must be conscious of important and exceptional insights, which he still keeps hidden; for nothing in metaphysics has recently come to my attention that could justify such a tone. But he does the world a great wrong by withholding his discoveries. For no doubt others are in the same position as I am: despite all the fine things written in this field for a long time, they have been unable to find that the science has advanced a finger’s breadth. One certainly still finds definitions sharpened, lame proofs fitted with new crutches, fresh patches or an altered cut given to the patchwork of metaphysics—but that is not what the world asks for. The world is sated with metaphysical assertions. What people want is the possibility of this science, the sources from which certainty in it can be derived, and secure criteria for distinguishing the dialectical illusion of pure reason from truth. The reviewer must possess the key to this, or he would never have spoken in so lofty a tone. But I begin to suspect that such a need of the science may never have entered his thoughts. Otherwise he would have directed his review to this point, and even a failed attempt in so important a matter would have earned his respect. If that is so, we are good friends again. Let him think his way as deeply into his metaphysics as he likes; no one shall hinder him. Only he cannot judge what lies outside metaphysics: its source within reason. That my suspicion is not groundless I prove by this: he did not mention a word about the possibility of synthetic a priori cognition, which was the real problem, on whose solution the fate of metaphysics entirely depends, and to which my Critique, just like my Prolegomena here, was wholly directed. The idealism he stumbled on and remained stuck at was incorporated into the doctrine only as the sole means of solving that problem, although it also received confirmation on other grounds. Here he ought to have shown either that the problem lacks the importance I assign it, as I do again now in the Prolegomena, or that my concept of appearances cannot solve it at all, or that it can be better solved in some other way. But I find not a word of this in the review. The reviewer therefore understood nothing of my work, and perhaps nothing of the spirit and essence of metaphysics itself—unless, as I prefer to assume, it was rather a reviewer’s haste, indignant at the difficulty of working through so many obstacles, that cast an unfavorable shadow over the work before him and made its fundamental features unrecognizable to him. A learned journal is still very far from being able to maintain in the field of metaphysics the standing it otherwise deserves, as it does in other fields, however well and carefully its contributors may be chosen. Other sciences and branches of knowledge do, after all, have their standards. Mathematics has its standard within itself; history and theology have theirs in secular or sacred books; natural science and medicine in mathematics and experience; jurisprudence in law books; and even matters of taste in the models of the ancients. But for judging the thing called metaphysics, the standard has yet to be found. I have attempted to determine both the standard and its use. What, then, is to be done until it has been discovered, if writings of this kind must nonetheless be judged? If they are dogmatic, one may proceed as one likes: no one will long play the master over another in this field without someone appearing to repay him in kind. But if they are critical, not with regard to other writings but to reason itself, so that the standard of judgment cannot already be assumed but is only now being sought, objections and censure need not be unwelcome. Yet a spirit of tolerance must underlie them, for the need is shared, and the lack of the requisite insight makes the authority of a judge delivering a verdict inadmissible. To connect this defense of mine with the interests of the philosophizing community as well, I propose a trial that will decide how all metaphysical inquiries must be directed toward their common goal. It is nothing other than what mathematicians have sometimes done to decide, in a contest, the superiority of their methods: a challenge to my reviewer to prove in his own way any single genuinely metaphysical proposition he asserts—that is, a synthetic proposition cognized a priori from concepts. It may even be one of the most indispensable: for example, the principle of the persistence of substance, or of the necessary determination of events in the world by their cause. But he must prove it, as is proper, on a priori grounds. If he cannot—and silence is an admission—he must concede that, since metaphysics is absolutely nothing without apodictic certainty in propositions of this kind, its possibility or impossibility must before all else first be settled in a critique of pure reason. He is therefore bound either to confess that my principles of critique are correct or to prove their invalidity. Yet I can already foresee that, however confidently he has so far relied on the certainty of his principles, when it comes to a rigorous test he will not find a single one in the whole compass of metaphysics that he can boldly put forward. I will therefore grant him the most favorable condition one could expect in a contest: I will relieve him of the onus probandi and take it upon myself. For in these Prolegomena and in my Critique, pp. 426–461, he will find eight propositions, arranged in mutually contradictory pairs, while every one necessarily belongs to metaphysics, which must either accept or refute it—although there is not one that has not at some time been accepted by some philosopher. He is now free to choose whichever of these eight propositions he likes and accept it without proof, which I waive for him—but only one, since wasting time will serve him as little as it serves me—and then to attack my proof of its opposite. If I can nevertheless uphold this proof, and thus show that, according to principles every dogmatic metaphysics must necessarily acknowledge, the opposite of his adopted proposition can be proved just as clearly, it is thereby established that metaphysics has an inherited defect. This defect cannot be explained, much less remedied, unless we ascend to its birthplace, pure reason itself. My Critique must then either be accepted or replaced by a better one, and thus must at least be studied, which is all I ask for now. If, on the contrary, I cannot uphold my proof, then a synthetic a priori proposition established from dogmatic principles stands firm on my opponent’s side; my charge against ordinary metaphysics was therefore unjust; and I offer to acknowledge his censure of my Critique as legitimate, although that would still be far from following. For this, however, I think he would need to emerge from his anonymity. Otherwise I do not see how to prevent unnamed and yet uninvited opponents from honoring or assailing me with several challenges instead of one. *Proposal for an investigation of the Critique after which judgment may follow.* I am also indebted to the learned public for the silence with which it has honored my Critique for a considerable time. For this at least shows a suspension of judgment, and thus some inkling that a work which leaves all the familiar paths and strikes out on a new one, on which one cannot immediately find one’s way, may nevertheless contain something capable of giving new life and fruitfulness to an important but now dead branch of human cognition. It therefore shows a caution against breaking off and destroying the still tender graft by a hasty judgment. An example of a judgment delayed on such grounds has just now come before my eyes in the Gothaische gelehrte Zeitung. Every reader will perceive its thoroughness for himself, without taking into account my praise, which is suspect in this matter, from its comprehensible and undistorted presentation of a part belonging to the first principles of my work. And now, since an extensive building cannot possibly be judged immediately as a whole by a cursory survey, I propose that it be examined piece by piece, beginning with its foundation, using the present Prolegomena as a general outline with which the work itself could then be compared as occasion arises. If this request rested on nothing more than my own fancy of its importance, which vanity ordinarily lends to all one’s own productions, it would be immodest and deserve to be rejected with indignation. But the affairs of all speculative philosophy now stand on the verge of complete extinction, although human reason clings to them with an inclination that never dies. Only because it is incessantly deceived does this inclination now try, though in vain, to turn itself into indifference. In our thinking age, we cannot suppose that many deserving men would fail to take every good opportunity to work together in the common interest of reason, which is becoming ever more enlightened, if only there is some hope of thereby reaching the goal. Mathematics, natural science, laws, arts, even morality, etc., still do not fill the soul entirely. There always remains a space in it marked out for mere pure and speculative reason. Its emptiness compels us to seek in grotesque fancies, trifles, or even visionary enthusiasm what appears to be occupation and entertainment, but is at bottom only distraction, to drown out the importunate call of reason. In accordance with its vocation, reason demands something that will satisfy it in itself, not merely keep it busy serving other purposes or the interests of inclinations. A consideration devoted solely to this sphere of reason subsisting for itself therefore has, as I have grounds to suppose, a great attraction for anyone who has even tried to extend his concepts this far, precisely because within this sphere all other branches of knowledge, and even ends, must converge and unite into a whole. I may well say that its attraction is greater than that of any other theoretical knowledge, which one would not readily take in exchange for it. But I propose these Prolegomena as the plan and guiding thread of the inquiry, rather than the work itself, for this reason. I am still entirely satisfied with that work as regards its content, arrangement, method of instruction, and the care spent on weighing and testing each proposition precisely before setting it down. For it took years to satisfy myself fully about the sources not only of the whole, but sometimes of even a single proposition. Yet I am not entirely satisfied with my exposition in some sections of the Doctrine of Elements, such as the deduction of the concepts of the understanding or the section on the paralogisms of pure reason. A certain prolixity in them impedes clarity. In their place, what the Prolegomena say here concerning those sections may serve as the basis of examination. The Germans are praised for their ability to advance further than other nations in whatever requires perseverance and sustained diligence. If this opinion is well founded, here is an opportunity to bring to completion an undertaking whose happy outcome can scarcely be doubted, in which all thinking people have an equal interest, but which has hitherto been unsuccessful, and thus to confirm that favorable opinion. This is especially so since the science concerned is of such a peculiar kind that it can be brought all at once to its entire completeness and to that *stable* condition in which it can be advanced no further in the least, nor enlarged or even altered by later discoveries. I do not count here embellishments through greater clarity here and there, or through added usefulness in various respects. This is an advantage that no other science has or can have, since none concerns a cognitive capacity so completely isolated, independent of others, and unmixed with them. Nor does the present moment seem unfavorable to my request. For in Germany now one scarcely knows what else to occupy oneself with, beyond the so-called useful sciences, that would not be mere play but also an undertaking through which a lasting end is achieved. I must leave to others the devising of means by which the efforts of scholars could be united for such a purpose. Meanwhile, I do not intend to demand that anyone merely follow my propositions, or even to flatter myself with the hope of it. Rather, attacks, repetitions, restrictions, or else confirmations, additions, and extensions may arise as circumstances warrant. If only the matter is investigated from the ground up, it can now no longer fail to produce a system of doctrine, even if not my own, that can become a legacy to posterity for which it will have reason to be grateful. What sort of metaphysics may be expected to follow from critique, once its principles have first been put on a sound footing, and how metaphysics, once stripped of its borrowed plumage, need by no means appear impoverished and reduced to an insignificant figure, but can in another respect appear richly and fittingly furnished, would take too long to show here. But other great benefits that such a reform would bring are immediately evident. Ordinary metaphysics was already useful in seeking out the elementary concepts of the pure understanding, in order to make them clear through analysis and determinate through definitions. It thereby became a means of cultivating reason, wherever reason might later see fit to turn. But that was all the good it did. For it nullified this merit in turn by encouraging conceit through rash assertions, sophistry through subtle evasions and glosses, and shallowness through the ease of disposing of the most difficult problems with a little school wisdom. This shallowness is all the more seductive the more it can choose to assume something of the language of science on the one hand and of popular accessibility on the other, and thus be all things to everyone, while in fact being nothing at all. Through critique, by contrast, our judgment is given a standard by which knowledge can be securely distinguished from sham knowledge. By being put fully into practice in metaphysics, critique establishes a way of thinking whose beneficial influence then extends to every other use of reason, and for the first time instills the true philosophical spirit. Nor should we underestimate the service it renders theology by making it independent of the judgment of dogmatic speculation, and thereby placing it in complete security against every attack from such opponents. For ordinary metaphysics, though it promised theology much support, could not afterward fulfill that promise. Moreover, by calling speculative dogmatism to its aid, it had done nothing but arm enemies against itself. Visionary enthusiasm, which in an enlightened age can gain ground only by hiding behind a school metaphysics, under whose protection it may dare to rave with reason, as it were, is driven by critical philosophy from this last hiding place. And beyond all this, it can hardly fail to matter to a teacher of metaphysics that he should at last be able to say, with general assent, that what he teaches is now finally *science* as well, and that a real benefit is thereby rendered to the community. [^373-1]: By no means *higher*. High towers and the metaphysically great men who resemble them, both usually surrounded by much wind, are not for me. My place is the fertile *bathos* of experience. The word transcendental, whose meaning I have indicated so often and which the reviewer has not even grasped—so cursorily has he looked at everything—does not mean something that goes beyond all experience, but something that does precede experience a priori, yet is destined for nothing more than making cognition of experience possible. When these concepts overstep experience, their use is called transcendent, as distinguished from their immanent use, that is, their use restricted to experience. The work has sufficiently guarded against all misinterpretations of this kind; but the reviewer found misinterpretations to his advantage. [^375-1]: Idealism proper has always had a visionary purpose and can have no other. Mine, however, serves solely to comprehend the possibility of our a priori cognition of objects of experience, a problem that has never yet been solved, or even raised. This brings down the whole of visionary idealism, which always inferred from our a priori cognitions, even those of geometry, an intuition other than that of the senses, namely an intellectual one, as can already be seen in *Plato*. For it never occurred to anyone that the senses too might intuit a priori. [^376-1]: The reviewer is mostly fighting his own shadow. When I contrast the truth of experience with a dream, it never occurs to him that the question here concerns only the familiar somnio objective sumto of *Wolffian* philosophy, which is merely formal. It has nothing to do with the distinction between sleeping and waking, nor can it have in a transcendental philosophy. Moreover, he calls my deduction of the categories and the table of principles of the understanding “commonly known principles of logic and ontology, expressed in an idealistic manner.” The reader need only consult these Prolegomena on this point to be convinced that a more wretched judgment, or even a more historically incorrect one, could not possibly be delivered. ## Translation — Mouseia > Mouseia’s complete Plain English edition, made independently and directly from the same complete German text. ### Title Prolegomena to Any Future Metaphysics That Will Be Able to Present Itself *as a Science*, by Immanuel Kant. ### Preface These Prolegomena are not for beginners. They are for future teachers, and even for them their purpose is not to organize the presentation of a science that already exists. Their purpose is to discover this science in the first place. Some scholars take the history of philosophy, both ancient and modern, as their own philosophy. These Prolegomena are not written for them. They must wait until those who try to draw from the sources of reason itself have settled their business. Then it will be their turn to report to the world on what has happened. Otherwise, nothing can be said that, in their opinion, has not been said before. In fact, this may count as an unfailing prediction of everything to come. Over countless objects and many centuries, human understanding has indulged in so many kinds of speculation that it would be hard not to find something old that bears some resemblance to anything new. My aim is to convince everyone who thinks metaphysics is worth working on that they absolutely must stop their work for now. They must treat everything done so far as if it had not been done, and first of all ask: is anything like metaphysics even possible at all? If it is a science, why can it not win general and lasting acceptance, as other sciences can? If it is not a science, how does it keep boasting under the appearance of one and keeping human understanding waiting, with hopes that never die but are never fulfilled? Whether we demonstrate our knowledge or our lack of knowledge, something definite must finally be settled about the nature of this supposed science. It cannot possibly go on as it has. While every other science constantly advances, it seems almost ridiculous that this one—which claims to be wisdom itself, whose oracle everyone consults—should keep turning in the same place without taking a single step forward. Its followers, too, have greatly dwindled. We do not see people who feel strong enough to excel in other sciences willing to risk their reputations in this one. Here anyone who is ignorant in every other field claims the right to give a decisive judgment. For in this territory there is still no reliable measure or weight for telling thorough work from shallow talk. Still, it is not unheard of that, after a science has been worked on for a long time, and people think they have made astonishing progress in it, someone should finally think to ask whether and how such a science is possible at all. Human reason is so eager to build that it has repeatedly put up its tower, then taken it down again to see what sort of foundation it might have. It is never too late to become reasonable and wise. But when insight comes late, it is always harder to put it into practice. To ask whether a science is even possible assumes doubt about whether it actually exists. Such doubt offends anyone whose entire wealth may consist in this supposed jewel. So whoever lets this doubt slip should be prepared for resistance from all sides. Some, proudly aware of their old possession, and considering it rightful for that very reason, will look down on him with contempt, their metaphysical textbooks in hand. Others, who see nothing anywhere except what is identical with something they have seen somewhere before, will not understand him. And for a while everything will remain as though nothing had happened to make anyone fear or hope for a coming change. Even so, I venture to predict that readers who think for themselves will not merely begin to doubt their previous science after reading these Prolegomena. They will eventually become fully convinced that no such science can exist unless the requirements stated here, on which its possibility rests, are met. And since that has never yet happened, they will be convinced that there is as yet no metaphysics at all. Yet demand for it can never disappear either,[^257-1] because the interests of universal human reason are so intimately bound up with it. Readers will therefore admit that a complete reform, or rather a new birth on a plan entirely unknown until now, is inevitably approaching, however much people may resist it for a while. Since the attempts of *Locke* and *Leibniz*, or rather since metaphysics first arose, as far back as its history reaches, no event has occurred that could have been more decisive for this science's fate than *David Hume's* attack on it. He brought no light into this kind of cognition. But he did strike a spark from which a light might well have been kindled, if it had caught suitable tinder and its glow had been carefully kept alive and strengthened. *Hume* started mainly from a single but important concept of metaphysics: the *connection of cause and effect*, and therefore also the resulting concepts of force, action, and so on. He challenged reason, which claims to have produced this concept from within itself, to explain by what right it thinks that something can be such that, once it is given, something else must necessarily be given through it. That is what the concept of cause says. He proved beyond dispute that reason cannot possibly think such a connection a priori, from concepts, because it contains necessity. We cannot see at all why, because one thing exists, something else must necessarily exist too, or how the concept of such a connection can be introduced a priori. From this he concluded that reason completely deceives itself with this concept. It wrongly takes the concept for its own child, when it is only an illegitimate child of the imagination. Made pregnant by experience, the imagination has brought certain representations under the law of association and passed off the resulting subjective necessity—that is, habit—as an objective necessity based on insight. He concluded that reason has no power even to think such connections in general, since its concepts would then be mere inventions. All its supposed a priori cognitions would be nothing but ordinary experiences falsely labeled. This amounts to saying that there is no metaphysics at all, nor can there be any.[^258-1] However hasty and mistaken his conclusion was, it was at least based on investigation. That investigation deserved to have the capable minds of his time join together to solve the problem more successfully, if possible, in the sense in which he presented it. A complete reform of the science would then soon have had to follow. But fate, always unfavorable to metaphysics, saw to it that no one understood him. It is painful to see how completely his opponents—*Reid*, *Oswald*, *Beattie*, and finally *Priestley*—missed the point of his problem. They always took for granted just what he doubted, while vehemently and usually with great arrogance proving what it had never occurred to him to doubt. They so misunderstood his hint toward improvement that everything stayed as it was, as though nothing had happened. The question was not whether the concept of cause is correct, useful, and indispensable for all cognition of nature. *Hume* had never doubted that. The question was whether reason thinks it a priori and whether it therefore has an inner truth independent of all experience, and thus perhaps a wider usefulness not restricted merely to objects of experience. That is what *Hume* wanted explained. The only issue was the origin of this concept, not its indispensability in use. Once its origin had been established, the conditions of its use and the range within which it could be valid would have followed on their own. To meet the problem, however, the famous man's opponents would have had to go very deeply into the nature of reason when it is occupied with pure thinking alone. That did not suit them. So they invented a more convenient way to act defiantly without any insight: an appeal to *common sense*. It is indeed a great gift from heaven to have sound—or, as it has recently been called, plain—common sense. But we must prove it by our actions, by the thoughtful and reasonable things we think and say, not by appealing to it as an oracle when we have nothing intelligent to say in our own defense. To appeal to common sense only when insight and science run out, and not before, is one of the subtle inventions of modern times. It lets the emptiest talker confidently take on the most thorough thinker and hold his own. But as long as a small remainder of insight survives, people will take care not to resort to this emergency measure. Looked at clearly, this appeal is nothing but an appeal to the crowd's judgment: applause at which the philosopher blushes, while the popular wit celebrates and acts defiantly. I should think that *Hume* had as much claim to sound understanding as *Beattie*, and also to something the latter certainly lacked: a critical reason that keeps ordinary understanding within bounds. It prevents ordinary understanding from getting carried away in speculation or, when speculation alone is at issue, from trying to decide anything, because it cannot justify its principles. Only in this way will it remain sound understanding. A chisel and mallet may serve perfectly well to work a piece of timber, but copper engraving requires an etching needle. In the same way, sound understanding and speculative understanding are both useful, but each in its own way. The first is useful for judgments that find their direct application in experience. The second is useful where judgments must be made in general from mere concepts, as in metaphysics. There, what calls itself sound understanding—often by a name that means the opposite of what it is—has no judgment at all. I freely admit that *David Hume's* reminder was precisely what first interrupted my dogmatic sleep many years ago and gave my investigations in speculative philosophy an entirely different direction. I was far from accepting his conclusions. They arose only because he did not see his problem as a whole, but struck upon just one part of it, which cannot provide an answer without considering the whole. If we start from a well-grounded thought that someone else has left us, even though it was not developed, we may hope that continued reflection will take us further than the perceptive man to whom we owe the first spark of this light. I first tried, then, to see whether *Hume's* objection could be stated in a general form. I soon found that the concept of the connection of cause and effect is far from the only one through which the understanding thinks connections between things a priori. Metaphysics consists entirely of such concepts. I tried to establish how many there were, and succeeded as I had hoped: from a single principle. I then set about the deduction of these concepts. I was now sure that they were not derived from experience, as *Hume* had feared, but arose from pure understanding. This deduction had seemed impossible to my perceptive predecessor. No one except him had even thought of it, although everyone confidently used the concepts without asking what grounded their objective validity. This, I say, was the hardest undertaking that could ever be attempted on behalf of metaphysics. Worst of all, metaphysics, however much of it existed anywhere, could not give me the slightest help in it, because that deduction must first establish the possibility of metaphysics. Once I had succeeded in solving Hume's problem, not merely in a particular case but with respect to the whole faculty of pure reason, I could take secure, though always slow, steps. In the end I could determine the whole range of pure reason completely, both its limits and its contents, according to general principles. That was what metaphysics needed in order to construct its system according to a reliable plan. But I fear that the *working out* of Hume's problem in its widest possible scope—namely, the Critique of Pure Reason—may fare just as the *problem* itself did when it was first presented. People will judge it incorrectly because they do not understand it. They will not understand it because they want to leaf through the book but not think it through. And they will not want to spend that effort on it because the work is dry, obscure, contrary to all their familiar concepts, and also long. I admit that I did not expect to hear a philosopher complain about a lack of popular appeal, entertainment, and ease when the very existence of a celebrated cognition indispensable to humanity is at stake. Its existence can be established only by the strictest rules of precise scholarly method. Popular appeal may follow these in time, but must never come first. Still, the complaint about a certain obscurity is justified. It comes partly from the wide scope of the plan, which makes it difficult to survey the main points that matter in the investigation. I will remedy that complaint through these *Prolegomena*. The other work, which presents the faculty of pure reason in its entire scope and limits, nevertheless remains the foundation. The Prolegomena refer to it only as preliminary exercises. For that critique must stand complete as a science, systematic down to its smallest parts, before we can even think of bringing metaphysics forward or entertain even a distant hope of it. People have long been used to seeing old, worn-out cognitions dressed up as new. They are taken out of their earlier connections and fitted with a systematic outfit of some arbitrary cut, under new titles. Most readers will expect nothing else from that critique either. But these Prolegomena will lead them to see that it is an entirely new science. No one had previously conceived even the thought of it; its very idea was unknown. Nothing previously given could be used for it except the hint that *Hume's* doubts could provide. He too had no inkling that such a formal science was possible. To bring his ship to safety, he ran it aground on the beach—skepticism—where it can lie and rot. My concern instead is to give it a pilot who can guide it safely wherever he thinks best. Equipped with a complete sea chart and a compass, he follows reliable principles of navigation drawn from knowledge of the globe. To approach a new science, entirely separate and the only one of its kind, with the prejudice that we can judge it through supposed knowledge acquired elsewhere achieves only one thing. We think we see everywhere what we already knew, perhaps because the expressions sound similar—even though the reality of just that supposed knowledge must first be thoroughly doubted. Everything must then seem terribly distorted, absurd, and unintelligible, because we take as our basis not the author's thoughts, but always our own way of thinking, which long habit has made second nature. Yet the length of the work, insofar as it belongs to the science itself rather than its presentation, and its unavoidable dryness and scholarly precision may be extremely advantageous to the subject itself while necessarily disadvantaging the book. Not everyone has the gift of writing as subtly and yet as engagingly as *David Hume*, or as thoroughly and yet as elegantly as *Moses Mendelssohn*. Still, I flatter myself that I could have given my presentation popular appeal if my only concern had been to draw up a plan and urge others to carry it out. But I cared about the welfare of the science that had occupied me so long. Besides, it took great persistence and no small self-denial to set aside the attraction of an early favorable reception for the prospect of approval that would be late but lasting. *Making plans* is often a self-indulgent and boastful activity of the mind. People give themselves the appearance of creative genius by demanding what they cannot themselves accomplish, criticizing what they cannot improve, and proposing something they do not themselves know where to find. Still, even a sound plan for a general critique of reason would have required more than one might suppose, if it was not to become merely the usual declaration of pious wishes. But pure reason is a sphere so separate, and so completely interconnected within itself, that we cannot touch one part without touching all the others. We can accomplish nothing without first determining each part's place and its influence on the others. Nothing outside this sphere can correct our judgment within it. So the validity and use of every part depend on its relation to the other parts within reason itself, just as in the structure of an organized body, the purpose of each member can be derived only from the complete concept of the whole. We can therefore say that such a critique is never reliable unless it is *complete* and *finished* down to the smallest elements of pure reason. In the sphere of this faculty, we must determine and settle either *everything* or *nothing*. Admittedly, a mere plan preceding the *Critique* of Pure Reason would be unintelligible, unreliable, and useless. But it is all the more useful when it follows the Critique. It enables us to survey the whole, examine one by one the main points that matter in this science, and arrange some things better in the presentation than was possible in the first version of the work. Here, then, is such a *plan*, after the completed work. It can now be organized according to the *analytic method*, whereas the work itself had to be written entirely according to the *synthetic method*. In that way the science could display all its joints in their natural connection, as the structure of a quite distinctive faculty of cognition. Anyone who finds this plan, which I offer as Prolegomena before any future metaphysics, obscure in its turn should consider the following. Not everyone needs to study metaphysics. Some talents do very well in thorough and even deep sciences that stay closer to intuition, but do not succeed in inquiries conducted entirely through abstract concepts. In that case, people must direct their mental gifts toward another subject. But anyone who undertakes to judge metaphysics, or even to write one, must fully meet the requirements made here. They may do so either by accepting my solution or by thoroughly refuting it and putting another in its place—for they cannot dismiss the requirements. Finally, the much-criticized obscurity, a usual cover for one's own laziness or poor vision, also has its use. People who maintain a cautious silence about all other sciences speak as masters and decide boldly on questions of metaphysics. Here their ignorance does not, of course, stand out clearly against others' knowledge. But it does stand out against genuine critical principles, which can therefore be praised in these words: they keep the drones, that idle herd, away from the hives. *Virg.* [^257-1]: The countryman waits for the river to finish flowing; but it flows and will keep flowing, rolling on forever. *Horat.* [^258-1]: Nevertheless, *Hume* himself called this very destructive philosophy metaphysics and valued it highly. “Metaphysics and morals,” he says (Versuche 4ter Theil, Seite 214, deutsche Übers.), “are the most important branches of science; mathematics and natural science are not worth half as much.” But this perceptive man was looking only at the negative benefit of moderating the exaggerated claims of speculative reason: completely ending the many endless and persecutory disputes that confuse humanity. In doing so, he lost sight of the positive harm that follows when reason is deprived of its most important prospects. Only through these can it set the highest goal for all the will's efforts. ### Preliminary reminder Prolegomena. Preliminary Reminder on *What Is Distinctive about All Metaphysical Cognition.* ### Section 1 § 1. *The Sources of Metaphysics.* If we want to present a body of cognition as a *science*, we must first be able to determine precisely what sets it apart: what it shares with no other science and what is therefore *distinctive* about it. Otherwise the boundaries of all the sciences run together, and none can be treated thoroughly in keeping with its nature. This distinctive character may consist in a difference in its *object*, its *sources of cognition*, its *kind of cognition*, or some if not all of these together. On it rests, first of all, the idea of the possible science and its territory. First, as for the *sources* of metaphysical cognition, its very concept implies that they cannot be empirical. Its principles—which include not only its basic propositions but also its basic concepts—must therefore never be taken from experience. It is meant to be not physical but metaphysical cognition, that is, cognition lying beyond experience. So it will rest neither on outer experience, the source of physics proper, nor on inner experience, the basis of empirical psychology. It is therefore cognition a priori, or cognition from pure understanding and pure reason. In this respect, however, it would have nothing to distinguish it from pure mathematics. It must therefore be called *pure philosophical* cognition. For the meaning of this expression I refer to Kritik d. r. V. Seite 712 u. f., where the difference between these two ways of using reason has been presented clearly and adequately.—So much for the sources of metaphysical cognition. ### Section 2 § 2. *The Only Kind of Cognition That Can Be Called Metaphysical.* a) *The Difference between Synthetic and Analytic Judgments in General.* Metaphysical cognition must contain nothing but a priori judgments. The distinctive character of its sources requires this. But whatever origin judgments may have, and whatever their logical form, there is a difference in their content. They are either merely *clarifying*, adding nothing to the content of cognition, or *extending*, increasing the cognition given. The first can be called *analytic* judgments, the second *synthetic* judgments. In their predicate, analytic judgments say only what was already actually thought in the concept of the subject, though not as clearly or with equal awareness. If I say that all bodies are extended, I have not increased my concept of body in the slightest. I have only broken it down into its elements. Extension was already actually thought in that concept before the judgment, even if not explicitly stated. The judgment is therefore analytic. By contrast, the proposition that some bodies are heavy contains something in its predicate that is not actually thought in the general concept of body. It therefore increases my cognition by adding something to my concept, and must be called a synthetic judgment. b) *The Principle Common to All Analytic Judgments Is the Principle of Contradiction.* All analytic judgments rest entirely on the principle of contradiction and are by their nature a priori cognitions, whether or not the concepts that supply their content are empirical. The predicate of an affirmative analytic judgment is already thought in the concept of the subject. It therefore cannot be denied of the subject without contradiction. Similarly, in an analytic but negative judgment, its opposite is necessarily denied of the subject, again according to the principle of contradiction. This is how it stands with the propositions: every body is extended, and no body is unextended (simple). For this same reason, all analytic propositions are a priori judgments even when their concepts are empirical. For example: gold is a yellow metal. To know this I need no further experience beyond my concept of gold, which would contain the fact that this body is yellow and is a metal. That constituted my concept itself. I only had to analyze it, without looking beyond it for anything else. c) *Synthetic Judgments Need a Principle Other Than the Principle of Contradiction.* There are a posteriori synthetic judgments, whose origin is empirical. But there are also synthetic judgments that are certain a priori and arise from pure understanding and reason. Both kinds agree in this: they can never arise from the principle of analysis, namely the principle of contradiction, alone. They require a quite different principle as well. Still, from whatever basic principle they are derived, they must always be derived *in agreement with* the principle of contradiction. Nothing may conflict with this principle, even though not everything can be derived from it. I will first divide synthetic judgments into classes. 1) *Judgments of experience* are always synthetic. It would be absurd to base an analytic judgment on experience. I do not have to go beyond my concept at all to make the judgment, and therefore need no testimony from experience for it. That a body is extended is a proposition established a priori, not a judgment of experience. Before I turn to experience, I already have all the conditions for my judgment in the concept. I need only draw the predicate from it according to the principle of contradiction. In doing so, I can also become aware of the judgment's *necessity*, which experience could not even teach me. 2) *Mathematical judgments* are all synthetic. This proposition seems until now to have entirely escaped the notice of those who analyze human reason. Indeed, it seems directly opposed to all their conjectures, though it is indisputably certain and very important in its consequences. People found that all the inferences of mathematicians proceed according to the principle of contradiction, as the nature of all apodictic certainty requires. They therefore persuaded themselves that the basic propositions too were known through the principle of contradiction. In this they were badly mistaken. A synthetic proposition can indeed be understood according to the principle of contradiction, but only by presupposing another synthetic proposition from which it can be inferred, never on its own. First, we must note that properly mathematical propositions are always a priori judgments, not empirical ones, because they carry necessity with them, and necessity cannot be taken from experience. But if anyone will not grant me this, then I will restrict my claim to *pure mathematics*. Its very concept implies that it contains no empirical cognition, but only pure a priori cognition. At first one might think the proposition 7 + 5 = 12 is merely analytic, following from the concept of a sum of seven and five according to the principle of contradiction. But closer inspection shows that the concept of the sum of 7 and 5 contains nothing more than the union of the two numbers into a single one. This does not at all tell us which single number contains them both. The concept of twelve is not already thought when I merely think of that union of seven and five. However long I analyze my concept of such a possible sum, I will not find twelve in it. We must go beyond these concepts by calling on the intuition that corresponds to one of them: for instance, our five fingers, or five points, as *Segner* does in his Arithmetic. Then we add the units of the five given in intuition, one by one, to the concept of seven. So the proposition 7 + 5 = 12 really does extend our concept. It adds a new concept to the first, one not thought in it at all. That is, an arithmetical proposition is always synthetic. We become all the more clearly aware of this if we take somewhat larger numbers. It is then obvious that, however we turn our concept this way and that, we could never find the sum merely by analyzing our concepts without calling on intuition. No basic proposition of pure geometry is analytic either. That the straight line between two points is the shortest is a synthetic proposition. My concept of straightness contains nothing about magnitude, only a quality. The concept of the shortest is therefore an entirely new addition. No analysis can draw it from the concept of a straight line. So here we must call on intuition, which alone makes the synthesis possible. Some other basic propositions that geometers presuppose are indeed analytic and rest on the principle of contradiction. But like propositions of identity, they serve only as links in the method, not as principles. For example: a = a, the whole is equal to itself; or (a + b) > a, that is, the whole is greater than its part. Yet even these, although valid through concepts alone, are admitted in mathematics only because they can be presented in intuition. What usually makes us believe here that the predicate of such apodictic judgments already lies in our concept, so that the judgment is analytic, is simply an ambiguity in the expression. We *must* add a certain predicate in thought to a given concept, and this necessity already belongs to the concepts. But the question is not what we *must think in addition* to the given concept. It is what we *actually*, even if only obscurely, *think in it*. And here we find that the predicate does indeed belong necessarily to that concept, but not directly. It belongs through an intuition that must be added. ### Section 3 § 3. Remark *on the General Division of Judgments into Analytic and Synthetic.* This division is indispensable for the critique of human understanding and therefore deserves an *established* place in it. Otherwise, I know of no considerable use it would have anywhere else. I also find here the reason why dogmatic philosophers neglected this division, which seems to offer itself naturally. They always sought the sources of metaphysical judgments in metaphysics itself, not outside it in the general laws of pure reason. Thus the famous *Wolff*, or the perceptive *Baumgarten* who followed in his footsteps, could look in the principle of contradiction for a proof of the principle of sufficient reason, which is plainly synthetic. By contrast, I already find a hint of this division in *Locke's* Essay Concerning Human Understanding. Earlier in the work, he had already discussed the different connections of representations in judgments and their sources. There he places one source in identity or contradiction (analytic judgments), the other in the existence of representations in a subject (synthetic judgments). Then, in the fourth book, third chapter, § 9 u. f., he admits at § 10 that our cognition (a priori) of the latter is very restricted and amounts to almost nothing. But there is so little that is definite or brought under rules in what he says about this kind of cognition that we need not be surprised that no one, not even Hume in particular, took it as an occasion to reflect on propositions of this kind. We do not readily learn principles that are both general and definite from others who had only an obscure sense of them. We must first reach them ourselves through our own reflection. After that we also find them elsewhere, where we certainly would not have discovered them first, because the authors themselves did not even know that such an idea underlay their own observations. People who never think for themselves nevertheless have the sharp sight needed to track down everything, once it has been shown to them, in what was said before—where no one could previously see it. ### General questions The General Question of the Prolegomena: Is Metaphysics Possible at All? ### Section 4 § 4. If a metaphysics that could establish itself as a science actually existed, we could say: here is metaphysics; you need only learn it, and it will convince you of its truth irresistibly and permanently. In that case this question would be unnecessary. Only one question would remain, testing our insight rather than proving that the thing itself exists: *how it is possible*, and how reason manages to attain it. But human reason has not been so fortunate in this case. We cannot point to a single book, as we can, say, to *Euclid*, and say: this is metaphysics; here you will find the chief aim of this science, cognition of a supreme being and a future world, proved from principles of pure reason. Many propositions can indeed be shown to us that are apodictically certain and have never been disputed. But they are all analytic. They concern the materials and equipment for metaphysics more than the extension of cognition that should be our real aim in it (§ 2. litt. c). You may also present synthetic propositions, for example the principle of sufficient reason, which you have never proved from reason alone, and therefore a priori, as you were obliged to do. People gladly grant them to you all the same. But when you try to use them for your main purpose, you fall into claims so inadmissible and uncertain that, at all times, one metaphysics has contradicted another, either in the claims themselves or in their proofs. It has thereby destroyed its own claim to lasting acceptance. Indeed, attempts to bring such a science into existence were undoubtedly the first cause of skepticism's early appearance. This is a way of thinking in which reason acts so violently against itself that it could have arisen only in complete despair of finding satisfaction in its most important aims. Long before people began to question nature methodically, they questioned only their isolated reason, which ordinary experience had already trained to some degree. That is because reason is always present to us, whereas laws of nature usually have to be laboriously sought out. And so metaphysics floated on top like foam. As soon as the foam that had been scooped up dissolved, new foam appeared on the surface, which some people always eagerly collected. Others, instead of looking in the depths for the cause of this appearance, thought themselves wise because they laughed at the first group's wasted effort. The essential feature that distinguishes pure *mathematical* cognition from every other a priori cognition is that it must proceed *not from concepts*, but always only through the construction of concepts (Kritik S. 713). Its propositions must therefore go beyond the concept to what is contained in the intuition corresponding to it. So its propositions can never, and must never, arise by analyzing concepts, that is, analytically. They are therefore all synthetic. I cannot help pointing out the harm done to philosophy by neglecting this otherwise easy and seemingly insignificant observation. *Hume* felt the calling, worthy of a philosopher, to look over the whole field of pure a priori cognition, where human understanding claims such large possessions. But he carelessly cut off an entire province, indeed the most important one: pure mathematics. He imagined that its nature and, so to speak, its constitution rested on quite different principles, namely on the principle of contradiction alone. Although he did not divide propositions as formally and generally, or under the names I use here, it amounted to saying that pure mathematics contains only *analytic* propositions, while metaphysics contains synthetic a priori ones. He was very much mistaken in this, and the mistake had decisively harmful consequences for his whole conception. Had he not made it, he would have extended his question about the origin of our synthetic judgments far beyond his metaphysical concept of causality, to the possibility of a priori mathematics as well. For he would have had to recognize mathematics as synthetic too. But then he could not possibly have based his metaphysical propositions on experience alone, since he would also have subjected the axioms of pure mathematics to experience. He was far too perceptive to do that. The good company in which metaphysics would then have found itself would have protected it from the danger of degrading mistreatment. The blows intended for metaphysics would also have had to strike mathematics, which was not and could not have been his intention. This would have drawn that perceptive man into reflections that would have had to resemble those occupying us now, but that would have gained immeasurably from his inimitably beautiful presentation. *Properly metaphysical* judgments are all synthetic. We must distinguish judgments *belonging to metaphysics* from *metaphysical* judgments proper. Many of the former are analytic, but they only supply the means for metaphysical judgments, which are the entire aim of the science and are always synthetic. If concepts belong to metaphysics—for example, the concept of substance—then judgments arising merely from their analysis necessarily belong to metaphysics too. For example: substance is what exists only as a subject, and so on. Through a number of such analytic judgments we try to approach the definition of the concepts. But the analysis of a pure concept of the understanding, of the kind metaphysics contains, proceeds no differently from the analysis of any other concept, even an empirical one that does not belong to metaphysics. For example: air is an elastic fluid whose elasticity is not canceled by any known degree of cold. So the concept is distinctively metaphysical, but the analytic judgment is not. This science has something special and distinctive in the production of its a priori cognitions. This production must therefore be distinguished from what the science shares with all other cognition through the understanding. Thus the proposition that everything in things that is substance persists is a synthetic and distinctively metaphysical proposition. Once we have collected, according to certain principles, the a priori concepts that supply the material and equipment of metaphysics, analyzing these concepts is of great value. This analysis can also be presented separately, as a special part—a philosophy of definitions, so to speak—containing nothing but analytic propositions belonging to metaphysics. It can be kept separate from all the synthetic propositions that make up metaphysics itself. In fact, these analyses have no considerable use anywhere except in metaphysics: that is, with regard to the synthetic propositions that are to be produced from the concepts first analyzed. The conclusion of this section is therefore that metaphysics properly concerns synthetic a priori propositions, and these alone make up its aim. For that purpose it certainly needs many analyses of its concepts, and thus analytic judgments. But here the procedure is no different from that in any other kind of cognition, where we seek merely to make our concepts clear by analyzing them. The essential content of metaphysics, however, is the production of a priori cognition, both in intuition and in concepts, and finally of synthetic a priori propositions within philosophical cognition. We are tired, then, of dogmatism, which teaches us nothing, and of skepticism, which promises us nothing at all—not even the peaceful state of an allowable ignorance. The importance of the cognition we need urges us on, while long experience makes us distrust every cognition we believe we possess or that offers itself under the name of pure reason. Only one critical question remains. Its answer will let us decide how to proceed in the future: *Is metaphysics possible at all?* But this question must not be answered by skeptical objections to particular claims of an actual metaphysics, for we are not yet accepting any. It must be answered from the still merely *problematic* concept of such a science. In the *Critique of Pure Reason*, I took a synthetic approach to this question. That is, I investigated pure reason itself and tried to determine, in that very source, both the elements and the laws of its pure use according to principles. This work is difficult. It requires determined readers to think their way gradually into a system that takes nothing as given for its foundation except reason itself. Without relying on any fact, it seeks to develop cognition from its original seeds. *Prolegomena*, by contrast, are meant to be preliminary exercises. Rather than present the science itself, they should indicate what must be done to bring a science into actual existence if possible. So they must rest on something already known to be reliable. From there we can start with confidence and ascend to sources we do not yet know. Discovering these sources will not only explain what we knew, but also reveal a range of many cognitions that all arise from those same sources. The method of the Prolegomena, especially of those intended to prepare for a future metaphysics, will therefore be *analytic*. Fortunately, although we cannot assume that metaphysics *actually exists* as a science, we can confidently say that certain pure synthetic a priori cognition is actual and given: *pure mathematics* and *pure natural science*. Both contain propositions that are recognized throughout: some as apodictically certain through reason alone, others through general agreement with experience, yet still as independent of experience. So we have at least some *undisputed* synthetic a priori cognition. We need not ask whether it is possible, since it is actual, but only *how it is possible*. From the principle of the possibility of the cognition given, we can then derive the possibility of all the rest. Prolegomena. General Question: How Is Cognition from Pure Reason Possible? ### Section 5 § 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. ### First part: How is pure mathematics possible? The Main Transcendental Question First Part. How Is Pure Mathematics Possible? ### Section 6 § 6. Here we have a great and well-established body of cognition. Its scope is already admirable, and it promises unlimited expansion in the future. Throughout, it carries apodictic certainty—that is, absolute necessity. So it rests on no grounds of experience and is therefore a pure product of reason. Yet it is also synthetic throughout. “How, then, can human reason produce such cognition entirely a priori?” Since this ability does not and cannot rest on experiences, does it not presuppose some deeply hidden a priori ground of cognition? Might that ground reveal itself through these effects, if we carefully traced their earliest beginnings? ### Section 7 § 7. We find that all mathematical cognition has this distinctive feature: it must first present its concept *in intuition*, and must do so a priori, hence in an intuition that is not empirical but pure. Without this means it cannot take a single step. Its judgments are therefore always *intuitive*. Philosophy, by contrast, must be content with *discursive* judgments, *from mere concepts*. It can certainly clarify its apodictic teachings through intuition, but can never derive them from intuition. This observation about the nature of mathematics already points us toward the first and highest condition of its possibility. Some *pure intuition* must underlie it, in which it can present all its concepts concretely yet a priori—or, as it is called, *construct* them.[^281-1] If we can discover this pure intuition and how it is possible, we can easily explain from it how synthetic a priori propositions are possible in pure mathematics, and therefore how this science itself is possible. Empirical intuition makes it possible without difficulty for us to extend synthetically, in experience, the concept we form of an object of intuition, by adding new predicates that the intuition itself supplies. Pure intuition will do the same. There is only this difference: in the latter case the synthetic judgment will be certain a priori and apodictically; in the former, only a posteriori and empirically. The empirically certain judgment contains only what is found in contingent empirical intuition. The a priori judgment contains what must necessarily be found in pure intuition. For this pure intuition, as an a priori intuition, is inseparably connected with the concept *before all experience* or individual perception. [^281-1]: See Kritik S. 713. ### Section 8 § 8. But with this step the difficulty seems to increase rather than decrease. For the question now is: *How is it possible to intuit something a priori?* An intuition is a representation of the sort that would depend directly on the object's presence. So it seems impossible to intuit *originally* a priori. The intuition would then have to occur without an object to which it referred being present either earlier or now, and thus could not be an intuition. Concepts, it is true, are such that we can quite well form some of them a priori without standing in an immediate relation to the object: namely, those that contain only the thought of an object in general, such as the concept of magnitude, of cause, and so on. But even these need a certain concrete use to give them meaning and sense: application to some intuition through which an object of these concepts is given to us. But how can the intuition of an object precede the object itself? ### Section 9 § 9. If our intuition had to be of a kind that represented things *as they are in themselves*, no a priori intuition would occur at all. It would always be empirical. For I can know what an object contains in itself only when it is present and given to me. Even then, admittedly, it is incomprehensible how the intuition of a present thing could let me know it as it is in itself, since its properties cannot migrate into my power of representation. But even if we grant this possibility, such intuition would still not occur a priori, that is, before the object was represented to me. Without that, no ground could be conceived for my representation's relation to the object, unless it rested on inspiration. So there is only one way my intuition can precede the object's actuality and occur as a priori cognition: *it must contain nothing but the form of sensibility, which in me as subject precedes all the actual impressions through which objects affect me*. For I can know a priori that objects of the senses can be intuited only in conformity with this form of sensibility. It follows that propositions concerning only this form of sensible intuition will be possible and valid for objects of the senses. Conversely, intuitions that are possible a priori can never concern anything other than objects of our senses. ### Section 10 § 10. It is therefore only through the form of sensible intuition that we can intuit things a priori. But through it we also know objects only as they can *appear* to us, to our senses, not as they may be in themselves. This assumption is absolutely necessary if synthetic a priori propositions are to be admitted as possible, or, where they actually occur, if their possibility is to be understood and determined in advance. Now space and time are the intuitions that pure mathematics takes as the basis of all its cognitions and judgments, which it also presents as apodictic and necessary. Mathematics must first present all its concepts in intuition, and pure mathematics in pure intuition: that is, it must construct them. Without this, it cannot take a step as long as it lacks pure intuition, in which alone the material for synthetic a priori judgments can be given. For it cannot proceed analytically, by analyzing concepts, but only synthetically. Geometry takes the pure intuition of space as its basis. Arithmetic produces its concepts of number by successively adding units in time. Above all, pure mechanics can produce its concepts of motion only through the representation of time. Both representations, however, are simply intuitions. If we remove everything empirical—everything belonging to sensation—from empirical intuitions of bodies and their changes, or motion, space and time still remain. They are therefore pure intuitions that underlie those empirical intuitions a priori and so can never themselves be removed. But precisely because they are pure a priori intuitions, they prove that they are mere forms of our sensibility. They must precede all empirical intuition, that is, the perception of actual objects. Objects can be known a priori in conformity with them, but of course only as they appear to us. ### Section 11 § 11. The problem of this part has therefore been solved. Pure mathematics is possible as synthetic a priori cognition only because it concerns nothing but objects of the senses. A pure intuition of space and time underlies the empirical intuition of these objects, and does so a priori. It can underlie that intuition because it is nothing but the mere form of sensibility, which precedes the objects' actual appearance and in fact first makes it possible. But this ability to intuit a priori does not concern the matter of appearance, that is, what is sensation in it, since that constitutes the empirical element. It concerns only the form of appearance: space and time. If anyone wanted to doubt in the slightest that space and time are not determinations belonging to things in themselves, but merely determinations belonging to their relation to sensibility, I would like to know how that person could find the following possible: to know a priori, and therefore before any acquaintance with things, before they are given to us, what the intuition of them must be like. Yet that is the case here with space and time. It is entirely understandable, however, as soon as the two are taken to be nothing more than formal conditions of our sensibility, and objects to be mere appearances. For then the form of appearance, that is, pure intuition, can certainly be represented from ourselves, that is, a priori. ### Section 12 § 12. To add some clarification and confirmation, we need only look at the usual and absolutely necessary procedure of geometers. All proofs of the complete equality of two given figures, where one can replace the other in every respect, ultimately come down to their coinciding with one another. This is plainly nothing but a synthetic proposition resting on immediate intuition. That intuition must be pure and given a priori. Otherwise the proposition could not count as apodictically certain, but would have only empirical certainty. It would say only that this is how we always observe it, and would hold only as far as our perception had so far extended. The claim that complete space—which is no longer itself a boundary of another space—has three dimensions, and that space in general cannot have more, is based on the proposition that no more than three lines can intersect at right angles at one point. But this proposition cannot be demonstrated from concepts at all. It rests directly on intuition, specifically pure a priori intuition, because it is apodictically certain. That we can require a line to be drawn on indefinitely (in indefinitum), or a series of changes—for example, spaces traversed through motion—to be continued indefinitely, presupposes a representation of space and time that can belong only to intuition, insofar as intuition is in itself limited by nothing. For it could never be inferred from concepts. So pure a priori intuitions really do underlie mathematics and make its synthetic, apodictically valid propositions possible. Our transcendental deduction of concepts in space and time therefore also explains the possibility of pure mathematics. Without such a deduction, and without assuming that “everything that may be given to our senses—to the outer senses in space, to the inner sense in time—is intuited by us only as it appears to us, not as it is in itself,” that possibility could be admitted, but could in no way be understood. ### Section 13 § 13. Those who still cannot give up the idea that space and time are real characteristics belonging to things in themselves can exercise their ingenuity on the following paradox. If they try to solve it without success, they may at least be free from prejudice for a few moments and suspect that perhaps there is some basis for reducing space and time to mere forms of our sensible intuition. If two things are completely alike in every respect that can ever be known in each taken by itself—in all determinations concerning magnitude and quality—then it must follow that one can replace the other in every case and relation, without this exchange producing the slightest discernible difference. This is in fact true of plane figures in geometry. But certain spherical figures, despite that complete internal agreement, show such a difference in their external relation that one cannot replace the other at all. For example, two spherical triangles in opposite hemispheres, with an arc of the equator as their common base, can be completely equal in both sides and angles. When either one is described on its own and completely, nothing is found in it that is not also in the description of the other. Yet one cannot be put in the other's place, namely in the opposite hemisphere. Here, then, is an internal difference between the two triangles that no understanding can identify as internal, and that reveals itself only through their external relation in space. But I will give more familiar cases that can be taken from everyday life. What could be more like my hand or my ear, and more equal to it in every respect, than its image in a mirror? Yet I cannot put a hand such as the one seen in the mirror in place of its original. If the original was a right hand, the one in the mirror is a left hand. And the image of the right ear is a left ear, which can never take the first one's place. Now there are no internal differences here that any understanding could even think. Yet the differences are internal as far as the senses teach us. Despite all the equality and similarity on both sides, the left hand cannot be enclosed within the same boundaries as the right: they cannot coincide. The glove for one hand cannot be used on the other. What, then, is the solution? These objects are not representations of things as they are in themselves and as pure understanding would know them. They are sensible intuitions, that is, appearances. Their possibility rests on the relation of certain things, unknown in themselves, to something else: our sensibility. Space is the form of this sensibility's outer intuition. The internal determination of any particular space is possible only by determining its external relation to the whole space of which it is a part—its relation to outer sense. That is, the part is possible only through the whole. This never occurs with things in themselves as objects of mere understanding, but it does occur with mere appearances. So we cannot make the difference between similar and equal but incongruent things—for example, snail shells spiraling in opposite directions—understandable through any single concept. We can do so only through their relation to the right and left hand, which refers directly to intuition. *Remark I.* Pure mathematics, and in particular pure geometry, can have objective reality only on the condition that it concerns solely objects of the senses. For these objects, however, the following principle is established: our sensible representation is in no way a representation of things in themselves, but only of how they appear to us. It follows that the propositions of geometry are not determinations of something merely created by our inventive imagination, so that they could not be reliably related to actual objects. Rather, they necessarily hold for space, and therefore also for everything that may be found in space. For space is nothing but the form of all outer appearances, under which alone objects of the senses can be given to us. Sensibility, whose form geometry takes as its basis, is what makes outer appearances possible. So these appearances can never contain anything other than what geometry prescribes for them. It would be quite different if the senses had to represent objects as they are in themselves. In that case, the representation of space that the geometer takes as a basis a priori, with all sorts of its properties, would not at all imply that all this, together with what is inferred from it, must hold in just the same way in nature. The geometer's space would be taken for a mere invention, with no claim to objective validity. For we do not see at all why things would necessarily have to agree with the image we form of them on our own and in advance. But suppose this image—or rather, this formal intuition—is the essential property of our sensibility, through which alone objects are given to us. And suppose this sensibility represents not things in themselves but only their appearances. It is then very easy to understand, and at the same time proved beyond dispute, that all outer objects of our sensible world must necessarily agree with the propositions of geometry in every detail. For sensibility, through its form of outer intuition—space, which the geometer studies—first makes those objects themselves possible as mere appearances. It will always remain a noteworthy phenomenon in the history of philosophy that there was a time when even mathematicians who were also philosophers began to doubt—not the correctness of their geometrical propositions insofar as these concerned space alone, but the objective validity and application to nature of this concept itself and all its geometrical determinations. They feared that a line in nature might after all consist of physical points, and therefore that true space in the object might consist of simple parts, even though the space the geometer has in mind cannot possibly consist of them. They did not recognize that this space in thought makes physical space, that is, the extension of matter itself, possible; that it is not at all a characteristic of things in themselves, but only a form of our sensible power of representation; and that all objects in space are mere appearances, that is, not things in themselves but representations of our sensible intuition. Space as the geometer thinks it is exactly the form of sensible intuition that we find in ourselves a priori, and that contains the ground of the possibility of all outer appearances as to their form. These appearances must therefore agree necessarily and with the greatest precision with the geometer's propositions. He draws them not from an invented concept, but from the subjective foundation of all outer appearances: sensibility itself. In this way, and no other, the geometer can be secured against all the quibbles of a shallow metaphysics regarding the undoubted objective reality of his propositions—however strange they must seem to that metaphysics, since it does not go back to the sources of its concepts. *Remark II.* Everything that is to be given to us as an object must be given in intuition. But all our intuition occurs only through the senses. The understanding intuits nothing; it only reflects. As we have just proved, the senses never let us know things in themselves in any respect at all, but only their appearances, which are mere representations of sensibility. “So all bodies, together with the space in which they are, must also be regarded as nothing but mere representations in us, and exist nowhere except in our thoughts.” Is this not obvious idealism? Idealism consists in claiming that none but thinking beings exist. The other things we believe we perceive in intuition are supposedly only representations in thinking beings, with no object outside those beings actually corresponding to them. I say, by contrast: things are given to us as objects of our senses located outside us. We know nothing about what they may be in themselves, but know only their appearances: the representations they produce in us by affecting our senses. So I certainly admit that there are bodies outside us. That is, there are things that are completely unknown to us as to what they may be in themselves, but that we know through the representations their influence on our sensibility provides us, and to which we give the name of a body. This word therefore means only the appearance of that object, which is unknown to us but nonetheless actual. Can this really be called idealism? It is precisely the opposite. We can say of many predicates of external things that they belong not to these things in themselves, but only to their appearances, and have no existence of their own outside our representation, without harming the actual existence of external things. This was generally accepted and granted long before *Locke's* time, but especially afterward. Warmth, color, taste, and so on belong here. For important reasons, I count the remaining qualities of bodies, called primary qualities, as mere appearances too: extension, place, and space in general, with everything attached to it—impenetrability or materiality, shape, and so on. Not the slightest ground can be given for declaring this inadmissible. Someone who will not accept colors as properties belonging to the object in itself, but regards them only as modifications belonging to the sense of sight, cannot be called an idealist on that account. Just as little can my doctrine be called idealist merely because I find that still more properties, indeed *all the properties that make up the intuition of a body*, belong only to its appearance. For this does not cancel the existence of the thing that appears, as actual idealism does. It only shows that through the senses we cannot know it at all as it is in itself. I would like to know what my claims would have to be like to avoid containing idealism. No doubt I would have to say not merely that the representation of space conforms perfectly to the relation our sensibility has to objects—for I have said that—but that it is even completely similar to the object. I can attach no meaning to that claim, any more than to the claim that the sensation of red resembles the property of cinnabar that arouses this sensation in me. *Remark III.* We can now easily dismiss an objection that is easy to foresee but groundless: “that the ideality of space and time would turn the entire sensible world into mere illusion.” All philosophical insight into the nature of sensible cognition had first been spoiled by treating sensibility as merely a confused way of representing things. On that view, we would still know things as they are, only without the ability to bring everything in our representation to clear awareness. We have proved, by contrast, that sensibility does not consist in this logical difference between clarity and obscurity. It consists in a genetic difference, one concerning the origin of cognition itself. Sensible cognition does not represent things as they are at all, but only how they affect our senses. Through it, therefore, only appearances, not the things themselves, are given to the understanding for reflection. After this necessary correction, an objection arises from an inexcusable and almost deliberate misinterpretation: that my doctrine turns all things in the sensible world into mere illusion. When an appearance is given to us, we are still entirely free in how we choose to judge the thing from it. The appearance rested on the senses, but this judgment rests on the understanding. The only question is whether the determination of the object contains truth or not. The difference between truth and dream is not settled by the character of the representations referred to objects, for these are alike in both cases. It is settled by their connection according to the rules that determine how representations are joined in the concept of an object, and by the extent to which they can or cannot stand together in one experience. It is not the appearances' fault at all if our cognition takes illusion for truth: that is, if the intuition through which an object is given to us is taken for a concept of the object, or even of its existence, which only the understanding can think. The senses represent the planets' course to us as sometimes forward and sometimes backward. In this there is neither falsity nor truth. As long as we acknowledge that this is, for now, only appearance, we make no judgment at all about the objective character of their motion. But a false judgment can easily arise if the understanding does not take care to prevent this subjective way of representing from being taken as objective. So we say: the planets seem to go backward. Yet the illusion is to be charged not to the senses, but to the understanding, which alone is responsible for making an objective judgment from the appearance. In this way, even if we did not reflect at all on the origin of our representations, but connected our sensible intuitions in space and time, whatever they might contain, according to the rules for joining all cognition in one experience, deceptive illusion or truth could arise, depending on whether we were careless or careful. This concerns only the understanding's use of sensible representations, not their origin. Likewise, suppose I take all representations of the senses, together with their form, space and time, as nothing but appearances, and take space and time as a mere form of sensibility that is not found at all in objects apart from sensibility. If I use those representations only in relation to possible experience, then taking them as mere appearances contains not the slightest temptation to error or any illusion. Despite this, they can still be correctly connected in experience according to the rules of truth. Thus all the propositions of geometry about space hold equally for all objects of the senses, and therefore for all possible experience, whether I regard space as a mere form of sensibility or as something belonging to things themselves. It is only in the first case, however, that I can understand how it is possible to know those propositions a priori of all objects of outer intuition. Otherwise, as regards all possible experience, everything stays just as it would if I had never departed from the common opinion. But suppose I dare to go beyond all possible experience with my concepts of space and time. This is unavoidable if I present them as characteristics belonging to things in themselves. For what would then prevent me from letting them hold for those very same things even if my senses were differently arranged, whether or not my senses fitted those things? Then a serious error can arise, based on an illusion. What was a condition of intuiting things that belonged only to me as subject, and certainly held for all objects of the senses and thus for all possible experience, I presented as universally valid. I did so because I related it to things in themselves instead of restricting it to the conditions of experience. So far from making the whole sensible world into mere illusion, my doctrine of the ideality of space and time is instead the only means of securing the application to actual objects of one of the most important bodies of cognition: the cognition that mathematics presents a priori. It also prevents this cognition from being taken as mere illusion. Without this observation, it would be entirely impossible to determine whether the intuitions of space and time—which we borrow from no experience, yet which lie a priori in our representation—might be mere fantasies of our own making. No object might correspond to them, or at least not correspond adequately, and geometry itself might therefore be a mere illusion. By contrast, we have been able to demonstrate its indisputable validity for all objects of the sensible world precisely because these are mere appearances. Second, these principles of mine are far from replacing the truth of experience with mere illusion because they make representations of the senses into appearances. Instead, they are the only means of preventing the transcendental illusion that has always deceived metaphysics and thereby led it into the childish pursuit of grasping at soap bubbles. This happened because appearances, which are mere representations, were taken for things in themselves. From this came all those remarkable episodes of the antinomy of reason that I will mention later. That antinomy is resolved by this single observation: as long as appearance is used within experience, it produces truth; but as soon as it goes beyond the limit of experience and becomes transcendent, it produces nothing but illusion. I therefore leave their actuality to the things that we represent through the senses. I only restrict our sensible intuition of these things: in no respect, not even in the pure intuitions of space and time, does it represent anything more than the mere appearance of those things, never their character in themselves. So this is no universal illusion that I have invented and attributed to nature. My protest against every accusation of idealism is so conclusive and clear that it would even seem unnecessary, were it not for unauthorized judges ready to replace well-defined concepts with their own delusions, thereby distorting and disfiguring them. They like to have an old name for every departure from their mistaken, though common, opinion. They never judge the spirit of philosophical terms, but cling only to their letter. The fact that I myself have given this theory of mine the name transcendental idealism gives no one the right to confuse it with the empirical idealism of *Descartes*. That was, however, only a problem; because he considered it insoluble, *Descartes* held that everyone was free to deny the existence of the bodily world, since the problem could never be adequately answered. Nor may my theory be confused with the mystical and visionary idealism of *Berkeley*. Against that and similar fantasies, our Critique in fact contains the proper antidote. What I called idealism did not concern the existence of things—doubting their existence is what actually constitutes idealism in the accepted sense. It never occurred to me to doubt that existence. It concerned only the sensible representation of things, to which space and time belong first and foremost. Of these, and therefore of all appearances in general, I have shown only that they are not things, but mere ways of representing, and are not determinations belonging to things in themselves either. The word transcendental, which in my usage never means a relation of our cognition to things, but only to the *faculty of cognition*, was meant to prevent this misinterpretation. But rather than let the term continue to cause it, I prefer to withdraw that name and call this idealism *critical*. If it is indeed a reprehensible idealism to turn actual things—not appearances—into mere representations, what name should we give to the idealism that, conversely, makes mere representations into things? I think we can call it *dreaming* idealism, to distinguish it from the former, which may be called *visionary* idealism, both of which, through my otherwise were supposed to have been kept from so-called transcendental idealism, better called *critical* idealism. ### Second part: How is pure natural science possible? *The main transcendental question* **Second Part.** *How is pure natural science possible?* ### Section 14 § 14. *Nature* is the *existence* of things insofar as it is determined by universal laws. If nature meant the existence of things in themselves, we could never know it, either a priori or a posteriori. We could not know it a priori. For how could we know what belongs to things in themselves? We can never find this out by analyzing our concepts, that is, through analytic propositions. I do not want to know what my concept of a thing contains, since that belongs to its logical essence. I want to know what the thing's reality adds to this concept, and what determines the thing itself in its existence outside my concept. My understanding, and the conditions under which alone it can connect the determinations of existing things, prescribe no rule to the things themselves. They do not conform to my understanding. Instead, my understanding would have to conform to them. They would therefore have to be given to me beforehand, so that I could learn these determinations from them. But then they would not be known a priori. Nor could we know the nature of things in themselves a posteriori. Suppose experience is to teach me the *laws* governing the existence of things. Insofar as these laws concern things in themselves, they would have to belong to them *necessarily*, even outside my experience. But experience does teach me what exists and how it is. It never teaches me that it must necessarily be this way and not otherwise. So it can never teach me the nature of things in themselves. ### Section 15 § 15. Yet we really do possess a pure natural science. It sets out a priori the laws governing nature, with all the necessity required for propositions that are absolutely certain. I need only point to the preparation for the study of nature known as general natural science. It comes before all physics, which is based on empirical principles. It contains mathematics applied to appearances. It also contains purely discursive principles, drawn from concepts, which make up the philosophical part of pure cognition of nature. Still, it also contains much that is not entirely pure and independent of experience. Examples are the concepts of *motion*, *impenetrability*—on which the empirical concept of matter rests—and *inertia*, among others. These prevent it from being called entirely pure natural science. Moreover, it deals only with objects of the outer senses. So it gives no example of a general natural science in the strict sense. Such a science must bring nature in general under universal laws, whether it concerns objects of the outer senses or of inner sense: the objects of physics as well as of psychology. But some principles of that general physics really do have the universality we require. Examples are the propositions *that substance remains* and persists, and *that everything that happens* is always *determined in advance* *by a cause* according to constant laws, and so on. These really are universal laws of nature that hold entirely a priori. So there actually is a pure natural science. The question now is: *how is it possible?* ### Section 16 § 16. The word nature has another meaning as well. This meaning identifies the *object*, whereas the earlier meaning indicated only the *lawfulness* of the determinations of the existence of things in general. Nature, then, considered in terms of its matter, is the *sum of all objects of experience*. This is all we are concerned with here. Besides, to know the nature of things that can never become objects of experience would require concepts whose meaning could never be given concretely, in any example of possible experience. We would therefore have to form concepts of their nature whose reality could not be decided at all. That is, we could not decide whether they really refer to objects or are merely things we think up. Knowledge of anything that cannot be an object of experience would be beyond physics. We are not concerned with anything of that kind here. We are concerned with cognition of nature whose reality can be confirmed through experience, even though it is possible a priori and precedes all experience. ### Section 17 § 17. The *formal aspect* of nature in this narrower sense is therefore the lawfulness of all objects of experience. Insofar as we know it a priori, it is their *necessary* lawfulness. But we have just shown that we can never know the laws of nature a priori in objects considered not in relation to possible experience, but as things in themselves. Here, however, we are not dealing with things in themselves. We leave their properties undecided. We are dealing only with things as objects of possible experience. Their sum is what we actually call nature here. Now I ask: when the issue is the possibility of knowing nature a priori, which way of putting the problem is better? Should we ask how it is possible to know a priori the necessary lawfulness *of things* as objects of experience? Or should we ask how it is possible to know a priori the necessary lawfulness *of experience* itself with respect to all its objects in general? On closer examination, the answer comes to exactly the same thing either way, as far as pure cognition of nature is concerned. That is the actual point of the question. For the subjective laws that alone make experiential cognition of things possible also apply to these things as objects of possible experience. Of course, they do not apply to them as things in themselves, but those are not under consideration here. The following two statements mean exactly the same thing. One says: a judgment of perception can never count as experience without the law that, whenever an event is perceived, it must always be related to something earlier which it follows according to a universal rule. The other says: everything that experience teaches us happens must have a cause. Still, it is more appropriate to choose the first formulation. We can know a priori, before any objects are given, the conditions that alone make experience of them possible. But we can never know what laws they may be subject to in themselves, apart from possible experience. So there is only one way to study the nature of things a priori. We must investigate the conditions and universal, though subjective, laws under which alone such cognition is possible as experience, as far as its mere form is concerned. Then we must determine from these the possibility of things as objects of experience. If instead I chose the second way of putting it, and sought the a priori conditions under which nature is possible as an *object* of experience, I could easily fall into a misunderstanding. I could imagine that I was speaking about nature as a thing in itself. I would then be driven into endless, fruitless efforts to find laws for things about which nothing has been given to me. So here we will deal only with experience and the universal conditions, given a priori, that make it possible. From these we will determine nature as the complete object of all possible experience. I think it will be clear that I do not mean rules for *observing* a nature that is already given. Those already presuppose experience. The question is not how we can learn nature's laws from it through experience. Those would not be a priori laws and would yield no pure natural science. The question is how the a priori conditions that make experience possible are also the sources from which all universal laws of nature must be derived. ### Section 18 § 18. We must therefore first note the following. All judgments of experience are empirical: they are based on immediate sense perception. But the reverse does not follow. Not all empirical judgments are therefore judgments of experience. Something must be added to what is empirical, and generally to what is given in sensible intuition. This addition consists of special concepts that originate entirely a priori in pure understanding. Every perception must first be brought under them. Through them it can then be turned into experience. *Empirical judgments, insofar as they have objective validity*, are **judgments of experience**. Those that *have only subjective validity* I call mere **judgments of perception**. The latter need no pure concept of the understanding. They need only the logical connection of perceptions in a thinking subject. The former, however, always require more than the representations of sensible intuition. They also require special *concepts originally produced in the understanding*. These are precisely what make the judgment of experience *objectively valid*. All our judgments are at first mere judgments of perception. They hold only for us, that is, for our subject. Only afterward do we give them a new relation, namely to an object. We then intend the judgment to hold for us at all times and equally for everyone. For if a judgment agrees with an object, all judgments about that same object must also agree with one another. So the objective validity of a judgment of experience means nothing other than its necessary universal validity. But the reverse is also true. If we find reason to regard a judgment as necessarily valid for everyone, we must also regard it as objective. This never rests on perception, but on the pure concept of the understanding under which the perception is brought. To regard the judgment as objective is to hold that it expresses a feature of the object, not just a relation between the perception and a subject. For there would be no reason why other people's judgments must necessarily agree with mine, if not for the unity of the object to which they all refer. They agree with that object and must therefore all agree with one another as well. ### Section 19 § 19. Objective validity and necessary universal validity for everyone are therefore interchangeable concepts. Although we do not know the object in itself, when we regard a judgment as universally valid and therefore necessary, this is exactly what we mean by its objective validity. Through this judgment we know the object by the universally valid and necessary connection of the given perceptions, even if the object otherwise remains unknown as it may be in itself. This holds for all objects of the senses. So judgments of experience do not get their objective validity from immediate cognition of the object, which is impossible. They get it only from the condition that makes empirical judgments universally valid. As we have said, this condition never rests on empirical conditions, or even on sensible conditions in general. It rests on a pure concept of the understanding. The object in itself always remains unknown. But suppose the concept of the understanding determines the connection of the representations that the object gives to our sensibility as universally valid. Then the object is determined through this relation, and the judgment is objective. Let us make this clear. Judgments that the room is warm, sugar is sweet, or wormwood is unpleasant[^299-1] have only subjective validity. I do not require that I should always find things this way, or that everyone else should find them as I do. They express only a relation of two sensations to the same subject, namely myself, and only in my present state of perception. They are therefore not meant to hold of the object. I call judgments like these judgments of perception. A judgment of experience is quite different. What experience teaches me under certain circumstances, it must teach me at all times and teach everyone else too. Its validity is not restricted to the subject or to the subject's state at that time. So I state all such judgments as objectively valid. For example, suppose I say that air is elastic. At first this is only a judgment of perception. I am merely relating two sensations in my senses to one another. If I want it to be called a judgment of experience, I require that this connection stand under a condition that makes it universally valid. I therefore intend that I, at all times, and everyone else too, must necessarily connect the same perception under the same circumstances. [^299-1]: I readily admit that these examples do not represent judgments of perception that could ever become judgments of experience, even if a concept of the understanding were added. They refer only to feeling. Everyone recognizes feeling as merely subjective, so it can never be attributed to the object. These judgments can therefore never become objective either. For now I only wanted to give an example of a judgment that is merely subjectively valid. Such a judgment contains no ground for necessary universal validity, and so none for a relation to the object. The next note gives an example of judgments of perception that become judgments of experience through the addition of a concept of the understanding. ### Section 20 § 20. We must therefore analyze experience in general. We need to see what this product of the senses and the understanding contains, and how a judgment of experience itself is possible. Its basis is intuition of which I am conscious, that is, perception (perceptio), which belongs only to the senses. But a second element is also needed: judging, which belongs only to the understanding. This judging can take two forms. First, I can merely compare perceptions and connect them in a consciousness of my state. Second, I can connect them in a consciousness in general. The first judgment is merely a judgment of perception and, to that extent, has only subjective validity. It is just a connection of perceptions in my state of mind, without reference to the object. So, contrary to the usual assumption, comparing perceptions and connecting them in one consciousness through judging is not enough for experience. This does not produce the universal validity and necessity of the judgment that alone can make it objectively valid and turn it into experience. So a quite different judgment must come first, before perception can become experience. The given intuition must be brought under a concept that determines the form of judging in general with respect to that intuition. This concept connects the empirical consciousness of the intuition in a consciousness in general, and thereby gives empirical judgments universal validity. Such a concept is a pure a priori concept of the understanding. All it does is determine for an intuition the general way in which it can serve in judgments. Suppose this concept is the concept of cause. It determines the intuition brought under it—for example, that of air—with respect to judging in general. Namely, the concept of air must stand to expansion as the antecedent stands to the consequent in a hypothetical judgment. So the concept of cause is a pure concept of the understanding. It is entirely different from any possible perception. It serves only to determine the representation contained under it with respect to judging in general, and thus to make a universally valid judgment possible. Before a judgment of perception can become a judgment of experience, then, the perception must first be brought under a concept of the understanding of this kind. For example, air comes under the concept of cause, which determines the judgment about it, with respect to expansion, as hypothetical.[^301-1] This means that expansion is represented not merely as belonging to my perception of air in my present state, or in several of my states, or in other people's state of perception. It is represented as belonging to it *necessarily*. The judgment that air is elastic becomes universally valid, and only then a judgment of experience, because certain judgments precede it. They bring the intuition of air under the concept of cause and effect. In this way they determine perceptions not merely in relation to one another in my subject, but with respect to the form of judging in general—here, the hypothetical form. This is how they make the empirical judgment universally valid. If we analyze all our synthetic judgments insofar as they hold objectively, we find that they never consist of mere intuitions connected in a judgment simply by comparison, as people usually think. Instead, these judgments would be impossible unless something were added to the concepts drawn from intuition: a pure concept of the understanding. The concepts drawn from intuition are brought under it, and only then are they connected in an objectively valid judgment. Even the judgments of pure mathematics in its simplest axioms are not exempt from this condition. The principle that a straight line is the shortest between two points presupposes that the line is brought under the concept of magnitude. This concept is certainly not a mere intuition. It belongs solely to the understanding. It serves to determine the intuition of the line in relation to the judgments that may be made about it, with respect to their quantity, namely plurality, as plural judgments (iudicia plurativa).[^302-1] For what these judgments mean is that a given intuition contains many things of the same kind. [^301-1]: For an example that is easier to grasp, take the following: when the sun shines on the stone, the stone becomes warm. This is merely a judgment of perception and contains no necessity, no matter how often I or others may have perceived this. The perceptions are only usually found connected this way. But if I say that the sun *warms* the stone, the understanding's concept of cause is added to the perception. This concept connects the concept of sunshine *necessarily* with that of heat. The synthetic judgment becomes necessarily universally valid, and therefore objective. It changes from a perception into experience. [^302-1]: This is the name I would prefer for the judgments that logic calls particularia, or particular judgments. For that latter expression already contains the thought that they are not universal. But if I start from unity, in singular judgments, and move on toward totality, I cannot yet mix in any reference to totality. I am thinking only of plurality without totality, not of an exception to totality. This is necessary if the logical functions are to serve as the basis for the pure concepts of the understanding. In their logical use, the old terminology can remain. ### Section 21 § 21. To show how experience is possible insofar as it rests on pure a priori concepts of the understanding, we must first present a complete table of what belongs to judging in general and of the understanding's different functions in it. The pure concepts of the understanding will correspond exactly to these functions. These concepts are nothing more than concepts of intuitions in general, insofar as the intuitions are themselves determined for judging in relation to one or another of these functions. They are therefore determined necessarily and with universal validity. This will also determine exactly the a priori principles that make all experience possible as objectively valid empirical cognition. For these principles are simply propositions that bring every perception, in accordance with certain universal conditions of intuition, under those pure concepts of the understanding. **Logical table** *of judgments.* | | 1. | | |---|---|---| | | *By quantity* | | | | Universal | | | | Particular | | | | Singular | | | 2. | | 3. | | *By quality* | | *By relation* | | Affirmative | | Categorical | | Negative | | Hypothetical | | Infinite | | Disjunctive | | | 4. | | | | *By modality* | | | | Problematic | | | | Assertoric | | | | Apodictic. | | **Transcendental table** *of concepts of the understanding.* | | 1. | | |---|---|---| | | *By quantity* | | | | Unity (the measure) | | | | Plurality (the magnitude) | | | | Totality (the whole) | | | 2. | | 3. | | *By quality* | | *By relation* | | Reality | | Substance | | Negation | | Cause | | Limitation | | Community | | | 4. | | | | *By modality* | | | | Possibility | | | | Existence | | | | Necessity. | | **Pure physiological table** of universal principles of natural science. | | 1. | | |---|---|---| | | *Axioms* | | | | of intuition | | | 2. | | 3. | | *Anticipations* | | *Analogies* | | of perception | | of experience | | | 4. | | | | *Postulates* | | | | of empirical thinking | | | | in general. | | ### Section 21a § 21. [a]. To bring everything said so far together in one concept, readers must first be reminded of something. The issue here is not how experience comes into being, but what it contains. The first question belongs to empirical psychology. Even there it could never be properly worked out without the second question, which belongs to the critique of cognition, and especially of the understanding. Experience consists of intuitions, which belong to sensibility, and judgments, which are solely the work of the understanding. But judgments that the understanding makes from sensible intuitions alone are still far from being judgments of experience. In the first case the judgment would only connect perceptions as they are given in sensible intuition. In the second case, however, the judgments are meant to state what experience in general contains, not what mere perception contains, whose validity is only subjective. So a judgment of experience must add something to sensible intuition and to its logical connection in a judgment, after comparison has made it general. This addition determines the synthetic judgment as necessary and therefore as universally valid. It can only be the concept that represents the intuition as determined in itself with respect to one form of judgment rather than another. That is, it is a concept of the synthetic unity of intuitions that can be represented only through a given logical function of judgments. ### Section 22 § 22. Here is the result. The task of the senses is to intuit. The task of the understanding is to think. Thinking means uniting representations in one consciousness. This unification either occurs only in relation to the subject, in which case it is contingent and subjective, or it occurs without qualification, in which case it is necessary or objective. A judgment is the unification of representations in one consciousness. So thinking amounts to judging, or to relating representations to judgments in general. Judgments are therefore either merely subjective, when representations are related only to a consciousness in one subject and united in it, or objective, when they are united in a consciousness in general, that is, necessarily in it. The logical functions of all judgments are so many possible ways of uniting representations in one consciousness. But when these same functions serve as concepts, they are concepts of the *necessary* unification of representations in one consciousness. They are therefore principles of objectively valid judgments. This unification in one consciousness is either analytic, through identity, or synthetic, through the combining and adding of different representations to one another. Experience consists in the synthetic connection of appearances, or perceptions, in one consciousness, insofar as that connection is necessary. Pure concepts of the understanding are therefore the concepts under which all perceptions must first be brought before they can serve in judgments of experience. In those judgments the synthetic unity of perceptions is represented as necessary and universally valid.[^305-1] [^305-1]: But how does the claim that judgments of experience must contain necessity in the synthesis of perceptions agree with my earlier, repeatedly stressed claim that experience, as a posteriori cognition, can provide only contingent judgments? When I say that experience teaches me something, I always mean only the perception it contains. For example, warmth always follows the sun's shining on the stone. To that extent, the proposition drawn from experience is always contingent. The judgment of experience does contain the claim that this warming follows necessarily from the sun's shining, by virtue of the concept of cause. But I do not learn this through experience. On the contrary, experience is first produced by adding this concept of the understanding—the concept of cause—to perception. For how perception comes to receive this addition, consult the Critique, in the section on the transcendental power of judgment, page 137 and following. ### Section 23 § 23. Judgments, considered merely as the condition for uniting given representations in one consciousness, are rules. Insofar as these rules represent the unification as necessary, they are a priori rules. Insofar as no higher rules exist from which they are derived, they are principles. Now consider the possibility of all experience, looking only at its form of thought. There are no conditions of judgments of experience higher than those that bring appearances, according to the different forms of their intuition, under pure concepts of the understanding. These concepts make the empirical judgment objectively valid. So these conditions are the a priori principles of possible experience. The principles of possible experience are also universal laws of nature that can be known a priori. This solves the problem posed by our present second question: *How is pure natural science possible?* For the systematic structure required for the form of a science is fully present here. Logic provides the formal conditions of all judgments in general, and therefore of all rules in general. No further such conditions are possible beyond these. They make up a logical system. The concepts based on them contain the a priori conditions for all synthetic and necessary judgments. For that very reason they make up a transcendental system. Finally, the principles through which all appearances are brought under these concepts make up a physiological system, that is, a system of nature. This system precedes all empirical cognition of nature and first makes it possible. It can therefore be called general and pure natural science in the proper sense. ### Section 24 § 24. The first[^306-1] of these physiological principles brings all appearances, as intuitions in space and time, under the concept of *magnitude*. To that extent it is a principle for applying mathematics to experience. The second does not directly bring the properly empirical element under the concept of *magnitude*. That element is sensation, which indicates what is real in intuitions. Sensation is not an intuition that *contains* space or time, although it places its corresponding object in both. Yet in time there is a difference with a magnitude between reality, represented in sensation, and zero, that is, the completely empty intuition. For between any given degree of light and darkness, between any degree of heat and complete cold, between any degree of heaviness and absolute lightness, and between any degree of filling space and completely empty space, we can always think of still smaller degrees. In the same way there are always smaller degrees even between consciousness and complete unconsciousness, or psychological darkness. So no perception is possible that would prove an absolute absence. For example, there is no psychological darkness that could not be regarded as a consciousness merely outweighed by another, stronger one. The same holds in every case of sensation. This is why the understanding can even anticipate sensations, which make up the actual quality of empirical representations, or appearances. It does this through the principle that all sensations, and therefore what is real in every appearance, have degrees. This is the second application of mathematics to natural science: the mathematics of intensities (mathesis intensorum). [^306-1]: These three consecutive sections can hardly be properly understood without consulting what the Critique says about the principles. But they may help readers survey the general outline more easily and attend to the main points. ### Section 25 § 25. When we consider the relation of appearances solely with respect to their existence, determining this relation is not mathematical but dynamical. It can never be objectively valid, and therefore suitable for experience, unless it stands under a priori principles that first make experiential cognition of this relation possible. So appearances must be brought under the concept of substance. The concept of substance, as a concept of the thing itself, underlies every determination of existence. Or, second, if a succession in time is found among appearances—that is, an event—they must be brought under the concept of an effect in relation to a cause. Or, if their coexistence is to be known objectively, that is, through a judgment of experience, they must be brought under the concept of community, or reciprocal interaction. In this way a priori principles underlie objectively valid, though empirical, judgments. That is, they underlie the possibility of experience insofar as experience is to connect objects in nature with respect to their existence. These principles are the laws of nature in the proper sense. They can be called dynamical laws. Finally, judgments of experience also include cognition of agreement and connection. This concerns not so much the agreement and connection of appearances with one another in experience, as their relation to experience in general. This relation involves either their agreement with the formal conditions that the understanding knows, or their connection with the material element of the senses and perception, or both united in one concept. It therefore involves possibility, actuality, and necessity according to universal laws of nature. This would make up the doctrine of method for natural science: the distinction between truth and hypotheses, and the limits of the reliability of the latter. ### Section 26 § 26. The third table, the table of principles, was drawn *from the nature of the understanding itself* by the critical method. It has a completeness that puts it far above any table ever attempted, though in vain, by the dogmatic method *starting from the things themselves*, or any that might be attempted in the future. It contains all synthetic a priori principles without exception. It was worked out from one principle: the capacity to judge in general, which makes up the essence of experience as far as the understanding is concerned. So we can be certain that there are no more principles of this kind. The dogmatic method can never give us this satisfaction. Still, this is far from the table's greatest merit. We must attend to the ground of proof. It reveals how this a priori cognition is possible. At the same time, it restricts all these principles to a condition that must never be overlooked. Otherwise the principles will be misunderstood and used beyond the original meaning the understanding gives them. This condition is that they contain only the conditions of possible experience in general, insofar as experience is subject to a priori laws. So I do not say that things *in themselves* have a magnitude, that their reality has a degree, or that their existence involves the connection of accidents in a substance, and so on. No one can prove this. For such a synthetic connection from mere concepts is absolutely impossible when there is no relation to sensible intuition on the one hand, and no connection of those intuitions in a possible experience on the other. The essential restriction on the concepts in these principles is therefore this: all things necessarily stand a priori under the stated conditions only *as objects of experience*. A second consequence is a distinctive way of proving these principles. The principles in question are not related directly to appearances and their relation to one another. They are related to the possibility of experience, of which appearances make up only the matter, not the form. That is, they are related to objectively and universally valid synthetic propositions. This is precisely what distinguishes judgments of experience from mere judgments of perception. The proof works as follows. Appearances, as mere intuitions *that occupy a part of space and time*, stand under the concept of magnitude. This concept synthetically unites their manifold content a priori according to rules. Further, perception contains sensation as well as intuition. There is always a transition by decrease between a sensation and zero, that is, its complete disappearance. So what is real in appearances must have a degree. Sensation itself *occupies no part of space or time*,[^309-1] yet the transition to it from empty time or space is possible only in time. Thus sensation, as the quality of empirical intuition, can never be known a priori with respect to what specifically distinguishes it from other sensations. Nevertheless, in a possible experience in general, as an intensive magnitude of perception, it can be distinguished from every other sensation of the same kind. This first makes possible and determines the application of mathematics to nature with respect to the sensible intuition through which nature is given to us. Above all, however, the reader must attend to the way the principles called analogies of experience are proved. Unlike the principles for applying mathematics to natural science in general, these do not concern the production of intuitions. They concern the connection of their existence in one experience. This connection can only be the determination of existence in time according to necessary laws, under which alone it is objectively valid and therefore experience. So the proof concerns not synthetic unity in the connection of things in themselves, but synthetic unity in the connection of *perceptions*. And it concerns these not in terms of their content, but in terms of their determination in time and the relation of existence within time according to universal laws. These universal laws therefore contain the necessity of determining existence in time in general, and thus according to an a priori rule of the understanding, if the empirical determination in relative time is to be objectively valid and therefore experience. In a work of prolegomena I cannot say more here. I can only offer a recommendation to the reader who has long been used to treating experience as a merely empirical combination of perceptions. Such a reader does not even consider that experience goes much further than perceptions can reach. It gives empirical judgments universal validity and, for that purpose, requires a pure unity of the understanding that precedes it a priori. I recommend paying close attention to this difference between experience and a mere collection of perceptions, and judging the method of proof from this point of view. [^309-1]: Heat, light, and so on are just as great in degree in a small space as in a large one. Likewise, inner representations, pain, and consciousness in general are no smaller in degree whether they last a short time or a long one. So here the magnitude at a point and at an instant is just as great as in a space or time of any size. Degrees are therefore greater not in terms of intuition, but in terms of mere sensation, or in terms of the magnitude of the ground of an intuition. They can be estimated as magnitudes only through the relation of 1 to 0. That is, each can decrease through infinitely many intermediate degrees until it disappears, or grow from zero through infinitely many increments to a determinate sensation within a certain time (quantitas qualitatis est gradus: the quantity of a quality is its degree). ### Section 27 § 27. This is the place to remove *Hume's* doubt at its root. He rightly maintained that reason gives us no way to understand the possibility of causality. That is, we cannot understand the relation of the existence of one thing to the existence of something else that is necessarily posited by it. I would add that we can no more understand the concept of subsistence: the necessity that the existence of things have an underlying subject that cannot itself be a predicate of anything else. Indeed, we cannot even form a concept of the possibility of such a thing, although we can point to examples of the concept's use in experience. The same incomprehensibility also applies to the community of things. We cannot understand at all how the state of one thing allows an inference to the state of entirely different things outside it, and reciprocally. Nor can we understand how substances, each of which has its own separate existence, are supposed to depend on one another, and necessarily so. Yet I am far from regarding these concepts as merely borrowed from experience. Nor do I regard the necessity represented in them as invented, a mere illusion that long habit presents to us. Instead, I have sufficiently shown that these concepts, and the principles drawn from them, are established a priori before all experience. They have unquestionable objective correctness, though admittedly only in relation to experience. ### Section 28 § 28. So I have not the slightest concept of such a connection of things in themselves: how they can exist as substances, act as causes, or stand in community with others as parts of a real whole. Still less can I think of properties of this kind in appearances as appearances. For these concepts contain nothing that lies in appearances, but only what the understanding must think. Yet we do have a concept of such a connection of representations in our understanding, specifically in judgments in general. In one kind of judgment, representations stand as subject in relation to predicates. In another, they stand as ground in relation to consequence. In a third, they are parts that together make up a whole possible cognition. Further, we know a priori that we could have no cognition that holds of an object unless we regarded the representation of that object as determined with respect to one or another of these functions. If we were dealing with the object in itself, however, no mark would be available by which I could recognize it as determined with respect to any of these functions. That is, I could not recognize it as coming under the concept of substance or cause, or, in relation to other substances, under the concept of community. For I have no concept of how such a connection of existence is possible. But the question is not how things in themselves are determined with respect to these functions of judgments in general. It is how experiential cognition of things is so determined. That is, how can and must things as objects of experience be brought under these concepts of the understanding? Here it is clear that I fully understand not only the possibility but also the necessity of bringing all appearances under these concepts. In other words, I understand using these concepts as principles of the possibility of experience. ### Section 29 § 29. Let us try this out on *Hume's* problematic concept, the concept of cause—his crux metaphysicorum, or stumbling block for metaphysicians. First, logic gives me a priori the form of a conditional judgment in general: to use one given cognition as the ground and another as the consequence. Now perception may present a rule of relation saying that one appearance constantly follows a certain other appearance, though not the reverse. This is a case for using a hypothetical judgment. For example: if a body is exposed to the sun long enough, it becomes warm. Of course, this does not yet contain a necessity of connection, and therefore not the concept of cause. But I go further and say: if that proposition, which is merely a subjective connection of perceptions, is to be a proposition of experience, it must be regarded as necessary and universally valid. Such a proposition would be: the sun, through its light, is the cause of heat. The empirical rule above is now regarded as a law. It holds not merely of appearances, but of appearances for the purpose of a possible experience. Such experience requires rules that hold without exception and therefore necessarily. So I understand very well the concept of cause as a concept necessarily belonging to the mere form of experience. I also understand its possibility as a synthetic unification of perceptions in a consciousness in general. But I do not understand at all how a thing in general is possible as a cause. The reason is that the concept of cause indicates no condition belonging to things, but only one belonging to experience. Namely, experience can be objectively valid cognition of appearances and their succession in time only insofar as the preceding appearance can be connected with the following one according to the rule of hypothetical judgments. ### Section 30 § 30. So pure concepts of the understanding also have no meaning at all if they leave objects of experience behind and try to refer to things in themselves (noumena). They serve, so to speak, only to spell out appearances so that we can read them as experience. The principles that arise from relating these concepts to the world of sense serve our understanding only for use in experience. Beyond that, they are arbitrary combinations without objective reality. We can neither know their possibility a priori nor confirm their relation to objects through any example, or even make it understandable. For all examples can be drawn only from some possible experience. So the objects of those concepts, too, can be found nowhere but in a possible experience. This complete solution of *Hume's* problem, though it runs against what its author expected, therefore preserves the a priori origin of the pure concepts of the understanding. It also preserves the validity of universal laws of nature as laws of the understanding. But it does so by restricting their use to experience, because their possibility is grounded solely in the understanding's relation to experience. It does not derive them from experience. Instead, it derives experience from them. This completely reversed way of connecting them never occurred to *Hume*. The result of all our investigations so far is therefore this: “All synthetic a priori principles are nothing more than principles of possible experience.” They can never refer to things in themselves, but only to appearances as objects of experience. So pure mathematics and pure natural science can never concern anything more than mere appearances either. They represent only what makes experience in general possible, or what is derived from these principles and must always be capable of being represented in some possible experience. ### Section 31 § 31. At last, then, we have something definite to hold on to in all metaphysical undertakings. Up to now these have ventured boldly enough, but always blindly, over everything without distinction. Dogmatic thinkers never imagined that the goal of their efforts would be set within such narrow bounds. Nor did those who defiantly relied on what they supposed was their sound reason. They used concepts and principles of pure reason that were legitimate and natural, but intended only for use in experience, to pursue insights for which they knew no definite limits and could know none. For they had never reflected, or been able to reflect, on the nature or even the possibility of such a pure understanding. Many a self-taught practitioner of pure reason—by this I mean someone who trusts himself to decide questions of metaphysics without any science—might well claim that he had long since done more than suspect what has been set out here with so much preparation, or, if he prefers, with such long-winded, pedantic display. Through the prophetic spirit of his sound reason, he had already known and understood it: “With all our reason, we can never go beyond the field of experiences.” But if one gradually asks him to state his principles of reason, he must admit that many were not drawn from experience. They are therefore independent of experience and valid a priori. How, then, and on what grounds, will he keep the dogmatist and himself within bounds? The dogmatist uses these concepts and principles beyond all possible experience precisely because they are known independently of it. And even he himself, this expert in sound reason, is not so safe from slipping unnoticed beyond objects of experience into the realm of fantasies, despite all the wisdom he claims to have gained so cheaply. In fact, he is usually deeply entangled in it. He merely gives his groundless claims a plausible appearance through popular language, presenting everything as only probability, reasonable conjecture, or analogy. ### Section 32 § 32. Since the earliest days of philosophy, investigators of pure reason have thought of special beings of the understanding (noumena), in addition to sensible beings or appearances (phaenomena), which make up the world of sense. These beings of the understanding were supposed to make up a world of the understanding. Because these thinkers treated appearance and illusion as the same thing—a mistake we can readily forgive in an age not yet fully developed—they granted reality only to beings of the understanding. In fact, if we rightly regard objects of the senses as mere appearances, we also admit that a thing in itself underlies them. We do not know what this thing is like in itself. We know only its appearance, that is, the way our senses are affected by this unknown something. So the understanding, precisely by accepting appearances, also admits the existence of things in themselves. To that extent we can say that representing beings of this kind, which underlie appearances and are therefore mere beings of the understanding, is not only permissible but unavoidable. Our critical deduction does not exclude such things (noumena) at all. Instead, it restricts the principles of the aesthetic: they must not extend to all things, which would turn everything into mere appearance. They must hold only of objects of possible experience. Beings of the understanding are therefore admitted, but with this rule firmly stressed, without any exception: we know, and can know, absolutely nothing definite about these pure beings of the understanding. For our pure concepts of the understanding, as well as our pure intuitions, concern nothing but objects of possible experience, and therefore mere sensible beings. As soon as we leave these behind, those concepts no longer retain the slightest meaning. ### Section 33 § 33. There is indeed something deceptive about our pure concepts of the understanding: they tempt us to a transcendent use. By this I mean a use that goes beyond all possible experience. Our concepts of substance, force, action, reality, and so on are completely independent of experience. They also contain no appearance of the senses at all. So they really do seem to concern things in themselves (noumena). But something else strengthens this supposition: they contain a necessity of determination that experience can never match. The concept of cause contains a rule according to which one state necessarily follows from another. But experience can show us only that one state of things often, or at most usually, follows another. So it can provide neither strict universality nor necessity, and so on. Concepts of the understanding therefore seem to have much more meaning and content than their mere use in experience could fully account for. So, without noticing it, the understanding adds a much larger annex to the house of experience. It fills this annex with nothing but beings of thought, without even realizing that, with its otherwise correct concepts, it has ventured beyond the limits of their use. ### Section 34 § 34. Two important, indeed indispensable, though extremely dry investigations were therefore needed. They are carried out in the Critique, page 137 and following, and page 235 and following. The first showed that the senses do not provide pure concepts of the understanding in concrete form. They provide only the schema for using them. The object corresponding to that schema is found only in experience, as the understanding's product from the materials of sensibility. The second investigation (Critique, p. 235) shows the following. Despite the independence of our pure concepts and principles of the understanding from experience, and even their apparently wider range of use, nothing at all can be thought through them outside the field of experience. For they can do no more than determine the logical form of judgment with respect to given intuitions. But beyond the field of sensibility there is no intuition at all. Those pure concepts therefore lack all meaning, since there is no way to present them concretely. Consequently, all such noumena, together with their sum, an intelligible[^316-1] world, are nothing but representations of a problem. Its object may well be possible in itself, but its solution is entirely impossible given the nature of our understanding. For our understanding is not a capacity for intuition. It is only a capacity for connecting given intuitions in one experience. Experience must therefore contain all the objects for our concepts. Outside it, all concepts will be without meaning, since no intuition can be supplied as their basis. [^316-1]: Not an *intellectual* world, as people usually put it. For *cognitions* through the understanding are *intellectual*, and these also concern our world of sense. But *objects* are called *intelligible* insofar as they can be represented *only through the understanding*, and none of our sensible intuitions can refer to them. Yet some possible intuition must correspond to every object. So we would have to think of an understanding that intuited things immediately. We have not the slightest concept of such an understanding, and therefore none of the *beings of the understanding* that it is supposed to concern. ### Section 35 § 35. We may perhaps forgive the imagination when it sometimes indulges in fantasies, that is, fails to stay carefully within the bounds of experience. At least such a free flight enlivens and strengthens it. It will always be easier to moderate its boldness than to remedy its weakness. But the understanding can never be forgiven for *fantasizing* instead of *thinking*, which is its task. For it alone provides all the help we have in setting limits to the imagination's fantasies where this is needed. Yet it starts out very innocently and modestly. First it clarifies the elementary cognitions it possesses before all experience, but which must still always find its application in experience. Gradually it drops these restrictions. And what should stop it, since the understanding has drawn its principles quite freely from itself? Now it turns first to newly invented forces in nature, and soon afterward to beings outside nature. In short, it turns to a world for whose construction we cannot lack building materials. Fertile invention supplies them in abundance, and experience never refutes them, though it does not confirm them either. This is also why young thinkers so love metaphysics in the genuinely dogmatic manner, and often sacrifice their time and otherwise useful talents to it. But it does no good to try to moderate these fruitless attempts of pure reason by all sorts of reminders about the difficulty of solving such deeply hidden questions, complaints about the limits of our reason, or reductions of assertions to mere conjectures. Unless their *impossibility* is clearly demonstrated, and reason's *self-cognition* becomes a true science, these vain efforts will never be fully stopped. In that science, the field of reason's correct use must be distinguished from that of its empty and fruitless use with, so to speak, geometrical certainty. ### Section 36 § 36. *How is nature itself possible?* This question is the highest point that transcendental philosophy can ever reach. It must also be led to this point as its boundary and completion. The question actually contains two questions. **First:** How is nature possible in the *material* sense, that is, in terms of intuition, as the sum of appearances? How are space, time, and what fills them both—the object of sensation—possible at all? The answer is: through the constitution of our sensibility. According to this constitution, sensibility is affected in its own distinctive way by objects that are unknown to it in themselves and entirely different from those appearances. The answer is given in the book itself in the transcendental aesthetic, and here in the Prolegomena through the solution of the first main question. **Second:** How is nature possible in the formal sense, as the sum of the rules under which all appearances must stand if they are to be thought as connected in one experience? The answer can only be this: nature is possible only through the constitution of our understanding. According to this constitution, all those representations of sensibility are necessarily related to one consciousness. This first makes possible our distinctive way of thinking, namely through rules, and through these rules, experience. Experience must be clearly distinguished from insight into objects in themselves. This answer is given in the book itself in the transcendental logic, and here in the Prolegomena in the course of solving the second main question. But how this distinctive property of our sensibility itself is possible cannot be further resolved or answered. Nor can we explain how the distinctive property of our understanding is possible, or the necessary apperception underlying it and all thinking. For we always need these again for every answer and for all thinking about objects. There are many laws of nature that we can know only through experience. But no experience can teach us the lawfulness in the connection of appearances, that is, nature in general. For experience itself requires laws that underlie its possibility a priori. So the possibility of experience in general is also the universal law of nature, and the principles of the former are themselves the laws of the latter. For we know nature only as the sum of appearances, that is, of representations in us. We can therefore obtain the law of their connection only from the principles of their connection in us. That is, we obtain it from the conditions of their necessary unification in one consciousness, which makes experience possible. Even the main proposition developed throughout this section—that universal laws of nature can be known a priori—already leads by itself to a further proposition. The highest legislation for nature must lie in ourselves, that is, in our understanding. We must not seek nature's universal laws from nature through experience. Instead, we must seek nature, with respect to its universal lawfulness, solely from the conditions that make experience possible and that lie in our sensibility and understanding. Otherwise, how would it be possible to know these laws a priori? They are not rules of analytic cognition, but genuine synthetic extensions of cognition. Such an agreement, and a necessary one, between the principles of possible experience and the laws that make nature possible can have only two causes. Either these laws are borrowed from nature through experience, or, conversely, nature is derived from the laws that make experience in general possible and is completely identical with experience's mere universal lawfulness. The first option contradicts itself. For universal laws of nature can and must be known a priori, that is, independently of all experience, and must underlie every empirical use of the understanding. So only the second option remains.[^319-1] But we must distinguish empirical laws of nature, which always presuppose particular perceptions, from pure or universal laws of nature. These latter laws are not based on particular perceptions. They contain only the conditions for the necessary unification of perceptions in one experience. With respect to these latter laws, nature and possible experience are exactly the same. In experience, lawfulness rests on the necessary connection of appearances in one experience. Without that connection we could know no object of the world of sense at all. So this lawfulness rests on the original laws of the understanding. Therefore, though it sounds strange at first, it is nonetheless certain when I say, with respect to these latter laws: *the understanding does not draw its laws (a priori) from nature, but prescribes them to nature.* [^319-1]: *Crusius* alone knew of a middle way: a spirit that can neither err nor deceive originally implanted these laws of nature in us. But deceptive principles often get mixed in as well, and this man's own system provides quite a few examples. We lack reliable criteria for distinguishing a genuine origin from a spurious one. So using such a principle is a very doubtful business. We can never know for certain what the spirit of truth, or the father of lies, may have instilled in us. ### Section 37 § 37. Let us clarify this seemingly bold proposition with an example. It will show that we ourselves already regard laws we discover in objects of sensible intuition, especially when we know them as necessary, as laws the understanding has put into those objects. This is so even though they are otherwise like the laws of nature we attribute to experience in every respect. ### Section 38 § 38. Consider the properties of a circle. This figure immediately unites so many arbitrary determinations of space within it under one universal rule that we cannot help attributing a nature to this geometrical thing. For example, take two lines that intersect each other and also cut the circle. No matter how randomly they are drawn, they always divide each other so regularly that the rectangle formed from the segments of either line equals the rectangle formed from those of the other. Now I ask: “Does this law lie in the circle, or does it lie in the understanding?” In other words, does the figure contain the ground of this law independently of the understanding? Or does the understanding, having itself constructed the figure according to its concepts, namely the equality of the radii, also put into it the law of chords intersecting in geometrical proportion? If we follow the proofs of this law, we soon see that it can be derived only from the condition the understanding made the basis of constructing this figure: the equality of the radii. Now let us extend this concept, to trace further the unity of the many properties of geometrical figures under common laws. We consider the circle as a conic section, which therefore stands under the same basic conditions of construction as other conic sections. We find that all chords intersecting inside these latter figures—the ellipse, parabola, and hyperbola—always do so in such a way that the rectangles formed from their segments, though not equal, always stand in equal ratios to one another. Let us go still further, to the basic teachings of physical astronomy. We find a physical law of mutual attraction extending throughout material nature. Its rule is that attraction decreases inversely as the square of the distances from each attracting point, just as the spherical surfaces over which this force spreads increase. This seems to lie necessarily in the nature of things themselves, and is therefore usually presented as knowable a priori. The sources of this law are simple: they rest only on the ratio of spherical surfaces with different radii. Yet its consequence is remarkable for the variety of the agreement it produces and for its regularity. It yields not only all possible paths of heavenly bodies in conic sections, but also such a relation of those paths to one another that no law of attraction other than the inverse-square law of distances can be conceived as suitable for a system of the world. Here, then, is nature resting on laws that the understanding knows a priori, chiefly from universal principles for determining space. Now I ask: do these laws of nature lie in space? Does the understanding learn them simply by trying to investigate the rich meaning contained in space? Or do they lie in the understanding and in the way it determines space according to the conditions of synthetic unity, toward which all its concepts are directed? Space is so uniform, and so undetermined with respect to all particular properties, that we certainly will not look in it for a treasure of natural laws. Instead, it is the understanding that determines space into the shape of a circle, a cone, or a sphere, insofar as it contains the ground of the unity of their construction. So the mere universal form of intuition called space is indeed the underlying basis of all intuitions that can be determined with respect to particular objects. And this basis certainly contains the condition of their possibility and variety. But the unity of objects is determined solely by the understanding, according to conditions that lie in its own nature. So the understanding is the source of the universal order of nature. It brings all appearances under its own laws and thereby first produces experience a priori as far as its form is concerned. Through this experience, everything that is to be known only through experience is necessarily subjected to the understanding's laws. For we are not dealing with the nature of *things in themselves*. That is independent of the conditions both of our sensibility and of our understanding. We are dealing with nature as an object of possible experience. Here the understanding, by making experience possible, also ensures that the world of sense is either no object of experience at all, or a nature. ### Section 39 § 39. Appendix *to pure natural science.* *On the system of categories.* Nothing can be more welcome to a philosopher than being able to derive a variety of concepts or principles from one a priori principle, and in this way bring them all together in one cognition. Before, they had appeared scattered as he used them in concrete cases. He merely believed that he had collected all the concepts left over after a certain abstraction, which, when compared with one another, seemed to make up a special kind of cognition. But this was only an *aggregate*. Now he knows that just this many concepts, no more and no fewer, can make up that kind of cognition. He understands the necessity of his division, and this is genuine comprehension. Only now does he have a *system*. Consider the task of picking out, from ordinary cognition, the concepts that are not based on any particular experience but still occur in all experiential cognition, of which they make up, so to speak, the mere form of connection. This required no more reflection or insight than picking out from a language the rules for the actual use of words in general, and so gathering elements for a grammar. Indeed, the two investigations are very closely related. Yet this does not let us explain why each language has precisely this formal constitution and no other. Still less does it show that precisely this many formal determinations of language in general can be found, no more and no fewer. *Aristotle* had collected ten such pure elementary concepts under the name of categories.[^323-1] Later he found it necessary to add five postpredicaments to these concepts, which were also called predicaments.[^323-2] Some of the additions, however, are already contained in the earlier ones, such as priority, simultaneity, and motion. But this patchwork could count as, and deserve approval as, a hint for future investigators rather than a systematically worked-out idea. So, as philosophy became more enlightened, it was rejected as entirely useless. While investigating the pure elements of human cognition, which contain nothing empirical, I succeeded only after long reflection in reliably distinguishing and separating the pure elementary concepts of sensibility—space and time—from those of the understanding. This excluded the 7th, 8th, and 9th categories from that list. The remaining ones were of no use to me. For there was no principle by which the understanding could be fully surveyed, and all its functions, from which its pure concepts arise, could be determined completely and precisely. To find such a principle, I looked for an act of the understanding that contains all the others. It would differ only through various modifications or functions in bringing the manifold content of representation under the unity of thinking in general. I found that this act of the understanding consists in judging. Here the logicians had already done the work, though it was not yet entirely free of defects. This enabled me to present a complete table of pure functions of the understanding, though these were undetermined with respect to any object. Finally, I related these functions of judging to objects in general, or rather to the condition for determining judgments as objectively valid. Pure concepts of the understanding arose from this. I could be certain beyond doubt that just these concepts, and just this many, no more and no fewer, can make up all our cognition of things through the understanding alone. I rightly called them by their old name, *categories*. I reserved the task of adding, under the name of *predicables*, a complete list of all the concepts that can be derived from them. These may be derived by connecting the categories with one another, with the pure form of appearance, space and time, or with its matter insofar as it is not yet empirically determined, that is, the object of sensation in general. I would add these as soon as a system of transcendental philosophy was achieved. In preparation for that system, my present concern was only with the critique of reason itself. But the essential feature of this system of categories, which distinguishes it from that old patchwork put together without any principle, is this: it made it possible to determine exactly the true meaning of the pure concepts of the understanding and the condition for using them. This is also why it alone deserves to count as philosophy. For it showed that, taken by themselves, the categories are nothing but logical functions. As such, they do not provide the slightest concept of an object in itself. Instead, they require an underlying sensible intuition. Then they serve only to determine empirical judgments with respect to the functions of judging. Otherwise those judgments are undetermined and indifferent with respect to all these functions. In this way the categories give them universal validity and, through them, make *judgments of experience* in general possible. Such an insight into the nature of the categories, which also restricted them to use in experience alone, never occurred either to their first author or to anyone after him. But without this insight, which depends entirely on their derivation or deduction, they are completely useless. They are a miserable list of names without explanation or a rule for their use. Had anything like this ever occurred to the ancients, there is no doubt that the whole study of pure rational cognition would have come down to us in a quite different form. Under the name of metaphysics, that study has spoiled so many fine minds over many centuries. Instead, it would have enlightened human understanding, rather than exhausting it in dark and futile brooding and making it unfit for true science, as actually happened. This system of categories now makes every treatment of any object of pure reason itself systematic in turn. It provides an unquestionable instruction or guide to how every metaphysical inquiry must proceed, and through which points of investigation, if it is to be complete. For it exhausts all the functions of the understanding under which every other concept must be brought. This is also how the table of principles arose. Only through the system of categories can we be certain that it is complete. Even in dividing concepts that are supposed to go beyond the understanding's use in natural science (Critique, p. 344, and also p. 415), it is always the same guide. It must always pass through the same fixed points, determined a priori in the human understanding. It therefore always forms a closed circle. This leaves no doubt that the object of a pure concept of the understanding or of reason, insofar as it is to be considered philosophically and according to a priori principles, can in this way be completely known. I could not even forgo using this guidance for one of the most abstract ontological divisions: the various distinctions between the concepts of something and nothing. Following it, I produced a systematic and necessary table (Critique, p. 292).[^325-1] Like every true system founded on a universal principle, this system also shows its inestimable usefulness by excluding all foreign concepts that might otherwise slip in among those pure concepts of the understanding, and by assigning each cognition its place. I had likewise used the categories as a guide to arrange in a table the concepts I called *concepts of reflection*. In ontology, these mingle with the pure concepts of the understanding without permission or legitimate claim. Yet the pure concepts of the understanding are concepts of connection, and through that, of the object itself. Concepts of reflection are only concepts for comparing concepts already given, and therefore have a quite different nature and use. My systematic division (Critique, p. 260) separates them out from this mixture. But the usefulness of that separate table of categories becomes still more obvious when we distinguish from it the table of transcendental concepts of reason, as we are about to do. Those concepts have a completely different nature and origin from the concepts of the understanding, so their table must also have a different form. Yet this absolutely necessary separation has never been made in any system of metaphysics. In such systems, therefore, ideas of reason and concepts of the understanding are mixed together without distinction, as if they were siblings in one family. Without a separate system of categories, this confusion could never have been avoided. [^323-1]: 1. Substance. 2. Quality. 3. Quantity. 4. Relation. 5. Action. 6. Being acted on. 7. When. 8. Where. 9. Position. 10. Having. [^323-2]: Opposition, priority, simultaneity, motion, having. [^325-1]: A table of categories, once presented, allows all sorts of interesting observations. 1) The third category arises from the first and second combined in one concept. 2) In the categories of magnitude and quality there is only a progression from unity to totality, or from something to nothing. For this purpose the categories of quality must stand in this order: reality, limitation, complete negation. These proceed without correlatives or opposites, whereas the categories of relation and modality do involve them. 3) Just as categorical judgments underlie all others in *logic*, the category of substance underlies all concepts of actual things. 4) Just as modality in a judgment is not a special predicate, modal concepts add no determination to things, and so on. All such reflections have great usefulness. If we also list all the *predicables*, which can be drawn fairly completely from any good ontology, for example *Baumgarten's*, and arrange them in classes under the categories, we must not neglect to add as complete an analysis of all these concepts as possible. This would produce a merely analytic part of metaphysics, containing no synthetic proposition at all. It could precede the second, synthetic part. Through its precision and completeness it would be not only useful but, because of its systematic character, would also possess a certain beauty. ### Third part: How is metaphysics in general possible? *The main transcendental question* **Third Part.** *How is metaphysics in general possible?* ### Section 40 § 40. Pure mathematics and pure natural science would not have needed the kind of justification we have so far provided for them *for their own security* and certainty. Mathematics rests on its own self-evidence. Natural science, though it comes from pure sources in the understanding, also rests on experience and its consistent confirmation. It cannot entirely reject or do without the testimony of experience. Despite all its certainty, as philosophy it can never match mathematics. These two sciences therefore needed this investigation not for themselves, but for another science: metaphysics. Besides concepts of nature, which always find their application in experience, metaphysics deals with pure concepts of reason. These can never be given in any possible experience whatsoever. It thus deals with concepts whose objective reality—that they are not mere fantasies—cannot be confirmed by experience, and with claims whose truth or falsity no experience can establish or expose. Moreover, this part of metaphysics is precisely its essential purpose. Everything else is only a means to that purpose. So this science needs such a justification *for its own sake*. The third question now before us concerns what is, so to speak, the core and distinctive feature of metaphysics. Reason occupies itself solely with itself. By brooding over its own concepts, it supposedly gains direct acquaintance with objects, without needing experience as an intermediary, or even being able to reach them through experience at all.[^327-1] Until this question is answered, reason can never satisfy itself. The use in experience to which reason restricts the pure understanding does not fulfill reason's entire purpose. Each particular experience is only a part of the whole range of its domain. But *the absolute whole of all possible experience* is not itself an experience. It is nevertheless a necessary problem for reason. Even to represent this whole, reason needs concepts quite different from those pure concepts of the understanding whose use is only immanent—that is, directed to experience insofar as it can be given. Concepts of reason, by contrast, aim at completeness: the collective unity of all possible experience. They thereby go beyond every given experience and become *transcendent*. Just as the understanding needed the categories for experience, reason contains the basis for ideas. By these I mean necessary concepts whose object nevertheless *cannot* be given in any experience. These ideas belong to the nature of reason just as much as the categories belong to the nature of the understanding. If ideas bring with them an illusion that can easily mislead us, this illusion is unavoidable. But we can certainly prevent it “from deceiving us.” Every illusion consists in taking a subjective ground of judgment to be objective. So the only safeguard against reason's errors will be its knowledge of itself in its transcendent use, which goes beyond experience. Reason falls into these errors when it misunderstands its purpose. It then applies to the object in itself, in a transcendent way, something that concerns only its own subject and the guidance of that subject in every immanent use. [^327-1]: We may say that a science actually exists, at least in everyone's idea of it, once the following is established: the nature of human reason presents everyone with the problems that lead to it, and so makes many attempts at it unavoidable at all times, even if those attempts are mistaken. If so, we must also say that metaphysics actually exists *subjectively*, and necessarily so. We are therefore right to ask how it is possible *objectively*. ### Section 41 § 41. *Ideas*, or pure concepts of reason, must be distinguished from categories, or pure concepts of the understanding. They are kinds of cognition with entirely different origins and uses. This distinction is crucial to laying the foundation of a science meant to contain the system of all these kinds of a priori cognition. Without it, metaphysics is simply impossible. At best, it is an unmethodical, amateur attempt to patch together a house of cards, without knowing the materials being used or their fitness for one purpose or another. Even if the Critique of Pure Reason had done nothing but bring this distinction to light for the first time, it would already have done more to clarify our concept and guide inquiry in metaphysics than all the fruitless efforts ever made to satisfy pure reason's transcendent demands. Those efforts never even suspected that they were in a field entirely different from that of the understanding. They therefore listed concepts of the understanding and concepts of reason together, as if they were all of one kind. ### Section 42 § 42. All pure cognition of the understanding has this feature: its concepts can be given in experience, and its principles can be confirmed by experience. Transcendent cognition of reason is different. Its *ideas* cannot be given in experience, and experience can never confirm or refute its propositions. Any error that may slip in can therefore be exposed only by pure reason itself. But this is very difficult. Through its ideas, this very reason naturally becomes dialectical. This unavoidable illusion cannot be kept within bounds by any objective, dogmatic investigation of things. Only a subjective investigation of reason itself, as the source of ideas, can do that. ### Section 43 § 43. Throughout the Critique, my main concern was not just to distinguish the kinds of cognition carefully. I also wanted to derive all the concepts belonging to each kind from their common source. Knowing where they came from would let me determine their use with certainty. It would also give me an invaluable advantage that no one had even suspected before: I could know, a priori and thus from principles, that my listing, classification, and specification of the concepts was complete. Without this, everything in metaphysics is just a disconnected collection. We never know whether what we have is enough, or whether something is still missing, or where. Admittedly, this advantage is available only in pure philosophy. But it also makes up the essence of pure philosophy. I had found the origin of the categories in the four logical functions of all judgments of the understanding. It was therefore quite natural to look for the origin of ideas in the three functions of rational inferences. For once such pure concepts of reason—transcendental ideas—are given, where else could we find them, unless we chose to regard them as innate? They must be found in the very same act of reason that constitutes the logical element of rational inferences when only its form is considered. When this act represents judgments of the understanding as determined a priori with respect to one form or another, it constitutes transcendental concepts of pure reason. Differences in the form of rational inferences require us to divide them into categorical, hypothetical, and disjunctive inferences. The concepts of reason based on these therefore contain, first, the idea of the complete subject (the substantial); second, the idea of the complete series of conditions; and third, the determination of all concepts within the idea of a complete totality of what is possible.[^330-1] The first idea was psychological, the second cosmological, and the third theological. All three give rise to dialectic, though each in its own way. This supplied the basis for dividing the entire dialectic of pure reason into its paralogism, its antinomy, and finally its ideal. This derivation makes us completely certain that all the claims of pure reason are presented here in full, and that not a single one can be missing. For it measures out the entire faculty of reason itself, from which they all originate. [^330-1]: In a disjunctive judgment, we regard *all possibility* in relation to a certain concept as divided up. The ontological principle of the thoroughgoing determination of a thing in general says that, of every pair of possible opposed predicates, one belongs to each thing. This is also the principle of all disjunctive judgments. It takes as its basis the totality of all possibility, within which the possibility of every thing in general is regarded as determinable. This helps explain the proposition above: the act of reason in disjunctive inferences has the same form as the act by which reason produces the idea of a totality of all reality. That idea contains what is positive in all mutually opposed predicates. ### Section 44 § 44. One further point deserves notice in this general discussion. Unlike the categories, the ideas of reason do not help us use the understanding in relation to experience. In that respect they are completely dispensable. They may even conflict with, and obstruct, the rules for rational knowledge of nature. Yet they are necessary for another purpose, which remains to be determined. Whether or not the soul is a simple substance can make no difference to our explanation of its appearances. No possible experience lets us present the concept of a simple being to the senses, and so make it intelligible in a concrete case. This concept is therefore completely empty when it comes to any hoped-for insight into the cause of appearances. It cannot provide a principle for explaining what either inner or outer experience supplies. Similarly, the cosmological ideas of the world's beginning or its eternity into the past (a parte ante) cannot help us explain any event within the world itself. Finally, a sound rule of natural philosophy requires us to avoid any explanation of nature's organization drawn from the will of a supreme being. That would no longer be natural philosophy. It would be an admission that we had reached the end of what we could do with it. These ideas thus have a purpose quite different from that of the categories and the principles built on them, which first made experience itself possible. Yet our laborious analysis of the understanding would also be entirely unnecessary if our aim were simply knowledge of nature as it can be given in experience. Reason does its work very well and securely in mathematics and natural science without any of this subtle justification. Our critique of the understanding therefore joins with the ideas of pure reason in an aim that goes beyond the understanding's use in experience. We said above, however, that using the understanding beyond experience is entirely impossible and has neither object nor meaning. Still, what belongs to the nature of reason must agree with what belongs to the nature of the understanding. Reason must contribute to the understanding's perfection. It cannot possibly throw it into confusion. The answer to this question is as follows. In its ideas, pure reason does not aim at special objects lying beyond the field of experience. It only demands completeness in the use of the understanding within the connected whole of experience. This completeness can only be a completeness of principles, not of intuitions and objects. Yet to represent it in a definite way, reason thinks of it as knowledge of an object, with that knowledge completely determined in terms of those rules. But this object is only an idea. Its purpose is to bring the understanding's knowledge as close as possible to the completeness that the idea indicates. ### Section 45 § 45. *Preliminary remark* *on the dialectic of pure reason.* We showed above, in sections 33 and 34, that the categories' freedom from any sensory determinations can tempt reason to extend their use entirely beyond all experience, to things in themselves. Yet the categories cannot find any intuition that could give them concrete meaning and sense. As mere logical functions, they do represent a thing in general. But by themselves they cannot provide a determinate concept of any thing. Objects of this exaggerated kind are what people call *noumena*, or beings of the pure understanding—better, beings of thought. Examples are a *substance* thought of *without persistence in time*, or a *cause* that does *not act in time*, and other such objects. We assign them predicates that serve only to make the law-governed character of experience possible. At the same time, we remove from them all the conditions of intuition under which alone experience is possible. This makes those concepts lose all meaning again. There is no danger, however, that the understanding will stray so recklessly beyond its limits into the field of mere beings of thought on its own, without being driven by laws foreign to it. But reason can never be fully satisfied with using the understanding's rules in experience, since that use is always still conditioned. When reason demands that this chain of conditions be completed, the understanding is driven out of its own domain. It must partly represent objects of experience in a series extending so far that no experience can encompass it. It must even, to complete the series, search for *noumena* entirely outside experience. Reason could attach the chain to these and thus finally give it complete support, independent of conditions of experience. These are the transcendental ideas. The true but hidden purpose of our reason's natural constitution may intend them not for concepts that go beyond all experience, but only for extending its use in experience without limit. Even so, through an unavoidable illusion, they lure the understanding into a *transcendent* use. Although deceptive, this use cannot be restrained merely by resolving to stay within the limits of experience. It takes scientific instruction, and considerable effort, to keep it within bounds. ### Section 46 § 46. I. *Psychological ideas.* (Kritik S. 341 u. f.) People noticed long ago that in every substance the actual subject remains unknown to us: what is left when all accidents, as predicates, have been removed. In other words, the *substantial itself* is unknown. They have often complained about these limits to our insight. But we should note carefully what the human understanding can be blamed for here. It is not that it fails to know the substantial in things—that is, to determine it by itself. Rather, it wants determinate knowledge of what is a mere idea, as though it were a given object. Pure reason requires us to seek, for every predicate of a thing, the subject belonging to it. For this subject, which is necessarily only a predicate in turn, we must seek its subject, and continue to infinity, or as far as we can reach. But it follows that we should regard nothing we can reach as a final subject. The substantial itself can never be thought by our understanding, however deeply it penetrates, even if all of nature were revealed to it. For the distinctive nature of our understanding is to think everything discursively: through concepts, and thus entirely through predicates. The absolute subject must therefore always be missing. So all the real properties through which we know bodies are merely accidents, even impenetrability. We must always represent that as the effect of a force whose subject is missing for us. Now it seems that in our consciousness of ourselves, the thinking subject, we have this substantial—and have it in an immediate intuition. For all predicates of inner sense refer to the I as their subject. And this I cannot in turn be thought as a predicate of any other subject. Here, then, completeness in relating the given concepts as predicates to a subject seems not merely to be an idea. The object, the *absolute subject* itself, seems to be given in experience. But this expectation is disappointed. The I is not a concept at all.[^334-1] It is only a designation for the object of inner sense, insofar as no predicate gives us any further knowledge of it. So it cannot itself be a predicate of another thing. But neither can it be a determinate concept of an absolute subject. As in all other cases, it is only the relation of inner appearances to their unknown subject. Still, through a quite natural misunderstanding, this idea gives rise to a very plausible argument. The idea is well suited, as a regulative principle, to destroy completely all materialistic explanations of our soul's inner appearances.[^334-1] But the argument uses this supposed knowledge of the substantial in our thinking being to infer its nature, insofar as knowledge of that nature lies entirely outside the whole range of experience. [^334-1]: If the representation of apperception, the I, were a concept through which anything was thought, it could also be used as a predicate of other things, or would contain such predicates. As it is, it is nothing more than a feeling of existence, without the slightest concept. It is only a representation of that to which all thinking stands in relation (relatione accidentis). ### Section 47 § 47. We may call this thinking self, the soul, a substance, since it is the final subject of thinking and cannot itself be represented as a predicate of another thing. Yet this concept remains completely empty and has no consequences unless we can prove that the self persists. Persistence is what makes the concept of substances fruitful in experience. But persistence can never be proved from the concept of a substance as a thing in itself. It can be proved only for the purposes of experience. This was sufficiently demonstrated in the first analogy of experience (Kritik S. 182). Anyone unwilling to accept that proof need only try it for themselves. Starting with the concept of a subject that does not itself exist as the predicate of another thing, can they prove that its existence is absolutely persistent? Can they prove that it can neither arise nor pass away, either on its own or through any natural cause? Such synthetic a priori propositions can never be proved in themselves. They can be proved only in relation to things as objects of possible experience. ### Section 48 § 48. Suppose, then, that we want to infer the soul's persistence from the concept of it as a substance. This can hold for the soul only for the purposes of possible experience. It cannot hold for the soul as a thing in itself, beyond all possible experience. Now life is the subjective condition of all our possible experience. We can therefore infer only that the soul persists during life. A person's death ends all experience as far as the soul as an object of experience is concerned, unless the opposite is demonstrated—and that is precisely what is in question. So we can demonstrate the soul's persistence only during a person's life, something no one will trouble us to prove. We cannot demonstrate it after death, which is what really matters to us. The general reason is this: insofar as the concept of substance is to be regarded as necessarily connected with the concept of persistence, this connection can rest only on a principle of possible experience, and so can hold only for the purposes of experience.[^335-1] [^335-1]: It is really very remarkable how carelessly metaphysicians have always slipped past the principle that substances persist, without ever attempting to prove it. No doubt this was because, as soon as they started with the concept of substance, they found themselves wholly without grounds of proof. Common understanding clearly recognized that without this assumption perceptions could not be united in one experience. It made up for the lack by a postulate. For it could never draw this principle from experience itself. For one thing, experience cannot follow materials, or substances, through all their changes and dissolutions far enough to find that the matter always remains undiminished. For another, the principle contains *necessity*, which always marks an a priori principle. Metaphysicians now confidently applied this principle to the concept of the soul as a *substance*. They inferred that it must continue to exist after a person's death. They felt especially secure against its destruction through dissolution because of this substance's simplicity, which they inferred from the indivisibility of consciousness. If they had found the true source of this principle—something requiring far deeper investigations than they ever cared to begin—they would have seen the following. The law that substances persist holds only for the purposes of experience. It therefore applies only to things insofar as they are to be known in experience and connected with other things. It can never apply to them without regard to all possible experience, and thus cannot apply to the soul after death. ### Section 49 § 49. We likewise cannot prove, as a connection between things in themselves, that something real outside us not only corresponds to our outer perceptions but must correspond to them. We can, however, prove it for the purposes of experience. This means that we can indeed prove that something exists outside us in an empirical way, as an appearance in space. For we are not concerned with objects other than those belonging to a possible experience. Such other objects cannot be given to us in any experience, and for that very reason are nothing for us. What is intuited in space is empirically outside me. Space, together with all the appearances it contains, belongs to our representations. Their connection according to laws of experience proves their objective truth just as the connection of appearances of inner sense proves the reality of my soul as an object of inner sense. Through outer experience, then, I am just as conscious of the reality of bodies as outer appearances in space as I am, through inner experience, of my soul's existence in time. I can know this soul only as an object of inner sense, through appearances that make up an inner state. Its being in itself, which underlies these appearances, remains unknown to me. *Cartesian* idealism thus distinguishes outer experience from dreams, and the law-governed character of experience, as a criterion of its truth, from the disorder and deceptive illusion of dreams. In both cases it assumes space and time as conditions of the existence of objects. It asks only whether the objects of the outer senses that we place in space when awake are really to be found there, just as the object of inner sense, the soul, really is in time. In other words, does experience carry reliable criteria for distinguishing it from imagination? This doubt is easy to remove. We do so constantly in ordinary life by examining the connection of appearances in both according to general laws of experience. If the representation of outer things agrees throughout with these laws, we cannot doubt that such representations make up genuine experience. Material idealism is therefore very easy to remove when appearances are considered, as appearances, only in terms of their connection in experience. It is just as certain in experience that bodies exist outside us, in space, as that I myself exist according to the representation of inner sense, in time. For the concept *outside us* means only existence in space. But the I in the proposition *I am* means not just the object of inner intuition, in time, but the subject of consciousness. Similarly, a body means not just outer intuition, in space, but also the thing *in itself* underlying this appearance. So we can unhesitatingly answer no when asked whether bodies, as appearances of outer sense, exist as bodies *outside my thoughts* in nature. But the same holds for the question whether I myself, *as an appearance of inner sense*—the soul in empirical psychology—exist in time outside my power of representation. That question must also be answered no. In this way, everything is settled and certain once it is given its true meaning. Formal idealism, which I have also called transcendental idealism, actually removes material or *Cartesian* idealism. For if space is nothing but a form of my sensibility, then as a representation in me it is just as real as I myself am. All that remains at issue is the empirical truth of the appearances in it. But if this is not so, and space and its appearances are something existing outside us, then no criteria of experience can ever prove the reality of these objects outside us, apart from our perception. ### Section 50 § 50. II. *Cosmological ideas.* (Krit. S. 405 u. f.) This product of pure reason in its transcendent use is reason's most remarkable phenomenon. Of all its phenomena, it also does the most to wake philosophy from its dogmatic sleep and move it to the difficult task of critiquing reason itself. I call this idea cosmological because it always takes its object only from the world of sense. It needs no idea other than one whose object is an object of the senses. To that extent it stays on familiar ground and is not transcendent, and thus is not yet an idea. By contrast, to think of the soul as a simple substance already means thinking of an object—the simple—of a kind that cannot be presented to the senses at all. Nevertheless, the cosmological idea extends the connection between what is conditioned and its condition, whether mathematical or dynamical, so far that experience can never match it. In this respect, then, it is always an idea whose object can never be adequately given in any experience. ### Section 51 § 51. Here, first of all, the usefulness of a system of categories becomes so clear and unmistakable that this alone would sufficiently demonstrate the categories' indispensability in the system of pure reason, even if there were no further proofs. There are only four such transcendent ideas, as many as there are classes of categories. In each class, they aim only at the absolute completeness of the series of conditions for something given as conditioned. Corresponding to these cosmological ideas, there are also only four kinds of dialectical claims of pure reason. Their dialectical character itself shows that each faces a contradictory claim based on equally plausible principles of pure reason. No metaphysical skill in making the subtlest distinctions can prevent this conflict. Instead, it forces philosophers to return to the very first sources of pure reason itself. This antinomy is not invented at will. It is grounded in the nature of human reason, and is therefore unavoidable and never-ending. It contains the following four propositions and their opposites. | | Proposition | Opposite proposition | |---|---|---| | 1. | The world has a *beginning* (a limit) in time and space | The world is *infinite* in time and space | | 2. | Everything in the world consists of what is *simple* | Nothing is simple; everything is *composite* | | 3. | There are causes in the world that act through *freedom* | There is no freedom; everything is *nature* | | 4. | In the series of causes in the world there is some *necessary* being | Nothing in it is necessary; in this series *everything is contingent*. | ### Section 52 § 52. Here we have the strangest phenomenon of human reason. No example of it can be found in any other use of reason. Suppose we think of the appearances of the world of sense as things in themselves, as people usually do. Suppose we also take the principles connecting them to hold universally for things in themselves, and not merely for experience. This is just as usual, and without our critique it is even unavoidable. An unexpected conflict then emerges. It can never be settled in the usual dogmatic way, because both the proposition and its opposite can be demonstrated by equally evident, clear, and irresistible proofs. I guarantee the correctness of all these proofs. Reason thus finds itself divided against itself. The skeptic celebrates this state of affairs, but it must make the critical philosopher thoughtful and uneasy. ### Section 52b § 52b. People can dabble in metaphysics in all sorts of ways without much fear of being caught saying something false. They need only avoid contradicting themselves, which is quite possible in synthetic propositions even when these are entirely made up. Whenever the concepts we connect are mere ideas that cannot be given in experience in their full content, experience can never refute us. For how could experience settle whether the world has existed from eternity or had a beginning? How could it settle whether matter is infinitely divisible or consists of simple parts? Concepts like these cannot be given in any experience, even the greatest possible experience. This test therefore cannot reveal the falsity of either the affirmative or the negative proposition. There is only one possible case in which reason would unwillingly reveal its hidden dialectic, which it falsely presents as dogmatic knowledge. It would have to base a claim on a universally accepted principle, and then infer the exact opposite from another equally accepted principle, using a completely correct form of inference. This case actually occurs here, with four natural ideas of reason. Four claims on one side, and as many counterclaims on the other, arise by valid inference from universally accepted principles. They thereby reveal the dialectical illusion of pure reason in its use of these principles—an illusion that would otherwise have to remain hidden forever. Here, then, is a decisive test. It must reveal to us an error hidden in reason's assumptions.[^341-1] Two contradictory propositions cannot both be false unless the concept underlying both is itself contradictory. For example, these two propositions are both false: a square circle is round; and a square circle is not round. In the first case, it is false that the circle in question is round, because it is square. But it is also false that it is not round—that is, that it is angular—because it is a circle. This is precisely the logical mark of an impossible concept: assuming it would make two contradictory propositions false at the same time. Since no third possibility can be thought between them, nothing at all is thought through that concept. [^341-1]: I therefore want critical readers to give their main attention to this antinomy. Nature itself seems to have set it up to make reason pause in its bold pretensions and force it to examine itself. I undertake to defend every proof I have given for both the thesis and the antithesis, and thereby to demonstrate with certainty that reason's antinomy is unavoidable. If this strange phenomenon leads readers back to examine the assumption underlying it, they will feel compelled to join me in investigating more deeply the first foundation of all cognition of pure reason. ### Section 52c § 52c. A contradictory concept of this kind underlies the first two antinomies. I call these mathematical because they concern the addition or division of things of the same kind. This explains how both the thesis and the antithesis in each are false. When I speak of objects in time and space, I am not speaking of things in themselves, since I know nothing of them. I speak only of things in appearance: of experience as the particular way of knowing objects that alone is available to human beings. When I think of something in space or time, I must not say that it is in space and time in itself, even without this thought of mine. That would contradict myself. Space and time, together with the appearances in them, are not something existing in themselves and outside my representations. They are themselves only ways of representing. And it is plainly contradictory to say that a mere way of representing also exists outside our representation. Objects of the senses therefore exist only in experience. To give them an existence of their own, subsisting by itself even without or before experience, is to represent experience as actual without or before experience. If I now ask about the world's magnitude in space and time, it is just as impossible, by all my concepts, to say that it is infinite as to say that it is finite. Neither can be contained in experience. Experience of an *infinite* space or an infinitely elapsed time is impossible. So is experience of the world's *limitation* by an empty space or a preceding empty time. These are only ideas. The world's magnitude, determined in either way, would therefore have to belong to the world in itself, apart from all experience. But this contradicts the concept of a world of sense. Such a world is only a totality of appearances, whose existence and connection occur only in representation, namely experience. It is not a thing in itself, but nothing more than a way of representing. It follows that the concept of a world of sense existing on its own is self-contradictory. Any attempted solution to the problem of its magnitude will therefore be false, whether it takes an affirmative or a negative form. The same holds for the second antinomy, which concerns the division of appearances. Appearances are mere representations. Their parts exist only in the representation of them, and thus in their division: in a possible experience in which they are given. Division therefore extends only as far as that experience extends. Suppose we assume that an appearance, such as that of a body, contains in itself, before all experience, every part that any possible experience could ever reach. We are then giving a mere appearance, which can exist only in experience, an existence of its own prior to experience. In other words, we are saying that mere representations exist before they are found in the power of representation. This is self-contradictory. So is every solution to this misunderstood problem, whether it claims that bodies in themselves consist of infinitely many parts or of a finite number of simple parts. ### Section 53 § 53. In the first, mathematical class of antinomy, the false assumption consisted in representing something contradictory—appearance as a thing in itself—as compatible in one concept. In the second, dynamical class, the false assumption consists in representing compatible things as contradictory. In the first case, both opposing claims were false. Here, by contrast, the claims that are opposed only through a misunderstanding can both be true. A mathematical connection necessarily assumes that what is connected is of the same kind, under the concept of magnitude. A dynamical connection does not require this at all. When the magnitude of something extended is at issue, all its parts must be of the same kind as one another and as the whole. In the connection of cause and effect, however, the two may be of the same kind, but need not be. The concept of causality, through which something posits something entirely different from itself, at least does not require it. If objects of the world of sense were taken for things in themselves, and the natural laws mentioned above for laws of things in themselves, the contradiction would be unavoidable. Likewise, if the subject of freedom were represented as a mere appearance, like the other objects, the contradiction could not be avoided. For the very same thing would be affirmed and denied at once of the same object in the same sense. But if natural necessity is referred only to appearances, and freedom only to things in themselves, no contradiction arises when we accept or allow both kinds of causality. This holds however difficult or impossible it may be to make causality of the latter kind comprehensible. In appearance, every effect is an event, or something that occurs in time. According to the universal law of nature, a determination of its cause's causality—a state of that cause—must precede it. The effect follows that state according to a constant law. But this determination of the cause to causality must also be something that occurs or *happens*. The cause must have *begun* to *act*. Otherwise we could not think of a succession in time between it and the effect. The effect would always have existed, just like the cause's causality. Among appearances, then, the *determination* of a cause *to act* must also have arisen. Like its effect, it must therefore be an event that in turn must have its cause, and so on. Natural necessity must consequently be the condition according to which efficient causes are determined. If, on the other hand, freedom is to be a property of certain causes of appearances, then in relation to those appearances as events, it must be a capacity to begin them *spontaneously* (sponte). That is, the cause's own causality would not have to begin, and so would need no other ground determining its beginning. But in that case the *cause*, in its causality, would not be subject to temporal determinations of its state. It would not be an *appearance* at all. We would have to take it as a thing in itself, and only its *effects* as *appearances*.[^344-1] If we can think without contradiction of such an influence of beings of the understanding on appearances, then natural necessity will indeed belong to every connection of cause and effect in the world of sense. Yet freedom can be allowed to the cause that is not itself an appearance, though it underlies appearance. Nature and freedom can thus be attributed without contradiction to the very same thing, but in different respects: once as appearance, and once as a thing in itself. We have a faculty within us that is connected not just with its subjectively determining grounds, the natural causes of its actions. In that respect it is the faculty of a being that itself belongs among appearances. But it is also referred to objective grounds that are mere ideas, insofar as they can determine this faculty. This connection is expressed by *ought*. The faculty is called *reason*. Insofar as we consider a being, a human being, solely in terms of this objectively determinable reason, we cannot regard it as a being of sense. Rather, this property belongs to a thing in itself. We cannot at all comprehend how such a property is possible: how the *ought*, something that has never yet happened, can determine its activity and cause actions whose effect is an appearance in the world of sense. Still, reason's causality in relation to its effects in the world of sense would be freedom, insofar as *objective grounds*, which are themselves ideas, are regarded as determining that causality. For its action would then not depend on subjective conditions, and thus not on conditions of time or on the natural law that serves to determine them. Grounds of reason give actions their rule universally, from principles, without the influence of circumstances of time or place. What I introduce here is only an example to make the matter intelligible. It is not a necessary part of our question. That question must be decided from mere concepts, independently of properties we find in the actual world. I can now say without contradiction: all actions of rational beings, insofar as they are appearances found in any experience, are subject to natural necessity. But these very same actions are free when considered only in relation to the rational subject and its capacity to act from reason alone. For what does natural necessity require? Nothing more than that every event in the world of sense can be determined according to constant laws, and thus has a relation to a cause in appearance. The underlying thing in itself and its causality remain unknown. But I say that *the law of nature remains*, whether the rational being causes effects in the world of sense from reason, and thus through freedom, or does not determine them from rational grounds. In the first case, the action follows maxims whose effect in appearance will always conform to constant laws. In the second case, where the action does not follow principles of reason, it is subject to the empirical laws of sensibility. In both cases the effects are connected according to constant laws. We require nothing more for natural necessity; indeed, we know nothing more about it. But in the first case reason is the cause of these natural laws, and is therefore free. In the second case, the effects follow the mere natural laws of sensibility because reason exerts no influence on them. Yet reason is not itself thereby determined by sensibility, which is impossible, and so is free in this case too. Freedom therefore does not obstruct the natural law of appearances. Nor does that law impair the freedom of the practical use of reason, which is connected with things in themselves as determining grounds. This saves practical freedom: the freedom in which reason has causality according to objectively determining grounds. It does so without the slightest harm to natural necessity in relation to those very same effects as appearances. This can also help clarify what we had to say about transcendental freedom and its compatibility with natural necessity in the same subject, though not considered in one and the same respect. For with regard to transcendental freedom, whenever a being begins an action from objective causes, this is always a first beginning in relation to those determining grounds. Yet in the series of appearances the same action is only a *subordinate beginning*. A state of the cause must precede it and determine it, and that state is itself likewise determined by a still earlier one. Thus, without contradicting natural laws, we can think of a capacity to begin a series of states spontaneously in rational beings, or in beings generally insofar as their causality is determined in them as things in themselves. For the relation of an action to objective grounds of reason is not a temporal relation. Here, what determines causality does not precede the action in time. Such determining grounds do not represent a relation of objects to the senses, and so do not represent causes in appearance. They represent determining causes as things in themselves, which are not subject to conditions of time. So the action can be regarded as a first beginning in relation to reason's causality, yet at the same time as a merely subordinate beginning in relation to the series of appearances. Without contradiction it can be considered free in the first respect, and subject to natural necessity in the second, where it is mere appearance. The *fourth* antinomy is resolved in a similar way to reason's conflict with itself in the third. We need only distinguish the *cause in appearance* from the *cause of appearances*, insofar as the latter can be thought as a *thing in itself*. Both propositions can then stand together. Nowhere in the world of sense is there a cause, under similar laws of causality, whose existence is absolutely necessary. Yet, on the other hand, this world is connected with a necessary being as its cause, though of a different kind and under a different law. These two propositions seem incompatible only because of a misunderstanding: what holds merely for appearances is extended to things in themselves, and the two are generally mixed together in a single concept. [^344-1]: The idea of freedom applies only to the relation of the *intellectual*, as cause, to *appearance*, as effect. We therefore cannot attribute freedom to matter in its ceaseless activity of filling its space, even though this activity proceeds from an inner principle. Nor can we find a suitable concept of freedom for pure beings of the understanding—for example, God—insofar as their action is immanent. For although his action is independent of externally determining causes, it is still determined in his eternal reason, and thus in the divine *nature*. Only when *something is to begin* through an action, and its effect is therefore to be found in the series of time and hence in the world of sense—for example, the world's beginning—does the question arise: must the cause's causality itself also begin, or can the cause initiate an effect without its causality itself beginning? In the first case, the concept of this causality is a concept of natural necessity; in the second, of freedom. Readers will see from this that, when I defined freedom as the capacity to begin an event spontaneously, I captured precisely the concept that poses the problem for metaphysics. ### Section 54 § 54. We have now set out and resolved the entire antinomy in which reason becomes entangled when it applies its principles to the world of sense. Even the first task, merely setting it out, would already be a considerable contribution to our knowledge of human reason. This holds even if the resolution of the conflict has not yet fully satisfied the reader. Here the reader must struggle against a natural illusion that has only recently been presented as such, after always having been taken for truth. Still, one consequence is unavoidable. There is absolutely no way out of reason's conflict with itself as long as we take objects of the world of sense for things in themselves, rather than for what they actually are: mere appearances. The reader will therefore be forced to return to the justification of all our a priori cognition and to reexamine the justification I have given, in order to reach a decision. I ask for nothing more at present. For if, in doing this, readers first think deeply enough into the nature of pure reason, they will already be familiar with the concepts through which alone the resolution of reason's conflict is possible. Without this familiarity, I cannot expect full agreement even from the most attentive reader. ### Section 55 § 55. III. *The theological idea.* (Kritik S. 571 u. f.) The third transcendental idea is the ideal of pure reason. It supplies material for reason's most important use. But if pursued merely speculatively, this use goes beyond experience, becoming transcendent and therefore dialectical. With the psychological and cosmological ideas, reason starts from experience. It is then tempted to ascend through grounds in search, if possible, of absolute completeness in their series. Here, by contrast, reason breaks off entirely. It starts from mere concepts of what would make up the absolute completeness of a thing in general. Through the idea of a supremely perfect original being, it then descends to determine the possibility, and thus also the actuality, of all other things. Here, then, we merely assume a being that is not in the series of experience, but is still thought for the purposes of experience, to make its connection, order, and unity comprehensible. This assumption, the idea, is easier to distinguish from a concept of the understanding than in the previous cases. The dialectical illusion could therefore be brought to light easily here. It arises when we take the subjective conditions of our thinking for objective conditions of things themselves, and a hypothesis necessary to satisfy our reason for a dogma. So I have nothing further to say about the pretensions of transcendental theology. What the Critique says about them is understandable, evident, and decisive. ### Section 56 § 56. *General remark* *on the transcendental ideas.* The objects given to us through experience are incomprehensible to us in many respects. Natural laws lead us to many questions that cannot be answered if pursued to a certain height, even when we always follow those laws. One example is why material things attract one another. But suppose we leave nature entirely, or go beyond all possible experience in tracing its connections, and thus immerse ourselves in mere ideas. Then we cannot say that the object is incomprehensible to us, or that the nature of things presents us with insoluble problems. For we are no longer dealing with nature, or with given objects at all. We are dealing only with concepts originating solely in our reason, and with mere beings of thought. Every problem arising from the concept of these beings must be capable of solution. For reason certainly can and must give a complete account of its own procedure.[^349-1] The psychological, cosmological, and theological ideas are all pure concepts of reason that cannot be given in any experience. So the questions reason puts to us about them are not imposed by objects. They are imposed by mere maxims of reason for its own satisfaction, and all must be capable of an adequate answer. This answer is given by showing that they are principles for bringing our use of the understanding to thorough agreement, completeness, and synthetic unity. To that extent they hold only for experience, but for experience as a *whole*. Although an absolute whole of experience is impossible, the idea of a whole of cognition according to principles in general is still essential. Only it can give cognition a special kind of unity: the unity of a system. Without this, our cognition is mere patchwork and cannot be used for the highest purpose, which is always only the system of all purposes. Here I mean not merely the practical purpose, but also the highest purpose of reason's speculative use. The transcendental ideas thus express reason's distinctive function: to be a principle of systematic unity in the use of the understanding. But we may take this unity in our way of knowing as though it belonged to the object known. We may regard it as *constitutive*, when it is really only *regulative*. We may persuade ourselves that these ideas let us extend our knowledge far beyond all possible experience, and thus in a transcendent way. In fact, this unity serves only to bring experience itself as close as possible to completeness. That means not restricting its progress by anything that cannot belong to experience. The mistake is therefore simply a misunderstanding of the actual purpose of our reason and its principles. It is a dialectic that partly confuses reason's use in experience and partly divides reason against itself. *Conclusion: determining the boundaries of pure reason.* [^349-1]: In his Aphorisms, § 728, 729, Mr. *Platner* therefore makes this perceptive remark: “If reason is a criterion, no concept can be possible that is incomprehensible to human reason.—Only in what is actual does incomprehensibility occur. Here incomprehensibility arises from the inadequacy of acquired ideas.”—So it only sounds paradoxical, and is otherwise not surprising, to say that much in nature is incomprehensible to us—for example, the power of reproduction—but that if we rise still higher, even beyond nature, everything becomes comprehensible to us again. For we then leave entirely behind the *objects* that can be given to us and occupy ourselves only with ideas. In these, we can very well comprehend the law that reason uses them to prescribe to the understanding for its use in experience, because that law is reason's own product. ### Section 57 § 57. After the very clear proofs given above, it would be absurd to hope to know more about any object than belongs to possible experience of it. It would be equally absurd to claim even the slightest knowledge determining the character of any thing as it is in itself, when we assume it is not an object of possible experience. For how would we make this determination? Time, space, and all concepts of the understanding have, and can have, no use other than to make experience possible. Still less can concepts drawn from empirical intuition or *perception* in the world of sense have any other use. If we remove this condition even from the pure concepts of the understanding, they determine no object at all and have no meaning whatsoever. On the other hand, it would be still more absurd to allow no things in themselves at all. It would be absurd to present our experience as the only possible way of knowing things, our intuition in space and time as the only possible intuition, and our discursive understanding as the model for every possible understanding. That would make principles of the possibility of experience into universal conditions of things in themselves. Our principles restrict reason's use to possible experience. Yet these very principles could become *transcendent* and present the limitations of our reason as limitations on the possibility of things themselves. *Hume's* Dialogues can serve as an example. This would happen unless a careful critique guarded the boundaries of our reason even in its empirical use and set a limit to its pretensions. Skepticism originally arose from metaphysics and its unregulated dialectic. At first, perhaps merely to favor reason's use in experience, it declared everything beyond that use empty and deceptive. But people gradually realized that the very same a priori principles used in experience also led unnoticed beyond the reach of experience, and apparently with just as much right. They then began to doubt even the principles of experience. There is no real danger here: sound understanding will surely always maintain its rights in this matter. Still, a special confusion arose in science, which cannot determine how far reason is to be trusted, or why only that far and no farther. Only a formal determination of the boundaries of our use of reason, drawn from principles, can remedy this confusion and prevent every future relapse. It is true that, beyond all possible experience, we can give no determinate concept of what things may be in themselves. Yet we are not free simply to refrain from asking about them. Experience never fully satisfies reason. In answering its questions, experience always sends it farther back and leaves it dissatisfied with the complete explanation of them. Anyone can see this clearly enough from the dialectic of pure reason, which for that very reason has a sound subjective basis. Who could bear reaching, in our inquiry into the nature of our soul, a clear consciousness of the subject and at the same time the conviction that its appearances cannot be explained *materialistically*, without asking what the soul actually is? And if no concept from experience is adequate, who could refrain from at least assuming a concept of reason—a simple immaterial being—for this purpose alone, even though we cannot demonstrate its objective reality? Who can be satisfied with mere experiential knowledge in all the cosmological questions about the world's duration and size, or about freedom and natural necessity? Whatever approach we take, each answer based on principles of experience always gives birth to a new question that equally demands an answer. This clearly shows that no physical explanation is adequate to satisfy reason. Finally, everything we can think and assume according to principles of experience is thoroughly contingent and dependent. Who does not see that it is impossible to stop there? Who does not feel compelled, despite every prohibition against losing oneself in transcendent ideas, to go beyond all concepts that experience can justify and seek rest and satisfaction in the concept of a being? We cannot gain insight into the possibility of the idea of this being in itself, though neither can we refute it, since it concerns a mere being of the understanding. Yet without it reason would have to remain forever dissatisfied. Boundaries, in the case of extended beings, always presuppose a space found outside a certain definite location and enclosing it. Limitations require no such thing. They are mere negations affecting a magnitude insofar as it lacks absolute completeness. Our reason, however, sees, as it were, a space around itself for the knowledge of things in themselves, although it can never have determinate concepts of them and is restricted to appearances alone. As long as reason's cognition remains of the same kind, no definite boundaries can be thought for it. In mathematics and natural science, human reason recognizes limitations but no boundaries. It recognizes that something lies outside it that it can never reach, but not that its own internal progress will ever be completed anywhere. The expansion of mathematical insight and the possibility of ever new discoveries go on to infinity. So do the discovery of new properties of nature, new forces and laws through continued experience, and reason's unification of them. Yet limitations here are unmistakable. Mathematics concerns only *appearances*. Whatever cannot be an object of sensory intuition, such as the concepts of metaphysics and morality, lies entirely outside its sphere, and it can never lead us there. But neither does it need them at all. There is thus no continuous progress or approach toward these sciences, nor even, so to speak, a point or line of contact. Natural science will never reveal to us the interior of things: what is not appearance but can nevertheless serve as the highest ground for explaining appearances. But it does not need this for its physical explanations either. Indeed, even if something of this kind were offered to it from elsewhere—for example, the influence of immaterial beings—it should reject it and not introduce it into the progress of its explanations. It should always base them only on what can belong to experience as an object of the senses and can be connected with our actual perceptions according to laws of experience. Metaphysics, however, leads us to boundaries in pure reason's dialectical attempts. These are not undertaken arbitrarily or recklessly; reason's own nature drives us to them. The transcendental ideas are unavoidable, yet can never be realized in objects. Precisely for that reason, they serve not only to show us the actual boundaries of pure reason's use, but also how to determine those boundaries. This is also the purpose and benefit of this natural disposition of our reason, which has given birth to metaphysics as its favorite child. Like every other birth in the world, its production must be attributed not to mere chance, but to an original seed wisely organized for great purposes. Perhaps more than any other science, metaphysics has been implanted in us in its basic features by nature itself. It cannot be regarded as the product of an arbitrary choice, or as an accidental extension in the progress of experience, from which it separates itself entirely. With all its concepts and laws of the understanding, which are sufficient for empirical use and thus within the world of sense, reason still finds no satisfaction in itself. Questions recurring to infinity take away all hope of a complete answer. The transcendental ideas, which aim at this completeness, are problems of reason of this kind. Reason now sees clearly that the world of sense cannot contain this completeness. Neither, therefore, can all the concepts serving solely to understand that world: space, time, and everything we have listed under the name of pure concepts of the understanding. The world of sense is nothing but a chain of appearances connected according to universal laws. It therefore has no independent subsistence. It is not really the thing in itself, and so necessarily refers to what contains the ground of these appearances: beings that can be known not merely as appearance, but as things in themselves. Only in knowing these can reason hope at last to satisfy its desire for completeness in progressing from what is conditioned to its conditions. Above, in sections 33 and 34, we indicated reason's limitations regarding all knowledge of mere beings of thought. Now the transcendental ideas nevertheless make progress toward them necessary. They have led us only, so to speak, to the point of contact between filled space—experience—and empty space—the noumena, of which we can know nothing. We can therefore also determine the boundaries of pure reason. For every boundary also contains something positive. A surface, for example, is the boundary of a solid space, yet is itself a space. A line is a space that bounds a surface. A point bounds a line, yet is still a location in space. Limitations, by contrast, contain mere negations. The limitations indicated in the sections cited are no longer enough, now that we have found that something lies beyond them, even though we will never know what it is in itself. For the question now is: how does our reason function in this connection between what we know and what we do not and never will know? There is an actual connection here between the known and something completely unknown, which will always remain so. Even if the unknown should not become in the least better known through this—as in fact we cannot hope it will—the concept of this connection must still be capable of being determined and made clear. We are therefore to think of an immaterial being, a world of the understanding, and a highest of all beings—all noumena. For only in these, as things in themselves, does reason find the completeness and satisfaction it can never hope for when deriving appearances from grounds of the same kind. Appearances really do refer to something different from them, and thus of an entirely different kind. For appearances always presuppose a thing in itself and therefore point to it, whether or not we know it more closely. Now we can never know these beings of the understanding in terms of what they may be in themselves: we can never know them determinately. Yet we must still assume them in relation to the world of sense and connect them with it through reason. At least, then, we can think this connection through concepts expressing their relation to the world of sense. If we think of a being of the understanding through nothing but pure concepts of the understanding, we really think nothing determinate, and our concept has no meaning. If we think of it through properties borrowed from the world of sense, it is no longer a being of the understanding. It is thought as one of the phenomena and belongs to the world of sense. Let us take the concept of the supreme being as an example. The *deistic* concept is a wholly pure concept of reason. But it represents only a thing containing all reality, without being able to determine any single reality within it. To do that, we would have to borrow an example from the world of sense. I would then always be dealing only with an object of the senses, not with something entirely different in kind that cannot be an object of the senses at all. For example, I would attribute understanding to it. But I have no concept of an understanding except one like my own: one to which intuitions must be given through the senses, and which works to bring them under rules of the unity of consciousness. The elements of my concept would then always lie in appearance. Yet it was precisely the inadequacy of appearances that forced me to go beyond them to the concept of a being that does not depend on appearances at all, and is not entangled with them as conditions of its determination. If, however, I separate understanding from sensibility to obtain a pure understanding, nothing remains but the mere form of thinking without intuition. Through that alone I can know nothing determinate, and thus no object. For that purpose I would have to think of a different understanding that intuited objects. But I have not the slightest concept of such an understanding, since the human understanding is discursive and can know only through general concepts. The same thing happens if I attribute a will to the supreme being. I have this concept only by drawing it from my inner experience. But underlying that experience is my satisfaction's dependence on objects whose existence we need, and therefore sensibility. This entirely contradicts the pure concept of the supreme being. *Hume's* objections to deism are weak. They never strike anything beyond its proofs, and never the deistic claim itself. But they are very strong against theism, which is supposed to arise through a closer determination of our concept of the supreme being, a concept that in deism is merely transcendent. Depending on how this concept is framed, his objections are irrefutable in certain cases—in fact, in all the usual cases. Hume always insists on this: the mere concept of an original being to which we attribute only ontological predicates—eternity, omnipresence, omnipotence—lets us think nothing determinate at all. Properties must be added that can supply a concrete concept. It is not enough to say that it is a cause. We must say what its causality is like, perhaps through understanding and will. At this point his attacks on the matter itself, namely theism, begin. Previously he had only attacked the grounds of proof for deism, which involves no particular danger. All his dangerous arguments concern anthropomorphism. He holds that it is inseparable from theism and makes theism self-contradictory. But if anthropomorphism were removed, theism would fall with it. Nothing would remain but a deism that we can do nothing with, that can be of no use to us and provide no foundation for religion and morals. If anthropomorphism really were unavoidable, then whatever proofs there might be of the existence of a supreme being, even if all were accepted, we could never determine the concept of this being without becoming entangled in contradictions. Let us combine the prohibition against making any transcendent judgments of pure reason with the apparently conflicting command to go out to concepts lying beyond the field of immanent, empirical use. We then realize that both can coexist, but only right on the *boundary* of all permitted use of reason. This boundary belongs as much to the field of experience as to that of beings of thought. At the same time, we learn how those remarkable ideas serve only to determine the boundaries of human reason. On one side, we must not extend knowledge from experience without limit, leaving nothing but the world for us to know. On the other side, we must not go beyond the boundary of experience and try to judge things outside it as things in themselves. We stay on this boundary when we restrict our judgment solely to the relation the world may have to a being whose very concept lies outside all the knowledge we are capable of within the world. For then we do not attribute to the supreme being *in itself* any of the properties through which we think objects of experience. We thereby avoid *dogmatic* anthropomorphism. But we still attribute these properties to its relation to the world. We allow ourselves a *symbolic* anthropomorphism that really concerns only language, not the object itself. When I say that we are compelled to regard the world *as if* it were the work of a supreme understanding and will, I am really saying no more than this: as a watch relates to its maker, a ship to its builder, or a regiment to its commander, so the world of sense—or everything making up the basis of this totality of appearances—relates to the unknown. Through this I do not know the unknown as it is in itself. I do, however, know it as it is for me: in relation to the world of which I am a part. ### Section 58 § 58. This kind of cognition is cognition *by analogy*. Here analogy does not mean, as the word commonly does, an imperfect similarity between two things. It means a perfect similarity between two relations involving things entirely unlike one another.[^357-1] Through this analogy, a concept of the supreme being remains that is sufficiently determinate for us, even though we have left out everything that could *determine it absolutely and in itself*. For we still determine it in relation to the world, and thus to ourselves. We need nothing more. *Hume's* attacks on those who want to determine this concept absolutely, borrowing the materials from themselves and the world, do not affect us. Nor can he accuse us of being left with nothing at all once objective anthropomorphism is removed from the concept of the supreme being. Suppose, at the outset, we are allowed the *deistic* concept of the original being as a necessary hypothesis. Hume himself allows this in his Dialogues, speaking as Philo against Cleanthes. In this concept, the original being is thought solely through ontological predicates such as substance and cause. *We must do this*, because reason is driven through the world of sense by nothing but conditions that are always conditioned in turn. Without this hypothesis it can find no satisfaction. And *we can quite legitimately do this* without falling into anthropomorphism, which transfers predicates from the world of sense to a being entirely distinct from the world. For those predicates are mere categories. They give no determinate concept of the being, but for that very reason give no concept restricted to conditions of sensibility either. Nothing, then, can prevent us from attributing to this being *causality through reason* in relation to the world, and so moving on to theism, without being compelled to attribute this reason to it *in itself* as an inherent property. On the first point, this is the only possible way to develop the use of reason in relation to all possible experience in the world of sense to the highest degree, in complete agreement with itself: we in turn assume a supreme reason as the cause of all connections in the world. Such a principle must benefit reason throughout, and can never harm its use in nature. *Second*, however, reason is not thereby transferred as a property to the original being *in itself*, but only *to its relation* to the world of sense. Anthropomorphism is thus completely avoided. For here we consider only the *cause* of the rational form found everywhere in the world. We attribute reason to the supreme being insofar as it contains the ground of this rational form of the world, but only by analogy. That is, the expression indicates only the relation that the supreme cause, unknown to us, has to the world, determining everything in it in the most rational way. This keeps us from using the property of reason to think of God. Instead, we use it to think of the world as we must in order to make the greatest possible use of reason regarding it, according to a principle. We thereby admit that what the supreme being is in itself is wholly beyond our investigation, and cannot even be thought *in a determinate way*. This keeps us from making a transcendent use of our concepts of reason as an efficient cause through the will. We do not determine the divine nature through properties that are always borrowed only from human nature, and so lose ourselves in crude or fanciful concepts. On the other side, it also keeps us from flooding our study of the world with explanations beyond nature, based on our concepts of human reason transferred to God. We do not divert this study from its proper purpose: to be a study of nature alone through reason, not a presumptuous derivation of its appearances from a supreme reason. The expression suited to our weak concepts will be this: we think of the world *as if* it derived its existence and inner determination from a supreme reason. In this way we partly come to know the character belonging to the world itself, without presuming to determine the character of its cause in itself. On the other side, we place the ground of this character—the rational form in the world—in *the relation* of the supreme cause to the world, without finding the world sufficient for it on its own.[^360-1] In this way, the difficulties that seem to stand against theism disappear. We combine *Hume's* principle—not to extend the use of reason dogmatically beyond the field of all possible experience—with another principle that *Hume* entirely overlooked: we must not regard the field of possible experience as something that sets its own boundaries in the eyes of our reason. Here the critique of reason marks out the true middle way between the dogmatism *Hume* opposed and the skepticism he wanted to introduce in its place. This is not like other middle ways, which people recommend determining for oneself in a more or less mechanical fashion—a little from one side, a little from the other—and which teach no one anything better. It is a middle way that can be determined precisely according to principles. [^357-1]: For example, there is an analogy between the legal relation of human actions and the mechanical relation of moving forces. I can never do something to someone else without giving that person a right to do the very same to me under the same conditions. Likewise, no body can act on another with its moving force without thereby causing the other to react against it with an equal force. Here, right and moving force are entirely unlike things, but their relation is perfectly similar. Through such an analogy I can therefore give a relational concept of things absolutely unknown to me. For example: promoting children's happiness = a relates to parental love = b as the welfare of the human race = c relates to the unknown in God = x, which we call love. This does not mean it has the slightest similarity to any human inclination. Rather, we can represent its relation to the world as similar to the relation things within the world have to one another. Here the relational concept is a mere category: the concept of cause, which has nothing to do with sensibility. [^360-1]: I will say: the causality of the supreme cause is to the world what human reason is to the works of its skill. The nature of the supreme cause itself remains unknown to me. I compare only its known effect, the order of the world, and that effect's rational character, with the known effects of human reason. I therefore call the supreme cause a reason. But I do not thereby attribute to it, as its property, the very same thing I mean by this expression in human beings, or anything else known to me. ### Section 59 § 59. At the start of this remark, I used the image of a *boundary* to establish reason's limitations regarding its appropriate use. The world of sense contains only appearances, which are not things in themselves. Precisely because the understanding recognizes objects of experience as mere appearances, it must therefore assume things in themselves, or noumena. Our reason encompasses both together. The question is: how does reason go about setting boundaries for the understanding with respect to both fields? Experience contains everything belonging to the world of sense, but it does not set its own boundaries. From anything conditioned, it reaches only something else that is conditioned. What is to bound it must lie entirely outside it. This is the field of pure beings of the understanding. But for us this is an empty space insofar as the *determination* of those beings' nature is concerned. Insofar as we seek dogmatically determinate concepts, then, we cannot go beyond the field of possible experience. Yet a boundary is itself something positive. It belongs both to what lies within it and to the space outside a given totality. So reason does gain actual positive cognition simply by extending itself to this boundary. But it must not try to go beyond the boundary, since it finds an empty space before it there. In that space it can think forms for things, but no things themselves. Still, at this standpoint reason retains knowledge of the *bounding* of the field of experience by something otherwise unknown to it. Through this knowledge, reason is neither shut up within the world of sense nor lost in fantasies outside it. Instead, as knowledge of the boundary requires, it restricts itself solely to the relation of what lies outside the boundary to what is contained within it. Natural theology is a concept of this kind on the boundary of human reason. Reason finds itself compelled to look beyond to the idea of a supreme being—and, in a practical respect, also to the idea of an intelligible world. It does not do so to determine anything about this mere being of the understanding, and thus anything outside the world of sense. It does so only to guide its own use within that world according to principles of the greatest possible unity, both theoretical and practical. For this purpose, it uses that world's relation to an independent reason as the cause of all these connections. In doing so, it does not simply *invent* a being for itself. Something that only the pure understanding thinks must necessarily be found outside the world of sense. Reason determines this something only in that way, though admittedly only by analogy. Our proposition above, the result of the whole Critique, therefore remains: “Through all its a priori principles, reason never teaches us anything more than objects of possible experience, nor anything more about these than can be known in experience.” But this restriction does not prevent reason from leading us to the objective *boundary* of experience: its *relation* to something that is not itself an object of experience, but must nevertheless be the highest ground of all experience. Reason teaches us nothing about this something in itself. It teaches us only in relation to its own complete use, directed toward the highest purposes, within the field of possible experience. But that is also all the benefit we could reasonably even wish for here, and we have reason to be satisfied with it. ### Section 60 § 60. We have thus given a detailed account of the subjective possibility of metaphysics as it actually exists in the *natural disposition* of human reason, specifically in what forms the essential purpose of working on it. We found, however, that the *merely natural* use of this disposition of our reason entangles reason in *dialectical* inferences that go beyond experience. Some are merely illusory, and some even conflict with one another. This happens unless a discipline, possible only through scientific critique, restrains reason and sets limits to it. We also found that this argumentative metaphysics is unnecessary for advancing our knowledge of nature, and may even harm it. So a problem still worthy of investigation remains: to discover the *natural purposes* that this disposition toward transcendent concepts in our nature may aim at. For everything in nature must originally be designed for some useful purpose. Such an investigation is certainly precarious. I also admit that what I can say about it is only conjecture, like everything concerning nature's first purposes. This alone may be permitted me here, since the question concerns not the objective validity of metaphysical judgments, but our natural disposition toward them. It therefore lies outside the system of metaphysics, in anthropology. Let me take together all the transcendental ideas. Their totality makes up the proper task of natural pure reason. This task forces reason to leave the mere study of nature, go beyond all possible experience, and in this effort produce the thing—whether knowledge or mere argumentation—that is called metaphysics. I then think I can see the aim of this natural disposition. It is to free our conception far enough from the bonds of experience and the limitations of merely studying nature that it at least sees a field opening before it. This field contains only objects for the pure understanding, which no sensibility can reach. The aim is not to occupy ourselves speculatively with them, since we find no ground on which to gain a foothold. Rather, it is to allow practical principles to expand to the universality that reason absolutely needs for moral purposes. Without such a space before them for their necessary expectation and hope, these principles could not reach that universality. I find, then, that the *psychological* idea at least shows clearly enough the inadequacy of all concepts from experience. It does so however little insight it gives me into the pure nature of the human soul, which stands above all such concepts. It thereby leads me away from materialism, a psychological concept that is unfit for explaining anything in nature and also narrows reason in its practical use. Similarly, the *cosmological* ideas show that all possible knowledge of nature is visibly inadequate to satisfy reason in its legitimate inquiry. They thus serve to keep us from naturalism, which would present nature as sufficient on its own. Finally, all natural necessity in the world of sense is always conditioned, since it always presupposes that things depend on other things. Unconditioned necessity must be sought only in the unity of a cause distinct from the world of sense. Yet if this cause's causality were in turn merely nature, it could never make the existence of the contingent comprehensible as its consequence. Through the *theological* idea, reason therefore frees itself from fatalism: both from blind natural necessity in the connection of nature itself without a first principle, and from blind natural necessity in the causality of that principle itself. It leads to the concept of a cause through freedom, and thus of a supreme intelligence. So although the transcendental ideas do not serve to teach us positively, they do serve to remove the bold claims of *materialism*, *naturalism*, and *fatalism*, which narrow the field of reason. They thereby make room for moral ideas outside the field of speculation. This, I think, would explain that natural disposition to some extent. The practical benefit that a merely speculative science may have lies outside that science's boundaries. It can therefore be regarded only as a supplementary remark. Like all such remarks, it does not belong to the science itself as one of its parts. Still, this relation lies at least within the boundaries of philosophy, especially philosophy that draws from pure sources of reason. There, reason's speculative use in metaphysics must necessarily form a unity with its practical use in morality. The unavoidable dialectic of pure reason, considered in metaphysics as a natural disposition, therefore deserves more than treatment as an illusion requiring resolution. As a natural arrangement, it also deserves to be explained in terms of its purpose, if that is possible. Admittedly, this work goes beyond what is required, and it is right not to demand it of metaphysics proper. A second supplementary remark, but one more closely related to the content of metaphysics, would be the solution to the questions pursued in the Critique from pages 642 to 668. Certain principles of reason are presented there that determine a priori the order of nature—or rather the understanding, which is to seek nature's laws through experience. They seem to constitute experience and legislate for it. Yet they arise from reason alone, which, unlike the understanding, must not be regarded as a principle of possible experience. This agreement may rest on the following. Nature does not belong to appearances or to their source, sensibility, in themselves. It is found only in sensibility's relation to the understanding. Similarly, the understanding may possess thoroughgoing unity in its use for the purposes of all possible experience, within a system, only through its relation to reason. Experience would therefore also be indirectly subject to reason's legislation. Whether this is so may be considered further by those who want to investigate reason's nature beyond its use in metaphysics—even in the general principles for making a natural history in general systematic. I presented this problem as important in the work itself, but did not attempt its solution.[^364-1] This completes my analytical solution to the main question I posed: How is metaphysics possible at all? I started from what actually shows its use, at least in its results, and worked back to the grounds that make it possible. [^364-1]: Throughout the Critique, my constant intention was to leave out nothing, however deeply hidden, that could make the investigation of pure reason's nature complete. Afterward, everyone is free to decide how far to pursue the inquiry, provided they have been shown what investigations might still need to be made. This can reasonably be expected of someone who has made it their task to survey this entire field, then leave it to others for future cultivation and division as they choose. This includes the two supplementary remarks. Their dryness is unlikely to recommend them to amateurs, so they have been included only for experts. ### Solution: How is metaphysics possible as a science? *Answer to the general question of the Prolegomena.* *How is metaphysics possible as a science?* Metaphysics exists as a natural tendency of reason. But on its own it also leads to dialectical conflict and deception, as the analytical answer to the third main question showed. So if we take our principles from this tendency and follow its natural but still false illusion when using them, we can never produce science. We can produce only an empty art of dialectic. One school may outdo another in this art, but none can ever win justified and lasting acceptance. For metaphysics to claim insight and conviction as a science, rather than deceptive persuasion alone, a critique of reason itself must set out everything in a complete system. This includes the whole stock of a priori concepts; their division according to their different sources in sensibility, understanding, and reason; a complete table of them; and an analysis of all these concepts, together with everything that can be inferred from them. Above all, it must establish the possibility of synthetic a priori cognition by a deduction of these concepts. It must also set out the principles for using them and, finally, the limits of that use. Critique, and critique alone, therefore contains the whole carefully tested and verified plan for establishing metaphysics as a science. It even contains all the means for carrying out that plan. No other ways or means can achieve this. So the question here is not so much how the task is possible. It is how to get it started, how to lead able minds away from their previous misguided and fruitless work toward a reliable approach, and how best to direct their combined efforts toward the shared goal. One thing is certain. Anyone who has once tried critique will always be disgusted by the dogmatic chatter he used to put up with out of necessity. His reason needed something, and he could find nothing better to occupy it. Critique stands in exactly the same relation to ordinary school metaphysics as *chemistry* to *alchemy*, or *astronomy* to fortune-telling *astrology*. I guarantee that no one who has thought through and grasped the principles of critique, even just in these Prolegomena, will ever return to that old, sophistical sham science. Instead, he will look forward with real pleasure to a metaphysics now fully within his reach. It needs no further preliminary discoveries, and it can for the first time give reason lasting satisfaction. Among all possible sciences, metaphysics alone can confidently count on this advantage: it can be completed and brought to a stable condition. It will then need no further changes and cannot be enlarged by new discoveries. Here reason finds the sources of its cognition not in objects and its intuition of them, through which it can learn nothing further, but within itself. Once it has completely set out the fundamental laws of its capacity, in a form secure against every misinterpretation, nothing remains that pure reason could know a priori, or could even reasonably ask. The sure prospect of knowledge so definite and complete has a special appeal, even if we set aside all its usefulness, which I will discuss later. Every false art and every empty claim to wisdom lasts for a time. Eventually it destroys itself, and its highest development is also the moment of its downfall. For metaphysics, that time has now come. Its decline among all learned nations proves this, despite the energy devoted to every other kind of science. The old organization of university studies still preserves its shadow. One academy of sciences still occasionally offers prizes to encourage one attempt or another in it. But metaphysics is no longer counted among the solid sciences. Readers can judge for themselves how an intelligent man might receive the well-meant but hardly enviable compliment of being called a great metaphysician. The time for the decline of all dogmatic metaphysics has undoubtedly arrived. But much is still missing before we can say that the time for its rebirth through a thorough and complete critique of reason has also arrived. Every transition from one inclination to its opposite passes through a state of indifference. This moment is the most dangerous for an author, but in my view it is the most favorable for science. Once former alliances have broken up completely and partisan feeling has died out, minds are in the best condition to listen gradually to proposals for joining forces on a different plan. I hope these Prolegomena may stimulate research in the field of critique. I hope they may give the general spirit of philosophy a new and promising subject to occupy it, since its speculative side seems to lack nourishment. When I say this, I can already imagine the question from everyone whom the thorny paths I led them along in the Critique have irritated and exhausted: What grounds do I have for this hope? My answer is: *the irresistible law of necessity*. We can no more expect the human mind to give up metaphysical investigations entirely than expect ourselves to stop breathing altogether to avoid always breathing impure air. So there will always be metaphysics in the world. More than that, everyone, especially every reflective person, will have one. Without a public standard, each will shape it in his own way. Now, what has so far been called metaphysics cannot satisfy a questioning mind. Yet giving it up altogether is also impossible. So in the end a critique of pure reason itself must be *attempted*. Or, if one already exists, it must be *examined* and submitted to general scrutiny. There is no other way to meet this urgent need, which is more than mere curiosity. Since becoming familiar with critique, whenever I have finished reading a work on metaphysics that both entertained and educated me through its precise concepts, its variety and order, and its easy presentation, I have been unable to avoid asking: *Has this author advanced metaphysics by even one step?* I ask the learned men whose writings have helped me in other ways, and have always helped develop my mental powers, to forgive this admission. Neither in their attempts nor in my own lesser ones, for which self-love still speaks favorably, have I found that the science has advanced in the slightest. The reason is quite natural. The science did not yet exist, and it cannot be assembled piece by piece. Its seed must first be fully formed in critique. To prevent any misunderstanding, we must remember what was said earlier. An analytical treatment of our concepts can certainly do a great deal for the understanding. But it does not advance the science of metaphysics in the slightest. These analyses of concepts are only materials from which the science must first be built. We may analyze and define the concept of substance and accident as well as we like. This is very good as preparation for some future use. But unless I can prove that substance persists in everything that exists and only accidents change, all that analysis has not advanced the science in the slightest. So far metaphysics has been unable to give a valid a priori proof of this proposition or of the principle of sufficient reason. Still less has it proved any more complex proposition, such as one belonging to the theory of the soul or to cosmology. In fact it has proved no synthetic proposition at all in this way. So all that analysis has accomplished, created, and advanced nothing. After so much activity and noise, the science is still where it was in *Aristotle's* time. Admittedly, the preparations for it are much better than before, if only a guiding thread for synthetic cognition had first been found. Anyone who feels offended by this can easily refute the accusation. He need only present a single synthetic proposition belonging to metaphysics that he offers to prove a priori by the dogmatic method. Only when he supplies that proof will I concede that he has really advanced the science, even if ordinary experience already confirms the proposition well enough. No demand could be more moderate or fair. And if he fails, as he inevitably and certainly will, no judgment could be more just than this: metaphysics as a science has so far never existed. If the challenge is accepted, I must exclude just two things. First, no playing with *probability* and guesswork. They are as inappropriate in metaphysics as in geometry. Second, no settling the matter with the divining rod of so-called *common sense*. This rod does not respond for everyone; its responses depend on personal characteristics. *As for the first*, nothing could be more absurd than trying to base judgments on probability and guesswork in metaphysics, a philosophy drawn from pure reason. To claim that something is known a priori is to claim that it is absolutely and necessarily certain. So it must also be proved in that way. We might just as well try to base geometry or arithmetic on guesses. Arithmetic's calculus probabilium does not contain probable judgments. It contains entirely certain judgments about the degree of possibility of particular cases under given conditions of the same kind. Across the total of all possible cases, these must unfailingly agree with the rule, even though the rule is not definite enough for each individual occurrence. Only in empirical natural science can conjectures be allowed, through induction and analogy. Even there, at least the possibility of what I assume must be completely certain. The *appeal to common sense* is, if possible, in an even worse position. Here we are talking about concepts and principles not insofar as they are supposed to hold for experience, but insofar as they are claimed to hold even outside the conditions of experience. What is *sound understanding*? It is *ordinary understanding* insofar as it judges correctly. And what is ordinary understanding? It is the ability to know and use rules in concrete cases, as distinct from *speculative understanding*, which is the ability to know rules in the abstract. Ordinary understanding will scarcely understand the rule that everything that happens is determined through its cause. It can never grasp it in this general form. So it asks for an example from experience. When it hears that the rule means nothing different from what it has always thought when a windowpane broke or a household item went missing, it understands and accepts the principle. Ordinary understanding therefore has no use beyond cases where it can see its rules confirmed in experience, even though it really does possess them a priori. Understanding them a priori and independently of experience belongs instead to speculative understanding. It lies entirely outside the range of ordinary understanding. But metaphysics deals solely with this latter kind of cognition. It is surely a poor sign of sound understanding to appeal to a witness who has no judgment at all in this matter. Usually this witness is looked down on, except when people find themselves cornered and cannot find any guidance or help in their speculations. These false friends of ordinary common sense praise it highly on occasion but usually despise it. They have a familiar escape. They say that there must ultimately be some propositions that are immediately certain. Not only do these need no proof; no account of them at all need be given. Otherwise we would never reach an end to the grounds of our judgments. But to justify this claim, they can never cite anything beyond mathematical propositions as unquestionably and immediately available to ordinary common sense. Examples are that twice two is four and that only one straight line lies between two points, and so on. They can also cite the principle of contradiction, but that is not enough to establish the truth of synthetic judgments. Mathematical judgments, however, are worlds apart from those of metaphysics. In mathematics I can make, or construct, through my own thinking everything that a concept represents to me as possible. I add another two to two step by step, and make the number four myself. Or I mentally draw all kinds of lines from one point to another. I can draw only one that is alike in all its parts, whether those parts are equal or unequal. But no amount of thinking can extract from the concept of one thing the concept of something else whose existence is necessarily connected with the first. I have to consult experience. My understanding does provide me a priori with the concept of such a connection, causality, though always only in relation to possible experience. Yet I cannot present this concept a priori in intuition, as I can with mathematical concepts, and so establish its possibility a priori. Instead, if this concept and the principles for applying it are to be valid a priori, as metaphysics requires, their possibility always needs a justification and a deduction. Otherwise we do not know how far the concept is valid, or whether it can be used only within experience or outside it as well. So in metaphysics, as a speculative science of pure reason, we can never appeal to ordinary common sense. We can do so, however, when we are forced to leave metaphysics behind and give up all pure speculative cognition, which must always be knowledge. That means also giving up metaphysics itself and its teachings on certain matters. In that case, reasonable belief may turn out to be all that is possible for us and enough for our needs—perhaps even healthier than knowledge itself. The situation is then completely different. Metaphysics must be science, not just as a whole but in every part. Otherwise it is nothing at all. As speculation by pure reason, it has no support anywhere except in universal insights. Outside metaphysics, however, probability and common sense can certainly have their useful and legitimate applications. But these follow principles of their own, whose weight always depends on their relation to practical matters. These are the requirements I consider myself entitled to set for the possibility of metaphysics as a science. ### Appendix *Appendix* *On what can be done to make metaphysics an actual science.* All the paths followed so far have failed to reach this goal. Nor is it ever likely to be reached without a prior critique of pure reason. So it seems reasonable to ask that the attempt now presented be examined closely and carefully. The alternative is to consider it wiser to give up all claims to metaphysics altogether. There is no objection to that course, provided people stay true to their decision. If we take events as they actually unfold, rather than as they should, there are two kinds of judgment. One is a *judgment made before the investigation*. In our case, this occurs when a reader uses his own metaphysics to judge the critique of pure reason, whose task is first to investigate whether metaphysics is possible. The other is a *judgment made after the investigation*. Here the reader can put aside for a while the consequences of the critical investigations, which may conflict sharply with the metaphysics he has accepted until now. He first examines the grounds from which those consequences may have been derived. If the claims of ordinary metaphysics were settled certainties, like those of geometry, the first kind of judgment would be valid. For if the consequences of certain principles conflict with established truths, those principles are false and must be rejected without further investigation. But suppose metaphysics does not possess a stock of indisputably certain synthetic propositions. Perhaps many of its propositions, as plausible as the best of them, even conflict with one another in their consequences. And suppose it has no reliable criterion at all for the truth of genuinely metaphysical, synthetic propositions. In that case, the first way of judging cannot apply. The principles of critique must be investigated before any judgment is made about their value or worthlessness. *Example of a judgment on the Critique made before investigation.* A judgment of this kind appears in the Göttingische gelehrte Anzeigen, the supplement's third issue, January 19, 1782, page 40 and following. Suppose an author knows the subject of his work well and has worked hard to bring his own thinking to every part of it. His work falls into the hands of a reviewer sharp enough to spot the points on which its value or worthlessness really depends. The reviewer does not get stuck on words but follows the substance of the argument, and does more than merely inspect and test the principles from which the author started. The author may dislike the severity of the judgment, but the public need not mind, because it benefits. The author himself can be glad to have an opportunity to correct or clarify his arguments after an expert has tested them early on. If he believes he is basically right, he can thus remove in good time an obstacle that might later damage his work. My situation with my reviewer is entirely different. He seems not to see at all what really mattered in the investigation I undertook, successfully or unsuccessfully. Perhaps he was impatient with thinking through a long work. Perhaps he was irritated by the threat of reforming a science in which he thought he had long since settled everything. Or perhaps, as I am reluctant to suspect, the cause was a genuinely limited outlook that keeps him from ever thinking beyond his school metaphysics. In short, he rushes aggressively through a long series of propositions that mean nothing to anyone who does not know their premises. He scatters criticisms here and there. The reader can no more see the grounds for these criticisms than understand the propositions they are directed against. So the review can neither usefully inform the public nor do me the slightest damage in the judgment of experts. I would therefore have ignored it altogether if it had not given me an opportunity for a few clarifications. These may protect the reader of these Prolegomena from misunderstandings in certain cases. The reviewer wants a viewpoint from which he can most easily present the whole work unfavorably to its author, without having to undertake any particular investigation. So he begins, and also ends, by saying: “This work is a system of transcendental—or, as he translates it, higher—[^373-1] idealism.” As soon as I saw this line, I knew what sort of review would follow. It was much as if someone who had never heard or seen anything of geometry found a copy of *Euclid* and was asked to judge it. After leafing through it and coming across many figures, he might say: “This book is a systematic guide to drawing. The author uses a special language to give obscure, unintelligible instructions. In the end they can achieve no more than anyone can manage with a naturally good eye, etc.” Still, let us see what kind of idealism this is. It runs through my whole work, although it is far from constituting the heart of the system. The position of all genuine idealists, from the Eleatic school to Bishop *Berkeley*, is contained in this formula: “All cognition through the senses and experience is nothing but illusion. Truth is found only in the ideas of pure understanding and reason.” By contrast, the principle that governs and defines my idealism throughout is this: “All cognition of things through mere pure understanding or pure reason is nothing but illusion. Truth is found only in experience.” But that is the exact opposite of idealism in the strict sense. How, then, did I come to use this expression for a completely opposite purpose? And how did the reviewer come to see it everywhere? The solution to this difficulty depends on something that could very easily have been understood from the work's context, if anyone had wanted to understand it. Space and time, together with everything they contain, are not things or their properties in themselves. They belong only to the appearances of things. So far I agree with those idealists. But they, especially *Berkeley*, considered space a merely empirical representation. On their view, space and all its determinations become known to us only through experience or perception, just like the appearances within it. I, by contrast, first show that we can know space and all its determinations a priori. The same holds for time, which *Berkeley* did not consider. Space, like time, is present in us before all perception or experience as the pure form of our sensibility. It makes every intuition of sensibility possible, and therefore every appearance as well. Truth depends on universal and necessary laws as its criteria. It follows that for *Berkeley* experience can have no criteria of truth, because he gave its appearances no a priori foundation. The result was that experience was nothing but illusion. On our account, by contrast, space and time, combined with the pure concepts of the understanding, prescribe their law a priori to all possible experience. This law also provides a reliable criterion for distinguishing truth from illusion within experience.[^375-1] My so-called idealism, which is properly critical idealism, is therefore of an entirely distinctive kind. It overturns ordinary idealism. Through it, all a priori cognition, even geometry, gains objective reality for the first time. Without the ideality of space and time that I have proved, even the most committed realists could not maintain that objective reality. Under these circumstances, I wish I could give my concept a different name to prevent every misunderstanding. But changing its name altogether would hardly be workable. So let me call it in future, as I already did above, formal idealism, or better still, critical idealism. This distinguishes it from *Berkeley's* dogmatic idealism and *Descartes's* skeptical idealism. I find nothing else worth noting in the review of this book. Its author judges everything in broad terms, a cleverly chosen method, since it does not reveal what he knows or does not know. A single detailed judgment, if it had properly addressed the main question, might have exposed my error. Or it might have exposed the extent of the reviewer's insight into this kind of investigation. His other tactic was also not badly devised. Some readers are used to forming their opinions of books solely from newspaper reports. To discourage them from reading the book itself right from the start, he races in one breath through a mass of propositions torn from their supporting arguments and explanations. In that state they must sound absurd, especially when they are as completely opposed to all school metaphysics as mine are. He assaults the reader's patience to the point of disgust. Then, after introducing me to the ingenious proposition that lasting illusion is truth, he ends with a blunt but fatherly lecture: “Why this dispute with generally accepted language? What is the point of the idealistic distinction, and where does it come from?” This judgment finally reduces everything distinctive about my book to a mere innovation in language, though earlier it was supposed to be metaphysical heresy. It clearly proves that my self-appointed judge has not understood the slightest thing about the book, and has not properly understood himself either.[^376-1] Yet the reviewer speaks like a man who must be aware of important and outstanding insights that he is still keeping hidden. I have recently learned of nothing in metaphysics that could justify such a tone. But he is very wrong to withhold his discoveries from the world. No doubt others have had the same experience as I have. Despite all the fine things written in this field for a long time, they have been unable to find that these writings have advanced the science by even a finger's width. We do still find sharpened definitions, new crutches for lame proofs, and new patches or a different cut for metaphysics' patchwork garment. But that is not what the world wants. The world has had enough of metaphysical assertions. People want to know whether this science is possible, what sources can provide certainty in it, and what reliable criteria distinguish pure reason's dialectical illusion from truth. The reviewer must possess the key to this. Otherwise he would never have spoken in such a lofty tone. But I begin to suspect that this need of the science may never have entered his mind. Otherwise he would have directed his review to this point. Even an unsuccessful attempt in so important a matter would have earned his respect. If that is the situation, we are good friends again. He may think his way as deeply into his metaphysics as he likes. No one should stop him. But he cannot judge what lies outside metaphysics: its source within reason. My suspicion is not groundless. The proof is that he said not a word about the possibility of synthetic a priori cognition. That was the real problem. The entire fate of metaphysics depends on its solution, and my Critique, just like these Prolegomena, was directed wholly toward it. The idealism he stumbled upon and then got stuck on was included in the doctrine only as the sole means of solving that problem, though it was also confirmed on other grounds. He should therefore have shown one of three things: that the problem is not as important as I claim, here in the Prolegomena as well; that my concept of appearances does not solve it at all; or that it can be solved better in some other way. But I find not a word of this in the review. So the reviewer understood nothing of my work, and perhaps nothing of the spirit and nature of metaphysics itself. Or perhaps, as I prefer to assume, his haste as a reviewer was irritated by the difficulty of working through so many obstacles. This cast an unfavorable shadow over the work before him and made even its basic outlines unrecognizable to him. A learned journal has a very long way to go before it can maintain in metaphysics the reputation it otherwise deserves, however carefully and well its contributors are chosen. Other sciences and branches of knowledge have their standards. Mathematics has its standard within itself. History and theology have theirs in secular or sacred books. Natural science and medicine have theirs in mathematics and experience. Law has its codes, and even matters of taste have the models of antiquity. But the standard for judging what is called metaphysics has yet to be found. I have tried to define both that standard and its use. Until it is established, what should we do when works of this kind still have to be judged? If they are dogmatic, people may do as they please. No one will play the master over someone else here for long without finding someone to pay him back in kind. But suppose the works are critical, and their criticism is directed not at other writings but at reason itself. The standard of judgment cannot then already be assumed; it must first be sought. In that case objections and criticism are welcome, but they must rest on mutual tolerance. The need is shared, and the lack of the necessary insight makes any claim to a judge's decisive authority inappropriate. To connect this defense of mine with the interests of the philosophical community, I propose a test. It will settle how all metaphysical investigations must be directed toward their shared goal. The proposal is simply what mathematicians have sometimes done to decide the superiority of their methods in a competition. I challenge my reviewer to use his own method to prove any single genuinely metaphysical proposition he asserts. It must, that is, be synthetic and known a priori from concepts. He may even choose one of the most indispensable, such as the principle that substance persists, or that events in the world are necessarily determined by their causes. But he must prove it properly, on a priori grounds. If he cannot do this—and silence counts as an admission—he must concede the following. Without the absolute certainty of propositions of this kind, metaphysics is nothing at all. So before anything else, its possibility or impossibility must be settled in a critique of pure reason. He is therefore bound either to admit that my principles of critique are correct or to prove them invalid. He has so far relied quite confidently on the certainty of his principles. But I can already foresee that, when it comes to a strict test, he will not find a single one in the whole range of metaphysics that he can confidently put forward. So I will grant him the most favorable terms anyone could expect in a contest. I will take the burden of proof off him and place it on myself. In these Prolegomena and in my Critique, pages 426–461, he will find eight propositions. The propositions contradict one another in pairs. Yet every proposition necessarily belongs to metaphysics, which must either accept it or refute it. In fact, every one of them has at some time been accepted by some philosopher. He is now free to choose whichever of these eight propositions he likes and accept it without proof. I waive that requirement for him. But he must choose only one, since wasting time serves him no better than it serves me. He can then attack my proof of the opposite proposition. Suppose I can still defend my proof and show in this way that the opposite of his chosen proposition can be proved just as clearly from principles that every dogmatic metaphysics must necessarily acknowledge. This will establish that metaphysics has an inborn defect. The defect cannot be explained, much less removed, unless we go back to its birthplace, pure reason itself. My Critique must then either be accepted or replaced by a better one. At the very least, it must be studied. That is all I ask for now. If, on the other hand, I cannot defend my proof, then a synthetic a priori proposition based on dogmatic principles is established on my opponent's side. My accusation against ordinary metaphysics was therefore unjust. I offer in that case to recognize his criticism of my Critique as justified, though this would still be far from following from the result. For this purpose, however, I think he would have to stop remaining anonymous. Otherwise I do not see how I could avoid being honored, or assailed, with several challenges instead of one by unnamed and still uninvited opponents. *Proposal for an investigation of the Critique that can be followed by a judgment.* I am grateful to the learned public even for the silence with which it has honored my Critique for a considerable time. This at least shows a suspension of judgment. It suggests some suspicion that a work which leaves every familiar path and takes a new one, not immediately easy to follow, may still contain something worthwhile. Perhaps it can give new life and fruitfulness to an important but now lifeless branch of human knowledge. The silence therefore shows caution about breaking off and destroying this still tender graft with a hasty judgment. Just now an example of a judgment delayed for such reasons has come to my attention in the Gothaische gelehrte Zeitung. Every reader will recognize its thoroughness from the clear and undistorted presentation of a point belonging to my work's first principles, without needing to consider my praise, which is suspect in this matter. A large building cannot possibly be judged as a whole straight away by a quick survey. So I now propose that my work be examined piece by piece, starting from its foundations. These Prolegomena can serve as a general outline, with which the work itself can be compared from time to time. If this request rested on nothing but my imagined sense of importance, which vanity usually gives all our own products, it would be immodest and deserve an irritated rejection. But the affairs of speculative philosophy as a whole are now on the point of dying out completely. Human reason still clings to them with an undying inclination. Only because it is constantly disappointed does that inclination now try, though in vain, to turn itself into indifference. In our reflective age, we cannot suppose that many deserving men would fail to use every good opportunity to work for the shared interests of reason as it becomes increasingly enlightened. They need only some hope that their efforts can reach the goal. Mathematics, natural science, laws, the arts, even morality, and so on, do not yet fill the mind completely. A space always remains in it, set aside for pure speculative reason alone. Its emptiness forces us to seek what looks like occupation and entertainment in absurdities, trivial games, or even wild fantasies. In reality, we seek only distraction to drown out reason's troublesome call. In keeping with its purpose, reason demands something that will satisfy it for its own sake, not merely keep it busy serving other goals or the interests of our inclinations. An investigation concerned solely with this sphere of reason, which stands on its own, therefore has a great appeal, as I have good grounds to suppose, for anyone who has tried to extend his concepts this far. For it is precisely within this sphere that all other knowledge, and even all goals, must meet and unite into a whole. I may even say that it has a greater appeal than any other theoretical knowledge, which one would not readily accept in exchange for it. I propose these Prolegomena, rather than the work itself, as the plan and guide for the investigation for the following reason. I am still quite satisfied with the work's content, its order and method of teaching, and the care given to each proposition. Before putting any proposition forward, I weighed and tested it carefully. It took years to satisfy myself fully about the whole, and sometimes even about the sources of a single proposition. But I am not fully satisfied with my presentation in some sections of the Doctrine of Elements, such as the deduction of the concepts of the understanding or the section on the paralogisms of pure reason. A certain wordiness in these sections gets in the way of clarity. Instead, the examination can be based on what these Prolegomena say about those sections. Germans are praised for being able to go further than other peoples in tasks requiring perseverance and sustained hard work. If this opinion is justified, here is an opportunity to confirm it. We can bring to completion a task that has not yet succeeded, although its successful outcome can hardly be doubted and all thinking people share an equal interest in it. This is especially so because the science concerned is of such a distinctive kind. It can be brought all at once to full completion and to that *stable* condition in which it cannot be advanced in the slightest. Later discoveries can neither enlarge it nor even change it. I am not counting refinements that increase clarity here and there, or added usefulness for various purposes. No other science has this advantage or can have it, because none concerns a cognitive capacity so completely isolated, independent of others, and unmixed with them. The present moment also seems favorable to my request. In Germany today, people scarcely know what they can occupy themselves with apart from the so-called useful sciences, if the occupation is to be not just play but also work that achieves a lasting goal. I must leave it to others to devise ways of combining the efforts of scholars toward such a goal. But I do not mean to ask anyone simply to follow my propositions, or even to flatter myself with the hope that they will. There may be attacks, repetitions, limitations, or, on the other hand, confirmation, additions, and extensions, as the case may be. If only the matter is investigated from the ground up, a system of doctrine can no longer fail to emerge, even if it is not mine. It can become a legacy to posterity for which later generations will have reason to be grateful. It would take too long to show here what kind of metaphysics we can expect from the principles of critique once those principles have been properly established. Removing its borrowed feathers need not leave metaphysics looking poor and insignificant. In another respect it can appear richly and suitably equipped. But the other great benefits of such a reform are immediately clear. Ordinary metaphysics already did some good by seeking out the elementary concepts of pure understanding, making them clear through analysis and precise through definitions. In this way it trained reason for whatever use reason might later choose to make of itself. But that was all the good it did. It canceled this merit by encouraging conceit through reckless assertions, sophistry through subtle evasions and excuses, and shallowness through the ease of disposing of the hardest problems with a little school learning. This shallowness is all the more seductive because it can choose to borrow something from the language of science on one side and from popular appeal on the other. It thereby becomes all things to everyone, but in reality is nothing at all. Critique, by contrast, gives our judgment a standard by which knowledge can reliably be distinguished from the mere appearance of knowledge. By being put fully into practice in metaphysics, critique establishes a way of thinking whose beneficial influence then extends to every other use of reason. For the first time, it instills the true philosophical spirit. Nor should we underestimate its service to theology. It makes theology independent of the judgment of dogmatic speculation and thereby secures it completely against all attacks by opponents of that kind. Ordinary metaphysics promised theology much support but could not later fulfill the promise. Worse, by calling speculative dogmatism to its aid, it had done nothing but arm enemies against itself. In an enlightened age, fanatical fantasy can arise only by hiding behind a school metaphysics. Under that protection it dares to rave with reason, so to speak. Critical philosophy drives it out of this last hiding place. Beyond all this, it must surely matter to a teacher of metaphysics that he can at last say, with general agreement, that what he teaches is finally *science*, and that it therefore provides a real benefit to the community. [^373-1]: Certainly not *higher*. High towers and the metaphysically great men who resemble them are not for me. There is usually plenty of wind around both. My place is the fertile *depths* of experience. The word transcendental does not mean something that goes beyond all experience. It means something that precedes experience a priori, but is intended for nothing more than making knowledge through experience possible. I have explained its meaning in so many ways, yet the reviewer has not even grasped it. That is how superficially he looked at everything. When these concepts go beyond experience, their use is called transcendent. This is distinguished from their immanent use, that is, use restricted to experience. The work has taken sufficient precautions against all misinterpretations of this kind. But the reviewer found misinterpretations to his advantage. [^375-1]: Idealism in the strict sense always has an aim of indulging in speculative fantasies, and can have no other. Mine, by contrast, aims solely to make sense of the possibility of our a priori cognition of objects of experience. This problem has never yet been solved, or even raised. This brings down the whole of that fanciful idealism. It always inferred from our a priori cognitions, even those of geometry, a different kind of intuition from sensory intuition, namely intellectual intuition. We can already see this in *Plato*. No one ever thought that the senses might also intuit a priori. [^376-1]: The reviewer is mostly fighting his own shadow. When I contrast the truth of experience with dreams, he never thinks that I am referring only to the familiar “dream taken objectively” of *Wolffian* philosophy. This is purely formal. It pays no attention at all to the distinction between sleeping and waking, nor can it do so in a transcendental philosophy. He also calls my deduction of the categories and my table of the principles of the understanding “commonly known principles of logic and ontology, expressed in an idealistic manner.” Readers need only look up these points in the Prolegomena to convince themselves that no more wretched judgment, or one more historically incorrect, could possibly be made.